Cheap, not Chinese, Voltage reference for DIY meter calibration

[dazz1]

Hi
I found the thread on cheap Chinese AD584 voltage references but then I started looking at DIY design and build.
When I did a search, I found very high performance devices and reasonable prices.  Just one example being the Max 6325.  1ppm temp coeff and 0.02% tolerance.  Far better than an AD584.
https://www.digikey.co.nz/en/products/detail/analog-devices-inc-maxim-integrated/MAX6341CSA/1867167

Just wondering if my next project should be to design and build a very cheap and very high performance voltage reference so I can check-cal my small collection of test equipment.

[tautech]

LOL, go for it Dazz !
Sorry if our emails had you tossing and turning about this last night. 

[nctnico]

Hi
I found the thread on cheap Chinese AD584 voltage references but then I started looking at DIY design and build.
When I did a search, I found very high performance devices and reasonable prices.  Just one example being the Max 6325.  1ppm temp coeff and 0.02% tolerance.  Far better than an AD584.
https://www.digikey.co.nz/en/products/detail/analog-devices-inc-maxim-integrated/MAX6341CSA/1867167

Just wondering if my next project should be to design and build a very cheap and very high performance voltage reference so I can check-cal my small collection of test equipment.

The long term stability and low temperature coefficient are great, but the tolerance is likely higher than you would want for DMM calibrations. A typical 4.5 digit DMM sits around 0.05% already.

[m k]

Where is tolerance?

I found only initial accuracy.

[BeBuLamar]

I don't understand! Cheap not Chinese?? The AD584 is Analog Device and it may be made in China but the MAX 6341 is by Maxxim Integrated which is part of Analog Device so how would you know that it is not made in China too?

[Kosmic]

I don't understand! Cheap not Chinese?? The AD584 is Analog Device and it may be made in China but the MAX 6341 is by Maxxim Integrated which is part of Analog Device so how would you know that it is not made in China too?

I guess the OP doesn't want people recommending an already made Chinese voltage reference and seem to be incline to go the DIY route?

[BeBuLamar]

The AD584 is not a premade voltage reference unit. It's just an IC like the MAX6341CSA. So the same amount of DIY work is needed for either option. Now I must admit that it's kind of difficult to determine where the IC is made. The same IC can be made in several different countries. What I know is that Analog Device is an American Company and it owns Maxim Integrated so it's quite possible both chips were made in the same factory.

[DimitriP]

. It's just an IC like the MAX6341CSA.

It's main feature is that it can provide 2.5 ,  5 ,  7,5  , 10 and -10 even using a single IC. Just one!
If you are not intrested or don't "need" all of the above with a single IC, there are better choices like the MAX series already mentioned.

[Martin72]

Just wondering if my next project should be to design and build a very cheap and very high performance voltage reference so I can check-cal my small collection of test equipment.

I will soon have my voltage reference ready so that I can send it off for calibration.
Calibration ? Yes, of course, you have to be sure that the reference is actually (significantly) more accurate than the device you want to calibrate with it.
Simply building a reference yourself is not enough.

[dazz1]

Hi
OK so by "not Chinese" I meant not from Aliexpress or Bangood.  I buy a lot off stuff of Aliexpress so I am not anti.

For me, calibration is about confidence, not certification.  After reading the thread about actual Chinese voltage reference modules, I wouldn't have confidence in the reference.  If I am going to spend $$$ on a reference it should be worth enough to justify a calibration.

The Maxim part I quoted was the result of a 2 minute search on a supplier's website.  No science behind that choice. 

One option is to buy the Aliexpress module, and fit a known reference part with providence.  I have done that before.  I anticipate a "proper" DIY design would get down to the tiny details like solid copper, unplated, connectors to avoid thermo-electric effects and other things.

I can't be the only one on the planet that has considered a DIY approach to a voltage reference using off the shelf parts.  I am interested to hear the views of others. 

[BeBuLamar]

Oh you meant a pre built module that uses the AD584 IC. So you could build yourself a module using the AD584 bought from an electronic dealer in Australia. Here in the USA I would buy it from Mouser, Digi Key or RS-Online.

[tautech]

Oh you meant a pre built module that uses the AD584 IC. So you could build yourself a module using the AD584 bought from an electronic dealer in Australia. Here in the USA I would buy it from Mouser, Digi Key or RS-Online.

Kiwi flag is ours. 
Not them Aussie rogues. 

[BeBuLamar]

Sorry about that. My apology.

[KungFuJosh]

For me, calibration is about confidence, not certification.  After reading the thread about actual Chinese voltage reference modules, I wouldn't have confidence in the reference.  If I am going to spend $$$ on a reference it should be worth enough to justify a calibration.

That's called false confidence. I am not opposed, but without calibration, you don't know what you're looking at. You can monitor stability from a non-certified reference, and that's useful, but it doesn't tell you about the accuracy of what you're using the reference for.

I bought this ref as recommended by another member: https://www.aliexpress.us/item/3256804431053626.html

I've found it to be very stable, and if I am interested in more than monitoring drift, I'll pay for proper certification.

I have a couple other refs I bought from ali, and the worst thing about them was the fake calibration. They were all fairly stable. I think that purple one was the best, but it also got the most use, and that on time leads to stability if it's not a total POS.

