My Tektronix®  SCOPE rebuild page.                                                             
Model  2430.  (non "A" version, only)  I  am not affiliated with Tektronix, but old logo's used with permission!
It is said , that this scope is the last scope , a tech or Ham guy can fix him self. ( as long as the CCD and Amp's custom are not bad , nor CRT bad)
Quote HP "It has a real-time useful storage bandwidth of 40 MHz for single-event acquisitions and an equivalent-time bandwidth of 150 MHz for repetitive acquisitions."
100 megasamples per second ! only. (no star performance here, expect aliasing to the max here.)   (Pay about $100 or less this old, less if dead)
To buy one is easy, the self tests pass and the extended tests!   (see my Calibration section below)
A far better old scope and low price $800, would be TeK, TDS784 . 4 Giga Sampes a second ,4 channels.!

Menu and Index to all the below topics

F269 +5vd fuse blew and all LEDs dead and dot center of the CRT (turn it off fast or damage CRT) (seen here)

How to do the full calibration lacking the PG506 rare generator.

How to find and remove power rail shorts (cap.'s love to short after 20 years life span)

Change out the large, NVRAM battery first (KEEPER-II)

Manuals free.

My new to me, OLD Tek 2430 (1986) (non "A") 
2013, purchased, a real DSO digital scope.   S/N 0116xx  options IR and 11.    
Do not miss this genius guys Pro grade hacking ,Dr. H. Holden (see how he captures the data on DALLAS NVRAM)
I bought it only for 2 reasons, nostalgia and it works. Cost me less than a set of new probes ! in pristine shape. (less that 2 potted plants a Lowes , LOL)
This is a very complex scope that may be near impossible to fix if  someone overheated the AMP/CCD modules inside. (ran it with dead fan or blocked fan ports or vents , boom)
The NVram battery going dead causes the scope to go nuts, all do that page 6-28, never fear this can be cured. If you see ChkSUM fail, 6000 stop now, do not read the other failures, see NVRAM.
Soon as you see CKSUM, know for a fact the battery is dead.
The second common fail (folks email-me) is the 1 electrolytic cap shorts and power is lost. (easy fix if you know how to isolate shorts on this complex daisy changed power system)
5 years later my ribbon cable pin1 insulation shorts to case blowing the +5vd fuse,  and NVram was scrambled, both and easy fix. (and I installed a new battery while inside the unit)
This SCOPE, shows TEK engineers (a pure marvel by me! and genius)  attempting the impossible then,  to make a DSO in 1986, (be my guess 2 years early this effort)  They DID IT.
Not only that, it is has self calibration features, in 1986.  (2 totally amazing feats for that year !)

See my non "A" slide show?

The comments below, in the REBUILD sections, are hands in side the scope, are with the power cord removed.
Plus stay out, of the High voltage section 2-14 KV and CRT leads and there is 400vdc in the left side of A16,  TAKE DUE CAUTIONS.
That is a Cathode Ray Tube and NOT some LCD screen. at 5vdc.

This next photo is just showing how nice this scope can present an asymmetrical signal.   Impossible  to do this magic sync up, below on my analog 465.!

That gap below causes my 465 scope  to  go crazy,  A new DSO will suck it out like a blood hound and so can the 2430 below . , I roll the horz knob and bingo.
a Ford  V8 engine crank sensor, I can see 36tooth signal.
The above signal (crank angle sensor)  or cam angle sensors are all easy to view here.

This DSO scope, THEORY of Operations:  

The best theory pages are in the ARMY manual. For sure.The  Military manual. (best book on theory)

MY REFURBISHED: 2430 - non-"A":  (and some diagnostic checks , if needed and some repair hints below)
See my slide show?

Rule 1:
Do no run this Scope, with the Case cover removed, with out a  12" min. External Fan  blowing on those many bottom card  heat sinks!!! (or it will melt down, and that is the END)
See those hot chips here
custom analog heat sinked chips

Rule 2: Do not run the scope with card A11 J100 removed or you will burn a hole in the screen phosphors. That also means, running scope with A11card, disconnected.
In fact pulling many connectors can do this burn up trick (I'm too chicken to try them all, as my scope runs like new)
To be safe, killing sweep generation like that, remove the A17 low voltage plug first. (lower left corner) P/J176
Rule 3: if getting your body near high voltage of CRT,  unplug P/J176 (A17)  BE AWARE that kills the built in fan too.  -12VDC is fan power.

For sure I blew out all dust inside, with low pressure air (9 PSI book says) then cleaned all dust off each heat sink.  (the latter , mandatory basic service and is in the SM  book)

Hard Failures.?   Diagnosis for dead screen or dead processors:
The +5v  Fuse F269 on the regulator card may  blow out and the CRT then shows a dot in the screen, do let it do that long or the phosphors can be damaged on the front of the CRT.
The fuse is hidden deep inside on card A16 Tab22 in book, behind the ribbon cable.
If the fuse blows again, you most likely have a shorted cap.  (Electrolytic Capacitor)  takes these actions. (and facts and skills)
The cards all run power through wide 3M standard ribbon cables.   
The 9 pins are even and odd side counted,  like this  1,3,5,7,----39   (pin one has square PC pad)
and ------------------------------------------------------------- 2,4,6,8-----40 
The tactic to use is two fold, that is pull the cables to find which card is shorted, first.  (  and second :know the paths and pull the correct cards in proper order)
To do this magic is not easy, you must read the card page schematics to see how any parallel or worse daisy chained connections are done,  (key fact find the farthest  PCB card and pull it first.)
The  +5v  power bus  can be seen on the tab3 , A13 card,  look there first, and all cards that use  plus 5vdc.
Most the cards all have 5vdc used (both +5vd or +5v (analog) )
Then pull the farthest connector first, any card farthest from the A16 reg card.
Once the card that is shorted is in your hand,  remove caps to find the one shorted.  (keep and open mind, any thing can short but Caps are like 100 fold more likely with a 20 year life span.
I will show more depth testing +5vdc shorts here.

The good news is that most failures can be fixed , except, the heat sinked parts blown up and VTI U670.
One could find a matching bad scope and cannibalize it?
The TTL logic chips are still sold and the Dac10s and the TL0 analog chips. (in a word, all small chips)

The real factory SM book: (one of the best books ever published on a topic, as complex as this)  Down load it and safe it. It's Free.
The Service Manual PDF is here. (30 mbyte size) (commented  battery issues , and power loss issues "OCR search able too !)

oops this page in the PDF  is blurry, see these 2 pages I scanned my self, 200 DPI resolution  Intra-connect A and   Intra B.

The only weak spots in the manual are lack of good  independent DC power distributions drawings (above is ok though)
  (or PCB fab drawings , so we can see all PCB traces that run power)
In both photos above. I  marked every negative supply with  the PDF manual, it's hard to tell minus from plus.
These 2 pages Intra connect show where all supplies start (LV or HV cards ) and all loads.

Hard failure like dead screen, or diagnostic errors?  I tried to help with the each class error below.
  • The CRT Screen is dead?
  • Diagnostic screen , errors?  1000, 2000... etc?  Do the relays inside, click as it boots up?
  • The Trigger, Ready, ARM, Plus, NEG. LED flash funny codes?
  • The same LED are stuck on, just one and flashes but is not code. This and the above is covered in the book.
  • One supply is dead, or the -15v is dead causing the others to be dead (normal that later)
  • One card "PCB" is shorted? A power buss feed is shorted? PCB means Printed Circuit card.
  • NVram battery failed. the NON "A" model has  simple battery external the "A" version as  better Dallas chip with deep hidden batteries (2) inside.

