Another Tektronix 465 Oscilloscope repair! (And a question)

[pez.diSpencer]

Well... forgive me if this is useless, but I was pretty happy to get this scope running again, so I'm sharing anyway. But I'll start with a question about the PS capacitors for the tek gurus-- if I've got clean DC off the supplies (ripple amplitudes w/in tektronix limits per service manual)... would you still recommend replacing the filter caps?? I know they're 30+ years old, but if it ain't broke? (😂)

Anyway, I scored this discrete Tektronix 465 on eBay for about the price of shipping. Completely dead, but in beautiful shape cosmetically and physically... For some reason, I kinda prefer equipment that's totally dead, just fingers crossed it's not the power transformer. Anyway, I knew that the PS in these things were built pretty well, with foldback protection and all... So I was feeling alright.

I snagged a manual and went straight to the troubleshooting for the power supplies-- I had 48V on the 55V regulated supply, and nothing below the 55v supply (15v and everything else was dead). The +15V supply troubleshooting chart lead me to Q1548, the foldover protection transistor, which showed 0.66Vbe, indicating the supply was in foldback... Something connected to the supply was shorted. So, I used the provided procedure for unloading the supply, and as soon as I disconnected the supply from the trigger board-- voila! All the supplies came back to nominal.

That left the culprit on the trigger board... Only 200 things it could be at this point! But before going into crazy troubleshoot mode, I decided to start with the tantalums... and the 3rd one I got to was dead short! Checked the rest, tacked in an electrolytic to test (no tantalum on hand), and she's back to life.

All I can say is that a PHYSICAL copy of the service manual is paramount for ease of repair!!

I think I will replace all the tantalums, unless anyone would suggest otherwise.

-Spencer

[tautech]

I think I will replace all the tantalums, unless anyone would suggest otherwise.

Do check the factory spec and uprate the voltage on all that you replace.
50V tants are pretty common these days and as a rule of thumb don't subject them to more than 66% of their voltage ratings.

Nice clean unit you have, it looks like it's worth some love so look after it and may it serve you well.

[Edwin G. Pettis]

I have several Tek scopes including old tube 500 series that are over 60 years old, I've only had to replace a very few electrolytics in them and then only when they were obviously going bad.  As long as ripple and voltages are within spec I see no reason to replace caps.  I just keep an eye on them, checking every 6 months or a year, when the ripple gets close to maximum, then I would replace.  While tantalums can be problematic after an extended period, I have seen them last much longer than 30 years, I have also seen them fail in much shorter times.  These caps often do not show any outward signs of pending failure, they just die and unfortunately shorting out is a fairly common death for them.  Many factors go into how long they are going to last, including circuit design, operating temperature and quality.

Tek usually doesn't underrate their components, I see no reason to change out the caps on that premise nor do I subscribe to some arbitrary operating voltage limit on caps, if they are within their ratings, there is no need to increase their voltage ratings, this is a fallacy running around, likely caused by people using cheap quality parts.  A rule of thumb of 66% is not only arbitrary but baseless, except for cheap quality parts, there is no need for such derating.  If the operating environment is on the hot side, then yes electrolytics in particular should be derated accordingly (check data sheets).

To illustrate this fallacy, lets say you have circuit operating at 60 volts and instead of using 75V caps, you install 100V caps thinking this will give you an extra margin of safety.....nope, in time the capacitor will reform itself to the lower operating voltage and will lose its 100V capacity, it will become the lower rated capacitor you were trying to 'protect' with higher ratings.  This is a fact of electrolytics, you're just wasting time and money putting in higher rated capacitors.

[Cyberdragon]

You need an ESR meter to really know if caps are good or bad. Also, using higher voltage caps is actually BAD! Higher voltage caps have higher ESR meaning they will be less effective in a circuit meant for a lower voltage.

[pez.diSpencer]

Thanks for the replies!

I do understand that about ESR vs. voltage rating, and I do have a meter; however, pulling the PS caps to check on a meter is a big task requiring near complete disassembly (though I think it may be possible to desolder the caps from the top of the interface board and fish them out the other side, I just don't really want to try unless it's absolutely necessary).... At which point I'd be tempted to install new caps anyway. But with a ripple less than 2mv p-p (well within the tek recommended limits), I know the filter caps are at least doing their jobs... Wouldn't an increase in ESR increase the ripple?

On another note, does anyone have any recommendations for cleaning the pots? I mean, they appear to be sealed. The response is a bit scratchy on the intensity and scale illum. controls-- I haven't yet thoroughly checked all the other controls/functions until I get the caps replaced on the trigger board. I plan on doing the complete performance check and what I can of the cal procedure (I don't have a standard amplitude generator, but I can at least calibrate time base) in the service manual.

