Analog Scope Vertical Drift

[rexxar]

Last year sometime, I picked up a used Leader LBO 502 analog scope on ebay for $17(!) After I'd used it for a while, I noticed that the trace drifted upward very slowly as the scope warmed up. This really didn't bother me, because I just got this for looking at waveforms, I didn't really need to take measurements. I did a cursory search online for service manuals, and didn't find anything.


Since I just started school to get my EE, I figured I'd have a go at fixing this thing. I've looked everywhere, and the service manual for this scope just doesn't exist online. Since the main board also goes to this scope's big brother LBO 505, I tried to find that manual as well. I managed to find it, but they had only uploaded the first ten pages... and the servicing started on page 40. 

Does anyone here happen to have a copy of this manual they could scan and upload?

Without a service manual, or even a schematic, I don't know where to start looking inside the scope. Absolutely nothing is labelled. Since everything else works fine, I didn't expect to see any blown up components, and of course there aren't any. (As an aside, though, all of the caps are Nippon Chemi-Con, except the HV ones, which are Marcon)

Not sure how to progress from here.   

[c4757p]

Start tracing back from the vertical deflection inputs from the CRT. Stick a voltmeter on things and see which voltages drift with the trace. When you've traced it back to a point where it no longer drifts you know you've found it.

Doesn't sound unusual for a cheapass scope to me, though. Some things are going to run hot, so you're going to be subject to temperature coefficients. I wouldn't use the thing for measurements anyway...

[rexxar]

Start tracing back from the vertical deflection inputs from the CRT. Stick a voltmeter on things and see which voltages drift with the trace. When you've traced it back to a point where it no longer drifts you know you've found it.

Sounds reasonable, but how do I know which pin is which?

Doesn't sound unusual for a cheapass scope to me, though. Some things are going to run hot, so you're going to be subject to temperature coefficients. I wouldn't use the thing for measurements anyway...

Heh, no I plan to get a 'real' DSO once I save up the money.

This oscilloscope definitely wasn't built with servicing in mind. To get the main board out, I'd have to de-solder the connections from the front panel 

edit: Would the deflection inputs be in the kV range? I wouldn't trust my cheap DMM with high voltages.

[c4757p]

Start tracing back from the vertical deflection inputs from the CRT. Stick a voltmeter on things and see which voltages drift with the trace. When you've traced it back to a point where it no longer drifts you know you've found it.

Sounds reasonable, but how do I know which pin is which?

Can you see through the glass? A lot of cheaper tubes have some clear sections where you can see the plates. Otherwise, find a manual for a similar model, they probably used the same CRT in all or most of their scopes at the time. I wouldn't go probing around, some of the voltages you find may make you sad.

edit: Would the deflection inputs be in the kV range? I wouldn't trust my cheap DMM with high voltages.

Hundreds of volts but not thousands. Use proper probing technique and you shouldn't die.

[rexxar]

Start tracing back from the vertical deflection inputs from the CRT. Stick a voltmeter on things and see which voltages drift with the trace. When you've traced it back to a point where it no longer drifts you know you've found it.

Sounds reasonable, but how do I know which pin is which?

Can you see through the glass? A lot of cheaper tubes have some clear sections where you can see the plates. Otherwise, find a manual for a similar model, they probably used the same CRT in all or most of their scopes at the time. I wouldn't go probing around, some of the voltages you find may make you sad.

Okay, the CRT was under a shield, and when I pulled it out, I can see something through the glass. I'm not sure what I'm looking at, though. I googled the part number (toshiba 130bhb1) and didn't immediately come up with anything. I'll attach some pics. The pics are sideways, the big white parts are at the top and bottom.

[edavid]

Usually this happens right at the front end, since later stages are differential.  I would start with the input JFET and see what its source is doing.  It may not really be possible to fix it.

Also, the Leader LBO-324 manual is on BAMA, can't hurt to look at it: http://bama.edebris.com/manuals/leader/lbo324/

[c4757p]

Not of much use, unless you can look down the tube. I'll have a look around tomorrow. It's pretty late where I am though so I am going to bed now...

[rexxar]

Usually this happens right at the front end, since later stages are differential.  I would start with the input JFET and see what its source is doing.  It may not really be possible to fix it.

