EEVblog #777 - Keithley 177 Microvolt DMM Repair

[boffin]

There is also a fair bit of psychology involved too. In the case the of the Keithley here, and most of my video repairs, I go into it secretly hoping it's going to be some elusive electronics fault that I can track down like an episode of Columbo. And that probably subconsciously influences how I do things.

Just one more thing 

While the rails gave +/- 15, how stable was it?  I would have been inclined to pop a scope on the power supplies to see if they had 2-3v of ripple.  If the bridge got ker-powed, I would look next door at the supply caps.

[John Coloccia]

Spraying cleaner on switches 2 minutes into the video would have been a pretty boring video. I kind of enjoyed seeing the troubleshooting if for nothing else than it's interesting to see the thought that went into writing the manual.  Trouble shooting manuals today are written for idiots servicing disposable items:

Problem: No Sound
1) check if the device is plugged in
2) check that the speakers are plugged in

If you still have no sound, throw it out.

Problem: No video
1) check that the device is plugged in
2) check that the monitor is plugged in

If you still have no video, throw it out.

It's nice to see some old school, nuts and bolts, grown up engineering!

[DanielS]

As an owner of a ~30 years old amplifier (the first stereo my parents bought when I was a kid that my mother gave to my grandmother and then inherited back after my grandmother passed), I am quite well acquainted with the effects of crusty old switches since I periodically have to cycle though most of the amplifier's switches when it starts popping or dropping out. As soon as I noticed that values tended to "come good" after switching ranges when Dave started playing with it, I suspected the switches.

I really should remember to try contact cleaner next time I open it for a tune-up, that could spare me a few minutes of frustration every couple of days. Now that I have a scope, I really should poke around and see if I can fix the gain mismatch between channels too.

[EEVblog]

As soon as I noticed that values tended to "come good" after switching ranges when Dave started playing with it, I suspected the switches.

It just as often didn't come good or go bad when the switches were changed. And often it would just come good all on it's own without touching anything. It seemed possessed!

[vlad777]

You can spray all the switches from the back side (plunger is visible).
Is it just me, or the two switches in question have wires on them that also should be sprayed.

Edit:
That plug is not "funny". It is a European 220V plug.

[EEVblog]

That plug is not "funny". It is a European 220V plug.

Yeah, like I said, funny.

[vlad777]

What IS funny is that unit has 220V plug and set for 110V ?

[phamuc]

Hey Dave,

You may want to check to bottom side of the PCB.
The gang of switches may have hairline solder cracks around some of the pins. Especially after years of use, and two Australians banging on the thing 
Maybe reheat the connections to be safe.


 I also didn't see you check the bottom of the PCB for blown lands, the thing did get a 30KV pulse that blew a resistor.

-Paul

[AF6LJ]

As an owner of a ~30 years old amplifier (the first stereo my parents bought when I was a kid that my mother gave to my grandmother and then inherited back after my grandmother passed), I am quite well acquainted with the effects of crusty old switches since I periodically have to cycle though most of the amplifier's switches when it starts popping or dropping out. As soon as I noticed that values tended to "come good" after switching ranges when Dave started playing with it, I suspected the switches.

I really should remember to try contact cleaner next time I open it for a tune-up, that could spare me a few minutes of frustration every couple of days. Now that I have a scope, I really should poke around and see if I can fix the gain mismatch between channels too.

The old gear is fun, I have plenty of it myself.
Just a word of caution regarding contact cleaner and switches. (post are a different story)
The insulating material on some switches will soak up the contact cleaner, that material will swell up and cause the rivits holding the contacts to become loose. This will make for a bigger, and more persistent problem. As a general rule switch contacts are self cleaning, especially rotary switches. A little (very little) lube on the shaft and detent will go a long way to keep those switches in working order.

[AF6LJ]

What IS funny is that unit has 220V plug and set for 110V ?

I think that would qualify as Scary.....
not funny.

[eV1Te]

What IS funny is that unit has 220V plug and set for 110V ?