-Spencer

[tautech]

While tantalums can be problematic after an extended period, I have seen them last much longer than 30 years, I have also seen them fail in much shorter times.  These caps often do not show any outward signs of pending failure, they just die and unfortunately shorting out is a fairly common death for them.  Many factors go into how long they are going to last, including circuit design, operating temperature and quality.

Tek usually doesn't underrate their components, I see no reason to change out the caps on that premise nor do I subscribe to some arbitrary operating voltage limit on caps, if they are within their ratings, there is no need to increase their voltage ratings, this is a fallacy running around, likely caused by people using cheap quality parts.  A rule of thumb of 66% is not only arbitrary but baseless, except for cheap quality parts, there is no need for such derating. 

I think many would challenge you about Tek's design and voltage rating in the use of Tantalum caps.

There are many mentions in Tek repair threads of their application of 15V tants on 12V rails. 
Go looking at old Tek schematics and you'll find examples of Tants being used too close to their voltage ratings.

Common consensus IS to derate tants, at least to a 50% higher voltage rating but the newer Tant technologies are much better but still not immune from failures.
Compulsory watch:
https://www.eevblog.com/forum/chat/capacitors-explained-by-james-lewis-of-kemet/

Good electronic design FOR equipment that is to se decades of use AND those that are not throw away consumer items still use conservative ratings for all componentry. In this day and age the $ and bean counters drive significant compromises in component spec at the peril of longevity of service. 

[pez.diSpencer]

Okay, so I've looked at all of the tants on the trigger board: they're all rated well above the circuit voltage EXCEPT for the one that failed short, which was 20V on a 15V rail. (They're pretty much all rated 20V, but the others are on 8 and 5 volts).

This makes me feel slightly better about leaving the others alone and just replacing the failed one with a capacitor of higher rating.

[Electro Detective]

If the scope is now working great, and within spec, I would leave it well alone and just use it. 

i.e. put off any pending repairs for another day (if or when it comes)

A smell and visual test, and a zap with a cheap Mini Infrared Temperature Thermometer is as far as I go if a unit is working ok.

If it hasn't been used for a while I would leave it running for a few hours with no trigger or trace and the intensity totally backed off,
to get all the inards stabilised again, before thinking about any calibrations.

These old school scopes still do the business, but being so old I don't leave them running/idle in the background unless I'm actually using them.
Otherwise I leave them switched off to minimize further wear and tear caused by heat and components slowly aging, especially in hot weather.

FWiW I haven't had to think about any of mine since I fixed/calD them 20 years ago,
if you need one on all the time in a workshop, better to have 2 or 3 oscilloscopes and alternate usage on/off times between them,

and the bonus is that both or 3 can't possibly be out of calibration at the same time,
so you have a cheap means of ensuring their specs are ok. 
 
If the sealed pots are scratchy it's more homework you don't need to source and replace  , if simply working them back and forth a bit gets you the result 

Hey, nice scope btw ! 

[tggzzz]

There are three forms of risk in this discussion:

• damage when removing/replacing a board
• damage when removing a capacitor from a board
• risk an unreplaced capacitor will fail

It is up to you to assess the relative risks and penalty if the risk materialises. Here's my approach...

Old X/Y mains input filter caps should be replaced, particularly if their skin is already cracked. I've had one cause a resistor and some prepreg to disappear. In some equipment these may be "hidden" inside mains lead socket, in which case the whole socket has to be replaced. (That added 33% to the cost of my 8digit DVM ).

Large 4-legged aluminium mains filter electrolytics are difficult to remove without damaging PCB tracks and PTHs. Replace them if the ripple is too high, but be careful to add bodge wires where necessary - including where the cap's case was acting as a jumper.

Tants are relatively easy to replace.
Don't bother replacing cylindrical "silver" tants unless they are faulty and/or have spewed acid across the tracks. If the latter has happened, consider replacing similar capacitors before they fail.
If a tant bead has failed, replace "similarly rated" tants - especially 15V tants on a 13V line - before they take out nearby components. (On 485s the 13V "rail" isn't on the main PSU schematic!).

If it it difficult/risky to remove/replace a board and you already have it out, consider wholesale capacitor replacement.

[pez.diSpencer]

If a tant bead has failed, replace "similarly rated" tants - especially 15V tants on a 13V line - before they take out nearby components. (On 485s the 13V "rail" isn't on the main PSU schematic!).
.

By "tant bead" you mean the little dipped tants? (In the case of the tek, the color coded dipped tants look like little pieces of candy all over the place). That's kinda my concern-- there's about eleven of the same tantalum on the trigger board. It's a 20V 15uF tant. One of them failed short; however, it was the only one that operated very near its maximum voltage (75%). The others are operating well below their max voltage. Initially I thought to replace all tants similar to the one that failed, but now I feel that it's best just to replace the one with a better rated cap, attributing its failure to stress.

