bad channel 1 on Tek 2445

[JLNY]

All,
I am currently working on repairing a Tek 2445. When I initially got it, it was not working and needed to have the low voltage power supply completely re-capped. I have done that and I appear to be getting good readings on all rails. The unit now powers on with no errors, and most of the functions I have tested appear to be working.

The main issue that I am having is that channel 1 seems to be displaying at roughly half the amplitude it should, and seems to have an issue with the frequency response (see image, channel 1 is the top trace). This appears to be the case across all volts/div settings and in DC coupled, AC coupled, and 50 Ohms modes. Adjusting the ch1 input capacitance seems to have no effect.  I have tried swapping the probes and also inputting directly from a cable and I do not believe there is any problem with the probe. Channel 2 (lower trace) appears to be normal.

Does anyone have any thoughts on what it might be or what might be the first thing to check? Although I have re-capped the PSU, the main board and the high voltage section have not been re-capped. I have noticed that a lot of the HV CRT adjustments seem to be pretty far towards their extremes, so I wonder if re-capping the HV section might help also be a good idea? It might just be due to drift on the components. Any suggestions would be greatly appreciated.

[MarkL]

Your Ch2 appears to be working, so the problem has to be something specific to Ch1 and not in common to both channels.  For that reason, it would not be the HV or CRT adjustments.  And it would be best not to mess with any adjustments until you figure out the Ch1 issue.

Offhand, it looks like something might not be making contact.  It could be a relay in the attenuator, or perhaps a bad contact on the Ch1 preamp socket.

Do you have another scope for troubleshooting?  If not, since your Ch2 appears to be working, you can use it to probe around in Ch1.

To start, I would suggest connecting the square wave signal to Ch1 and look at the Ch1 attenuator output, labeled "B" on the schematic.  You'll need to take off the small shield to access that area.  See if the output has the same shape as what you're seeing on the screen.  And use a 10x 10Mohm probe since the impedance is high at that test point.

If the waveform has the same curved shape, the problem is in the attenuator.  If not, look further downstream in the signal chain.

You could try reseating the Ch1 preamp, and/or looking at its output to see if the problem starts in the preamp.

You may also find the test setups, test points, and waveforms in the schematics to be useful for this problem.

If you have another scope, an easy to way pinpoint the problem is to feed the same square wave to both Ch1 and Ch2 and then start comparing the same test points in each of the channels.  Find out where the waveform starts being different between the two.

[Tom45]

When you change the input mode (DC, AC, Gnd, etc) you hear a relay clicking. I have a 2445 that intermittently has the same symptom as yours. I think it might be related to the relay.

Since I have a 2465 that I prefer to use I haven't looked into the 2445 input intermittent failure.

[JLNY]

Hmm, these all sounds like good suggestions. I can definitely hear relays clicking when I change input modes, but my limited testing so far doesn't seem to show the issue as intermittent, but if one of the relay contacts has significant corrosion on it, it may be making bad contact more times than not. I'll try cycling the input modes a bit and see if I can get it to work intermittently. Is there any easy way to take the cover off the attenuator assembly? I might try giving it a shot of contact cleaner if I can reach it.

I do have another scope. I'll see if I can set it up later today and do as you suggest. I did try probing channel 1 a bit with the channel 2, but even with a 10x probe, it seemed like the probe was having an effect on the response of ch 1. I think that might have been with ch 1 in one of the 1M imput impedance modes, so I might try that again in 50 Ohm mode.

The possibility of a bad preamp is another thought that occurred to me. It seems that spare preamp hybrids aren't too difficult to come by ebay, but it would be annoying to have to order one just to test. If nothing else pans out, I wonder if it might be possible to swap the ch 1 and ch 2 hybrids and see if that isolates the issue?

[MarkL]

Hmm, these all sounds like good suggestions. I can definitely hear relays clicking when I change input modes, but my limited testing so far doesn't seem to show the issue as intermittent, but if one of the relay contacts has significant corrosion on it, it may be making bad contact more times than not. I'll try cycling the input modes a bit and see if I can get it to work intermittently. Is there any easy way to take the cover off the attenuator assembly? I might try giving it a shot of contact cleaner if I can reach it.

I don't know of an easy way to get the cover off.  Looking at the cover screws, it appears the attenuator assembly has to be removed, which is not trivial.  There are removal instructions on page 6-27.  I would be sure it's the attenuator first before diving into removal.

I do have another scope. I'll see if I can set it up later today and do as you suggest. I did try probing channel 1 a bit with the channel 2, but even with a 10x probe, it seemed like the probe was having an effect on the response of ch 1. I think that might have been with ch 1 in one of the 1M imput impedance modes, so I might try that again in 50 Ohm mode.

