Tektronix TDS420 Oscilloscope smells like a fish market

[OptoIsolated3]

If anyone can help or lend some insight. I have a TDS420a. When I originally got it and powered it up there was a strong fish odor. The odor has went away and the scope still functions properly. I have opened it up and haven't seen any leakage but it wasn't a through inspection . My professor thinks it could be an old transformer off gassing Should I look into recapping the scope? and if so Does anybody have any tips or advice about recapping? Any help is appreciated and let me know if I'm doing anything wrong this is my first post, Thank you!

[TheMG]

Fishy smell in my experience comes from leaking electrolytic capacitors.

Look very carefully around the bottom of every electrolytic cap, sometimes it's not very obvious. Wouldn't hurt to poke around with an ESR meter if you have one too.

I would check this even if it appears to be working properly, electrolyte leaking onto a PCB can over time corrode traces.

[Smoky]

Here's my guess.

On the main board there are voltage regulators near small groups of electrolytic capacitors.

The caps have leaked and fluid has crept under the regulators.

When you turn on the oscilloscope, the regulators heat up and bring up that smell.

If this is the case, eventually, a hole will burn into the PC board.

I'd clean each board and test a few capacitors even though the leaking capacitor issue was more of a problem in the earlier TDS line of scopes.

[shakalnokturn]

Wouldn't the TDS420A be one of the "earlier line of TDS scopes" considering that some of the TDS500, TDS600 also suffered from poor capacitors?

[Smoky]

I own a TDS420 and it's dated 1992. I believe the "A" versions were dated around 1996.

As for OptoIsolted3's problem, since he wasn't soldering anything at the time when he smelled that foul odor, that to me means, a leaky capacitor is, or located near a component that is, getting hot enough to make it smell like that.

[CaptDon]

Don't know what world that professor lived in to think a fishy smell
would be an off-gassing old transformer??? 'Green' uncured transformers
smell like hot pitch or asphalt and do indeed stink. Sylvania had that
problem with some of their televisions that used a constant voltage
transformer which runs very hot and some slipped out without going
through the curing process. That sickly fish smell is INDEED bad capacitors
leaking electrolyte. You may even have had one that partially shorted
and popped and cleared the fault and no longer smells. Most of the 400,500
and 600 as well as 400A, 500A and 600A series had faulty caps and
absolutely need replaced if still original. I just got done doing a 644A
and a 644B. By the time they got to the 'C' series they got it figured out.
They even went to the 'Sure to Short' tantalum series in later models.
Oh, and you'll just love the polyfilm (RISA?) exploding capacitors across
the A.C. mains. They should have included a built-in fire suppression
system in the power supply section. Oh, and pray the 420 power supply
never eats the 'green unobtainium hybrid' during a power surge or
lightning tickle. I have a 420 (non-A model) about to hit the dumpster.
No floppy, No USB, it is a worthless hulk now that the power supply
won't run.

[Smoky]

Excuse me, but CaptDon, before you decide on dropping your TDS420 in the dumpster, I'll gladly pay shipping to have it brought to my doorstep

[james_s]

It's the surface mount electrolytic capacitors on the boards, particularly the acquisition board but there are at least a few of them on almost every PCB in there. First thing to do is unplug it NOW because the leaking electrolyte is damaging the boards and electrolysis rots out traces and vias even faster. You have to wash the boards thoroughly and replace every one of those capacitors, I've done two TDS400 series scopes so far, the job is not too bad if you catch it early but ignored it creates a HUGE mess.

The fishy odor is an immediate tipoff, that's the classic stench of leaky caps, I thought that was common knowledge but maybe those stupid SMT electrolytics aren't used so much anymore. I always replace them with tantalum chip types.

[james_s]

Oh, and pray the 420 power supply
never eats the 'green unobtainium hybrid' during a power surge or
lightning tickle. I have a 420 (non-A model) about to hit the dumpster.
No floppy, No USB, it is a worthless hulk now that the power supply
won't run.

The hybrid has been reverse engineered and somebody designed a replacement board for it, I forget who but I came across it a while back on this forum. There's not really much in it, some op amps, a handful of transistors, diodes and a few passives. Nothing exotic.

That power supply was a real mystery for a long time but then I finally got my head around how most of it works. There's a schematic for a variation of it in a Lecroy scope manual. The Lecroy version has different polarities of some of the outputs but the PCB is identical and there are only minor changes to the output section.

[madao]

TDS420A is too new about problem with SMD eletrolytics.  Problwm with  SMD eletrolytics is solved since 1994/5. But probably:  Eletrolytics on PSU  make a problem with leaking caps.
I have a  TDS430A corpse,  SMD elektrolytics is fine, but not PSU.


