(SOLD) DE/EU Simple LM399 Reference Board

[RoadRunner]

Last week i posted in open source forum about this simple little board i made. Many people have asked if i have them for sale. i don't wanted to make open source section a buy sell section. So here i it.
i have few boards available, Checked for stablity and Measured using my new keysight 34465A. I will provide CSV file with actual reading over a hour period. Board has LM399AH and LT1001ACN8 on board with nice wago clamp connectors.
Price 40 EUR including shipping worldwide. Paypal as friend or SEPA.
Please send  PM.

I am doing few experiments with LTC2500 so i needed stable reference voltage, i quickly made this little PCB which some may find interesting. It is open source , Schematic, Gerber and PCB files are available in my github account.

https://www.circuitvalley.com/2018/08/diy-open-source-lm399-voltage-reference-10v-dmm-calibrate.html

[HighVoltage]

Why did you not make it adjustable for 10.0000 V?
 

[RoadRunner]

Why did you not make it adjustable for 10.0000 V?

Resisitors used for gain setting has 10PPM tempco,
if you want it tunable to 10V, Either have multiple fixed value resistors in series or parallel, or use Variable resistor. Variable resistor in 10PPM tempco are going to cost too much. i was more concerned about stablity compare to absolute value.
i have footprint for multiple 0603 resistor, so if absolute value matters, user may use these to get value he is looking for.

[guenthert]

Well done!   

Why did you not make it adjustable for 10.0000 V?

Resisitors used for gain setting has 10PPM tempco,
if you want it tunable to 10V, Either have multiple fixed value resistors in series or parallel, or use Variable resistor. Variable resistor in 10PPM tempco are going to cost too much. i was more concerned about stablity compare to absolute value.

Your concerns are well noted.  I think it is customary to use a combination of fixed and variable resistors for the voltage devider, with the bulk of the voltage dropping over the (temperature stable) fixed value resistor, so that the large tempco of the variable resistor has little influence.
 

i have footprint for multiple 0603 resistor, so if absolute value matters, user may use these to get value he is looking for.

Can I see a high-res picture of that?

[guenthert]

'nother note:  while most DVM have some 'over-range' (I've seen 3 to 100%), I got here a PCI based DAQ, which refuses even a 10.001V source in the 10V range.  I don't know, how common that is, but there might be more such devices.  So I think it would be advantageous to err on the low side, i.e. chose resistors, so that the output is just below, instead of above 10V.

[RoadRunner]

Well done!   

Why did you not make it adjustable for 10.0000 V?

Resisitors used for gain setting has 10PPM tempco,
if you want it tunable to 10V, Either have multiple fixed value resistors in series or parallel, or use Variable resistor. Variable resistor in 10PPM tempco are going to cost too much. i was more concerned about stablity compare to absolute value.

Your concerns are well noted.  I think it is customary to use a combination of fixed and variable resistors for the voltage devider, with the bulk of the voltage dropping over the (temperature stable) fixed value resistor, so that the large tempco of the variable resistor has little influence.
 

i have footprint for multiple 0603 resistor, so if absolute value matters, user may use these to get value he is looking for.

Can I see a high-res picture of that?

Thanks of suggestion.

Here is layout.

'nother note:  while most DVM have some 'over-range' (I've seen 3 to 100%), I got here a PCI based DAQ, which refuses even a 10.001V source in the 10V range.  I don't know, how common that is, but there might be more such devices.  So I think it would be advantageous to err on the low side, i.e. chose resistors, so that the output is just below, instead of above 10V.

I was lacking experience and also did not paid much attention to the fact that not all instrument will be happy to accept little over the 10V range.

Any way changing resistor will fix this issue.

Regards

[bingo600]

May i suggest some reverse polarity input protection.

I have lost 2 references on that account 

/Bingo

[RoadRunner]

May i suggest some reverse polarity input protection.