(see page 37 block diagram  part 2 )   all pages refer to this book here.

The Pages to read first are about boot diagnostics. Chapter 6 page 27.
Then Page 6-34 failure logic table ,   if not signs of life then see Procedure 8. on page 6-93 CALLED dead start checks.
 It tells you to check all supplies (D.C. power) first.  This scopes power supply converts 120vac to 400DC first, then many other supplies next.
The scope will not run with any DC supplies bad. If the base line DC supplies are good the digital signal called, DCOK goes to TRUE state. (logic)
If DCOK fails, the scope will flash error codes. (that is, if +5vdc is still good, processor power)   << no magic

System Diagnostics.  (on board diagnostics one of the first systems to have it !)
The IS THERE LIFE test, is in the SM book Check  all  LED trigger errors and the GPIB LED set,  first, if screen is dead. (the LEDs talk to you, by flashing errors or are DEAD)
If the LED's send blink codes 6-28   and if it can't blink codes the trigger section just blinks dumb like no code. this too has meaning as does dead GPIB LEDs.
IF the trigger LED's act dumb (oddly) (boot up  flashing is dead) or do not flash out codes but in the stead, they just blink with no code, that is shown on page 6-94   (flash but no code page)
So lets say you expect a bad power supply.?  I have updated my slide show , so you can see real photos of the TEST points (side card), and how to access the LV card. (low voltage supply)

The steps to TILT up the top board is shown in the slide show. The 2 cam locks (plastic) are real tricky buggers. (go easy there)  The last chapter in the SM book shows this too.
Yes, you can run the scope with the top card tilted up and fans blowing on the bottom side (I  use  2"x4"s (wood) to get a path under the scope for air flow)  Think about how to do that!

Before working an any scope, we check that all supplies are in spec.  (or the Calibration can and will fail)
The supplies (some are load voltage sensed,  (kelvin connects)  and must be terminated.  (at one  end  or the other) as seen on our +5v power feed.
 ( These Supply sense lines do  have  100 terminators, to allow open loads, but do be careful)

If you have, "Is there LIFE" failures (in the book tests) the first place  to measure power is here. The A13 card (right side card)  check them all.

If some are way low, they are bad, so the next stop is the LV supply, well buried A16 here, after tilting up the top PCB board.
There is a well hidden fuse  on the rear right of A16 behind that fat ribbon cable.  (must read 0 ohms with any resistance meter)
The slide show, shows the way to get there.
If you need to go deeper to the front UNREG side of the A16 card, the metal cover needs to be removed. 
Warning 1, there is LETHAL 400 vdc under that cover! DO NOT TOUCH THIS until power off and grounded with a HV grounding  probe. LEARN HOW 1st.

A dead supply is easy to find, just don't touch fingers there, as you are doing LIVE testing, and probe with meters and scopes. (yes , it may take a scope to fix a scope)
Any short on a supply bus, will cause that  regulator to shut down. (via over current and heat sensors inside the IC, that does this regulation)
Connect the meter, (using ball clip leads) then with hands out of the scope, turn it on (hands free and live)  <<< safety101
This is standard procedures working on any High voltages. (or near) Anything over 36v is high voltage.(UL rules)

All comments below are from the above book and exact PDF pages, (adobe reader page numbers) or in this fashion chapter 6- page 27.  6-27)
The scope if it fails, the boot up Diagnostics  it flashes the Trigger LED in a binary sequence to tell you, which test fails. PDF page 208 shows you how to do this. (the 8 steps)
PDF page means Adobe reader page number.
This scope must have 2 running processors (main +display) to flash those codes  (read, has some smarts) See PDF page 205+
In most cases the flash codes work , if you have a working +5vdc supply inside. (A16)

If the Scope is dead, see page PDF p214,  See photo 2 here, for a normal screen.
First check the Trigger LEDs, and if the scope has GPIB option it too can help with the diagnosis, in fact, you can scan the SCOPE using the GPIB bus. (say with screen dead?)
GPIB means General Purpose Interface bus, a standard. (Its just a parallel strobed bus structure with special connectors and runs in parallel with others , with addresses)
They will mention GPIB (IEEE488 std.)  this is a standard parallel communications port that has vast features. (RTM read the manual) and has LED status lamps to watch.
Those GPIB LED are at top right of CRT bezel. (lock, Srq, Addr.) PDF page 215.   (they blink at boot up time)  This scope boots from  fixed ROM. (EPROM actual)
These LED's must be ALIVE too. (if not?  you lost internal power , most likely)
I think the GPIB LEDs are last in the chain to work right, in the boot diagnostic process.

If you lose a power supply , of the many, you need to measure them all on the right side card or  the A16  LV card,  so use  any DMM to check the  REG pins, then last the UNREG side .
The Unreg pins are just checking the huge CAP.'S on the UNREG sections of A16 card. (under metal panels "danger")  Unreg is the raw DC side.
This rear A16 side has 400vdc there, do not be casual.
For sure, check that if in USA you are not running in 120v  not the  240vac rear  LINE switch selection mode.    (120 vac is USA voltage)
Let say the low voltages supplies are all good:
If all those are good, check all the voltages on the  A17 High voltage card (left side , under cover with 1 screw), (with CARE) Per the SM, service manual.
CRT rules (generic)  1kV per diagonal inch of screen is  rule of thumb.
The CRT runs at lethal  7-14,000 volts. (page 3-75)  at CR565 and CRT neck ! Don't touch this, it is no LCD here, no liquid Crystal Display.  It has a cathode ray tube.
Even the LV board Low voltage A16 card has near 400vdc there under a metal cover. (respect it or get hit)
Read the service book  if playing near the CRT cards, or under the A16 metal dog house hatch. Read the warnings.
Some scopes and most equipment like this , have bleeder resistors, a smart tech, never trusts his life to a stupid $1 buck bleeder resistor, ever !  <<< live to tell?
All voltages over 36v , you should use meter ball clips, then connect the meter up, with power cord removed, then turn on the scope for live testing (when done, power off first)

Check DC power supplies levels first, if all signs of life are dead, to all sections of the main boards, Mother boards.  (I check voltage outputs  at A16/A17 cards and side card A13)
If 5vdc is dead the scope fails all sings of live but 1, the dot in the center if the CRT doing damage to it.
There are over 10 DC voltages to check.
 Just because say, +5vdc is present at the LV card, that it makes it to all other boards.  (ribbon wires do break)
 (keep an open mind and let the readings dictate tests and directions)
Now HINT # 1:  Many DC supplies have what is called Dependencies.  For example the -15vdc supply needs 3 other supplies working first. Measure them all first., and win.
The same is true of -5vdc. it needs -15v to work.
If one supply is dead, then remove its loads, 1 by 1 to find out if that line is shorted.
Measure all LV (low voltage) power supply  bus levels and , for excess 60hz RIPPLE levels (read the manual, it shows that, and the spec. tolerances.)   (takes a scope to fix a scope,sorry)
If ripple tests fail, that means those old 20 year life rated electrolytic caps are history.   Ripple spec.'s are on the order of 50mV or less. See chapter 6 page 10.