Good call, that seems to be where it is.

I took measurements of the source voltage over about an hour and got some interesting results. It started out at 7.23V, and after 30 seconds it was up to 7.27V, and the trace had already gone down 1/5cm. At 10 minutes it was 7.41V and -1.5cm. At 30, it was 7.46V and -2cm, and at 1 hour it was 1.5V and -2.5cm.

The JFET is on a little V-Amp board which has a plus and minus 27 volt rail, both of which were rock steady the whole time.

Does this sound like a fixable problem?

[dfmischler]

Does this sound like a fixable problem?

Maybe.  Got any circuit freeze spray?  It's useful for locating temperature sensitive components.  See if there is a component that either works correctly or fails completely when you freeze it.  If so, see if you can get a replacement for it.

[edavid]

Does this sound like a fixable problem?

It depends on the rest of the circuit.  Does it use a dual JFET, 2 discrete JFETs, or just 1 JFET?  Does it have a DC balance adjustment like the LBO-324?

[rexxar]

Does this sound like a fixable problem?

It depends on the rest of the circuit.  Does it use a dual JFET, 2 discrete JFETs, or just 1 JFET?  Does it have a DC balance adjustment like the LBO-324?

There are two identical JFETs strapped together in a socket (I guess to keep them at the same temp?). I've not looked to closely at how the second one is connected. There is a DC balance adjustment, but I think I've found the problem.

I neglected the first law of troubleshooting: thou shalt check voltages There's a handy testpoint here that's supposed to be at 12V. It's closer to 14. It's attached to a TO-220 device, which I'm guessing should be a regulator. The device is really hot, 40C with the case open and the air on in my apartment.

I happened to have some heatsinks lying around, so I put one in the freezer, and attached it to the regulator. Keeping the temperature down keeps the voltage-- and the trace-- stable, but the voltage is still wrong.

I'm assuming it's a regulator, just based on the location, but I can't find any concrete information about it. I think it's Toshiba branded, and the part number is D235. Some places say it's an audio transistor, some it's an NPN amp, but I swear I saw somewhere call it a 12V regulator, but I can't find that site again.

[edavid]

Does this sound like a fixable problem?

It depends on the rest of the circuit.  Does it use a dual JFET, 2 discrete JFETs, or just 1 JFET?  Does it have a DC balance adjustment like the LBO-324?

There are two identical JFETs strapped together in a socket (I guess to keep them at the same temp?). I've not looked to closely at how the second one is connected. There is a DC balance adjustment, but I think I've found the problem.

I neglected the first law of troubleshooting: thou shalt check voltages There's a handy testpoint here that's supposed to be at 12V. It's closer to 14. It's attached to a TO-220 device, which I'm guessing should be a regulator. The device is really hot, 40C with the case open and the air on in my apartment.

I happened to have some heatsinks lying around, so I put one in the freezer, and attached it to the regulator. Keeping the temperature down keeps the voltage-- and the trace-- stable, but the voltage is still wrong.

I'm assuming it's a regulator, just based on the location, but I can't find any concrete information about it. I think it's Toshiba branded, and the part number is D235. Some places say it's an audio transistor, some it's an NPN amp, but I swear I saw somewhere call it a 12V regulator, but I can't find that site again.

If it's really marked D235, that would be a 2SD235 transistor, but I wonder if there's another digit in there that you missed.  They wouldn't mark a regulator with a transistor-like code like that.

[c4757p]

There are two identical JFETs strapped together in a socket (I guess to keep them at the same temp?). I've not looked to closely at how the second one is connected. There is a DC balance adjustment, but I think I've found the problem.

Probably matched JFETs. You'll need a big batch to pick from and a curve tracer (I suppose you could do without the curve tracer but you'll have to take a tedious series of measurements) to match a pair.

I neglected the first law of troubleshooting: thou shalt check voltages There's a handy testpoint here that's supposed to be at 12V. It's closer to 14. It's attached to a TO-220 device, which I'm guessing should be a regulator. The device is really hot, 40C with the case open and the air on in my apartment.

"Closer to 14" may be close enough, if it's just an auxiliary 12V rail, and 40C isn't really hot. The more important part is stability, both long term and short term.

