Resistive divider + calibration request

[GregDunn]

A couple of my DMMs require a 1VDC reference for calibration, and I'm sure I am not the only one who has a 10V ref but not a 1V ref.  I see lots of posts about arbitrary divider networks, but surely there's a simple 'fixed' 10:1 divider which will minimize the load on a reference, not degrade the accuracy, and not require re-calibration?  The best resistors I have in my lab are PRC "standards" (the MC-7) with 0.01% rated accuracy, so I recognize I may need to buy specially graded resistors for the build.

On a related note, is there anyone in the central Indiana area who might be able to measure my MC-7 and give me 5-digit or better values for the elements?  I'd like to be able to at least compare my DMM ohms measurements with a little more accuracy than the marked values, and some testing I've done already hints to me that a couple of them are borderline.

[The Soulman]

A Hamon divider is what you are looking for:
http://conradhoffman.com/HamonResistor.html

What output resistance do your 10V reference's have?

I never heart of PRC, funky little product you have.
https://www.precisionresistor.com/MC-7-Calibrator.html

If it was me I'd put those resistors in a solid housing with nice binding posts before shipping them anywhere. 

[GregDunn]

The only references I'm sure about are branadic's LTZ1000 (which uses a LTC1052 as an output buffer) and RoadRunner's LM399 with a LT1001 buffer.  The others are AD584 refs from the Bay and I haven't taken them apart to find out what or if there's a buffer at all.  Those might be problematic.  In any event, I'd want to optimize the divider's resistance so loading and noise are not excessive.  The first two are my go-to references since I know their provenance exactly.

I'd forgotten about Hoffman's Hamon divider.  That looks like just the ticket, and easy to tweak!

Yeah, I'm thinking about putting the MC-7 in a proper box so I don't chew up the leads from frequent use.

[beanflying]

Depending on the DUT watch the loading of your resistor network. If you are wanting to check 10M meters it becomes an issue at 4.5 digits and up in particular at the LSD. So it is more than just the references ability to remain stable under load. With my HP 735A for example I run my 34401A at 10M to set it before transferring it to handheld meters.

By raising the input voltage in the order of about 10mV (more like 5mV into a 10K dfivider) on 10V into the network shown below you can overcome that loading when using it as a transfer to 10M. Column 1 is sort of an idea on a 10G meter for loading 2 is the same 10V fed into 10M and the third is 10.01V fed into 10M.

The values shown are what I am waiting on from Edwin for my Hamon divider. I have selected fairly low values to minimize this loading issue.

[Edwin G. Pettis]

As an example, a 100K load (such as a 100K KVD) will load the output of a Fluke 731B 0.5 PPM, it is consistant with that circuit and a know commodity, about 1Meg drops it .005 PPM.  Your LTZ circuit will vary with topology of the output.

[David Hess]

The Hamon divider looks ideal.

I was at first thinking a charge-pump capacitive divider but it would be more complex and expensive although not require calibration.

[Kleinstein]

A charge pumped divider in a real implementation is not ideal, as charge injection and parasitic capacitance can add offsets. So the charge pump can provide very stable dividers the absolute divider ration can be off by a few ppm and might need calibration.###

Depending on the implementation a precision transformer (inductive divider) with synchronous rectification might be more accurate without calibration. However some circuits (e.g. like the divider / rectifier in the Datron 1281) can still be slightly off from the ideal turns ratio.

[GregDunn]

My most precise meter is only spec'ed to tenths of a mV accuracy on a 10V source (24 hours) so an error of even 50µV is not going to matter when doing a check.  I'm thinking the Hamon divider is good enough for what I'm trying to do, as long as I keep relative impedances in mind.
 

[Inverted18650]

If you don't go with the Hamon Divider (and you should) maybe a Fluke 731B will work for you? Bring the meter and we can get it done too brotha.