Keithley 199 System DMM/Scanner

[dml]

This week's test gear repair short story:

Scraped up a Keithley 199 System DMM (without the scanner board) for $50 plus shipping. Not in great shape, but still...




"does not power on" ...spin the wheel, roll the dice!




Looks nice and clean inside. No obvious destruction.




But turning it over....



Looks promising! In a bad, yet good way.


Fix fix...




And a bit of recapping here and there. All of the small electrolytics were reading 5-10 Ohms at a range of frequencies. Dry. The big ones were all ok.


Quick check that things are going to plan...




A good clean and reassembly next. Some basic tests for ohms, volts.




Overall, worth the small amount of cost and trouble. And more fun than the other things I had to do today.


A few notes:

The board and pads were damaged by overheating pins, caused by dry solder joints. These pins were badly oxidized and would not take solder. Using flux wasn't enough - I had to grind the sides of the exposed pin. The board in this unit is also very easily damaged by heat and I found acetone quickly strips the green coating from the board so only ISO is safely usable. This all took longer than it really should have and wasn't as tidy as I would have liked. I'll need to look into better ways of recovering oxidized/corroded/burned pins as it has become a bit of a theme in recent repairs. I have been dealing with a Tek 7613 PSU with similar problems on the solder joints of some header pins.

[macboy]

The 199 is a nice unit, I have two of them. Both of mine have the fairly common scanner card, as well as the rare connector paddle boards for the scanners. This was an instrument worthy of repair, good work.

It doesn't use a heated reference so it is fairly accurate right away but it needs a long warm up for minimum drift. Even then, you really need to enable the MUX (a.k.a. auto-zero) or it drifts quite a lot quite quickly. With the MUX off, the display updates too quickly to see (>40 readings/second). With the MUX on, the meter will read the zero reference, the voltage reference, and the input in order to calculate every reading. It has no advanced math but the dB feature can be very useful. GPIB provides 6.5 digits (3 million counts) and even the last digit is noise free with the filter enabled, but the drift prevents true 6.5-digit-class use.

If you have another trustworthy instrument or references of the same or better accuracy then you can very easily "calibrate" (actually, just adjust) individual ranges as desired. I've cal'd mine to match my K 2001 meters.

[dml]

The 199 is a nice unit, I have two of them. Both of mine have the fairly common scanner card, as well as the rare connector paddle boards for the scanners. This was an instrument worthy of repair, good work.

Thanks! I have seen the scanner cards but not the paddle boards. I probably won't need the scanner any time soon but if I do pick up one of the boards it sounds like I'd probably have to construct an interface for the inputs to make it useful.

It doesn't use a heated reference so it is fairly accurate right away but it needs a long warm up for minimum drift. Even then, you really need to enable the MUX (a.k.a. auto-zero) or it drifts quite a lot quite quickly.

I had a play with the setup menu and found the MUX was configured on, so it's been left that way.

I haven't tested it properly after warmup but at least within the first 20 minutes it disagrees with my two other meters in the ~20k resistance range, beyond 2 decimals. I'll need to do more thorough tests and likely make some adjustments.

With the MUX off, the display updates too quickly to see (>40 readings/second). With the MUX on, the meter will read the zero reference, the voltage reference, and the input in order to calculate every reading. It has no advanced math but the dB feature can be very useful. GPIB provides 6.5 digits (3 million counts) and even the last digit is noise free with the filter enabled, but the drift prevents true 6.5-digit-class use.

If you have another trustworthy instrument or references of the same or better accuracy then you can very easily "calibrate" (actually, just adjust) individual ranges as desired. I've cal'd mine to match my K 2001 meters.

Thanks for all the info - will prove quite useful! I have the service/user manual and it's going to take some time to get through it.