I happened to have some heatsinks lying around, so I put one in the freezer, and attached it to the regulator. Keeping the temperature down keeps the voltage-- and the trace-- stable, but the voltage is still wrong.

The trace stabilizes if you cool the regulator? Try replacing it.

I'm assuming it's a regulator, just based on the location, but I can't find any concrete information about it. I think it's Toshiba branded, and the part number is D235. Some places say it's an audio transistor, some it's an NPN amp, but I swear I saw somewhere call it a 12V regulator, but I can't find that site again.

It's probably an NPN transistor acting as a regulator. Only takes a small handful of discrete parts to make a regulator (resistor, Zener diode, and one or two transistors), so it used to be quite common. Any old TO-220 NPN transistor should do as a replacement.

[rexxar]

Well, I dropped a new transistor in, and nothing's changed except the 12V rail is a little lower and the transistor runs cooler. I think the improvement I was seeing earlier was because I had left the case open; when I put it back on and let it sit, the trace crept back to where it was before.

So, something's overheating. A few of these big electrolytic caps look a bit sick, could they be the culprit?

[edavid]

Well, I dropped a new transistor in, and nothing's changed except the 12V rail is a little lower and the transistor runs cooler. I think the improvement I was seeing earlier was because I had left the case open; when I put it back on and let it sit, the trace crept back to where it was before.

So, something's overheating. A few of these big electrolytic caps look a bit sick, could they be the culprit?

Why do you think it's overheating?  Try adjusting the DC balance.  Be sure to mark the original position in case you make it worse.

[rexxar]

Why do you think it's overheating?

Because when the case is open, and the scope has good ventilation, it behaves normally. When I put the case on, and things heat up, the trace wanders away.

Try adjusting the DC balance.  Be sure to mark the original position in case you make it worse.

Okay, what's it supposed to do?

[c4757p]

Why do you think it's overheating?

Because when the case is open, and the scope has good ventilation, it behaves normally. When I put the case on, and things heat up, the trace wanders away.

That is a possibility. Shitty capacitors and semiconductors often become shittier as they heat up. You could try changing some of the sick capacitors.

Interestingly, many items run hotter when the case is open because you've disrupted planned ventilation paths. My very tightly packed Tektronix 2445A is an example. I doubt this one is like that - I don't think it even has a fan - but it's something to keep in mind.

Try adjusting the DC balance.  Be sure to mark the original position in case you make it worse.

Okay, what's it supposed to do?

The DC balance has to do with the vertical offset between scales on each channel. I don't think it'll cause drift like that, unless I'm missing something?

[rexxar]

I can't seem to find the right value caps on Mouser or Digikey. The caps are 4.7uf at 500v, and I can't find those as axial caps. This is in the HV rectifier section, so how critical are the actual values?

edit: There's also four huge oil capacitors on the same board. How prone are these to failure? Should they be replaced after the 30 years they've been around?

[vk6zgo]

I can't seem to find the right value caps on Mouser or Digikey. The caps are 4.7uf at 500v, and I can't find those as axial caps. This is in the HV rectifier section, so how critical are the actual values?

edit: There's also four huge oil capacitors on the same board. How prone are these to failure? Should they be replaced after the 30 years they've been around?

Oil filled capacitors ,unless they are leaking,are usually good for 30+ years,so I wouldn't worry a lot about them.

Older test equipment,especially it originally was inexpensive, usually needs to warm up for a while before it stabilises.
if you are taking the position of the trace when cold as correct,then you will see drift--does it ever stop drifting?

[GK]

Last year sometime, I picked up a used Leader LBO 502 analog scope on ebay for $17(!) After I'd used it for a while, I noticed that the trace drifted upward very slowly as the scope warmed up. This really didn't bother me, because I just got this for looking at waveforms, I didn't really need to take measurements. I did a cursory search online for service manuals, and didn't find anything.

This is 100% normal behavior for any old solid state 'scope; particularly for designs that have discrete component amplifiers and/or jfet buffers at the front-end. Trust me - I have and maintain quite a collection. Tube scopes are even worse. My "hybrid" Tek 422 is the version with a nuvistor front-end and that is positively awful for both short term an long term drift. The drift generally settles down almost completely after five for ten minutes, but there almost always is some degree of long term drift that requires periodic re-calibration of the ground reference position when taking accurate measurements.