I have lost 2 references on that account 

/Bingo

Thank you for suggestion.
Many Things are missing , I even missed to put polarity and input voltage on silkscreen. So i just put some printed label.
It was a quick project to test LTC2500 32bit ADC board i made for experimenting with LTC2500. 

[Jon.C]

Hello @RoadRunner

received, works perfect and good packaging

Thank you!   

[RoadRunner]

Hello @RoadRunner

received, works perfect and good packaging

Thank you!   

I am glad it fulfills your needs.

Few boards Still available.

[GigaJoe]

" Variable resistor in 10PPM tempco are going to cost too much."
(all below IMPE, )

not exactly.. you do trimming of existed value to a very small amount, a fraction of ohms, so let say  adding a resistor in a sequence , 10ohm 5ppm + 100 (200 -500) ohm trimpot at 100 (200-300) ppm ...  you may calculate influence by trim-pot, it will be pretty small ...

option2:  you may adjust offset of Lt1001,  as it in range of microvolts, it doesnt affect opamp TC much. Yet, it not a recommended , but in a millivolt range.

fancy thing:  opamp divider to get 10V, TC does influence on each shoulder differently , so if you have resistors with positive and negative TC yo may compensate it place a specific resistor in a specific shoulder. you can do a super fine tuning,  a resistor made from a long copper wire, as copper has very high positive TC.

[RoadRunner]

" Variable resistor in 10PPM tempco are going to cost too much."
(all below IMPE, )

not exactly.. you do trimming of existed value to a very small amount, a fraction of ohms, so let say  adding a resistor in a sequence , 10ohm 5ppm + 100 (200 -500) ohm trimpot at 100 (200-300) ppm ...  you may calculate influence by trim-pot, it will be pretty small ...

option2:  you may adjust offset of Lt1001,  as it in range of microvolts, it doesnt affect opamp TC much. Yet, it not a recommended , but in a millivolt range.

fancy thing:  opamp divider to get 10V, TC does influence on each shoulder differently , so if you have resistors with positive and negative TC yo may compensate it place a specific resistor in a specific shoulder. you can do a super fine tuning,  a resistor made from a long copper wire, as copper has very high positive TC.

Thank you for suggestion. if i ever get around making second revesion , i will definately put them into design.

I also though that both side of a pot will have same TC influence but then i though about wiper. i was not sure what happen to wiper Contact and as it was a day project, i did not get into details.

[GregDunn]

Received mine today and gave it a chance to warm up.  Looks good on all my meters - the Keithley 197a reads exactly the listed voltage.  Not too surprising, it seems to be the most accurately calibrated of all my meters.  I didn't see any drift at all - again, not surprising; I don't have a meter good enough to measure drift on this board.  I have made my own foam "hat" for the LM399 and will probably put it all in a custom box eventually.

Thanks for the board!

[RoadRunner]

Received mine today and gave it a chance to warm up.  Looks good on all my meters - the Keithley 197a reads exactly the listed voltage.  Not too surprising, it seems to be the most accurately calibrated of all my meters.  I didn't see any drift at all - again, not surprising; I don't have a meter good enough to measure drift on this board.  I have made my own foam "hat" for the LM399 and will probably put it all in a custom box eventually.

Thanks for the board!

Thank you all for nice feedback. i may make second batch with new design taking all suggestions into account.

[guenthert]

Received mine today and gave it a chance to warm up.  Looks good on all my meters - the Keithley 197a reads exactly the listed voltage.  Not too surprising, it seems to be the most accurately calibrated of all my meters.  I didn't see any drift at all - again, not surprising; I don't have a meter good enough to measure drift on this board.  I have made my own foam "hat" for the LM399 and will probably put it all in a custom box eventually.

Thanks for the board!

Thank you all for nice feedback. i may make second batch with new design taking all suggestions into account.

Yes, please do.  You might want to post design ideas to the project or metrology section in order to solicit more qualified suggestions.