I didn't think of that during the video, but I can assure you that the DMM was set for 230 V before I sent it. It must have flipped during shipping or when unpacking... A reminder that one should always check such switches before plugging in a device 

[MBY]

Hey Dave,

You may want to check to bottom side of the PCB.
The gang of switches may have hairline solder cracks around some of the pins. Especially after years of use, and two Australians banging on the thing 
Maybe reheat the connections to be safe.


 I also didn't see you check the bottom of the PCB for blown lands, the thing did get a 30KV pulse that blew a resistor.

-Paul

Yes, it surprised me too that Dave didn't check the bottom side of the PCB. The PCB is two sided and some soldering joints on my unit around the switches indeed needed a reflow.

And I can attest that I did go through the same path as Dave despite of suspecting the switches in the first place. No amount of cleaning, rubbing and switching did help at first and the problem were popping in and out of existence even if I wasn't anywhere near the unit. So I checked the rails, reference, chopper, divider and pots until I completely disassembled the switches for thorough cleaning.

And the switches are still somewhat unreliable. The wipers are okay, I'm sure. The problem is the pins that defies visual inspection (but allows scrubbing, spraying and so on).

As I mentioned in my earlier post, my unit quite recently fell out of calibration after many years of reliable operation. This time I suspect high humidity and I will wait for cooler and dryer weather before I do a recal. It is currently not possible to calibrate as the trimmers doesn't allow the range. Worst case scenario is to replace with new pots. And yes, it's time for another round of contact spray I would think...

[EEVblog]

You may want to check to bottom side of the PCB.

I did that. Wasn't in the video.

[DanielS]

The old gear is fun, I have plenty of it myself.
Just a word of caution regarding contact cleaner and switches. (post are a different story)
The insulating material on some switches will soak up the contact cleaner, that material will swell up and cause the rivits holding the contacts to become loose.

Yeah, there is that possibility too, depending on what sort of material is used for the insulation and the specific nature of the contact cleaner, much like how spraying fine petroleum-based lubricants on rubber is usually a bad idea since the lubricant will often either dissolve the rubber or cause it to dry out.

One of the switches is a fancy four positions linear  selector driven by a flat open-frame mechanical transmission cable which connects to a rotary knob on the front. Instead of bringing weak input signals to the front, they bring the mechanical input all the way to the rear to keep the signal path as short as possible.

[Old Fart Analog Engineer]

Doctors don't do well diagnosing themselves, Lawyers don't do well representing themselves, Architects do poorly at construction and Engineers do not troubleshoot well.  I too, was talking to the screen while going down this bizarre expedition.  When we started measuring film capacitors rated >100 V I started smiling (a wry, condescending smile, I'm afraid). Toward the end of nearly an hour, I was expecting an analysis of the fiber "orientation" of the fiberglass PCB. 

If engineers actually spent several years fixing the the things they design, they might find the weak points affecting durability and long life.  It took over 40 years for people to figure out that if they did not use thick gold for low current switches & contacts, they must use sealed relays or electronic cross-points. 

The problems encountered with old test equipment (& other commercial / medical electronics): 1) CONTACTS (connectors, switches, relays, potentiometers)  2) SOLDER CONNECTIONS  3) LOOSE GROUNDS, i.e  "grounding screws" from terminals or PCB to ground   4) Electrolytic Capacitors: the higher the voltage and larger the capacitance, the more suspect.   This is based on 100s of thousands of repairs.  No troubleshooting manual or schematic is needed for checking these.  Years of experience is.  Common sense of what is the most likely cause also helps.