   

   

[rexxar]

Last year sometime, I picked up a used Leader LBO 502 analog scope on ebay for $17(!) After I'd used it for a while, I noticed that the trace drifted upward very slowly as the scope warmed up. This really didn't bother me, because I just got this for looking at waveforms, I didn't really need to take measurements. I did a cursory search online for service manuals, and didn't find anything.

This is 100% normal behavior for any old solid state 'scope; particularly for designs that have discrete component amplifiers and/or jfet buffers at the front-end. Trust me - I have and maintain quite a collection. Tube scopes are even worse. My "hybrid" Tek 422 is the version with a nuvistor front-end and that is positively awful for both short term an long term drift. The drift generally settles down almost completely after five for ten minutes, but there almost always is some degree of long term drift that requires periodic re-calibration of the ground reference position when taking accurate measurements.

It might stop eventually, I've never really let it run for a long time to see. Earlier in this thread I monitored the scope for an hour, and it was mostly stable then at almost 3cm below the starting point, but it was still drifting. I would think that is a significant amount of drift, and indicative of something gone wrong.

This scope might be a bit older than I thought. the date code on these transistors seems to be 1976.

[GK]

I would think that is a significant amount of drift, and indicative of something gone wrong.

That is a significant amount of drift but by no means necessarily indicative of anything actually wrong with the instrument - especially so for the "state of the art" of a design as ancient as <=1976. You can only do good for it by replacing all of the old electrolytic capacitors but don't waste too much time chasing ghosts out of lack of experience with the performance limitations of such old gear. Just reminisce the good old days of drifty scopes!
 

[vk6zgo]

With respect,a Leader LBO 502 wasn't anything like "state of the art" for 1976.
It was a reasonably inexpensive 'scope,& was appropriately,built to a budget.

[rexxar]

I would think that is a significant amount of drift, and indicative of something gone wrong.

That is a significant amount of drift but by no means indicative of anything actually wrong with the instrument - especially so for the "state of the art" of a design as ancient as 1976. You can only do good for it by replacing all of the old electrolytic capacitors but don't waste too much time chasing ghosts out of lack of experience with the performance limitations of such old gear. Just reminisce the good old days of drifty scopes!

I think replacing the caps is about as far as I'll go with this thing. It's not a huge deal that the trace wanders away, as long as the thing still works. I'm actually going to get a Rigol DS1052E soon, which I'll use for actual measurements.

Since I can't find all the right value caps, how close is 'close enough' for the HV rectifier circuit?

[GK]

With respect,a Leader LBO 502 wasn't anything like "state of the art" for 1976.
It was a reasonably inexpensive 'scope,& was appropriately,built to a budget.

Well that just further makes my point. The cheap/simplistic designs generally had worse drift performance than the expensive scopes. Just saying that the drift issue as reported isn't out of the ordinary for something that old even if it is in pristine working condition. 
 

[GK]

I think replacing the caps is about as far as I'll go with this thing. It's not a huge deal that the trace wanders away, as long as the thing still works. I'm actually going to get a Rigol DS1052E soon, which I'll use for actual measurements.

Since I can't find all the right value caps, how close is 'close enough' for the HV rectifier circuit?

If you can replace all of the old electros and actually establish that the regulated supply rails are within specification and stable then you are pretty much 95% of the way there.

As for the uF values of psu caps, one size up (or even more) generally isn't a problem at all and almost always preferable than going less in value.
 

[rexxar]

I think replacing the caps is about as far as I'll go with this thing. It's not a huge deal that the trace wanders away, as long as the thing still works. I'm actually going to get a Rigol DS1052E soon, which I'll use for actual measurements.

Since I can't find all the right value caps, how close is 'close enough' for the HV rectifier circuit?

If you can replace all of the old electros and actually establish that the regulated supply rails are within specification and stable then you are pretty much 95% of the way there.

Oh, that's another thing. How can I probe the HV stuff without blowing up my cheap DMM? I think the highest rail is 2kV

[vk6zgo]

Some of the older 'scopes had a voltage divider off the EHT,so you could read a lower voltage proportional to the high one.

The test point was marked "HV test" or something like that.