Video Teardown, Repair and Cal. of a Keithley 220 Programmable Current Source

[Hugoneus]

In this episode Shahriar investigates the cause of failure of a Keithley 220 Programmable Current Source. Despite being manufactured in the mid-1980's, the Keithley 220 is still a very popular programmable current source. Alongside the repair efforts, the schematic of various analog, digital and power supply boards are also presented. By interpreting the failure mechanism, the fault is traced and corrected. The Keithley 220 is then calibrated and the calibration procedure of the sub-100nA range is explained. Finally, the performance of the calibrated unit is verified using a Rigol DM3068 Multimeter.

You can see the video here: [56 Minutes]
http://thesignalpath.com/blogs/2012/11/26/teardown-repair-and-calibration-of-a-keithley-220-programmable-current-source/

More videos at The Signal Path:
http://www.TheSignalPath.com

[peter.mitchell]

Great video!
About the power switch; it can't be helped, no matter how careful you are, something like that always happens.
No harm in replacing caps that are 20 years old either, that's just one less thing you have to worry about if it plays up again

[free_electron]

a pointer on those 100 M resistors : never -ever- touch those with your bare hands ! wear gloves.
The same goes for the circuitboard around the entire sesing system. the grease that occurs naturally on your skin is enough to cause massive leakage across the body of those resistors throwing the machine way off course...
The same goes for those teflon turrets. After any kind of work has been done on those they need through cleaning using flux-off and a second dip in alcohol (IPA) to make sure everything is gone. And don't use your mouth to blow 'air' on the alcohol to evaporate it faster. the humid air you exhale will condensated due to the cold produced by the evaporating alcohol... as our salive contains acidic products ... you basically blow another layer of conductive crap on the just cleaned turret ...

The 220 is a very sensitive instrument only surpassed by their electrometer. That thing is a horror to fix... after soldering on the most sensitive part ( the charge baancer ) you need to let the board lay down for a few hours .... before attempting calibration...

Otherwise excellent video ( apart from discharge capacitors that should be discharge capacitors and transformers that make 30 ohms instead of volts and 150 ohms instead of volts.. you may want to dub that ... )

[sorin]

thanks for your "WORKS"

[robrenz]

Great video. The step by step troubleshooting analysis is good for wannabee repairmen like myself.

[Hugoneus]

a pointer on those 100 M resistors : never -ever- touch those with your bare hands ! wear gloves.
. . .
Otherwise excellent video ( apart from discharge capacitors that should be discharge capacitors and transformers that make 30 ohms instead of volts and 150 ohms instead of volts.. you may want to dub that ... )

Thank you for your contribution. I mentioned the crux of your comments during the video. I also did a full IPA bath and bake which I did not show in the video (it was already almost one hour long). I own the Keithley Electrometer also by the way.

I don't fully understand your last sentence however...

[quarks]

 I like your reviews very much and this one was especially interesting, because you showed some good details about things to watch out with high ohm resistances.   

Thanks a lot.

[free_electron]

a pointer on those 100 M resistors : never -ever- touch those with your bare hands ! wear gloves.
. . .
Otherwise excellent video ( apart from discharge capacitors that should be discharge capacitors and transformers that make 30 ohms instead of volts and 150 ohms instead of volts.. you may want to dub that ... )

Thank you for your contribution. I mentioned the crux of your comments during the video. I also did a full IPA bath and bake which I did not show in the video (it was already almost one hour long). I own the Keithley Electrometer also by the way.

I don't fully understand your last sentence however...

when talking about the bleed resistors across the caps int he powersupply you say 'capacitor' instead of resistor.
same with the transformer. instead of saying 150 volts and 30 volts you say 150 ohms .. and 30 ohms.

[Hugoneus]

a pointer on those 100 M resistors : never -ever- touch those with your bare hands ! wear gloves.
. . .
Otherwise excellent video ( apart from discharge capacitors that should be discharge capacitors and transformers that make 30 ohms instead of volts and 150 ohms instead of volts.. you may want to dub that ... )

Thank you for your contribution. I mentioned the crux of your comments during the video. I also did a full IPA bath and bake which I did not show in the video (it was already almost one hour long). I own the Keithley Electrometer also by the way.

I don't fully understand your last sentence however...

when talking about the bleed resistors across the caps int he powersupply you say 'capacitor' instead of resistor.
same with the transformer. instead of saying 150 volts and 30 volts you say 150 ohms .. and 30 ohms.

Oh yes, you are right. Thanks for pointing it out! 

[IanB]

You mentioned in a YouTube comment about measuring the voltage drop across a 1 Gohm resistor using a 10 Gohm impedance voltmeter, but I didn't understand your answer. Normally we would expect the voltmeter to bypass some current around the resistor and affect the voltage reading (by about 10% or so?), but you said this would not be a problem. I could only understand this not being a problem if the voltage was buffered in some way using an even higher impedance sensor, but then why would you need to remark about the voltmeter being >10 Gohm impedance? Please would you explain this in more detail?