The switches and relays used in this type of equipment can be several different types, but all have some factors common.  Low current contacts (aka "signal" / "dry" / "logic level") have either brass, alloy, silver or gold plating on the contact points.  Most are silver.  If the visible parts of the contact are gray or black, the silver is oxidized. Silver Oxide is an insulator.  Certain contact cleaners will clean these well.  The "grease" you may find is not there primarily as a mechanical lubricant, but to keep oxygen away from the silver.  After cleaning with a spray liquid cleaner, the "grease" is washed away, ensuring future contact problems.  To make a lasting repair, use an electric grade of silicone grease on both the moving and stationary contacts.  If there are still issues, disassemble the contacts, clean them (IPA works) and look at them under well lit magnification.  If it has the silver or gold worn away by mechanical friction from use, your only option is to replace the switch or relay (or even pot).  This is usually impossible.  Another alternative is to add and wire-in relays or semiconductor switches. 

This is the major issue with used test equipment from eBay.  Even if it works flawlessly, or is "not right", "out of calibration", or the dreaded "intermittent", clean all contacts (switches, relays, connectors, pots) first! CAIG makes some great products for this http://store.caig.com/s.nl/sc.2/category.292/.f   http://store.caig.com/s.nl/sc.2/category.188/.f.  Don't use any cleaner which is flammable, because these may harm / swell the plastics holding the contacts.  When buying test equipment without being able to have it inspected by someone who knows what to look for prior to purchase, assume that it will need 'fixin' and that parts may not be available and that whatever you pay for the item and shipping will be lost.

OFAE

[AF6LJ]

The old gear is fun, I have plenty of it myself.
Just a word of caution regarding contact cleaner and switches. (post are a different story)
The insulating material on some switches will soak up the contact cleaner, that material will swell up and cause the rivits holding the contacts to become loose.

Yeah, there is that possibility too, depending on what sort of material is used for the insulation and the specific nature of the contact cleaner, much like how spraying fine petroleum-based lubricants on rubber is usually a bad idea since the lubricant will often either dissolve the rubber or cause it to dry out.

One of the switches is a fancy four positions linear  selector driven by a flat open-frame mechanical transmission cable which connects to a rotary knob on the front. Instead of bringing weak input signals to the front, they bring the mechanical input all the way to the rear to keep the signal path as short as possible.

That is some fine design, and you use to see that or variations of that used in quality stereo / Hifi equipment of a time gone by. Then there is this....
A marginal mechanical design in a piece of amateur radio gear built to a cost.

Underside view of the Galaxy V transceiver. Sorry for the crappy picture.
It is the only one I had and I couldn't find any of the underside on the net.

I might add this must have been made back in the days when Men were Men, because it takes about 5 foot pounds of to9rque to turn the band switch even it is clean, lubed and all the switches and couplings are aligned to limit binding. otherwise the radio is a real POS....

[Tothwolf]

How about fixing the Tektronix scopes that you got in a previous mailbag?

Those scopes are notoriously horrible to fix. Uses a ton of custom parts. My spidy sense tells me it would just end up like the DSA repair.

Nope, sorry Dave. Those Tektronix 24x5 scopes would be far easier to repair than the DSA. I'm active in the tekscopes community and I own one of these scopes. The two most common problem areas are the aluminum electrolytics in the PSU failing due to age and the SMD aluminum electrolytics on some of the later processor boards can leak and do minor board damage. Another somewhat common problem area are the versions of these scopes with use a Dallas battery backed memory module. When those fail, the scope looses its cal data. All of these issues can be overcome without that great of difficulty. While these scopes can have failed hybrid modules, those faults are not very common, and those parts can also be found with some asking around. Restoring one of these scopes is very labor intensive, but well worth it. The only real challenge I saw with your scopes is the missing covers, rear panels, and hardware. All of those parts are out there (I had to replace a number of broken/missing structural/mechanical parts on mine), but I suspect replacing all those missing parts on your scopes would cost somewhere in the neighborhood of $100 USD per scope.

As for the switches in the Keithley 177, those are a problem area in any gear of this vintage. 1980s Fluke and Simpson meters which use these type of push button (slide contact) switches tend to have the same sort of issues. Schadow (ITT) sold millions of these style of switches and they were used everywhere. Even my early 1980s vintage Tektronix 2213 scopes used a single push on/push off Schadow switch for the power button, which after 30+ years, tend to stick, and had to be serviced. What complicated matters with the 2213s, is they trimmed away the top connection pins on the switch and insulated it with a slide-on plastic cover. The cover cannot be removed to allow the switch to be cleaned without first desoldering it from the board, which involves a lot of disassembly.