If power is good?,  and all fuses, including the 2 hidden fuses. ?
The next thing to check is are the 2 processors.  (main and display)
Make sure the power pins are hot and at spec. on all  power pins of the processors.
Chapter 6 covers vast numbers of checks and there are even special diagnostic jumpers to do special checks.  (even data and address line checks !)
Make sure line DCOK is true.J102 pin 28 A16 and outward.
Then check that the master clocks are good to both processors.
This scope has a full blown embedded main processor "uP", with ROM and RAM  and ports. 
The processor in this scope  Motorola®  68K, that  has code (hard machine code, called Firmware) stored in the old very small  UV EPROMS. (a bank of them)
(these can be copied easy and reprogrammed) The 68k is same as used on all Apple computers. (long ago)
This scope has at least 4 processors,  MAIN, WAVEFORM and Display Processor for Status LEDs and buttons) A time base processor too, not sold , by VTI VF4083 (rare to extreme now).
If there is no power here, it can not process or do anything.   
If the screen is dead, you will never know if the processor, even boots up.   (you are now blind, but the trigger LED's and GPIB LEDs do prove signs of life.)
There are  life tests, then also other tests, in ROM, one is the  Display uP test,  with jumper J155 (PDF page255)
If your shop has  PC with GPIB card , you can run extended diagnostics, that way, with a dead scope display.  Most shops today lack this simple tool and cable. (it's a  generic IEEE-488 bus)

The screen goes dead for NVRAM errors,  in fact, the screen will show CRC 6000  fails for that loss or corruption.. or the LED flash codes, if there is still life in the processors.
If the screen is dead, then the screen power A7 board is bad, or lost DC power A16, or the CPU is dead (lost +5v)  or the EPROM code bank is dead/ scrambled (rare that).
If the LEDs work and screen is dead, suspect the A17 board. (screen HV board) (the LED flash code win, if they flash codes)
In many cases, it takes an extra good bench scope  to test a  bad scope (diagnose it)
  •  In fact, that extra scope you can look at serial data to the CRT video drive grid, and Horz and Vertical  plate sweep levels to the same CRT.
  • See all master clocking inputs and action, on the MOT 68K pins.   4mhz , 8mhz and others. 
  • See that the uP (microprocessor) main is cycling  many of the ROM bank address pins.  A0 and up. If not the uP is dead, for some reason, it's not reading op codes from memory. (It must)
    There is a jumper pin to run data and address line scan checks in chapter 6.
  • You can check power lines for excess ripple (60hz) and all master clocks.
This is just some of the checks , there are many tests and checks to find the true cause of failures, use the SM book , chapter 6,section with diagnostic, flow charts.

See (PDF) page 273  for DEAD start.
See page 274 for trigger, LED flashing errors, codes and meanings.

  The key to simple diagnosis is : (just for dead scope or failing DIAGNOSTICS.)
  • Is the main processor dead? "Signs of LIFE"  CRT dead, or  (trigger LEDs) yes dead, then check all power supplies and clocks to the main processor.
  • If the trigger LEDS are flashing out codes, then look them up in the manual and do what it says to do. (Ch.6-pg28)
  • No, not above?,  then is JUST the screen dead?   yes, then check power supplies input to HV card.
  • No matter what, do check all supplies , at the test points on the side card. or on A16
  • Then, if all power is good and processor above is ok then check the A11 display and video card and then the HV CRT card A17 per the manual.  (do so with great care)
  • IF the CRT is dead and all voltages check out (DC) and  IS THERE Life tests  pass?, then check CRT neck filament 6v  voltage. if good  then power down and check filament for 0 ohms short. (unplugged) if way more ohms (infinity is bad dead open), then CRT is dead. The heater is just a standard. 6.3v (like many vacuum  tubes of olden days)
Those are just some common failures, there are many more possible.  
This page section is for person asking, specific type failures.
Next is NVram issues
and last is RIK's failure.

2430 (non-"A" only) The early scopes have this amazing 30 year life battery that never leaks acids causing damage.  BENEFIT 3 is , the battery is not hidden and easy to replace, unlike version "A".
NVRAM:   NON volatile Random Access Memory. (the battery makes this RAM, not forget data at power off time)
The only down side with this external battery version of the scope is to not let pin 24 on the SRAM go below 3v and for sure 2vdc.  (by any means and is easy to  test !)
The newer scope hides the RAM (better) and is more protected, but you can't test the battery in then new scope, so my scope wins ! In my opinion is the better scope.
The ram never fails ever, just the battery, that was yellow cased new, seen below.  (there are good substitutes, and brands all made for medical instruments)
t's fast SRAM memory, that does not forget.  (with battery  good)
The below are my theories on how to backup this data.  (this SRAM retains data even at 2.0v VCC(VDD))
ISSUES of all kinds and both versions,  A and non  A versions.  Do read the (service manual first)
Error code 4720 is low battery , below 2.4vdc. (time to get busy get a new battery now)
Error code 6000 is NV CRC failure  , almost always caused by that yellow Keeper-II dead battery seen here.  (this is a fix this first  error) You can't go beyond this error.
Error code 6100 is CAL constants in NV lost, same deal, battery dead.

NV means Non Volatile.  SRAM means Static RAM.  (This NVRAM has 2 sections, Factory CAL section, and user settings, at least)
This scope running writes to NVRAM  all the time and fast. 
In lay terms, it means this memory that does not forget. (at power loss) 

NV-Ram battery has failed?,  is usually the end of the scope, if you can not do the expensive factory external CAL.
The key to future life as  legacy scope, is to test the voltage at CR801 cathode for above 2.4v,  (3.5 is normal or even 3.7 so if lower that 3v the battery is slowly dying)

This  cave man style NVRAM is like FLASH memory today, only  it's really battery powered CMOS SRAM.  (battery + SRAM = quasi Flash?) dig?
One trick if you can backup the data is to replace U664 with a Dallas DS1220 NVRAM,  preprogrammed and in new fancy ZIF socket,
There were no fast to program (write) EEPROMs back then, nor FLASH.  ONLY this part, that  could work, back then circa 1984 drawing board ! this changed in 1987 with the new 2430A scope.
This battery is rated for 30 year life, mine hit 32 today. (the case is SST and is solder closed up and can never leak acids or caustics.)
The SRAM does not wear out like FLASH EEPROMs do. (can , for  sure older versions !)
The below model scope has the most easy battery service , the newer Scope 1987?+  you must remove the Dallas DS1230 chip to renew the battery.
Later below I will cover, how to backup first your data, but for now, the first topic is how to get the battery replaced first. (see R764 1k ohm resistor ?) with no data loss.

 What we do is boost the power with a power pack of some kind , seen below and then replace the battery with a new KEEPER-II battery.

