WTB [AU] Portable Voltage Reference based on LM399

[enut11]

WTB: a portable 10v reference based on the LM399 chip.

I am in the process of assembling a travelling metrology kit
https://www.eevblog.com/forum/metrology/mini-metrology-kit-for-travelling/new/#new

EBAY offerings so far may or may not be aged or calibrated.

Unboxed reference is OK but I need it to be calibrated with a decent meter.

[KungFuJosh]

You should talk to @TylerPeppy and see if he can sell you one of his ADR1399 refs, or if you can make your own.

[Kean]

I have a CircuitValley ADR1399 one from @RoadRunner - worth it to check but it is much more expensive than ebay/AliExpress.

I also have one of the AliExpress ones (Emeco) which seems to track well, although I have not yet done any proper comparison or logging.  I've just left them both to burn on for 12+ months (more like 20?).

I have multiple 6.5 digit meters (all out of cal), so of course I have no idea of their actual voltage...

[TylerPeppy]

WTB: a portable 10v reference based on the LM399 chip.

I am in the process of assembling a travelling metrology kit
https://www.eevblog.com/forum/metrology/mini-metrology-kit-for-travelling/new/#new

EBAY offerings so far may or may not be aged or calibrated.

Unboxed reference is OK but I need it to be calibrated with a decent meter.

If you have the meters in cal, I have extra boards I can ship, and you can assemble. The CircuitValley 1399 refs would be similar price to what it would cost me to assemble and ship to you. But if you have in cal meters, then assembling yourself makes it cheaper.

Tyler

[Kean]

I presume that is the one described in this post?
https://www.eevblog.com/forum/projects/udr1399-adr1399-based-10v-reference-module/

[TylerPeppy]

I presume that is the one described in this post?
https://www.eevblog.com/forum/projects/udr1399-adr1399-based-10v-reference-module/

Correct. BOM without connectors is about 150$ CAD.

[HairyWombat]

If you have the meters in cal, I have extra boards I can ship, and you can assemble. The CircuitValley 1399 refs would be similar price to what it would cost me to assemble and ship to you. But if you have in cal meters, then assembling yourself makes it cheaper.

Tyler

Tyler

I would be interested in one of those also.
I have a 34401A I just got calibrated so this may be my best option.

[Kean]

I have a 34401A I just got calibrated so this may be my best option.

Can you share who you used for calibration?  I really should get at least my 34465A cal'd.  Then I can cross check all my 34401A's and 34970A's.

[HairyWombat]

I used Ciscal in Epping but they have offices everywhere including internationally.

Dropped off a 34401A, DT4282 and an Agilent UT1252A (Felt lazy with the handhelds )
34401A and Hioki came back nice and matched my Yoko TY720 which was nice.
The UT1252A went back as it was out across the range and is still with them, have read they can be pesky.

I first contacted Rapid-Tech as I need a Yoko part but 4 people got involved and never heard back about the part.
They also quoted AUD$300 plus to calibrate the 34401A.....Yeah Nah

Ciscal quoted me:
For the 34401A and Keithley 2001 AUD$210 Plus GST Each
For the Handhelds AUD$85 Plus GST each

I thought this was quite reasonable.

[Kean]

I used Ciscal in Epping but they have offices everywhere including internationally.

Great, there is one in Seven Hills, NSW just around the corner from a client of mine.  When things are quieter I should really organise to cal my 34465A and maybe my newest 34401A.  Or maybe I'd be better off getting them to do my K2400 SMU.

I'll take another look at the local options on the NATA list.  I know of one quite close to where I live, but their uncertainty is too high for 6.5d.  NMI in Lindfield has a JVS, but cal against that is probably going to be unnecessarily costly.  They may not even deal with my small company, even though I've gone on a tour of the facility.

[enut11]

Reply to all the relevant posts above.

We are travelling in a caravan so I do not have access to my lab to build and calibrate a voltage reference.

I am after a completed pcb with a cal check with at least a 6.5 digit meter.
enut11

[KungFuJosh]

Reply to all the relevant posts above.

We are travelling in a caravan so I do not have access to my lab to build and calibrate a voltage reference.

I am after a completed pcb with a cal check with at least a 6.5 digit meter.
enut11

You should specify your budget for a completed / calibrated ref.