[dr.diesel]

Speaking of, I need to renew my stock of contact cleaner/lube.  Is there a US version of that EML - Contact Cleaner Lubricant under a different name?  Doesn't appear to be readily available here in the US.

[EEVblog]

Restoring one of these scopes is very labor intensive, but well worth it.

That's why I'm not doing it, I don't have the time. Murphy will ensue that they will have all sorts of issues making any repair repair attempt even longer. It will turn into the DSA repair all over again, I know it.

The only real challenge I saw with your scopes is the missing covers, rear panels, and hardware. All of those parts are out there (I had to replace a number of broken/missing structural/mechanical parts on mine), but I suspect replacing all those missing parts on your scopes would cost somewhere in the neighborhood of $100 USD per scope.

You also have to try and get them in Oz.

[AF6LJ]

How about fixing the Tektronix scopes that you got in a previous mailbag?

Those scopes are notoriously horrible to fix. Uses a ton of custom parts. My spidy sense tells me it would just end up like the DSA repair.

Nope, sorry Dave. Those Tektronix 24x5 scopes would be far easier to repair than the DSA. I'm active in the tekscopes community and I own one of these scopes. The two most common problem areas are the aluminum electrolytics in the PSU failing due to age and the SMD aluminum electrolytics on some of the later processor boards can leak and do minor board damage. Another somewhat common problem area are the versions of these scopes with use a Dallas battery backed memory module. When those fail, the scope looses its cal data. All of these issues can be overcome without that great of difficulty. While these scopes can have failed hybrid modules, those faults are not very common, and those parts can also be found with some asking around. Restoring one of these scopes is very labor intensive, but well worth it. The only real challenge I saw with your scopes is the missing covers, rear panels, and hardware. All of those parts are out there (I had to replace a number of broken/missing structural/mechanical parts on mine), but I suspect replacing all those missing parts on your scopes would cost somewhere in the neighborhood of $100 USD per scope.

I would agree, we might have to talk about my two Tek 475s I plan on making one work... I know about the power supply issues they are legendary in Tek Scopes going way back.

As for the switches in the Keithley 177, those are a problem area in any gear of this vintage. 1980s Fluke and Simpson meters which use these type of push button (slide contact) switches tend to have the same sort of issues. Schadow (ITT) sold millions of these style of switches and they were used everywhere. Even my early 1980s vintage Tektronix 2213 scopes used a single push on/push off Schadow switch for the power button, which after 30+ years, tend to stick, and had to be serviced. What complicated matters with the 2213s, is they trimmed away the top connection pins on the switch and insulated it with a slide-on plastic cover. The cover cannot be removed to allow the switch to be cleaned without first desoldering it from the board, which involves a lot of disassembly.

That switch design is old as dirt. I have seem them in comm gear going back to the late 1960s.
They are reliable to a point, once the plating wears off they are Gonski...

[John Coloccia]

Speaking of, I need to renew my stock of contact cleaner/lube.  Is there a US version of that EML - Contact Cleaner Lubricant under a different name?  Doesn't appear to be readily available here in the US.

It's not the same stuff, but I've has excellent luck with Deoxit. I mainly use it because it doesn't generally damage guitar finishes, but it seems to work pretty well regardless.

[Tothwolf]

Restoring one of these scopes is very labor intensive, but well worth it.

That's why I'm not doing it, I don't have the time. Murphy will ensue that they will have all sorts of issues making any repair repair attempt even longer. It will turn into the DSA repair all over again, I know it.

The only real challenge I saw with your scopes is the missing covers, rear panels, and hardware. All of those parts are out there (I had to replace a number of broken/missing structural/mechanical parts on mine), but I suspect replacing all those missing parts on your scopes would cost somewhere in the neighborhood of $100 USD per scope.

