Datron 1281 DC Amplifier

[splin]

A couple of questions:

1) From the Service manual (my bold):

Basic Design
The required input characteristics are achieved by
using a differential FET input to give low input
current and high frequency response characteristics,
coupled with a synchronous chopping amplifier
to reduce offset and low-frequency noise. A
second amplifier stage provides most of the forward
gain, with the frequency/gian compensation
necessary for an effective amplifier bandwidth of
1 MHz.

Given that:

a) the DC Amplifier is not used in the AC measurement path
b) the maximum reading rate is 1000 samples/s (for the 1271, 150 samples/s for the 1281)
c) the DC Voltage input settling time (to 10ppm step size), with filter out, is specfied as < 500us (50ms for the 1281),

why would a 1MHz bandwidth 1MHz be deemed necessary for the DC amplifier?

FWIW I've attached the DC Amp schematics.

Also, does anyone know the chopping frequency?

[Kleinstein]

The DC amplifier looks like it might reach about an 1 MHz GBW. This might be a good idea to get reasonable fast settling if a DC gain of maybe 100 is needed. Settling to 10 ppm needs something like 10 time constants. For 500 µs settling this would be a 50 µs time constant and thus about 3.3 kHz required BW. This is not much from a 10 kHz BW derived from 1 MHz GBW and a gain of 100.

[splin]

Thanks Kleinstein. It seems quite obvious when you put it like that!

I was tending to think of the 1281 as a bench instrument which wouldn't be measuring rapidly changing voltages on the DC ranges, but of course the 1271 is designed for systems use where this is very likely to be the case.

I don't suppose anyone has any good references as to how the autocal works on the 1281 as there doesn't seem to be anything in the service manual? Is that big transformer for AC calibration only?

[Dr. Frank]

I was tending to think of the 1281 as a bench instrument which wouldn't be measuring rapidly changing voltages on the DC ranges, but of course the 1271 is designed for systems use where this is very likely to be the case.

I don't suppose anyone has any good references as to how the autocal works on the 1281 as there doesn't seem to be anything in the service manual? Is that big transformer for AC calibration only?

I suppose, that 1271 and 1281 are practically identical, apart from the double references.
There's a complete service manual available for the 1271, and the complete schematics also.

I did not find the description of this selfcal /autocal circuit, but if I remember correctly, T701 is used for precision DC ratios of 9.93: 7.23 : 1.83 : 1.00 : 0.19.
This is accomplished by a precision synchronous rectification scheme, delivering these precise DC ratios as listed.

transformer ratios can be made extremely precise and stable, so this will be the key transfer method inside the 1271/1281, but I can't estimate the precision of this transfer method.

Anyhow, this selfcal circuit is omitted in the FLUKE 8508A, so it was not that convincing, obviously.

Frank

PS: Just discovered the full service manual, on ko4bb.com, including schematics for the 1281.. this ratio selfcal module is similar/identical, but also not described in detail.
There's a brief hint, that this selfcal module is capable of resetting the calibration to the 90day uncertainty.
That's the 'extended specification', having half the 1 year uncertainty.

[guenthert]

[..]
Anyhow, this selfcal circuit is omitted in the FLUKE 8508A, so it was not that convincing, obviously.
[..]

Well, either that or Fluke figured that they're happy to make an additional buck or two calibrating the device for you periodically.  Don't want to come across as excessively paranoid, but I see there a conflict of interest.

[Kleinstein]

If the ADC is much better in the 8508A there might be no more need for the heavy and likely kind of expensive special transformer. The alternative transfer is using a short time stable divider and relying on the INL of the ADC. This way could still be kind of slow, but starting up the transformer and filter circuit would also take quite some time. It looks like the transformer is not running all the time, as there is a control signal with cal name. The dual  reference (and AFAIK even an extra LM399) in the Datron 1281 is also kind of having a backup to get a really in depth self test capability.  The long time stable transformer would also alow for a good linearity test.

[e61_phil]

I've never heard of an approach using a transformer for stable and precise DC ratios. Is there any kind of document to start reading something about this technique?

And does anyone know if there are still two references in the Fluke 8508A?

[mycroft]

Take a look at https://www.eevblog.com/forum/metrology/datron-calibrator-7v-to-10v-conversion-with-a-transformer/?all (Datron Calibrator 7v to 10v conversion with a transformer).