The steps to replace  any yellow block colored, BT800, even sold and   I got mine from
Keep in mind, my KEEPER battery is still at 3.7vdc after 32 years (amazing no?)
The schematic looks easy, but not Physically at the PCB. (do not rush, look at the PCB seen in my slide show)

 A: the Scope is turned off and unplugged from AC wall jack , if not the new booster battery blows up. (or the scope overheats doing this)
 B: When my battery hits 3.0vdc I'd replace it fast.  (hint, 1 volt lower and the game is lost)

  1. Not letting letting chip U644 VCC (power pin 24) drop below 2.0vdc ever (below 2v the chip forgets all it ever learned)
  2. Buying real KEEPER -II (tm) battery is 1st choice, one that is fresh date coded. (or use cheap non chargeable Lithium 3.5v batteries for lower life spans)
  3. First clip in the booster battery, a coin cell, 2032 sold in all stores, even walmart or  with tabs or with a holder for same to node X. cathode of CR802
  4. Measure pin 24 now, with DMM meter set to 20vdc range (or 5 or 10v)see about 3.3vdc. (even 3v is ok)
  5. Now unsolder the old BT800 KEEPER from the PCB. (mark the PCB so that PLUS pad is known.)
  6. Next take the new  battery,  and solder it to the 2 BT800 PCB pads, now. (any why you can, polarity correct)
  7. Remove the booster, and you are done.
The original battery. The maker claims 30 year life and never to leak. (mine passed this test) a.k.a COMP-88. SEE more details on this topic here.
This battery is used in many medical instruments and in water and gas meters for home and business. (long service life)
The OLD Original Keeper-II, had 4 pins.
3 are Jumper'd inside to minus pin   and  the last pin is Plus pin, and is plastic cased in potted epoxy, unlike the  above newer and modern SST metal case.
The new case minus is shorted to the case inside, so it must not touch anything on the PCB. I put a plastic standoff glued to bottom of the Case above.
The new battery is 2 pins and to my horror, the PCB needs the 2 extra "VIA" right side  pads, Jumper'ed to the new keeper minus pin. Using jumper  wires.See photo below.

The next horror is the new battery leads are too big in diameter, (and are per data sheets) the original battery is no longer sold, horror 3,
OK,  happy feet ,my battery fits and works.  The new battery is a better design, it can never leak and damage the PCB below or other parts there.  (the jumper on the right simulates the old battery wiring)
New Battery , details

end battery horrors, all.

The data in the NVRAM can be backed up 2 ways.  (or NOT)
2340 (non-A) very very difficult.  Using a programmer attached directly to the SRAM.  (it will fail due top 5mA, IoH current draw in the CS pin. 1k pull up there)
2340A, is easy , remove it and back up the data before the data is lost. (before battery inside is dead.) Using any programmer, even this TLS866 for $42 cost ,off fleabay
The reason to do that is moot, the calibration takes 15 to 30 minutes, and is fast and makes the above never needed.  (a secret no in the HP manual)
If the battery dies, CKSUM 6000 fails.
You then replace  the battery, (or Dallas DS chip on the "A" model) and then do the 3 phase calibration (level,trigs,repet)
Then turn it off, and like magic CKSUM errors end, only after the 3 phases of calib., show, Pass, Pass and Pass.
Put back the J156 jumper and you are good to go.

As you can see my 1986 Tek 2430 non "A" slide show.

The never do section: ( all 2430s made)  Unless you must (lost base CAL.)
Avoid touching jumper J156.  (until you) Read the Service Manual,  read page PDF  page 179 on it's proper usage.
This jumper prevents the External CAL from activating  (doing so by accident destroys the NBS ( now called NIST) NVRAM calibration) pg 182.

After a loss of all data in the above NVram

More warnings and actions:
Do not touch the HV high voltage CRT or Cards that run it.!

  1. Clean (blow out with 9PSI air,  the scope make sure the side 12v fan works.
  2. Check all Power Supplies for correct spec. voltages. here.
  3. Check for excessive supply ripple (120Hz or noise of any kind). (bad CAP.'s) page 6-10 in your manual (free) covers this spec. or see it here.
  4. If you see blue/green dot on the CRT face, turned on,  turn off power now,  do not burn the screen,  even the F269 blew.
  5. One more hidden fuse is, F961, -15vdc , and the CRT card goes dead as do other things. Hidden on A16 again.
Do  not overheat your scope.
Do not block any vents.
"Do not run it with the case off, until you put   35-CFM or larger rated fan blowing on the A10 card, directly at the CCD bottom heat sinks, or they will be damaged." 
Use a cheap desk fan 10" to 12" will be good directed the hot  running bottom PCB and all will be OKAY.
READ the manual, then take it apart.
 Do not get near the CRT power rails , they can be very lethal touched.   the CRT does have high voltage pins, this is no LCD.
In the olden days, this was called the "HighVoltage  DOG House" for good reason.
The LV card (and oxymoron) has 400VDC and is lethal too. ( no second chances here, its not LCD tech)
Do not let the outer case , sliding  back on wreck any parts for sure J148  top rear (5 pin ribbon) nor J152/J05 small ribbons bottom left near front,  J152 pin 1 is +5vd  and if it hits case the fuse blows fast.
Do  not damage the bottom center rear switch (power) or its frail  ribbon connector there, we have  our ribbon cable there, DUCT taped down.

Do not allow the integrated  side fan, to stall OR FAIL for any reason. Even while doing the above, service with power! (if you pull the HV A17 card J176 connector to be safe the FAN SHUTS OFF.)
Do not allow the fan to go dead, ever.

If autocal passes, you are good, but there is 1 more of 3 , calibrations that can be done easy.    (do so only if unhappy with vertical accuracy)
The extended diagnostic, is good, run it any time.  J156 protects you from accidently calibration (in the menu's)
There is a  0.2, 2.0, & 20 volt DC external calibration, that you can do.

DMM = digital multimeter.
CCD = Teks name for analog charge coupled devices, (custom to Tek, not sold at radio shack or totally unrelated to modern Camera CCD photo array's.
HV = high voltage do not touch me, nor expect it to discharge turned off or unplugged from the wall AC power. (we use a discharge probe via a resistor to do that as the factory manual tells you to do )
CRT, a cathod ray tube just like old relic TVs hadm but smaller and 100x more accurate here.
LCD, the liquid crystal device , screens you don't have on this scope.
Lethal, a voltage so high ,  that touching it , you can die easy. (like walking across a freeway blind folded, with same risk and pure luck , results)
SCOPE = O'Scope, or oscilloscope. (or DSO or DPO)
Dog House, a box to keep the neophyte away from death. (or HV cage)
J156 =  Calibration mode disabled (jumper in place)
DC = Direct current , not AC , alternating current.
CFM = cubic feet a minute, air flow, (the fans data sheet tells you this as does the TEK manual tells you how much you need for safe service.

Links: (good)

Dr. H. Holden (see how he captures the data on DALLAS NVRAM)

Best page on Probe selection, no more guessing.... (he makes it look so easy...)

Hacking ++  (I'd not do this to the NVRAM , but if compelled by demons... or? )

THE crème de la crème ,  Edward Kujawski 
I LOVE YOU,  MR BAMA,  work, He has my 456 PDF with the calibration in volume 2.  

How I got both service manuals:
At Tektronix's hidden pages,  you must register then login , and last Go here. for the 2430
2430A at Tek is here.

Tek allows you to share manuals on obsolete equipment,  nice people there.  And you will get addicted to TEK, they are the best.
I used Google site search, to find these 2 books.  Tek does not INDEX them "Support", but they are present, (ask Google xxx site:www......)
No need to buy CDroms, off fleabay. (rip offs)   See my OCR versions of these manuals.
Tek is smart, they want people leveraged in to new products.
This is a great way for the young (?) to get a cheap scope and save up for a real DSO.
This equipment is very old and is totally , a GOBsmack, that this equipment runs after 34 years, 3 long decades,  Go TEK !  