You also have to try and get them in Oz.

I had to import the mechanical parts for my 24x5 from Greece and Israel... 

[Tothwolf]

The switches and relays used in this type of equipment can be several different types, but all have some factors common.  Low current contacts (aka "signal" / "dry" / "logic level") have either brass, alloy, silver or gold plating on the contact points.  Most are silver.  If the visible parts of the contact are gray or black, the silver is oxidized. Silver Oxide is an insulator.  Certain contact cleaners will clean these well.  The "grease" you may find is not there primarily as a mechanical lubricant, but to keep oxygen away from the silver.  After cleaning with a spray liquid cleaner, the "grease" is washed away, ensuring future contact problems.  To make a lasting repair, use an electric grade of silicone grease on both the moving and stationary contacts.  If there are still issues, disassemble the contacts, clean them (IPA works) and look at them under well lit magnification.  If it has the silver or gold worn away by mechanical friction from use, your only option is to replace the switch or relay (or even pot).  This is usually impossible.  Another alternative is to add and wire-in relays or semiconductor switches. 

Don't use any cleaner which is flammable, because these may harm / swell the plastics holding the contacts.

While I agree with a large portion of what you said, there are some things which need to be pointed out.

First, silver oxide is not an insulator. Silver oxide is very conductive and tarnished silver contacts by themselves are usually not a problem. If the tarnish is very "grainy" and the contacts are used in a low voltage circuit without debounce circuitry, usually only then is silver oxide a problem. The best way to clean these sort of contacts is a strip of thin cardboard or thick paper, pulled lightly through a closed set of contacts.

Second, silicone grease should never be used anywhere near electrical switch and relay contacts. Western Electric / Bell System and others all learned this the hard way back in the day with their electromechanical telephone exchange switches. When used in or near switch contacts which interrupt current, any arcing can cause the silicone to break down into silicon-carbide, which is not only electrically insulative, but also extremely hard and abrasive, which is very damaging to contacts. I personally use Sanchem "NO-OX-ID A-Special" grease when I need a grease for switch contacts. [Note however, this is NOT to be confused with "Ideal NOALOX" aluminum wire compound, which is a flammable abrasive paste which is supposed to penetrate aluminum oxide (it actually doesn't work very well since aluminum oxide is harder than the zinc particles that compound contains).]

As for flammable vs plastic-safe contact cleaners, many popular contact cleaners are flammable, and many non-flammable cleaners are not plastic safe.

Isopropyl Alcohol (isopropanol, IPA) for example, is extremely flammable, but plastic safe, and is one of the most common first-choice cleaners people use (not "rubbing alcohol", as this contains oils and a large percentage of water).

Cleaning solvents which should be avoided for electronics work contain such things as Acetone (Benzene), Trichloroethane, Tetrachloroethylene (Perchloroethylene), and similar. Acetone is flammable, Trichloroethane is low-flammability, and Tetrachloroethylene is non-flammable. All will readily dissolve plastics. Chlorinated cleaning solvents will also react with aluminum, and will also leach into and damage aluminum electrolytic capacitors (most capacitor manufacturers today explicitly warn about this in their datasheets).

I've been using CRC QD Contact Cleaner (02130) and CRC QD Electronic Cleaner (05103) to clean electrical contacts, potentiometers, and as a board wash for really stubborn stuff which doesn't completely clean off with Isopropyl Alcohol. Both are plastic-safe and contain Isohexane, Difluoroethane, and n-Hexane. The only difference I can tell between the current formulations of the two is QD Contact Cleaner contains Isopropyl Alcohol while QD Electronic Cleaner contains Ethanol. (Older formulations of QD Electronic Cleaner contained Naphtha and Methanol instead of the Ethanol.)

For switches and potentiometers, I usually follow these up with CRC 2-26 Plastic Safe Multi-Purpose Precision Lubricant (02004) which is basically Mineral Oil and some thickeners. This is pretty much the same stuff which used to be in TV tuner cleaners and contact cleaners which contained a lubricant. An added benefit to a separate lubricant is that it doesn't end up all over a pc board and everywhere else you really don't want it.

