Condition of electrolytic capacitors in older oscilloscopes ?

[lordvader88]

How often do people replace these in older scopes? I have Tektronix 2430A so 1989 or later.

There's not that many in the PSU so maybe I'll pop out a few and test since its currently apart.

Do people replace some as a precaution just for the hell of it on older scopes ?

[tautech]

As its to do with this thread:
https://www.eevblog.com/forum/testgear/tekronix-2430a-can't-figure-out-vertical-position-setting/

For the PSU if the rails are within spec leave things as they are.
For the rest of the caps in the scope, careful inspection for leaking caps would be recommended but before you spend too much money and time on it get the primary fault with Ch 1 fixed first.

[tggzzz]

How often do people replace these in older scopes? I have Tektronix 2430A so 1989 or later.

Whenever necessary. Some old capacitors are very reliable and it can be very difficult to find a replacement

There's not that many in the PSU so maybe I'll pop out a few and test since its currently apart.

Why not replace a few other random components at the same time?

Bear in mind that replacing anything brings risks.

Do people replace some as a precaution just for the hell of it on older scopes ?

Yes and no respectively.

Most people do their research, by looking for other people's experience with their scope and similar scopes - thus 2445==2465 and 2445B==2465B, but 2445!=2445B.

[Alex Eisenhut]

Tektronix 547. Over 50 years old. Not once.

[SingedFingers]

I've had a Tek 453 in about 2009 that was fine still after 40 odd years. Same with my current Philips from '81. It's a function of quality of capacitor and how hot they get. No problems, no replace.

The ONLY thing I will blindly replace is no brand ones and those horrible RIFA X2's which make a right mess when they go boom as I found out recently.

My whitelist is BC, Philips, Vishay, Rubycon, Sprague and that's about it. Usually replace with Vishay/BC ones.

[T3sl4co1l]

More likely over time.

I recapped my Tek 475 (vintage 1973) some years ago.  Supply ripple was so bad, it was modulating the vertical amplifiers, and causing bias shifts.

Did my TDS460 a few years ago.  This is the version with the shitty SMT caps that leaked and corroded traces.  The TDS460A supposedly is fine (but will still be prone to age and temperature related failures, as all electrolytics are).

A Tek 2430 is certainly of the age where you need to check quality of the caps.  They might not need to be replaced, but they might not be fine, either.

Often, general system diagnostics will fail, signaling capacitor failure.  (In the case of my TDS460, errors popped up on the Acq, CPU and interface boards -- basically everything.  Scopes that don't have diagnostics, YMMV.)  When this point comes, replace everything.

Tim

[lordvader88]

ok so, so far people haven't told horror stories of Tek. of that era having problems, and that in general the caps in oscilloscopes last well enough then

[tggzzz]

ok so, so far people haven't told horror stories of Tek. of that era having problems, and that in general the caps in oscilloscopes last well enough then

Except when they don't.

I've replaced many tant beads after they have shorted, and some prophylactically where the initial rating was "optimistic" (15V rated on a 13V line).

I've had an X2 short circuit and vapourise a series resistor. Visible cracking in the translucent exterior is a big warning sign! SingedFingers will appreciate that, and has the right attitude!

I've had silver "liquid" tants spew acid across traces, destroying them.

But in general I'll replace things when they cause problems on a PSU rail, and I'll consider replacing nearby caps when they are already accessible.

[barry14]

I have had a Tektronix 453 since 1972 (it was built in 1967).  It has worked perfectly all these years without any parts replacement and calibration adjustments.  Its timebase is still well within the specified accuracy. I recently measured its bandwidth to be  about 60 MHz (well above the specified 50 MHz).  Its build quality is second to none and that includes military hardware (I was a defense electrical engineer for over 42 years). I also have a Tektronix 465B which is a few years newer.  It also works fine and I have never replaced any parts in it either.  Tektronix in the 60's and 70's built the finest equipment with high quality parts (many made by Tektronix including CRTs, resistors, capacitors, and transformers).  I don't think any equipment is built today to the same standards probably because it would be too expensive.  Also, today's equipment is not intended to last as long since technology improvements occur more rapidly which makes everything obsolete in a few years.

[dan3460]

There are many histories of scopes not needing parts as they are as needing many parts. Depending on the model of the scope, and it has to do with how hard is to get to the parts, I would do a shot gun replacing caps. For example, on a 475 a shot gun replacement takes me one or two hours. Why I do it? because early on I got a 475, chased the problem to one shorted cap (tantalum), disassemble change the cap, assemble back in and did not work.  Chase the problem to another tantalum, and repeated the same for 3 or 4 more caps. I spent more that 10 hours working on this thing.
On 2465s I would shot gun all the power supply caps.
On most TDSs I would replace all the SMD caps.

[cdev]

What should I use to glue electrolytic capacitors back to the PCB after removing them to test? The sticky stuff (silicone?)

[dan3460]

Normally, if I have removed a cap to test I will put a new one in its place. If you have to put a cap back, the solder will hold it in place. I have never used glue to put something back. I have used silicone type glue, to squirt some around components that could move or vibrate.

[cdev]

The goop around the bases of these caps was probably in there because they were mounted sideways.

 They all were 22000 uf Sprague electrolytics.  The ESR was approximately the same on all of them, just slightly over 1 ohm.

[dan3460]

Don't know about the goop.
As per ESR, seems to me that would be ok.
Here is a question, what prompted you to de-solder the caps to test them?
De-solder put a lot of stress on the board and on the part you are taken out. Many times with old solder you have to apply heat longer and that heat is transferred to the part. Only de-solder part that you are almost sure that are bad. Don't get me wrong, in my years of fixing things I have chased many rabbits that turned out to be ok.

[cdev]

Age.