TDS420 , pleas don't trow it away, despite low  100Ms/s  samplerate. In germany, i sell always with sucess him for over 100€. I wan't keep him, but  repair: YES


Greetings
Matt

[CaptDon]

Smokey, The unit was working perfectly with one slight weirdness.
When you powered it on it would run about 10 minutes exactly, then
power down. When you hit the power button again it would power on
and run the rest of the day. We took a huge power surge at my workplace
when a car hit the utility pole outside. The scope was not powered on
at the time, but the power supply portion is always on and it smoked.
I made an attempt at repairing the supply but the hybrid is bad. I am
sure the rest of the scope is perfect and I had even replaced both coin
cells about 2 years back. I could put the scope back together and wrap
up the power supply separate. Nothing is butchered. I would suspect
because of the weight it would be around $50 to ship. I used the corrected
LeCroy power supply schematics to troubleshoot. It is a TDS420 and
not an A model. They only sell for about $150 on Ebay and the power
supplies when listed sell for about $125 so we decided to just shove
it under the bench. I can put it back together and send you the serial
number. What city and state do you live in? I'll try to get a shipping
estimate. B.T.W. it is super clean with no physical damage and the C.R.T.
is bright and well focused. It was a pleasure to use this scope until
the power supply smoked with a shorted switch transistor.

[Smoky]

Thank you CaptDon 

PM on the way!

[james_s]

TDS420A is too new about problem with SMD eletrolytics.  Problwm with  SMD eletrolytics is solved since 1994/5. But probably:  Eletrolytics on PSU  make a problem with leaking caps.
I have a  TDS430A corpse,  SMD elektrolytics is fine, but not PSU.

I just went and looked, the one I have is a TDS410A, it was not leaking nearly as bad as the TDS460 but it certainly had some bad SMD caps, and now that you mention it the output filter capacitors in the PSU were leaking too. It's best to replace them all, even good quality SMD electrolytics don't last forever and 1995 was 25 years ago.

[OptoIsolated3]

Thanks for all the reply's, I'm going to replace all the smd electrolytic capacitors in the scope. Should I replace the lithium batteries as well? Also if any one can point me in the right direction finding a procedure to replace the batteries, I hear its a tricky thing to do and there is a potential to lose the chips memory but I haven't found much googling things like "Oscilloscope lithium battery change" or any other combination of search terms I can think of. Is there a name for this?

[Smoky]

On the back of the PC board you will see the two solder joints where the battery connector ties in.

Solder two short jumper wires to those solder joints and then attach a new temporary battery in parallel.

You can then un-clip the old battery harness and replace the cell without losing power to the board.

When you're finished, de-solder the temporary jumper wires/battery.

[james_s]

I unplugged the batteries in mine and it sat that way for months. As far as I could tell, the only thing those batteries keep alive are saved waveforms, settings, time/date and SPC constants. I ran SPC on my TDS420 and it passed, everything was then good to go, only takes a few minutes and you should run SPC occasionally anyway.

There is no critical calibration data stored in NVRAM that I can see.

[Smoky]

James, I was careful when changing-out the batteries because the TDS420 that I have has the FFT option installed. I'm not sure how that option is stored but I didn't want to take a chance on losing it. Here's a shot I took today when I fed a 12kHz signal into it:



I plan to work on CaptDon's TDS420 and I was thinking about trying to activate options 1M and 2F within it. Has there been a clear path to success in doing that?

Anyway, since I have plenty of extra parts to do so, would anyone know which board the "options" are stored on?

For the audio work that I do, the TDS420 fits the bill 
 

[OptoIsolated3]

Update: Upon further inspection nearly all smd electrolytic capacitors have leaked. The THT electrolytic capacitors seem fine. The time in between the initial inspection was about 5 months ago and the scope was powered on for about ~40 min. This caused corrosion in multiple places. My digi-key order came and i'm replacing all of those smd caps. I'm going to desolder them, then wash the boards in with a light detergent and water, then rinse with distilled water. After that ill let it dry and then do a pass with IPA and compressed air. Here is some Pics of the board!  https://imgur.com/a/oudEQJf

[james_s]

James, I was careful when changing-out the batteries because the TDS420 that I have has the FFT option installed. I'm not sure how that option is stored but I didn't want to take a chance on losing it. Here's a shot I took today when I fed a 12kHz signal into it:

(Attachment Link)

I plan to work on CaptDon's TDS420 and I was thinking about trying to activate options 1M and 2F within it. Has there been a clear path to success in doing that?

Anyway, since I have plenty of extra parts to do so, would anyone know which board the "options" are stored on?

For the audio work that I do, the TDS420 fits the bill 

The options are stored the same way the other TDS scopes and can be enabled via GPIB or serial commands. Note that the TDS400 series does not support all of these options and some require physical hardware to be present but those it does support can be enabled with the same commands as the bigger models. Also there are two dip switches on the CPU board you have to flip instead of the single switch as in the TDS500/600/700.