The 2430 End Zone :    Most of the below is moot , if the RAM is already, Dumped (lost all its data, offsets, gains, etc)
Batteries: 1986 2430 only. not the newer scope.
My 2430 (non -"A" early )
 Scope uses a very high quality  LITHIUM / THIONYL CHLORIDE ,packed in side Stainless steel case (today)with retention, after 15years at room temperature!
Mine has Lithium "Keeper ®" battery. and can be measured easy with a  DMM meter. (the diagnostics fails if below 2.5v (or 2v)  or above 3.7v (a secret) < estimates.
My scope came from a Lab, and is 28 years old and reads 3.7v in 2014 (Aug)  (no joke, those are real date codes)  And amazing product for sure.  TEK RULES !
34 years old now, and the battery is still good.
By the way , still sells them. at $15 (must have been a genius idea to last up to 2018 , huh?)

Procedure: (a new battery, with little effort, leaving the old one, still planted)  NUT SHELLED STEP!  The battery is on the Processor board A (M)68K.
  1. Remove the Scope main case outer housing (it's all here, in the SM).(Service Manual) (side and rear screws) < 30mega byte
  2. The Top of scope, is  a main board, is now tilted UP 3 screws out first; this is the horizontal sweep logic and below it is the processor board.
  3. Then, remove the 3x  Torx® #T15 screws centered in  middle of the sweep  board (top board) or it will never ever tilt on the hinges.
  4. Next,  turn the 2 top  black plastic cam lock screws (left front and rear facing right to left at side) 1/4 turn CCW  (ignore centered black screw) Be gentle take your time here.
  5. Lift the whole hinged DOOR up, gently and remove 1 ribbon cable on back of processor board (under at about 45deg angle , go easy)  it's free to hinge out now.
  6. Behold the yellow battery now.

The battery is yellow packaged flat mounted item, lower left of card.   We bost how wire CR802 with another battery then remove the old one first.
To the left of  the battery "Keeper ®"  p/n #LTC-7pn-s4 is a ground TP (test point) . on the PCB, in plain site. Eye ball it 1 time.  $15-18 cost new at
I have programmer that can read this SRAM and save it to a file savedSN.bin

The first test:
Connect  a DMM minus volts lead (black) to TP ground there next to battery, and the  Red DMM lead to Anode of CR802,  I get 3.702 volts exactly. 
My battery date code shows, "3284",  the 32nd work week of 1984. (amazing that is. running so long , 32 years now 2018)
It's a runner, if you read closer to 2.5v it's about to die. 2.0 volts is factory spec, on for sure die, or tad less, 10% less? For sure don't let it go there.
But you must boost the ram first. (unlike the newer "A" scope never need to do that 1987+)

I hope I saved you long wasted hours, on Extended J156 calibration steps. (and impossible to find calibration instruments and tools)  SG-5?? etc.

A slide show of my 2430 scope, Insides. with nice comments on chips too.

BOOT up Diagnostics:  See chapter 6 page 28.

The Trigger set LEDs are,  Trigger, Ready, ARM, Plus, NEG.
These LED glow or flash a codes, that tells  you about catastrophic errors, (the SM book cover this) for example if DCOK  is false it will flash codes for that error. (in this example DCOK is false)
The scope sets DCOK true, as soon as the supplies go active, then the main processor reads this status bit,  if false, it then sets a TRIGGER LED error code.
There  are many boot up tests. not just DCok.
You must read the manual to understand the 3 ways these LED can glow,  (normal, flash codes, or stuck on codes)
Let the on board processor tell you what test fails first.
If all the LEDs  are dead,  you lost power to one of the DC power rails. +5vd tops this list as does J156.

Example 1:  DC SHORTS.  Cap.s means Capacitors, the ones with large uF markings (electrolytic) 1uF and up(typ)
 (the failure is super easy , a shorted cap. but finding it is much harder.)

Of a dead short to any  power bus .
(shorted tantalum caps and electrolytic caps top the list for shorts , as do any chip that is super hot)
In all cases, like this power supply shorts, find the shorted card first (sorry, if obvious)  This can be hard to do, but I do have lots of advice on this topic, (I never give up. either)
I use a clamp on meter, that measures amps from 1ma to 40amps DC. this finds which wire is shorted easy.
but wait, some are in ribbon cables.(with power feeds in them , not fun this... this was a bad idea at HP, for sure, but this is what you have here.)
What I do, is to make up an adapter for the cable with power missing (ground still intact)I remove this power feed, totally from the chain,
I then add my own power feed, direct to the boards, with jumper wires.
With the power bus no broken out of that silly ribbon mess,  I now can use modern tools to find shorts. (called the AMP CLAMP meter DC)
HP and Tek both make current probes, that are a probe, small with a mag tip that tells you the current and its direction, for finding shorts on PC traces or a single wire, or even ribbon wire stripes.
This tool is very rare and was and is very expensive, so lets skip this fancy tool. But this tool finds the short with out me, breaking out the power bus. (the shorted one)
After finding which  card is shorted first, you can remove the caps one by one on this  card, to find which cap was shorted.
I can  test them all, one side of cap. lifted.
If all pass tests, then the short  must be a chip shorted on this PCB. (or other part ) the schematic tell you all parts that land on this shorted feed.
Find the bad chip , is best done with a thermal gun /camera first.  The camera method is the best way to find too hot parts.
Ohms laws:  P = power = watts, I = current amps, R = resistance ohms,  E= volts (or V)
Keep in mind at all times 0 ohms shorts  (rare) make no heat (P = I² x R) power (heat) is equal to current squared, times R,  if "I" is 10amp and R = 0 , then P =0.
Anything times zero, is zero, and no HEAT emits from the perfectly shorted PCB  !!! but we never fail to see if parts are hot, first. ever.
Most times a short  will be like 3 or 4 ohms and this fact makes finding it vastly more easy.(for sure thermal tests) (P=I²xR = 100 watts at 5amps)
Best is to use bench supply set at 1amp and work up to 5 amps to find what gets hot first.
The most hardest short to find is on power planes. (3 or 4 or more layer PCB)  PCB means printed circuit boards not, toxic "
Polychlorinated biphenyls"
However the part will get hot but the voltage drops on the power plane will be the same everywhere on the plane, making tests harder. (not impossible,see this....)

I have cheap camera that is IR (infrared) sensitive, I shoot (click  the Camera.) this PCB card, while powered up and hot, to find anything on the card that seen HOT. (ignore normal warm parts)
I set my bench supply to 5 amps, to see make lots of heat.
This is not easy to find, best is just lift all polarized caps first to find the shorted cap.  (most easy, below is hard way) Many tech.s in fact Shot gun caps to cure them. (after all they are end of life at 30 years)
In a factory the bad card gets sent to rework department, (but this is no factory) nor do we have the $1000 tool to make this easy.
Best is to start with common cap failures, then chips. to find the short.
I tell the full story here, as I have worked in all service  modes, of employment , factory tech , field-tech. , engineering tech, and in a calibration lab (USN)
In all cases in any system complex as this we must find which card or harness is shorted first.  Job 1. then card repair happens.
See my full example of shorts in this further  section below:
Card shorts: no Harness or Jack ribbon shorts like I had.
  • Caps shorted, all kinds of electroltics short this old, think common as nails.
  • Chips short.
  • It is not solder bridges (as some call whisker shorts) this is rare and bad service caused.
I always hope for CAPs bad, after all they only have 2 pins to unsolder and are cheap and super easy to buy and replace.
Warning this HP scope has daisy  chained power, and  goes in one end to the next PCB, from other end, so be very sure which wire is shorted first.
The power goes out W1000 ribbon harness, the power is on up to 9 pins (+5vd) and is in parallel to all cards.
The the power is cross feed between cars, yes,  tricky as tricky can be, this structure, and personally never seen on any system ever.
Most systems use mecca grounds and power rails (aka:star ground) this prevents ground loops, etc.
Running power feeds a long way invites killer voltage drops (VCC must not drop more than .25volts, for sure dynamically) to do this would take heavy power cables.
This scope uses those find #28 Gauge ribbon wires for power, something I don't like at all ,but must live with.
These facts alone, make this scope hard to diagnose shorts.

