MI 1340A precision divider question

[Overspeed]

Hello

MI ( Canada ) propose a 1/10000 divider built by using 1/10 divider in serie with quite impressive performance

Question : when the 1/10000 ratio is used ( 4 1/10 in serie ) how the resistor noise is managed ?

This sound interesting to generate very low voltage for testing , does someone know what circuit / components they use ?

4 circuits with the same design and the same components .

Regards
OS

[Kleinstein]

They don't show much details, but chances are this is  1 resistor chain, like  9M , 900 K, 90 K and 10 K. The resistance values may also be lower (like 1/10) - have not found a value.

Depending on the resistance there will be resistor noise, but not much to be done about it. The resistor chain is made to start with a high voltage, like 100-1000 V. Starting with a lower voltage (e.g. 10V) one has the choice of high resistance and rel. high noise with the full chain and starting in between with less resistance, but also less divider steps then. So the divider would not be a good solution to go from 10 V to 1 mV.
In addition to the resistor noise there may also be thermal EMF at the terminals. There are other dividers for very low voltages: power coefficient and leakage is less critical, but thermal EMF and proper drive sense (quad junctions for steps in between) more important.

[Overspeed]

They don't show much details, but chances are this is  1 resistor chain, like  9M , 900 K, 90 K and 10 K. The resistance values may also be lower (like 1/10) - have not found a value.

Depending on the resistance there will be resistor noise, but not much to be done about it. The resistor chain is made to start with a high voltage, like 100-1000 V. Starting with a lower voltage (e.g. 10V) one has the choice of high resistance and rel. high noise with the full chain and starting in between with less resistance, but also less divider steps then. So the divider would not be a good solution to go from 10 V to 1 mV.
In addition to the resistor noise there may also be thermal EMF at the terminals. There are other dividers for very low voltages: power coefficient and leakage is less critical, but thermal EMF and proper drive sense (quad junctions for steps in between) more important.

Hello

Thanks for your answer , I search a solution without rocket science to generate 1 mV or lower value with acceptable precision compare to a Ref 54 at 5V so not drowned in noise . most articles / schematics are for HV voltage unfortunately.
Do you have a track to follow ?

Regards
OS

[StacysMom]

The lab I work at has one of these but it is currently out for calibration, so I am unable to get you the actual resistance value of each stage.

We have noticed significant noise when starting from the highest stage and dividing down to very low voltages.

Once the unit arrives back calibration, I can provide more details.

Edit:
Just looked back on some of my notes and found that the 5 resistors in series are 4MΩ, 400kΩ, 40kΩ, 4kΩ, and 400Ω.

[bastl_r]

Once the unit arrives back calibration, I can provide more details.

Oh yes!
We would of course be delighted to receive a detailed teardown.   

Regards 

[Overspeed]

The lab I work at has one of these but it is currently out for calibration, so I am unable to get you the actual resistance value of each stage.

We have noticed significant noise when starting from the highest stage and dividing down to very low voltages.

Once the unit arrives back calibration, I can provide more details.

Edit:
Just looked back on some of my notes and found that the 5 resistors in series are 4MΩ, 400kΩ, 40kΩ, 4kΩ, and 400Ω.

hello
Thanks as any additional informations is useful even if that not for low voltage , the concept to stack 1/10 divider is interesting .

Regards
OS