Tektronix 475 twitching display

[Gazucha]

Hi all.

 I'm writing here as I don't have the time right now to go deeply into diagnosing the problem which began yesterday (when something went pop) with a TEK 475.

Perhaps someone recognises the symptoms and can point me in the right direction?

 What is happening is that with both input channels 1 & 2 connected to the CALIBRATOR, and with the 'ALT' option selected, along with 'AC' in COUPLING, (and only with the 'SOURCE' in 'NORM'), one of the channels 'twitches' horizontally. Which channel twitches, depends on whether the 'A Trigger' SLOPE switch is switched left or right.

Settings right now are .2 Volts/Div and .2 ms/Div

 If you switch the SOURCE to CH 1 or CH 2, it stops twitching.
 If you take it out of ALT, it stops twitching.
 If you switch the COUPLING to 'LF REJECT', it stops twitching
 If you vertically adjust the channels so they are both in the same half (upper or lower) of the display, it stops twitching.

When it went crack/pop/bang, there was no smoke, and there is no obvious component damage, even under a microscope.

Any pointers would be very appreciated.

We're in the middle of works at home and my spending a day reading manuals and testing electronics won't go down well with the wife...

[T3sl4co1l]

Would look around for suspiciously exploded tantalum in the trigger circuit.

Funny, I never used NORM trigger source, I probably wouldn't notice this problem.. (unless it's just that it's most apparent there.)

Tim

[Gazucha]

Thanks Tim.

 I have looked in the trigger circuit and was expecting to find half a tantalum but nothing obvious.

Don't forget that the trigger circuit works fine when taken out of ALT.

It's gonna be a case of a good look at the circuit and following it through. Just stretched for time right now.

[bd139]

Tants are evil and sometimes go with a bang and blow a hole in their butt where you can't see it. Had that in a Tek SA a while ago. Right in the middle of the damn thing inside an RF box. I managed to find that without even having to power the plugin up.

Worth check them all for shorts with a DMM. Literally 90% of problems with these Teks are tants, bad leaf switches and dying electrolytic caps.

[Gazucha]

But where?

 Considering the trigger circuit seems fine when taken out of ALT or with both traces in the same half of the display.

[bd139]

The scope for problems here are quite large. The trigger pathway spans the rear half of the vertical amp, the interface board next to the timebase and a chunk of the timebase. I would start at the vertical board at or around the channel switch myself and work along with the service manual.

Do you have another scope to signal trace as well?

I wouldn't power it up at the moment until you've eliminated all the tants as some of them are in RC decoupling networks and the resistors tend to burn after a few minutes if you fire them up with shorted tants.

[Gazucha]

Do you have another scope to signal trace as well?

No other scope here.

I wouldn't power it up at the moment until you've eliminated all the tants as some of them are in RC decoupling networks and the resistors tend to burn after a few minutes if you fire them up with shorted tants.

Well, I have had it powered up and nothing else smoked, so perhaps that would at least indicate it is not a shorted tant?


Unless I am very mistaken, as the case was already off, the 'sound' appeared to come from the right hand side, perhaps underneath. I shall check it all when possible, but it didn't appear to come from the vertical board.

[bd139]

If it came from underneath, might be one of the bridge rectifiers. They tend to die noisily and result in rails jumping around causing triggering problems.  I'd check power supply voltage/ripple first. Should be able to get good neough ripple measurements with a DMM in AC volts.

[Gazucha]

 Shall do.

  It'll have to be this evening.

Working on a floor and off to the builders now...

 But thank you

 

[Gazucha]

 Are you talking about the low-voltage rail?

 +110V
 +50v
 +15v
 -15v
 -8v
 +5v

If so, there are just fluctuating readings on all except the +5v which goes to 0.000v

[bd139]

Yes. You’re saying the 5v goes down to zero occasionally?

If so axial electrolytic on main board slap bang in the middle is probably duff.

[Gazucha]

Yes. You’re saying the 5v goes down to zero occasionally?

If so axial electrolytic on main board slap bang in the middle is probably duff.

Goes to zero, full stop.

Don't suppose you have the number for that electrolytic?

On the A9 main interface board?

I'm only seeing two axials. One is under the high voltage cover.
The other is C1505

[bd139]

Sorry was thinking of Tek 465 there. I'm in front of SMM now.

Look for C1458. It's a tant next to the fuse on the big interface / power supply board near the rear of the scope. I had that short on the last one I had which eventually caused it to blow up.

[Gazucha]

Thanks for the suggestion .


Found and tested that.

No holes in it. Not burnt.

 Capacitor C1458 - 33µF, 10 volt (20% tolerance)
 
Capacitance reading was 35.2µF and  the ESR was 1.3 Ω


Gonna have to have a look for the test points in the schematic and have a closer look


 

[Gazucha]

Following the idea that something is wrong in the +5v line, let's see where that circuit goes.

 In the service manual, on the Power Supply & Distribution schematic (attached), there is a boxed section in the upper right corner which is labeled, 'Partial A8, Trig Gen and Z Axis Logic Board' and contains a small circuit with the 'Low Line' lamp. 

 Maybe I'm being a bit thick (and maybe these are irrelevant), but I can't see any of those numbered components (*1492 -*1499) on any photo, nor grid locator, nor amendment, nor on my scopes A8 board, even though I can see the Low Line lamp!