Jump to +5vd shorts now. for real examples and tests.


I have 3 free manuals here, both are OCR rebuilt (allows searching; try "Trigger LED") Tek allows me to source them here. (relics, only)

My copies with permission.
The Non "A" Original manual: (now search-able)

Version "A" 2430A only: (with Dallas NVRAM)

The  Military manual. (best book on theory)

The new Maxim pages show  what is made still: NVRAM.(old Dallas)

Tek Direct manuals:

2430 (non A)

Calibration must happen if the NVRAM dumps all data on the floor (lost, for any reasons)  A.K.A, "all data landed in the byte bucket" (lost)
The NVRAM stores all configurations and calibration tables and offsets for the gains of on board Amplifiers.
I most cases pages 5-6 to 5-9 will not be needed to be done. Nor the screen dot calibration (pure screen issues )
This fact makes calibration child's play (DIY perspectives) you only need  simple home made or imported precision voltage reference.

The procedure begins on chapter paged 5-10 SELF CALIBRATION..
Chapter 5-10 to 5-12 covers this fast  cure, in the free manual above Non-"A"
Page 6-28 and 28 tell you why NVRAM causes the scope to go berserk. Read carefully here, about bad flags and  if you sees the words Check Sum or CRC , ignore all other errors.
The cure is a cold boot and full chapter 5 re-Calibration.
The Chapter 5 calibration has 3 phases.
Phase 1 is Analog  20vdc, 2.0v and 0.2vdc using  precision analog voltage reference. (DIY or pro made) No need for NASA, NIST, NBS or ISO quality here. (±0.1% is easy at home)
Phase 2, is Trigger checks using the PG-506 , this is again using only pure DC clean levels, there are no signals generated here.
Phase 3 is automatic, Repet cal. (no tools or what I means is no SiG Gen  needed here either)
Read chapter 5 and learn the calibration is super easy if you an accurate voltage source of any kind. (home made and calibrated by hand to your most accurate DMM meter is ok)
I extracted just chapter 5 here.

The Scope manual calls out a PG-506 Calibration generator. (1980 first one, ouch) The free book is here.(OCR updated and annotated)
Finding the spec or book on such a thing from 1980 proves,  hard.
I read the 506 manual and the schematic and this mode is pure DC mode. "STD OUTPUT"   todays kids say , sexually transmitted disease output, huh?
Accuracy is crude, 0.25% ± 1uV (amplitude) my 4-1/2 meter does .05% , so I can beat this number easy. Using any modern DMM to set an exact voltage reference I made.
The output is only a square wave.  (or DC) (and signals are used here at all)
Frequency or   period of ,  1uS to 10 mS  5% accuracy in decade steps. period knob below. (any SIG gen can do this)
6v to 60v into a non 50ohm terminated load. but into such a load 5v min. (means they assure 5v only )
The only true advantage of the 506 is the word "LEVELED" in Tek Parlance that means the SIG Gen levels hold at your setting as frequencies changes,  level regulation is a better phrase.
In truth it is only Convenience  or expeditious. (DYI folks are not in hurry) (we have good meters and scopes )
The Tek word STD , means DC output only, and what a pain to find and learn that, in the lofty year 2018.

Now the silly HP manual 2430, vagueness,  there are 4 BNC jacks the 506, the bottom right is (or STD) Output. (okay bingo)
It never mentions the PERIOD knob.  or what frequency to use ever. {omg, they never use  this mode generator mode}
Move the top rocker switch to STD AMPL.  This turns off the Square wave generator and runs only the DC amp card. I now get it ,not owing a 506 how can that be clear.?
The bottom period knob is never mentioned because the GEN is turned off (no square waves)
Only amplitude is switch on, is accurate by indexed knob steps into any HiZ scope input. (we use 0.2v 2v and 20v.)  For sure one can make up a bench supply to do that with a nice DMM to prove it is correct.

It turns out (had to read the schematic to learn this) STD AMPL mode is only DC mode.
The  output jack is only connected to PCB  card 2, labeled Standard Amplitude
There is no chopper on that card, it's pure DC there.  with 18 knob settings at .25% accuracy.  (the module 2, has 4  calibration pots and all other voltage are created from precision resistors)
 (var is set to calibrated on both turn pot knobs, skipped in the manual but is important but all labs know that....)

There is no need to buy one of these today, ever.

So  making up a cheap bench calibrator is easy. (DIY not commercial usage )
Low noise Power supply. (bench grade PSU regulated.)
Low drift (Tempco) low PPM drift.
Very good regulation. (both thermally and line regulation very good)
Adjustable would be nice. 20v down to 0.2vdc ?  Here is China solution that might work.DPS-3003 ($25) (I will test it for noise)
The supply must at least not have voltage drift in time  (for any reasons).  (say stable for 1 hour work or less)
I have 2 resistors in my junk box that  are low PPM drift devices.  Metal film resistors, about $1 each not 3 cents like carbon resistors are.
10k ohms and 200.  I used it to get 2v, and  0.2v and 0.5 voltrages. super easy and stable.

At no time do we need a signal generator at any time to calibrate using the External calibration.
The attenuation uses only DC voltages , the trigger calibration the same.
and the REPET needs nothing , it is automatic. (nothing added)
The Check sum (CHKSUM) failure 6000 will not clear off the board until the full external calibration completes and you turn off the scope and everthing gets stored in NVRAM with the new ChkSUM.
Do not forget to replace jumper P156 (calibration disable) last, (yes after turned off or soon later in the day, but yes, before sun set, LOL)

Using pair of stable resistors I was able to calibrate the scope in 15 minutes work (once connected and setup) (chapter 5-10 to 5-12)
My scope is ready for 30 more years, LOL !

Power rail shorts to +5vd. (failure diagnosis and tests , all ways from Sunday!)  
First off I'm no stranger to shorts, it is all part  of a systems tech's life. (50 years here on the job, but retired no)
There is also +5v and a 5v unreg line, but is not +5vd (I guess d= digital)

This failure is common due  to CAP shorts. (they love to do that THIS OLD NOW the life of a cap is rated at 20 years, (no, not ceramics , just electrolytic)
The first step to finding shorts is the technique called isolation. (we can't just try all new boards 1 by 1 and learn which one shorted  can we?)
The second problem is having good tools ! A short list of tools, cheaper to expensive. (say tools make the tech ! in most cases the tools make a tech fully effective)

  • A finger to feel something is too hot (and bravery, ouch !)  An IR thermal GUN is like $15 and saves burned fingers, no?  or a camera that can see IR light spectrum. (some can)
  • A cheap DMM (digital multimeter) (to measure ohms shorted, and to measure voltage drops under power)
  • An Ammeter (wired or current clamp type or both) (the ribbon cables here will make this near useless ,see the video below.)
  • A bench power supply that has current limit and a voltage limit,  (no tech works lacking this very important tool)  0-20v and 0 to 5amps is best
  • The current probe seen in this video. (HP 547a relic at the least)
A shorted part, any part that is say near 3 ohms will produce 3 watts heat at 1amp bench supplied and limited current.  (at 3amps, I² x R =  3 squared time 3 or 27 watts ,yah hot it will be.)
Btw my A12 card measure 80 ohms . (in hand) +5vd
These pages below show the odd nature of power on this TEK scope. But wait, the power is also cross fed, see how to deal with that horror below.
 Intra-connect A and   Intra B.  (I made these scans using 200DPI resolution.)