I can't be the only one on the planet that has considered a DIY approach to a voltage reference using off the shelf parts.  I am interested to hear the views of others.

You're certainly not. There are many threads dedicated to DIY refs, including this one: https://www.eevblog.com/forum/testgear/elektor-voltage-current-calibrator-project/

It's not a cheap project (build cost probably still above $500 per unit), and it should get calibrated too.

Whether you go the DIY route, or get a premade ref, they all need significant burn-in time for stability. Once they're stable, they're worthy of calibration.

[luma]

How about https://dmmcheckplus.com/

Solid reference, but the key value is the cheap and fast calibration service in the future.  $20 service + $7 shipping (and free for the first two years) makes a solid reference into a damn useable standard for the home gamer.

[tautech]

How about https://dmmcheckplus.com/

Solid reference, but the key value is the cheap and fast calibration service in the future.  $20 service + $7 shipping (and free for the first two years) makes a solid reference into a damn useable standard for the home gamer.

Certainly one of the best and trusted solutions....and we discussed these at length privately but $ more that Dazz wants to outlay.

[Martin72]

How about https://dmmcheckplus.com/

I have the device, it's good for a quick check(therefore it´s name) to see if a meter is working.
As a reference it is OK for 3.5 digit meters.
The LCR option is somewhat sobering.
The capacitors and coils are not exact types, but simply measured.
But as I said, it's good for testing and comparing.

[dazz1]

You're certainly not. There are many threads dedicated to DIY refs, including this one: https://www.eevblog.com/forum/testgear/elektor-voltage-current-calibrator-project/

I have made Elector stuff years ago and it was always nice but it sounds like their quality control is not so good for this project. I see you spent $USD366 on the Elector calibrator. 

I note there is a commercial calibrator on that thread that was "only" £350  https://www.ianjohnston.com/index.php/onlineshop/handheld-precision-digital-voltage-source-2-mini-detail  Looks good but too many £££ for my budget.

It's not a cheap project (build cost probably still above $500 per unit), and it should get calibrated too.

For that money it would only make sense to get it properly calibrated.

Whether you go the DIY route, or get a premade ref, they all need significant burn-in time for stability. Once they're stable, they're worthy of calibration.

Looking at a few data sheets, it appears that the initial value of voltage references can be quite different to the nominal output voltage.  Maybe 1% or so.  That would make it fairly essential to get it calibrated.

[dazz1]

Hi
I have just found the LB02A process calibrator which looks promising.
https://www.aliexpress.com/i/1005001632714207.html
 reviewed here:


It is Chinese and not lab quality substitute for a good volt reference but its usefulness could compensate for that. 

The LB06 is the highest priced version.

[dazz1]

Hi
It appears the LTZ1000 was replaced by the ADR1000 that was then replaced by the ADR1001.
At about $USD100 definitely not the cheapest solution but maybe the best accessible for a DIY solution.
https://www.digikey.co.nz/en/product-highlight/a/analog-devices/adr1001-precision-voltage-reference

Marco Reps channel did a ADR1000 reference

[BennoG]

what about this one.

https://voltagestandard.com/

I have looked at it but not yet ready to buy one.

Benno

[Kosmic]

Hi
It appears the LTZ1000 was replaced by the ADR1000 that was then replaced by the ADR1001.

Please note that the LTZ1000 is still produced and available (so not replaced by the ADR1000).

ADR1000 is the newer version and one of the improvement is lower noise. Still not readeally available though (last time I checked).

ADR1001 is the SMD variant (so do not replace the ADR1000).

[dazz1]

what about this one.

https://voltagestandard.com/

I have looked at it but not yet ready to buy one.

Benno

They look nice but they are in imperial units because these are made in the US of A.  I don't know anyone who has a thermometer to measure 68°F.
When I go to checkout, none of the addressable states are outside of the USA.
Previous experience tells me that shipping anything across the US border is excessively expensive, especially if a courier service is used.

It is cheaper to buy a metric voltage standard.

So yes, but no.

[dazz1]

Hi
It appears the LTZ1000 was replaced by the ADR1000 that was then replaced by the ADR1001.

Please note that the LTZ1000 is still produced and available (so not replaced by the ADR1000).

ADR1000 is the newer version and one of the improvement is lower noise. Still not readeally available though (last time I checked).

ADR1001 is the SMD variant (so do not replace the ADR1000).

That's what happens when I skim across the Internet.

I could not find a supplier of the ADR1000 but Arrow and Digikey sell the ADR1001EBZ device and  ADR1001AEZ eval board.
Not the cheapest device but I'd say good value for the $$.  Spend more and get more.

[schmitt trigger]

what about this one.

https://voltagestandard.com/

I have looked at it but not yet ready to buy one.

Benno

They look nice but they are in imperial units because these are made in the US of A.  I don't know anyone who has a thermometer to measure 68°F.
When I go to checkout, none of the addressable states are outside of the USA.

There is a pull down menu with a list of the World’s countries. New Zeland is there. Mexico is also there and if selected the Mexican state menu will appear.
With respect to the 68F label, it means the unit was calibrated at that temperature, meaning 20C.