Scoping sample & hold voltage

Use a 10x probe and a 1 nF capacitor in series.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC / Hobbs ElectroOptics 
Optics, Electro-optics, Photonics, Analog Electronics 
Briarcliff Manor NY 10510 

http://electrooptical.net 
http://hobbs-eo.com

Thanks, should I hook that up at the 4051's output or the top of the 6800p?

I've replaced the 4051 with a known-good unit so I'm fairly sure the problem I'm having is either a bad joint somewhere or bad S&H cap in one of the networks.

If the voltages to be measured aren't in the millivolt range and you need DC accuracy you could use a 10M or higher series resistor, then crank up the scope channel's gain.

Or ... plain old dirt and grime on the board. Sometimes a good circuit board cleaing can fix S&H issues.

--
Regards, Joerg 

http://www.analogconsultants.com/

I'll try that, too. I forgot to mention that the original ~30 year old

4051 had some bad channels and needed to be swapped also; the S&H outputs are feeding some high-freq VCOs on a custom chip. It's working better now but some thing's still not quite right one of the VCOs is "sweeping" down in frequency so one of the S&H isn't holding. I don't have a pinout for the custom chip so I can't say immediately which channel of the multi is feeding which VCO, so I have to test them all. ARGH.

This board is full of that 4051 + S&H arrangement because DACs probably cost ten bucks at the time but it seems prone to reliability problems. I notice 30 y/o 4000 series logic seems to fail fairly often, too. A lot more than 30 y/o LS series for sure.

Also, stand the board on a corner and squish flux cleaner under the

4051s until it comes out the other side, then clean that off.

If this tends to happen after some years or decades of use that is a concern. It could point to overvoltage or ESD issues with this design.

If you have a lot of them just whip up a source follower with some FET you have. Measure its Vgs threshold when operating and add that to your measured values. To determine sag you wouldn't even need DC accuracy. You could also use a CMOS opamp as a follower which would be just one single part. Ok, plus a bypass cap.

If it's a one-off job you might also be able to tap all the VCOs to ground via a resistor. That makes the frequency jump and then you'd see when you touched the sagging one, gives you the pin-out via the Inspecteur Clouseau method.

The DAC08 series was cheap back then. The reason for semi-discrete solutions like yours was often that the granularity of cheaper DACs and ADCs wasn't sufficient, not enough bits.

That isn't my experience. However, 30+ years ago some 4000 logic devices had weak or almost no ESD protection so they can become damaged easily. Muxes can become partly damaged where they still kind of work but not very well.

--
Regards, Joerg 

http://www.analogconsultants.com/

Yup, you're right, the DAC on this board is in fact an DAC0812 12-bitter.