At least in my case, option 1M required installing 6 additional SRAM chips on the DSP board, I did it on mine just for fun, cost about $13 from Digikey. Look on the board and see if there are vacant spots, I added the chips and it just worked, I didn't have to enable anything in software for that.


Power on
Set calibration protect switche to unprotected
Send "password pitbull" via the GPIB.
Send:
WORDCONSTANT:ATPUT 327686,1 Option 1M (130,000 record length)
WORDCONSTANT:ATPUT 327687,1 Option 05 (Video trigger) *
WORDCONSTANT:ATPUT 327688,1 Option 13 (RS-232/Centronics Hardcopy Interface Ports) *
WORDCONSTANT:ATPUT 327689,1 Option 2F (Advanced DSP math)
WORDCONSTANT:ATPUT 327690,1 Option 1F (Floppy disk) *
WORDCONSTANT:ATPUT 327692,1 Option 2C (Communication Signal Analyzer)
WORDCONSTANT:ATPUT 327693,1 Option 3C (P6701B with system calibration) *
WORDCONSTANT:ATPUT 327694,1 Option 4C (P6703B with system calibration) *
WORDCONSTANT:ATPUT 327695,1 Option 2M (8 M acquisition length) *
Return the calibration switch the normal working position.
Cycle the power and see the initial banner.

Edited to make the strings you actually send to the scope green so it's more clear.

[james_s]

Update: Upon further inspection nearly all smd electrolytic capacitors have leaked. The THT electrolytic capacitors seem fine. The time in between the initial inspection was about 5 months ago and the scope was powered on for about ~40 min. This caused corrosion in multiple places. My digi-key order came and i'm replacing all of those smd caps. I'm going to desolder them, then wash the boards in with a light detergent and water, then rinse with distilled water. After that ill let it dry and then do a pass with IPA and compressed air. Here is some Pics of the board!  https://imgur.com/a/oudEQJf

Yep, looks about right, glad you caught it before things got too bad. I have a TDS460 that was more severely damaged, I still don't have that fully working, SPC fails on it. The newer scopes are not quite as prone to leakage as the older ones but those SMD lytics *always* go bad. They were notorious in 90s camcorders, you'd pull it out of the closet, open the tape door and the smell of rotting fish was a dead giveaway. Those things contained dozens and dozens of them and were a real nightmare to work on. 68k Macs are full of them too, I re-capped a lot of IIci's, LC's, SE's, etc. They leak and it makes an awful mess, you really have to be sure you get all the electrolyte out from underneath ICs and out of vias.

There's a crazy sounding trick someone showed me for removing the old caps that works amazingly well. Don't try to desolder them, you'll rip up pads for sure. Instead grab the cap with needle nose pliers, push down slightly and twist, it will pop right off and then you can separately wipe each pin off the pad with a soldering iron. I've changed hundreds of them with that technique and have yet to damage a board.

Oh, also replace them a few at a time so you don't lose track. I remember there being some confusing values on there, like some are 10V at one value, and some are 10uF at whatever voltage and they look similar, you've got a bunch of detailed pictures at least which should help keep track.

[Smoky]

Honestly, I don't like what I'm seeing.

Btw, under this ribbon cable are more electrolytic capacitors:



Take a very close look at this voltage regulator. These will burn a hole through the board in no time:



The voltage regulator on the front panel board looks like it has caused some damage due to the leaking electrolyte. There are traces under it that can short/rot:



In almost 100% of the cases, the traces that flow under those small tantalum caps and SMD resistors lose continuity. The bubbly look around them appears to be extreme heat damage.

Now that I've seen this crazy leakage in the "A" version scopes, check the capacitors mounted on the backplane board too. They are through-hole type capacitors but I've seen some ugly damage done by them.

Thanks for the tips James! I have two extra display boards to scrounge extra memory from 

[Smoky]

I looked through my repair photographs and found one that may help troubleshoot your attenuator board.

Here is your picture of U134, the 14051B CMOS Multiplexer:



Here is a picture I took when I repaired a trace on my attenuator board. Follow the traces and check to see if you lack continuity under that chip, if so, it may need to come off or be jumped:

[TheDefpom]

That smell sounds like the main IEC input filter, they smell like that when they fail, it could still be a leaking capacitor but I have had a few of those IEC filters go on me and they stink...(the fun of repairing OLD gear and upping the input voltage from 115VAC to 240VAC)

[james_s]

That smell sounds like the main IEC input filter, they smell like that when they fail, it could still be a leaking capacitor but I have had a few of those IEC filters go on me and they stink...(the fun of repairing OLD gear and upping the input voltage from 115VAC to 240VAC)

No, it's the SMT electrolytic capacitors all over the boards, this has already been confirmed.

I've had those line filter capacitors burn up a couple of times and it's a whole different matter. They sizzle and smoke and stink and then fizzle out, it's nothing like the fish smell of leaky electrolytics.

[madao]

Pic with bad SMD caps, it came from  TDS420 and not TDS420A, correct ?