Solartron 7081 noise
Hi everyone,
I recently purchased a Solartron 7081 DMM. It▓s one of the first model, dated Apr. 1984.
I suspect that it is not as close to accurate as it could be due to aging and
storage in a humid environment. My test showed good accuracy, but the huge input
bias current and suspiciously large noise. After 8 hours of heating at 25
degrees Celsius input current was 330 pA causing a zero shift in the
short-circuited input √ 12 uV (internal resistance at the Hi input is 36000
Ohms). Standard warm-up procedure after prolonged low-temperature storage did
not help. One of the problem was found in the input amplifier and DC selector:
surface and volumetric leakage in the PCB from the relays power supply channel
(- 10 V), especially in places that are not protected by guard traces. After
checking each critical item in the input circuits (attenuator, DC ranging, input
amplifier with MDM channel) and installing the PTFE bushings and insulators
(http://img14.imageshack.us/img14/3432/pcb6top.jpg ,http://img825.imageshack.us/img825/3896/pcb6bottom.jpg )
input bias has been reduced down to 1-2 pA!
After all, S7081 was warmed up again for more than 12 hours and has been tested
by drifts through the RS232 interface and a terminal on a PC:
- Short inputs voltage drift while 6 hours preheating (http://img109.imageshack.us/img109/2384/7081zeropreheating.jpg);
- Short inputs voltage drift in DC 10V range after preheating (http://img198.imageshack.us/img198/4846/708110vzerofront.jpg);
- ADC zero test result, used by S7081 in the selftest procedure, periodical
(every 15 minutes) drift correction and while 6..8-1/2 resolution selection
(http://img155.imageshack.us/img155/2254/7081test0.jpg); - Internal reference test result, used by S7081 in the selftest procedure (http://img694.imageshack.us/img694/3379/7081test10.jpg).
After analyzing of the results it seems that the next big problem is the ADC
noise, since its amplitude does not decrease even with a short ADC inputs during
the test of zero (TEST0). Maybe I should change the integrators op-amp ICL7650S.
Mickle Timofeyev (Russia)
Mickle,
I am admiring your plots of data about your meter. How did you do that?
I recently purchased a 7081 (SN 001523) and would like to do a similar assessment of my meter. I have been able to use HyperTerminal to download the CAL constants and I just turned on the 'Output,RS232,ON' command and it is happily moving the readings into HyperTerminal. Now, it would appear, I need to find a way to move the data into Excel then find a way to 'plot' it. Perhaps it would be easier to do this via the GPIB?
Any thoughts from anyone?
Joe
-----Original Message-----
From: volt-nuts-bounces@febo.com [mailto:volt-nuts-bounces@febo.com] On Behalf Of Ìèõàèë
Sent: Sunday, September 11, 2011 7:48 AM
To: volt-nuts@febo.com
Subject: [volt-nuts] Solartron 7081 noise
Hi everyone,
I recently purchased a Solartron 7081 DMM. It▓s one of the first model, dated Apr. 1984. I suspect that it is not as close to accurate as it could be due to aging and storage in a humid environment. My test showed good accuracy, but the huge input bias current and suspiciously large noise. After 8 hours of heating at 25 degrees Celsius input current was 330 pA causing a zero shift in the short-circuited input √ 12 uV (internal resistance at the Hi input is 36000 Ohms). Standard warm-up procedure after prolonged low-temperature storage did not help. One of the problem was found in the input amplifier and DC selector: surface and volumetric leakage in the PCB from the relays power supply channel
(- 10 V), especially in places that are not protected by guard traces. After checking each critical item in the input circuits (attenuator, DC ranging, input amplifier with MDM channel) and installing the PTFE bushings and insulators (http://img14.imageshack.us/img14/3432/pcb6top.jpg ,http://img825.imageshack.us/img825/3896/pcb6bottom.jpg ) input bias has been reduced down to 1-2 pA! After all, S7081 was warmed up again for more than 12 hours and has been tested by drifts through the RS232 interface and a terminal on a PC:
- Short inputs voltage drift while 6 hours preheating (http://img109.imageshack.us/img109/2384/7081zeropreheating.jpg);
- Short inputs voltage drift in DC 10V range after preheating (http://img198.imageshack.us/img198/4846/708110vzerofront.jpg);
- ADC zero test result, used by S7081 in the selftest procedure, periodical (every 15 minutes) drift correction and while 6..8-1/2 resolution selection (http://img155.imageshack.us/img155/2254/7081test0.jpg);
- Internal reference test result, used by S7081 in the selftest procedure (http://img694.imageshack.us/img694/3379/7081test10.jpg).