Any tip here would be greatly appreciated.


On a different tangent, why would a failed cap cease to distort the trace when both traces are in the same half of the display?

This is gonna be a great learning curve...

[David Hess]

If you switch the SOURCE to CH 1 or CH 2, it stops twitching.
If you take it out of ALT, it stops twitching.
If you switch the COUPLING to 'LF REJECT', it stops twitching
If you vertically adjust the channels so they are both in the same half (upper or lower) of the display, it stops twitching.

That says it all but you have to know something about the design history of oscilloscopes.  What you are seeing is completely normal.

When the trigger source is set to NORMal (sometimes called VERTical mostly on older oscilloscopes) and ALTernate is selected, the trigger source alternates between channel 1 and 2 depending on which is being displayed.  If the channels do not overlap on the CRT, then the jump in level when the vertical channel switch selects the other channel can be picked up by the trigger circuit causing a false trigger.  Aligning the traces on the same vertical level removes the jump in level preventing this problem.  The fact that this configuration works at all is amazing.

Historically this was referred to as the VERTical trigger source and physically the connection was made to some point after the vertical channel switch as with the 465/465B/475/475A so the trigger source was exactly what is seen on the CRT.  For some very early tube oscilloscope, this was the only trigger source.  VERTical in this case means literally the vertical CRT amplifier signal.  Slightly more modern oscilloscopes including the 485 and 22xx series which replaced the 100MHz 465 series include a completely separate trigger channel switch for the trigger circuit path which duplicates this behavior by deliberately switching between the channels in ALTernate mode following the vertical channel switch.  The difference on these newer oscilloscopes is that if CHOP mode is selected, which would make no sense on an earlier oscilloscope like the 475 because the trigger signal would include the asynchronous chopping waveform, the trigger channel switch adds the channels together instead which seems a little weird but allows proper triggering.

Alternate triggering is one of those functions which is very useful in specific applications but missing in practically all DSOs; the Tektronix 22xx DSOs can do it because of their analog heritage and I think the obsolete Rigol 1000E series supports it although I have no idea why.  For instance it can be used to stably display two asynchronous signals at the same sweep speed in real time.

[Gazucha]

Hi David.

  Thanks for the lesson, and putting my mind at rest (even though I am sure it wasn't twitching before the 'pop').

 Still leaves a couple of questions, such as, "what went pop?" and even accepting that a DMM is a poor AC ripple tester, "why is there no ripple on the +5v test point?"


To be honest, I shall probably go no further with unnecessary probing and repairs,  seeing that it is behaving as designed.

Great! 

 

[Gazucha]

I have a confession to make.

 The source of the 'pop' (which led to this thread) has been traced... And it wasn't a component.

 The table upon which the scope is sitting has a wicker laundry basket leaning against it. When knocking the table just now, the same ' cracky pop'  was heard.


 Bollocks!!!!



 Which explains why there was no smoke, no fried component etc.

 So I apologise for wasting everyones time.

 Remember kids, check the laundry basket first before starting new threads.

[bd139]

Hahaha at least you know now! Better to be safe than sorry with these things though.

[Gazucha]

I know one thing...

 It could be painfully hilarious to have an EEVblog 'Confession Box'

 

[David Hess]

Thanks for the lesson, and putting my mind at rest (even though I am sure it wasn't twitching before the 'pop').

The alternate triggering function on these oscilloscopes is sensitive to conditions so performance varies over time and between instruments and models.  Generally it works pretty well when the two channels are aligned vertically and not so well otherwise.

Tektronix made a deliberate effort over time to keep the alternate trigger functionality which was a side effect of how early oscilloscopes were designed but they did not make any extra effort to perfect or improve it so it is more of a bonus feature.

To be honest, I shall probably go no further with unnecessary probing and repairs,  seeing that it is behaving as designed.

If you think the alternate triggering is not working as well as it should, then you could do the trigger calibration procedure in the service manual and check for excessive noise and ripple on the power supply outputs.  Other than those two things, I would not mess with it for fear of damaging something for little gain.

[David Hess]

It could be painfully hilarious to have an EEVblog 'Confession Box'

Forgive me Dave for I have sinned ...

Or we could post selfies with a sign tied around our neck saying what we did.

[tggzzz]

I have a confession to make.
 The source of the 'pop' (which led to this thread) has been traced... And it wasn't a component.
 The table upon which the scope is sitting has a wicker laundry basket leaning against it. When knocking the table just now, the same ' cracky pop'  was heard.
 Bollocks!!!!
 Which explains why there was no smoke, no fried component etc.
 So I apologise for wasting everyones time.
 Remember kids, check the laundry basket first before starting new threads.

Ah. A new mistake.

I always told my daughter "make new mistakes".

[bd139]

I told mine to carefully look for other people’s mistakes and learn from them 

[tggzzz]

I told mine to carefully look for other people’s mistakes and learn from them 

People infrequently learn from their own mistakes, and rarely learn from other people's mistakes. I'll settle for the lower hanging fruit.

Having said that, the principle was clearly important when she was learning to fly gliders, because it is quite difficult to find new ways to kill yourself gliding.