[Hugoneus]

You mentioned in a YouTube comment about measuring the voltage drop across a 1 Gohm resistor using a 10 Gohm impedance voltmeter, but I didn't understand your answer. Normally we would expect the voltmeter to bypass some current around the resistor and affect the voltage reading (by about 10% or so?), but you said this would not be a problem. I could only understand this not being a problem if the voltage was buffered in some way using an even higher impedance sensor, but then why would you need to remark about the voltmeter being >10 Gohm impedance? Please would you explain this in more detail?

Sorry for the confusion.

When using resistors smaller than 1M, the output voltage is directly monitored across the resistors using the multimeter. I set the multimeter to >10GOhm to ensure its internal resistance does not influence the measurements.

When using resistors larger than 1M, the entire setup is placed in the dual-cage and the 'guard' output port is used to monitor the voltage (i.e. like a buffered output as you mentioned). In this case the resistance of the multimeter is not in parallel with the load. However, I set it to >10GOhm anyway, but technically I didn't need to.

[EEVblog]

when talking about the bleed resistors across the caps int he powersupply you say 'capacitor' instead of resistor.
same with the transformer. instead of saying 150 volts and 30 volts you say 150 ohms .. and 30 ohms.

Happens to me almost every video 
The joys of having to think ahead, talk, and present info at the same time.

Dave.

[robrenz]

I think almost every video I have posted has at least one text box correction of a verbal blunder.

[Hugoneus]

I think almost every video I have posted has at least one text box correction of a verbal blunder.

Yes... There is a lot of multitasking involved. The videos are also not scripted, so naturally mistakes happen!

[free_electron]

If you ever need the special dual jfet for the electrometer input stage let me know .i have some. probably the last that were ever made ..

[Hugoneus]

If you ever need the special dual jfet for the electrometer input stage let me know .i have some. probably the last that were ever made ..

That is interesting. My unit is a Keithly 6517. It is a more modern unit. I should do a video about it... It can count electrons!

[muvideo]

That is interesting. My unit is a Keithly 6517. It is a more modern unit. I should do a video about it... It can count electrons!

That would be very interesting, I'd like to see some real world example of use for a modern electrometer, just to
see what can be done with it.
And thanks for the video, I think it's one of the few long videos on youtube I've watched entirely 

Fabio.

P.S. Wonderful equipment set you have access to!

[poodyp]

I also did a full IPA bath and bake which I did not show in the video (it was already almost one hour long).

Do you mean you literally dumped the board in a bucket of IPA? Or you poured it over? I've always wondered about the proper way to clean electronics. I've tried with cotton balls and cotton swabs, but they always get caught on everything, fall apart, and leave cotton strands everywhere.

[T4P]

A full IPA bath is soaking the whole thing in IPA  easy to do if you have a huge vertical container lying around or an ultrasonic cleaner with 50 IPA/50 DI Water

[Hugoneus]

I also did a full IPA bath and bake which I did not show in the video (it was already almost one hour long).

Do you mean you literally dumped the board in a bucket of IPA? Or you poured it over? I've always wondered about the proper way to clean electronics. I've tried with cotton balls and cotton swabs, but they always get caught on everything, fall apart, and leave cotton strands everywhere.

I have access to an ultrasound bath. I used that for this purpose. Although I don't own that unit, plus with so much surface area, you don't want something like that full of IPA at your house!

[robrenz]

Probably everyone knows this but just in case.  As soon as you put a board in a bath of solvent (ultrasonic or not) the bath is contaminated to some level by the materials that disolve or are dislodged by solvent. A thorough final rinse with virgin solvent is necessary to remove the contaminated solvent from the board.  Vapor phase cleaning gets around this solvent contamination problem.

[BravoV]

As soon as you put a board in a bath of solvent (ultrasonic or not) the bath is contaminated to some level by the materials that disolve or are dislodged by solvent.

...and its like applying a perfect coating with the dissolved unwanted material/contaminant "evenly" through out the board. 

[Hugoneus]

As soon as you put a board in a bath of solvent (ultrasonic or not) the bath is contaminated to some level by the materials that disolve or are dislodged by solvent.

...and its like applying a perfect coating with the dissolved unwanted material/contaminant "evenly" through out the board. 

The never ending cycle!

[SeanB]

A faulty 741...... I have changed hundreds of them for parametric shift, it is quite common. Still have a few dozen TO100 can versions around.

[Hugoneus]

A faulty 741...... I have changed hundreds of them for parametric shift, it is quite common. Still have a few dozen TO100 can versions around.

My main concern was 'how' it died, hopefully it just aged...