After analyzing of the results it seems that the next big problem is the ADC noise, since its amplitude does not decrease even with a short ADC inputs during the test of zero (TEST0). Maybe I should change the integrators op-amp ICL7650S.
Mickle Timofeyev (Russia)
volt-nuts mailing list -- volt-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
and follow the instructions there.
Excellent work, Mickle!
Nice touch adding the PTFE standoff insulators. How did you make connections
on the trace-side to the PTFE stand-offs? (To reduce leakage and dielectric
absorption, I assume you ran hard-wires instead of using existing PCB
traces?)
Your PCB leakage problem reminds me of a similar problem Sony had with a
high-impedance circuit back in the seventies. The problem began showing-up
in older units of that model after years of use. Technicians were able to
temporarily eliminate the electrical leakage currents by spraying the area
with PCB cleaner solutions. But the problem would return in a few months.
Finally Sony determined the problem was caused by bacterial growth in higher
humidity environments due to a contamination problem at the factory when the
products were made. Interestingly, Sony then released a Service Note,
recommending that you spray the area with super-cold Freon to kill the
bacteria by freezing it. ...I tried that approach and it worked. I've never
forgotten that situation, because it was so counterintuitive and strange!
Apparently according to the 7081 manual, its integration time is determined
by the Scale Length you select, hence number of digits displayed? So
apparently your 7 1/2 digits resolution corresponds to 3.2 seconds
integration time. Do I have that right?
It might be interesting if some other 7081 users will duplicate your tests
for comparison.
Cheers,
Greg
----- Original Message -----
From: "Михаил" timka2k@yandex.ru
To: volt-nuts@febo.com
Sent: Sunday, September 11, 2011 6:47 AM
Subject: [volt-nuts] Solartron 7081 noise
Hi everyone,
I recently purchased a Solartron 7081 DMM. It▓s one of the first model,
dated Apr. 1984.
I suspect that it is not as close to accurate as it could be due to aging
and
storage in a humid environment. My test showed good accuracy, but the huge
input
bias current and suspiciously large noise. After 8 hours of heating at 25
degrees Celsius input current was 330 pA causing a zero shift in the
short-circuited input √ 12 uV (internal resistance at the Hi input is 36000
Ohms). Standard warm-up procedure after prolonged low-temperature storage
did
not help. One of the problem was found in the input amplifier and DC
selector:
surface and volumetric leakage in the PCB from the relays power supply
channel
(- 10 V), especially in places that are not protected by guard traces. After
checking each critical item in the input circuits (attenuator, DC ranging,
input
amplifier with MDM channel) and installing the PTFE bushings and insulators
(http://img14.imageshack.us/img14/3432/pcb6top.jpg
,http://img825.imageshack.us/img825/3896/pcb6bottom.jpg )
input bias has been reduced down to 1-2 pA!
After all, S7081 was warmed up again for more than 12 hours and has been
tested
by drifts through the RS232 interface and a terminal on a PC:
- Short inputs voltage drift while 6 hours preheating
(http://img109.imageshack.us/img109/2384/7081zeropreheating.jpg); - Short inputs voltage drift in DC 10V range after preheating
(http://img198.imageshack.us/img198/4846/708110vzerofront.jpg); - ADC zero test result, used by S7081 in the selftest procedure, periodical
(every 15 minutes) drift correction and while 6..8-1/2 resolution selection
(http://img155.imageshack.us/img155/2254/7081test0.jpg); - Internal reference test result, used by S7081 in the selftest procedure
(http://img694.imageshack.us/img694/3379/7081test10.jpg).
After analyzing of the results it seems that the next big problem is the ADC
noise, since its amplitude does not decrease even with a short ADC inputs
during
the test of zero (TEST0). Maybe I should change the integrators op-amp
ICL7650S.
Mickle Timofeyev (Russia)
volt-nuts mailing list -- volt-nuts@febo.com
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
and follow the instructions there.
