General > General Technical Chat
building hobbiest grade cal standards
(1/2) > >>
Dan Moos:
I am repairing an hp 3468a bench meter. I already know the cal data is lost. (battery died)

I have no need for a lab grade cal, I'd just like it to be in agreement with, say, my Siglent 3055.

Lets take a resistance standard. Would it be reasonable to use a network of same value resistors in a series/parallel combination that on paper should equal the resistance of 1 of those resistors? Basically using many of them to create an average that is close enough? Using as many as is practical?

bdunham7:

--- Quote from: Dan Moos on May 21, 2023, 03:18:03 am ---I have no need for a lab grade cal, I'd just like it to be in agreement with, say, my Siglent 3055.

Lets take a resistance standard. Would it be reasonable to use a network of same value resistors in a series/parallel combination that on paper should equal the resistance of 1 of those resistors? Basically using many of them to create an average that is close enough? Using as many as is practical?

--- End quote ---

No, that theory would require the resistors to have a standard-ish distribution centered exactly around the nominal value.  In reality, they usually are not.  So if you have 10 100R 1% resistors, they all very well might be 99R.  Or, if you get lucky, they might have a standard distribution with a standard deviation of 0R2 and a center of 100R0.  Do you feel lucky?





beanflying:
There is a heap of info to be had in the Metrology section.

Highly recommended starting point however is Conrad Hoffmans articles https://www.conradhoffman.com/mini_metro_lab.html

For a basic Resistance 'standard' then buying a few Vishay metal foil resistors at 0.005% isn't a bad place to start.
ebastler:

--- Quote from: Dan Moos on May 21, 2023, 03:18:03 am ---I have no need for a lab grade cal, I'd just like it to be in agreement with, say, my Siglent 3055.

--- End quote ---

If you are happy with your Siglent 3055 as the "calibration reference", why can't you use any old resistor, as long as it is reasonably stable over time? E.g. take that nominally 10k resistor, measure it as 10.04k on the Siglent, then confirm that the HP gives you the same value (within tolerances). Why would you need a resistor with the exact nominal value?

Just make sure that you don't touch the resistor in the process to avoid temperature-induced drift, i.e. mount it on a suitable base.
EPAIII:
If you can go for a bit less accuracy, 0.05% resistors are very reasonable in price. Not exactly up to digital standards, but can be used for analog meters.

https://www.mouser.com/ProductDetail/IRC-TT-Electronics/WIN-T1206LF-11-1001-A?qs=pD8CpA3T84ph5%2Fg8bDLJgw%3D%3D




--- Quote from: beanflying on May 21, 2023, 04:40:42 am ---There is a heap of info to be had in the Metrology section.

Highly recommended starting point however is Conrad Hoffmans articles https://www.conradhoffman.com/mini_metro_lab.html

For a basic Resistance 'standard' then buying a few Vishay metal foil resistors at 0.005% isn't a bad place to start.

--- End quote ---
Navigation
Message Index
Next page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod