Maker Pro
Maker Pro

Circuit selector

Status
Not open for further replies.

Rocksteady

Aug 29, 2022
11
Joined
Aug 29, 2022
Messages
11
One possibility to get rid of all the switching and severe MOSFET requirements :



A compendium of articles on active resistors (see references at page bottom) :

https://link.springer.com/article/10.1007/s10470-009-9351-5

One example using OTA :



Regards, Dana.
Thank you Dana, that is one I have not yet considered. I will look into that.
 

hevans1944

Hop - AC8NS
Jun 21, 2012
4,878
Joined
Jun 21, 2012
Messages
4,878
I am building an ohm meter calibrator and I do not want a uC that will interfere with the signal.
That would be the LEAST of your potential problems... switching transients.

You want to build a precision resistance calibrator that "dials in" and digitally displays resistance from one ohm to 1.9996x10^8 ohms? Good luck with that! I used to work as a technician in an electronics laboratory that maintained a metrology facility for checking the electrical calibration of voltmeters, ammeters, and ohmmeters... among other more exotic instruments. We maintained various resistance "standards" that had to be periodically sent out to an independent laboratory for measurement of their "nominal" resistance. It was important to have a history of these measurements and the temperature and humidity conditions under which the measurements were made. That historical record could tell us immediately if something was amiss with a particular resistance standard. All of our resistance standards were of the four-terminal Kelvin arrangement: two wires in for excitation, two more wires that connected to the standard resistance for measurement. Here is a picture that shows a resistance standard similar to the half-dozen or so our metrology lab maintained:

1662340102910.png
Note the holes in the outer cylinder. This was to allow the standard to be immersed in a temperature-controlled oil bath. Several days were required before the resistance element reached equilibrium temperature and the resistance no longer changed. Planning ahead was encouraged. Now how, exactly, will you ensure that your "calibrator" is accurate and remains accurate after it is tossed around on your bench for awhile? You want to maintain the same precision over more than eight orders of magnitude change in resistance? You want to (somehow) place switch contacts in series and/or parallel with your twenty-eight decade-configured calibration resistors without that having any measurable effect on the resistance? I suggest you go to this EEVblog Electronics Community Forum for a fascinating thread on DIY ohmmeter calibrators.
 

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,700
Joined
Nov 17, 2011
Messages
13,700
The only problem is I can not have ANY interference on the output signal.
Why do you think my circuit produces interference? As long as no button is pressed all signals are static, no interference to be expected from static signals.

Btw: Expecting"no Interference" is illusory. There's so much electrromagnetic noise in the environment, you'd need a completely shielded chamber to achieve a very low level of interference - that's why such chambers are used foe e.g. emc measurements.
I need a precision of less than 0.1 % on the resistance.
What kind of switch do you intend to use? A typical quality signal relay (link to example) can have an initial contact resistance up to 0.1 Ω, that is already 10 % of the lowest value in your setup (1 Ω).
 

crutschow

May 7, 2021
839
Joined
May 7, 2021
Messages
839
One way to switch the resistors with low error is to use power MOSFETs, which can have an on-resistance of 0.5mΩ (example) with no intrinsic offset voltage.
That would give a error of 0.05% in series with a 1 ohm resistor.
 
Status
Not open for further replies.
Top