Hi Mickle,
As it seems it was stored in poor conditions it would need to have all the PCB's and wires washed carefully first.Try and get hold of some pure isopropanol, be careful as many of the easily obtained samples of this often contain some methanol, which WILL craze any polystyrene capacitors. The best is HPLC grade Isopropanol, but usually much too pricey. So first test your IPA for compatibility with some spare polystyrene capacitors.Then wash all PCB's carefully using a stiff brush, then use a hairdrier to dry out the boards, leave at least overnight before powering up again, Also check for IPA getting into the connectors, use an air line if possible to get them dry. I used to use that technique to clean PCB's that had a 30-100G ohm feedback resistor around an opamp.Then once you have it safely powered up again it will possibly need a month to stabilise again?
Good luck, and if any of my instructions are unclear please ask for clarification, as it is late and I am in a hurry for bed.
Mike
From: gbusg@comcast.net
To: volt-nuts@febo.com
Date: Sun, 11 Sep 2011 12:40:31 -0600
Subject: Re: [volt-nuts] Solartron 7081 noise
Excellent work, Mickle!
Nice touch adding the PTFE standoff insulators. How did you make connections
on the trace-side to the PTFE stand-offs? (To reduce leakage and dielectric
absorption, I assume you ran hard-wires instead of using existing PCB
traces?)
Your PCB leakage problem reminds me of a similar problem Sony had with a
high-impedance circuit back in the seventies. The problem began showing-up
in older units of that model after years of use. Technicians were able to
temporarily eliminate the electrical leakage currents by spraying the area
with PCB cleaner solutions. But the problem would return in a few months.
Finally Sony determined the problem was caused by bacterial growth in higher
humidity environments due to a contamination problem at the factory when the
products were made. Interestingly, Sony then released a Service Note,
recommending that you spray the area with super-cold Freon to kill the
bacteria by freezing it. ...I tried that approach and it worked. I've never
forgotten that situation, because it was so counterintuitive and strange!
Apparently according to the 7081 manual, its integration time is determined
by the Scale Length you select, hence number of digits displayed? So
apparently your 7 1/2 digits resolution corresponds to 3.2 seconds
integration time. Do I have that right?
It might be interesting if some other 7081 users will duplicate your tests
for comparison.
Cheers,
Greg
----- Original Message -----
From: "Михаил" timka2k@yandex.ru
To: volt-nuts@febo.com
Sent: Sunday, September 11, 2011 6:47 AM
Subject: [volt-nuts] Solartron 7081 noise
Hi everyone,
I recently purchased a Solartron 7081 DMM. It▓s one of the first model,
dated Apr. 1984.
I suspect that it is not as close to accurate as it could be due to aging
and
storage in a humid environment. My test showed good accuracy, but the huge
input
bias current and suspiciously large noise. After 8 hours of heating at 25
degrees Celsius input current was 330 pA causing a zero shift in the
short-circuited input √ 12 uV (internal resistance at the Hi input is 36000
Ohms). Standard warm-up procedure after prolonged low-temperature storage
did
not help. One of the problem was found in the input amplifier and DC
selector:
surface and volumetric leakage in the PCB from the relays power supply
channel
(- 10 V), especially in places that are not protected by guard traces. After
checking each critical item in the input circuits (attenuator, DC ranging,
input
amplifier with MDM channel) and installing the PTFE bushings and insulators
(http://img14.imageshack.us/img14/3432/pcb6top.jpg
,http://img825.imageshack.us/img825/3896/pcb6bottom.jpg )
input bias has been reduced down to 1-2 pA!
After all, S7081 was warmed up again for more than 12 hours and has been
tested
by drifts through the RS232 interface and a terminal on a PC:
- Short inputs voltage drift while 6 hours preheating
(http://img109.imageshack.us/img109/2384/7081zeropreheating.jpg); - Short inputs voltage drift in DC 10V range after preheating
(http://img198.imageshack.us/img198/4846/708110vzerofront.jpg); - ADC zero test result, used by S7081 in the selftest procedure, periodical
(every 15 minutes) drift correction and while 6..8-1/2 resolution selection
(http://img155.imageshack.us/img155/2254/7081test0.jpg); - Internal reference test result, used by S7081 in the selftest procedure
(http://img694.imageshack.us/img694/3379/7081test10.jpg).
After analyzing of the results it seems that the next big problem is the ADC
noise, since its amplitude does not decrease even with a short ADC inputs
during
the test of zero (TEST0). Maybe I should change the integrators op-amp
ICL7650S.
Mickle Timofeyev (Russia)
volt-nuts mailing list -- volt-nuts@febo.com
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
and follow the instructions there.
volt-nuts mailing list -- volt-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
and follow the instructions there.