Calibration of an ANENG AN870

[fant]

Good morning;

anybody knows how to calibrate the ANENG AN870?

I have the multimeter that with a DC input voltage of 10,00000 V, read 10,008 V and I would like to calibrate it.

Thank you in advance

[ebclr]

Are you kidding?

[radiolistener]

I have the multimeter that with a DC input voltage of 10,00000 V, read 10,008 V and I would like to calibrate it.

try to put it near the heater or piece of ice 

[Fungus]

anybody knows how to calibrate the ANENG AN870?

I have the multimeter that with a DC input voltage of 10,00000 V, read 10,008 V and I would like to calibrate it.

Try cooling it down a bit.

(Seriously: Put it in the fridge for an hour then measure again...see the problem?)

[MiroS]

anybody knows how to calibrate the ANENG AN870?

I have the multimeter that with a DC input voltage of 10,00000 V, read 10,008 V and I would like to calibrate it.

Try cooling it down a bit.

... or put sticker on LCD 10,00000 V It will be rock solid.

[pickle9000]

[floobydust]

OP I think you are 0.08% out on one range of a 0.05%+3sd spec'd multimeter    If you did calibrate it perfectly, that is not long-term accuracy because it uses vanilla 1% tolerance resistors and an ICL8069, both of which will drift with temperature and age.

It is not easy calibrating an ANENG multimeter. With older builds, the cal co-efficients are stored in EEPROM with a Write-protect.
You have to take the meter apart and jumper the write-protect, put it in cal mode. The known procedure is horribly translated from chinese and might not even apply exactly to the AN870 firmware. Otherwise, you have to hack the EEPROM. Newer multimeter pcb builds have no external EEPROM and I have not looked into how calibration is done there.

https://www.eevblog.com/forum/testgear/cheap-aneng-meter-calibrationreset-procedure/msg1969907/#msg1969907

[fant]

Thank you for the answer, floobydust.

I know that such multimeters are not HP, but I want to try anyway.

BTW, this is my calibrated reference, and this is HP.

[Fungus]

Those bench meters have an internal heater and thermostat for their critical components.

[fant]

I think to know how they are made, having three 3458 in the lab ( one of them, the one on the picture, powered 24/7 ).
My request was on how to calibrate the unit, then if it is close, far or completely out of order it is my problem.
And only one guy (that I thanks) helped me, the others just used bits to move air.

Best regards

[bdunham7]

My request was on how to calibrate the unit, then if it is close, far or completely out of order it is my problem.
And only one guy (that I thanks) helped me, the others just used bits to move air.

For someone with 3 HP3478s, you are asking a rather imprecise question.

You appear to be actually asking "how do I change the gain constants in the calibration memory" or perhaps "what is the manufacturer's calibration procedure", the second of which might be a better question if it weren't for the fact the the product is an unsupported basement-bargain POS.  In either case, user floobydust apparently has some experience or knowledge and it looks like it would be a challenge, given the complete lack of support that you generally find for this level of product.

However, if you are motivated by anything other than idle curiosity as to how to enter new gain constants and you actually care whether the meter is accurate or not, you would want additional data, such as what does it read at 5.00000 V and 15.00000V?  You certainly have the means to check that.  And even if you only want to measure 10 volts for some reason, how consistent is that 10.0008V and under what conditions? Absent a specific manufacturer procedure, unless you take the time to characterize the meter for linearity, tempco and stability you aren't 'calibrating' anything, you are just doing the modern version of twiddling the knobs without a plan.

[fant]

Good morning;

I use all the meters in the lab, temperature from 20 to 23 °C, look here: https://www.eevblog.com/forum/metrology/vintage-weston-element-is-it-any-good/msg3101976/#msg3101976, first picture.
Sorry if I have forgotten the words "calibration memory" but, if a person has already done the work, knows that there is an EEPROM inside.
Anyway, this is the workhorse we have (actually 5 units) in the lab and all are with a reading little bigger than the correct one.
For this reason I wanted to test on one of them the possibility to correct the reading.
For the official use in front of the customers we use the Agilent 34401 that is periodically calibrated by an external company.

best regards

[2N3055]

I think to know how they are made, having three 3458 in the lab ( one of them, the one on the picture, powered 24/7 ).
My request was on how to calibrate the unit, then if it is close, far or completely out of order it is my problem.
And only one guy (that I thanks) helped me, the others just used bits to move air.

Best regards

LOL.

No, other didn't just move air. They went out of their way to politely explain you that your question is, well, nonsense.. Nad wanted to save you grief of lost time for no particular result.
It makes no sense to adjust that very cheap instrument to be better than 500 ppm accuracy. Because it is not stable enough to keep that across even few °C around 23°C. It will also drift with humidity changes and generally drift because of components used.
It is a miracle that they are as accurate and stable as they are..

They merely wanted to point out it is not useful to do so.. If you want to do it for fun, without expectations that 2 weeks later it will hold adjustment, then by all means..

[bdunham7]

Anyway, this is the workhorse we have (actually 5 units) in the lab and all are with a reading little bigger than the correct one.
For this reason I wanted to test on one of them the possibility to correct the reading.
For the official use in front of the customers we use the Agilent 34401 that is periodically calibrated by an external company.