If the fuse for +5vd blows F269, we put back a new fuse (or 10amp ATC Circuit  breaker that I use for testing) and then pull cable P166 - p176 to the A17 HV card.
F269 is hidden on the A16 behind that wide ribbon cable, and is a PITA to change out, I used pig tail automotive ATC inline fuses, RED code, or CB.
Did the fuse hold now?
As you can see the A12 card is the first landing point next for +5vd.  ("d" means the digital 5v not the analog or unreg. rails.)
To your horror, it's not simple at all, it has a grand level power rail parallelism here. (the way to get fooled here is the cross feeds will find the path to the short too, so you must pull all ribbon cables first )
See that 3M™ 50  pin ribbon cable?(top right side of PCB A12), W1000 is its name, and it goes to many places at once.  9 power pin wires are used to handle all that heavy current . (10AMPS)
See sheet 3 (5E coordinates)  (pin-4,6,10,12,16,18,20,22,24) the 9 pin list.( ground has more)
That w1000 cable has 4 ends, and lands on A11, A12 and A13 cards and A13 cross feeds A10  card.
What's worse is the other  cross feeds, See A12 card cross feeding A11 at J120 pin one (oh boy, tricky this is.)

The way to  solve this (theory and practice) is to remove all cables from a12, see if the problem clears.
I can not use my Amp-clamp inductive meter to see which line is shorted, due to ground currents in the cable same, cancels out my meter readings to zero. (Faraday's laws are not repealed)

If you had a  guided tip ammeter$$$$, that lets you find shorts on a PCB trace (an expensive but loved tool by me) I could find which cable wire is too much current using that tool.
One cheap tactic is to make up a ribbon cable adapter   male to female 50 pins, then break out the 9,  5v wires, and then use the amp clamp to find which card is shorted.(9 wires paired up)
NEMA spec. 9 times 1.4amps =  12.4amps max "free air spec",  I De'rated that to 10.  Gee the fuse  is 10amp,  no surprises!

So one could check the current at each w1000 wire set (9 wires broken  out bundled to test +5vd line) then do the cross connects.
Eg. +5vd runs the main processor and digital,  but +5v bus runs analog chips, we tend to call this quiet 5volts.  It is standard practice not to mix analog supplies with digital so is no surprise here.
If you look deeper  the +5v supply (not Vd)  has  sense line, that terminates at remote point at sheet A10,sheet 10.  F/7 J111 pin 31 = sense. or w101- pin44 to the reg card.
The +5Vd line has  more simple senses feed back all contained on the A16 card.
Once you know for a fact one card is bad (PCB): (or is a card for sure) we remove it.
Card tests:
One other way to test is to remove each card, and hot wire each card to a current regulated Power Supply, dial in amp to 1 to 5amps and set voltage to 5v and see if the card is shorted.
Easy to do with a spare ribbon connector and stubbed off cut ribbon, skinned (stripped) back for 5v and ground pins  ( using any 50 pin  3M connector in fact access to first 24 pins does this)
In fact (my junk boxes are full)
 Make  a JIG.
Plug in say any old floppy cable 34pin cut in half , hack it up and you have  test jig for power. (Hard drive 40  pin connectors are hard keyed (block pin) and do not work here !)
I will do the jig way  on mine, and report results here. (if the fuse keeps blowing)
I strip all 9  power wires and all 8 grounds as shown in Intra-connect A link just  above   to w1000 power bus connector on card under test.  Odds are ground ,even pins are +5vd
I did it now:
A12 example card. Isolated reads  (data: mine reads 2amps @5vdc)  (equal to 2.5 ohm load normal)  My A11 card, isolated reads 1.2amps @ 5vd.  Using my bench supply .
Step 1, remove all 3m cables from all major boards. (eg. J103 -A12) 
 Step 2, hot wire the card using my ribbon JIG to Bench power,   plug it in to card A12.  Justify pin 1 to pin 1 if using say an old floppy drive cable cut in half.
Be real sure polarity is correct. pin 4 must be at 5vdc, not pin  1 or 3 ,etc  check with a voltmeter at the 3M connector end, before using it in any PCB.
Step 3: Start at 1 amp  and set 5vdc.  ( I never need to go over 2 amp) 
Step 4,  turn on PSU (bench power supply, PSU)
 if the card is shorted, current goes over 2  amp  and 5vdc is not attained, the card is sorted , if 5vdc is attained it is not shorted (TEK has no spec here)
Lets say voltage reached, 0.5 volts (1/2 a VOLT and for sure not 5vdc)
 Raise current to 3 amps max, now. (mine was 2amps)
If voltage hits 5v the card may be good.
If the voltage again is very low, not 5v then the card is shorted,  I do not think any one card ever use over 3amps, in fact not over 1amp,but,  TEK is mute here for this spec.
I discovered the A12/11 cards never exceed 2 amps, (by bench testing the good cards)
I did not test the main A10 board for current  draw.  The only hot heat sink  fitted chips (only 2) are U370/470 at rear and only then use +5vd power,(and some 74LS low power logic.)
I say A10 uses very little of this bus power. (I did not check this card, I am chicken of those very rare chips seen there under heat sinks)
If this card passes the test then do all cards, on the W1000 ribbon bus. 1 by 1 , only 1 connected.
BINGO,? Photo's next of me doing this, and finding a real short.

I also learned that as the case slides closed  it drags on J152, and it cut pin 1 insulation!; guess what?, that IS +5VD pin,  BOOM fuse blew.
I had to move the ribbon to a new location so it can not drag the case again.
The +5vd pin reads 80 ohms to ground A12 on DMM meter (3 meters tried) (take great care sliding the case closed)
There are other caps that can short on all power rails.
To find those will not be easy. (most shops just"shot  gun" the caps , Electrolytic and Tantalum.) <<< means replace all at one time.
In a factory that made 1000s of cards a month, we did the following steps (none skipped below)  From easy to harder and more expensive or difficult tests.
  1. Look at all caps and all chip for polarization errors (assembly or rework caused) chip pin 1 to pad pin one and all caps  Plus to Plus pad.  (consider shot gunning the caps)
  2. Look for solder bridges, a rare find on field scopes, but shoddy work happens.
  3. Feel all parts with wet finger tip, for over heating (brave, I've felt chips so hot the burned spit off tip of finger)
  4. Skip step 3 and use a modern $15 IR thermal gun, Yah ! best of best idea yet. (bang for the buck cheap  tools)
  5. Use  thermal IR camera, may cheap cameras today have that mode, use it. see heat.
  6. Next to last and better yet use the very expensive guided amp probe tool shows current and , wait for it....... DIRECTION.!! It even shows what path the current flows and the direction.
  7. Last if not 3 + layer PCB you can cut power traces, the one shorted to find what end of board is shorted (even down to 1 trace) (not on production cards just use #6 above)
  8. Some times you find cheap PCB with very limited small power feeds and you can measure voltage drops on that trace ; telling you easy what traces has the  short. It can work this method.
One web site says you cant measure current in trace(yes you; can make a guided probe ammeter, and is not clamp on device, that I am talking about.) Or the guided tone tracer,
 A  PCB  trace does have a magnetic filed and the direction can be measured !!!   The   HP 547A (schematic) or manual.   (old and rare too bad it is not cloned after 48 years)
What I learned is the tip design is secret and well protected by patents and very hard to manufacture (tiny magnetometer)
The probe is: Current-STEP SENSOR. The Current-Step Sensor comprises a pickup core and winding, and eddy-current shield. The signal to be traced is sensed by the pickup winding and core.
See this newer tool video, not cheap but shows you it does work well.
In the case of step 8 we set the power supply to 10-amps max, limiting current to that amount, I'd use no more than 5 amps, as no one card is allowed to use 10.
Keep in mind you do not want to blow traces off any board. (there are 3 classes here, A power trace power bus, 2 matrix buss, or a power plane) use more current is you get to power plane.
I do no have the PCB fabrication drawing , if I did , I'd calculate all power feeds maximum current limits. easy using 1oz copper PCB rules.(copper sq. mil areas)
A meter DMM can resolve 1mV changes easy so is a very good tactic to find shorts.  As you get to the short the mV drop goes to zero. (as you slide your meter probe point to pin , pad to pad)
All you need is a current regulated Power Supply, (V and A) and limit the current below 5 amps, Heck start at 1 amp and work higher. V set to 5v (or what it should be)
A  good DMM (heck spend $40)  can to this,  4 and 1/2 digital meter.  200mV/2V/20V/200V ±(0.05%+5) ( the 200mV range is good here)
A super short is only a low ohm resistor, say even  0.1 ohm.
Warning Algebra and Ohms laws shown below, (do not read if it scares you , please)
How to measure a short and discover it's location:
 So using ohms laws  (E=I x R,  5amps  = 0.5vdc) you then walk the voltmeter probe from the card 5v input and across the cards power bus feed, and see the voltage drop as you near the short.  
or using V instead of classic E , V = I x R. x  means times. "I", means current flow.
In grade school it can be said,  voltage is equal to current times resistance and avoid Algebra totally.
Using simple tests and simple math we can find a short.
We must first, use a bench power supply to set and limit current, so we can measure voltage drops.
One good starting point is 2 amps. (my 2 cards (not CCD bottom card) draw 1 to 2 amps, normally on +5vd rails. (rails means the power rail 5v and ground, it takes two rails to have current flow)
I set my PSU (bench power supply unit) to 2amps and 5vdc.
I measure the power rail at every chip that uses this rail.  the chip or cap that reads the lowest voltage is shorted,

I've repaired cards with at once 5 chips or more shorted by LIGHTENING. (many times)
The PCB trace voltaged drops can be computed easy if you know 2 things , thickness (1oz copper is typ) and the width. This is super useful information finding any short of any kind.
Just like knowing a harnesses, wire gauge would be , finding shorts there.
Know that a good tech, may discover the trace width, and the thickness of 1oz for /ft² the copper resistivity tables tell you the ohms and E= IxR.  easy. using simple ohm laws on pure copper.
The calculator above shows mils, a mil is , here,  .001" inches, (we use imperial relic linear scales in USA) it does not mean mm , millimeter, at all. sorry, USA we still are cave men. (I do  both)
eg., 30mil wide ,1oz copper 10" long (trace) is 0.162 ohms.   so say a cap shorts at the far end of that tiny trace (theory and ohms law)  and say said cap is 0.16 ohm shorted.
And you feed 1amp to the  far end.
What happens?, well easy,  you get .324volts at the input, and 1/2 that on the shorted cap, the heat is half on the trace and half on the cap.
Now what else can you learn, wire equivalence ,, what gauge wire is near this 30mil trace?,  so I can use standard (STD) wire current rules.
That .162 ohms value,  makes this so easy it is 33AWG wire  (the circular square mils, sets the rules) max current is 0.072 amps. 72 milliamp.
That is a poor example but is correct for any data lines. (super narrow)
A power feed trace would be huge. For sure on any 5vdc TTL bus structures for power are.
A 1 inch wide  trace , 1000mil "
The ohms is tiny .00486 (rounded to .005 ohms for about 1 foot)    (or 6 ohms per 1000 feet , as wire is specified)
One way to think of traces is in terms of WIRE GAUGE:
This above 1 inch trace (wide) equates to 17 AGW wire and rated to 3 amps.
So knowing the trace width is very important to any tech, wanting to find a short and not blow up a trace. (like a doctor , first do no harm?)
Rationalities based on crude facts:
Obvious to me is that a set of boards  (5 here?) at 10 amps total,  that 2 amps per card might be normal.?(as was true)but if one card draws, much over that the fuse blows at 10amps.
But I bet the fuse is set 2 times normal power usages.
My guess is the normal currents are near 1 amp per card and 5 amps max.  so if one card started drawing 5 amps total,  the fuse might just hold. (toss a coin)
This is why I think (don't know for sure) that 2 amps is a safe current for testing for shorts.
The truth is one card used 1amp and the other 2amps. (found out later)
I look, at my A12 card and can see there  are power and ground planes in the PCB, at least 4 layers, using my flashlight. I bet both planes can handle 3 amps easy. 2 for sure.

Now the ribbon wiring,

Now back to RIBBON W1000:
J100 ?(this is W1000 cable 1 end)  see those even pins for 5vd J100,  and odd  pin grounds there 1 to 23, ?
Connect up (soldered) together 9 power pins and 9 grounds to say a 18 gauge wire,  2 wires.
Connect said wires to a bench supply that one can measure current ( at least 3 ways possible , Internal , or external series ammeter or the great amp clamp meter))
Manual page 399: below drawing.
Card A11,  +5vd comes from sheet 22 and 23, shows 5.12vdc , the +5v reg is on sheet 23 a different regulator Q879, for testing only use ground pin 1 to 23 never higher#

If all w1000 cable tests pass , then add  the other cables 1 by 1 to the card under test and if one shows a short, BINGO.

Finding a short on the -15vdc supply would be way more hard, the traces are small.  I'd error to the side of caution with these and use no more that say 500mA current testing them.
The -15v is good for 1amp, I think. limited by the regulator.
The -5v rail Is limited by the 2SB826 transistor with 10amp rating , VCE = 1.3v so an source huge currents Pd-13 watts at 10amps, max. (IMO)

This is my personal journal, it is not published nor offical nor for prime time usage of any kind, this is for me, only, it's to remember what I learned at last time the scope died.
But everyone is free to read my rambling comments.
I hope my page helps others fix their scope.
If I had nothing to do with $800 bucks. I'd get the TDS784 scope
Do not buy or own any vacuum Tube scopes. CRT are ok, IMO.  73's to one and all ! 

rev 18++++ 8-27-2013  (last edit, 9-6-2016) revised again Jan,29-2018 (battery swaps) and fixed ribbon cable shorting out +5vd and its fuse.

I'm on,  K5JXH.

Warning , This page (nor any others) is not affiliated to the company Tektronix, in any way, shape or form, nor to any other company.
This page is only commentary on their very fine products. The best of the best ! 
The Tek logo is the property of Tektronix.  I use it, only to show their old cool logo. 

Long live TEK !