You should see some loading from the probe, so the amplitude will go down 10% or so.  But the basic shape should remain consistent, unless the issue is near the output of the attenuator in which case the additional capacitance from the probe may be adding to the problem.

Can you post some screen shots?

50ohms would only appear at the input of Ch1.  The "B" signal is already high impedance at that point, so it shouldn't matter.  Wrong!  See below.

One thing you could try is measure the input resistance of Ch1 with a DMM.  It should be 1Mohm on DC on any V/div setting (try them all), infinite on AC, and 50ohms on 50ohms.

The possibility of a bad preamp is another thought that occurred to me. It seems that spare preamp hybrids aren't too difficult to come by ebay, but it would be annoying to have to order one just to test. If nothing else pans out, I wonder if it might be possible to swap the ch 1 and ch 2 hybrids and see if that isolates the issue?

Yes, you can swap them.  It's fairly easy to do, but I would do a little more testing first before taking a soldering iron to anything.  It's up to you.


EDIT: Actually, the impedance at "B" depends on the V/div setting.  I was looking at the input to the preamp where it is high impedance through R100.  So, there may be a different influence probing at "B" when 50ohm termination is on depending on the V/div setting.

[JLNY]

Progress!

I tried toggling the input mode and volts/division settings back and forth, and it seems that the issue is indeed in the relays. The issue seems to have been in one of the attenuator relays rather than the input coupling relay. The more I toggle it between the different volts/div ranges, the more they seem to come up good, possibly from breaking up the corrosion a bit. After toggling it back and forth maybe 50 times or so, it seems to be working almost every time. Perhaps it was just suffering from a bit of underuse?

I'm a bit embarrassed that the issue turned out to be so simple, but it certainly beats having to buy a new hybrid. If the issue returns, I might look into opening up the attenuator assembly and seeing if I can get some contact cleaner on it, but in the mean time I may just close it up and see how it goes. Tom, if yours was having a similar issue, I guess you might be able to try this out for yourself and see if it works for you.

[Tom45]

Progress!

Good news. I thought that might be it as your photo showing the channel one rounded and attenuated square wave is what I would see on my 2445 when it was acting up.

Yes, cycling the input mode switch does indeed help with my 2445. But it usually came back after awhile. It has been several years since I fired it up, so I don't know what it is like now. Looking into it for a more permanent fix is low on my todo list as I have other scopes available. And plenty of more pressing work.

Lets hope that regular use is all that your scope needs.

One other problem with my 2445 is that apparently a previous owner overloaded the 50 ohm input at one time. So I need to open it up anyway to work on that.

Tom

[MarkL]

Progress!

I tried toggling the input mode and volts/division settings back and forth, and it seems that the issue is indeed in the relays. The issue seems to have been in one of the attenuator relays rather than the input coupling relay. The more I toggle it between the different volts/div ranges, the more they seem to come up good, possibly from breaking up the corrosion a bit. After toggling it back and forth maybe 50 times or so, it seems to be working almost every time. Perhaps it was just suffering from a bit of underuse?
...

Great - I'm glad the problem was simple.  Contacts can get finicky from tarnish and other contaminants when sitting unused too long.

I haven't done it, but I know others have been able to remove the attenuator assembly and clean the contacts.  Use small strips of paper soaked in isopropyl and pull them out from under the contacts when closed.

With other attenuators, I've been successful at beating the contamination off the contacts by connecting the relay coils up to a 10Hz square wave from a signal generator and letting it run for a minute or two.  Use a diode in parallel with the coil to prevent damage to the sig gen from inductive kick-back.

Also, you mentioned you adjusted the Ch1 input capacitance.  I assume you mean C105.  Unless you know you got it back to where it was, you should probably readjust it.  The adjustment is described on page 5-8 and it affects probe compensation for the 50mV/div and less ranges.  You could also use an LCR meter to match the input capacitance between 0.1V and 50mV ranges, if the meter has enough resolution (0.01pF would be good).

[JLNY]

Yes, I probably plan to go through and calibrate the unit as much as I can before closing it up.

I'll definitely look into how to remove and clean the attenuator assembly if the issue comes back, but for now just toggling it a few times seems to have worked alright. Thrashing it at 10Hz sounds like a fun idea, too, and if I get really desperate, I have a can of industrial-grade contact cleaner at work that would really do a good job, but it has a tendency to damage certain plastics, so I wouldn't do that unless I took it all the way out of the case.