[Tothwolf]

Restoring one of these scopes is very labor intensive, but well worth it.

That's why I'm not doing it, I don't have the time. Murphy will ensue that they will have all sorts of issues making any repair repair attempt even longer. It will turn into the DSA repair all over again, I know it.

Well, Murphy or not, if you do eventually tackle these scopes PSU and processor boards, and you do some really careful, methodical step-by-step how-to videos covering the restoration process, I guarantee you that those repair videos will be some of your most popular. There are a ton of 24x5 scopes out there and people are constantly looking for repair information about the really common faults.

If you do decide to tackle them, there is one gotcha to be on the lookout for. The service manual has some component values wrong for some of the aluminum electrolytic capacitors on the PSU board. Many people discovered this the had way with their replacement electrolytics exploding. The Rifa brand class-X and class-Y safety capacitors (translucent yellow plastic) are also a must-replace item. All of this stuff has been well documented in the tekscopes email list archives though.

[marcan]

I think Dave's biggest troubleshooting fail here was not following a divide and conquer approach. If only he'd thrown a multimeter on the ADC input terminal, he could've ruled it out from the very beginning. After the obligatory troubleshooting steps (#0: look for blown stuff, #1: check the rails, #2: wiggle stuff), the right approach is to binary search the signal path until you narrow down the culprit, instead of guessing what the faulty component might be and dive right into troubleshooting it (making mistakes and creating red herrings for yourself along the way) before confirming that it really is borked to begin with.

Personally, though, as soon as I saw the creeping display and the correlation with changing ranges, I was thinking "faulty switches". But that's because I've seen a multimeter with a dodgy range switch do the same exact thing to me, so I'm lucky there (it was a cheapie chinese one, but the same principle applies). Experienced troubleshooters (especially of test equipment) will have the right hunches to save time, and you can't blame Dave for not having that. But knowing how to divide and conquer doesn't require experience, it's the basics, and I really feel that's where he messed up the procedure.

On an unrelated note, I don't think the blown resistor has anything to do with the fault, or anything to do with the blown diode bridge. Even if you short out the output, there's nothing driving it that could cause that resistor to blow. That resistor goes off to ground. We all know how you blow devices through the ground pin: by forgetting about earth reference. That thing was hooked up to a scope or something that is mains earth referenced, and someone tried to measure AC mains with phase on the negative input terminal (or something equally nasty), which is internal ground. Poof goes the resistor. Shouldn't have caused any other damage though, I think.

Then again, we all make dumb mistakes. Just last week I tried using a charger for a specific device. Wouldn't charge. Measured the output. 0V. It's rated for 230V range only but I'm on 100V - I could've sworn it worked on 100V even though it wasn't specced to (it's switchmode and not a chinese cheapie), but just in case I pulled out my step-up transformer. Nothing. Took it apart. Measured the output. 0VDC. Measured the input. 0VAC. Well, I had replaced the AC mains plug on this one, so take that apart, maybe one of the wires was loose. Nope, all solid. Then I turned on the power strip I was running this off of. Cue facepalm.

Well, at least I used this excuse to replace the mains plug again with a different style. Yes, I've lived in countries with 3 different kinds of AC mains plugs now. And sorry to break it to you, world, but European Schuko (the kind with a built-in safety shutter) and its unearthed Europlug variant is the best and most balanced home plug (and if you care about phase/neutral you're designing things wrong), American/Japanese 2-pin unearthed are the best travel plug (just because they're small and you can fit a bunch into a small power strip), the British design is an overengineered workaround for their shitty ring mains circuits and they need to stop pretending like the rest of the world hasn't figured out safety shutters yet, Ireland (where I have lived) is, as usual for Ireland, bass-ackwards in copying the British but then being more European (thus making the British design unnecessary with European wiring standards that don't require it), and I have no idea what you Aussies are up to but I haven't heard of any redeeming qualities for it. There, I said it