Fascinating that you have not one, but three reference meters that I can't possibly justify even one of, but your everyday 'workhorse' , in a use where precision appears to matter, literally costs less than my test leads.  You have to understand that many will find that difficult to understand!

So, as for getting your meters to read accurately, is the high reading consistent over time and scale?  Do you get 10.0008  10.008 volts from all of them every time you try?  If not, how much variance?  And how about 5.00000 and 15.00000?  Is it 5.004 and 15.012?  If not, what is changing the gain constant really going to get you?  If so, and assuming you don't solve the issue of changing the calibration memory, perhaps you could simply try adding 8K resistance to the test leads and see if the results are acceptable.  If so, find the 10M input resistor and add 8K to it.  Or, remove and measure it and sub in a low-tempco resistor with about 8K more resistance.   From another thread, I think that the 10M input is R29 + R30, but I'm not at all sure since I don't have an example here.

[fant]

Nothing fascinating.
On the normal work, when you have to see the voltage of a power supply, you use the 870, when you have to see the reference for a 18 bit A/D converter to read altitude you use the 3458.
And during the test with customers only 34401 are allowed.
Anyway, the discussion is not going in the direction to solve the problem and for me is closed.

Best regards

[coromonadalix]

something like this ?

https://www.eevblog.com/forum/testgear/cheap-aneng-meter-calibrationreset-procedure/

http://www.kerrywong.com/2016/03/19/hacking-dtm0660l-based-multimeters/

https://www.eevblog.com/forum/testgear/the-official-an8008-meterthread/

youll have to sniff the eeprom and decypher it to get the cal constants  or use some of the links methods ??

[Fungus]

And only one guy (that I thanks) helped me, the others just used bits to move air.

Moving air is the real problem. Even if you calibrate it it might be wrong in a few hours because the sun changed position and temperature changed by one degree.

So... before you even attempt to adjust this meter, it's a good idea to find out how much it's affected by temperature. ie. put it in the fridge for a while. Put it in a warm place for a while. See how much it changes.

After that you'll need to find a small EEPROM inside it that holds calibration data for each range and figure out how to read/modify it.

[Trader]

Sorry about my naive question, I know the way to calibrate those DMMs is through the EEPROM, but since they use the ICL8069 voltage reference, why not discover that resistor responsible to the output voltage and just adjust it?

https://datasheets.maximintegrated.com/en/ds/ICL8069.pdf

[Fungus]

Sorry about my naive question, I know the way to calibrate those DMMs is through the EEPROM, but since they use the ICL8069 voltage reference, why not discover that resistor responsible to the output voltage and just adjust it?

Simple: Because each range has a separate calibration. If you change that then you make everything change.

Plus: I don't think there's a potentiometer. The reference is set approximately then everything else is done in software.

[Trader]

Simple: Because each range has a separate calibration. If you change that then you make everything change.

If you did calibrate it perfectly, that is not long-term accuracy because it uses vanilla 1% tolerance resistors and an ICL8069, both of which will drift with temperature and age.

I thought the DMM lost the calibration because of the ICL8069 and the adjust-resistor. So, fixing them will restore the initial characteristics when EEPROM was set up.

[floobydust]

It's not just the reference voltage that affects a function's calibration.
Most of the DMM functions include the reference voltage plus additional divider or shunt resistors, so the CAL values end up covering for all of these parts.
You can try do a manual CAL with the multimeter, or read/modify the external EEPROM if it has one.

OP was trying to super-calibrate his DMM, which I think is a poor idea because plain 1% resistors are not 0.01% resistors afterwards.

[ialbert]

just for info.

my an870 suddenly gives bad value. DC and Resistance.
I shorted a jumper inside to activate CAL.
Turn on the DMM. Set it to res/diode.
Press the Select to SKIP those hex value.
then for ex. you want to calibrate DC. set the knob there.
then get a supply like for AN870. on DC set external supply to 1v, put the probe then it will show the reading. you can press the hold button to calibrate it for the right value.
same with resistance.
after that you can remove the jumper inside.
and it works. now my an870 gives right value again.

[ivo]

just for info.

my an870 suddenly gives bad value. DC and Resistance.
I shorted a jumper inside to activate CAL.
Turn on the DMM. Set it to res/diode.
Press the Select to SKIP those hex value.
then for ex. you want to calibrate DC. set the knob there.
then get a supply like for AN870. on DC set external supply to 1v, put the probe then it will show the reading. you can press the hold button to calibrate it for the right value.
same with resistance.
after that you can remove the jumper inside.
and it works. now my an870 gives right value again.

Thanks for this, now my AN870 reads 5.000V on my standard instead of 5.005V! Although the 'menus' were a bit f***y. Operator beware, it wants exact 1 x 10ⁿ values only (I think)! I tried to get it to cal 5V (thinking maybe the hold let you move the cal offset value around somehow) and almost stuffed my main voltage reading! It thought everything was half its value (5 is half of 10 I guess), and it really struggled to hold and take a 10V back exactly as a new cal reading. Got there in the end though.