ADCMT 7480T 8.5 digit multimeter

[TheSteve]

I also wonder what this magic board with SOT23-3 ic is for.

-branadic-

I assume it is the multislope converter.

[TiN]

Similar to Keithey 2001/2002 ADC, that module with epoxied discrete SOT23's JFETs likely to be part of ADC charge switches. Epoxy and separate substrate to help with isothermal construction and stable leakage conditions. You can also sorta guess presence of 0402 or 0603 SMD resistors in between the active parts, perhaps setting ratio currents for multislope ramp operation. Resistors near the module look like normal 5% 0603's, so those likely just digital control lines pullups/series.

[TheSteve]

A better closeup of it.

[Kleinstein]

I don't think the extra substrate is for the ADC - my guess is more like the main multiplexer. The part on the right of the ref. board looks more like the input stages. Just on top of the od board there seem to be 4 LM399 - like used to control a bunch of JFETs. This would be too slow if used for switching the ADC.

The ADC is more like on the other side of the reference board. There are silk screen prints with slope an integ and comp.

[Mickle T.]

This is a 12:1 (or so) guarded input multiplexer.

ADC is here:

[TheSteve]

Thanks for all of those pics!
Only one question: what is a lower limit of the integration time in this meter (how fast is it)?

Fastest is 1 millisecond, slowest is 100 NPLC.

[Mickle T.]

Fastest is 1 millisecond, slowest is 100 NPLC.

Thank you.

As far as I can see, 7480T Input amplifier and ADC are simpler than 6581(D,T) due to lack of high speed modes. 6581 have a second, wideband Inp. Amp. channel, second pair on Cint/Caz capacitors with selector, HS x1 mode of the slope amplifier e.t.c. Both 6581 and 7840T have a lack of input precharge compensation, have the same ADC and potentially the same INL and TC problems. 7480T don't have a DCI measurement mode, all of the DCI shunts are used as artifacts only for Ohms calibration.

[martinr33]

Advantest units have always had funky resistor arrays. Here's one from an old 6.5 digit unit.

It is in the reference amplifier section.

[zhtoor]

Advantest units have always had funky resistor arrays. Here's one from an old 6.5 digit unit.

It is in the reference amplifier section.

what are those coiled copper (?) wires, temperature compensators?

regards.

[HalFET]

Advantest units have always had funky resistor arrays. Here's one from an old 6.5 digit unit.

It is in the reference amplifier section.

True, but their RF gear its build quality is pretty ludicrous. I have really old units from them that internally look as if they were designed yesterday with modern CAD software.

[chickenHeadKnob]

what are those coiled copper (?) wires, temperature compensators?
regards.

My guess, and it is only a guess, is that it is to  mechanically divert  stress from expansion or contraction of the leads. So distantly a type of temperature comp. Zlymex worked out a system to provide temperature compensation to wire wound resisters which utilized trim resisters and short sections of copper wire for final trim. You will have to search his posts to find it.  I think if they wanted to do that type of T.C. they would have included those trim components on the hybrid itself and trim it while it is still separate from the main board for ease of manufacture.

[Mickle T.]

Good news: 7480T hybrid MUX is pin-to-pin compatible with the R6581 one 
Bad news: it does not have one FET switch