[tautech]

If a tant bead has failed, replace "similarly rated" tants - especially 15V tants on a 13V line - before they take out nearby components. (On 485s the 13V "rail" isn't on the main PSU schematic!).
.

By "tant bead" you mean the little dipped tants?

They are the same.

The others are operating well below their max voltage. Initially I thought to replace all tants similar to the one that failed, but now I feel that it's best just to replace the one with a better rated cap, attributing its failure to stress.

Yes stress but understand what that might be.
Voltage rating is only part of what has led to failure.
Different technologies are used in Tantalum manufacturing today that make them much more reliable however there are some that have had too many bad experiences with Tants that they will no longer use them but they do have some great attributes, namely high density and speed like ceramics.

For better understanding watch the vid I linked and have a study of the last half of this page:
https://en.wikipedia.org/wiki/Tantalum_capacitor

[tggzzz]

If a tant bead has failed, replace "similarly rated" tants - especially 15V tants on a 13V line - before they take out nearby components. (On 485s the 13V "rail" isn't on the main PSU schematic!).
.

By "tant bead" you mean the little dipped tants? (In the case of the tek, the color coded dipped tants look like little pieces of candy all over the place). That's kinda my concern-- there's about eleven of the same tantalum on the trigger board. It's a 20V 15uF tant. One of them failed short; however, it was the only one that operated very near its maximum voltage (75%). The others are operating well below their max voltage. Initially I thought to replace all tants similar to the one that failed, but now I feel that it's best just to replace the one with a better rated cap, attributing its failure to stress.

They aren't colour coded in the way resistors are colour coded. In the pictures above they are the chocolate brown with a yellow blob (but verify that before taking any destructive action!)

My phrase "similarly rated" might have been confusing; it would have been better to say "similarly underrated". Hence your qualification of 75% of maximum voltage is a clearer statement.

Personally I have no strong opinion about what percentage implies the capacitor is "at danger", but when replacing a tant bead I try to have the voltage rating at least twice the applied voltage. OTOH, if the only available tant beads had a lower spec, you can still use those to replace faulty caps - the new ones would probably work until you dispose of the scope! It is all a probability, not an absolute.

BTW, with care I have found it possible to remove some tant beads without removing the boards. That does, of course, increase the risk of solder splatters ending up in an unfortunate place.

As a separate issue, if you need to clean the switch contacts on the boards, check other topics to see how to do it safely: right/wrong paper, chemicals and directions of movement.

[Armadillo]

Can't remember who said it, maybe Dave...
"if it isn't spoil, don't touch it".   

IMHO, manufacturers always have the best quality parts, tested/run-in. They have the QC departments and they reject components by the "LOTS/BATCH" and these ultimately ended up in the market. These must go somewhere right.?

How sure are we, that we getting correct parts? So as the saying goes, "if it isn't spoil, don't touch it".

For example, you won't want to bend even a single strand of bonding wire in a YIG oscillator right? I think is the same as matched pair theory. It somehow alters the calibration/performance somewhat. That includes opening up and invading into the instrument.

We made impression ourselves that by changing new components, that it ought to be better. Well some human always think that by adding a personal touch onto the instrument, the instrument will be instantly better if not better than the original without any further verifying tests. LOL. 

you get the picture....

.... I heard Dave saying......"No Touchy"... 

[floobydust]

Here is link to catalog of Tektronix part number to vendor part number cross reference 27MB .zip from Dave Miller AA4DF. Very useful for repairing 400 series. Note Tek bin ranked many parts; transistors by beta and breakdown voltage etc.

[pez.diSpencer]

Here is link to catalog of Tektronix part number to vendor part number cross reference 27MB .zip from Dave Miller AA4DF. Very useful for repairing 400 series. Note Tek bin ranked many parts; transistors by beta and breakdown voltage etc.

Thanks! That might be very helpful!

Anyway, I've ordered a handful of 35V 15uF tants from mouser. Once the new one is in place, I will double-check ripple from the PS's and voltages around the tants to verify they are not "seeing" anything they shouldn't be. Otherwise, I'll run the performance check after full warm-up and post about any other issues I may come across. Otherwise, with a bit of cleaning of the knobs, this thing will look like brand-new (and hopefully work like it too). I don't think it saw a lot of use-- I don't see a lot of the typical signs of wear. Perhaps this cap failed fairly early in the scope's life and it just never got repaired? Hard to believe, but ya never know.

[PartialDischarge]

What's so special about this scope?

I just got one for free, it has an issue with the sweep, trace looks like a dot most of the time, sometimes disappears...
It's very clean on the inside and outside, tube is bright as new, but I don't know if I should spend time to repair it 

[David Hess]

What's so special about this scope?

It has the right feature set and performance and it is easy to maintain.

[Old Don]

Spencer, I would button it back up and start putting it to use.

[grtyvr]

That was exactly what failed in mine.  Nice looking scope you have!