[enut11]

Budget?
Not an easy answer.
I already have very good ovenised references at home.

I was hoping to contact someone willing to sell me a portable calibrated reference so I can play metrology games on-the-go.

If nothing shows up here I will look elsewhere.
enut11

[TylerPeppy]

Budget?
Not an easy answer.
I already have very good ovenised references at home.

I was hoping to contact someone willing to sell me a portable calibrated reference so I can play metrology games on-the-go.

If nothing shows up here I will look elsewhere.
enut11

It becomes difficult to reccomend if people aren't aware of the requirements. Sounds like CircuitValley's Ten399 reference fits your bill. Additionally, I would not go with an LM399 based reference compared to the ADR1399, as many users have covered improvements that the 1399 has that are worthwhile.

On-the-go calibration labs are difficult though, as the reference will go through more temperature changes. Even with my own reference, I have shown that the voltage will indeed drift with temperature changes of 2-3C.

[enut11]

@TylerPeppy
I cannot be too specific as this is a work-in-progress. I also know that it is difficult to do any serious met experiments in an uncontrolled environment. It is mostly about keeping myself busy and learning stuff.

I did say in my metrology post that I had no budget but now realise that this was a rash statement. My current environment does not justify spending too much unless the item has a future use when we finish travelling.
enut11

[KungFuJosh]

I did say in my metrology post that I had no budget but now realise that this was a rash statement. My current environment does not justify spending too much unless the item has a future use when we finish travelling.
enut11

A budget simply means what you want to spend on it. It doesn't matter what it is. You want something good, it's going to be at least $200 or significantly more. If you want something cheap, then get something cheap. Go ahead and get the $50 LM399 in the plastic case if that fits your current need. Pick a number or a range.

Since this is about learning more than anything else, it doesn't matter what you get. If you buy the LM399 in the plastic case (don't waste your time with anything not in a case, unless you're planning to improve it and put it in a case), then you can use the mediocrity of it as a learning experience. You can document the noise, stability, and drift over time and different environmental conditions. Maybe you'll be pleasantly surprised, and it will be awesome. Who knows.

I'm personally curious about that cheapo LM399 ref (in the plastic). I wonder how it is, and how it would be if it was upgraded to ADR1399. And/or if other improvements can be made (low cost/effort) to improve it.

Back to budgeting, "spending too much" is entirely relative. If you think $500 is cheap, then you have a lot of options. If you think it's too much, then don't spend that much. If you say you want to spend $50, then somebody willing to get rid of a ref might contact you. If you say up to $200, then more people will be likely to contact you.

[enut11]

Point taken.
I have just purchased the cheap chinese LM399 in the plastic case. Realistically, this will most likely more than challenge the 4.5 digit meter that I will end up with. (4.5D seems to be the upper limit for compact portable meters.)

My biggest problem is calibration.
The seller claims 10v and Vz are checked with a Keithley 7510. We will see.
enut11

[KungFuJosh]

If you get a good/new 4.5D meter, it should have its own calibration. Hopefully, that will let you know if the ref is close enough for your needs to the claimed value.

[Kean]

Point taken.
I have just purchased the cheap chinese LM399 in the plastic case. Realistically, this will most likely more than challenge the 4.5 digit meter that I will end up with. (4.5D seems to be the upper limit for compact portable meters.)

My biggest problem is calibration.
The seller claims 10v and Vz are checked with a Keithley 7510. We will see.
enut11

FYI, as a comparison of the above mentioned references.

I have the 10V out of my CircuitValley ADR1399 ref connected to the front inputs on my Keysight 34465A (last cal in 2015 at factory, 9.999,985,9) and my more trusted 34401A (Agilent version with round buttons).  The Emeco LM399 ref 10V output is connected to the rear inputs.

The CV ADR1399 ref reads 10.000,274 on the 34465A and 10.000,16 on the 34401A.  On my older Agilent 34401As it reads 10.000,86 and 10.000,33.  The 34465A using stats, 10V range, and 10,000+ readings at 10PLC, std.dev 0.000,002.
The Emeco LM399 ref reads 10.000,111 on the 34465A, and 10.000,01 on the 34401A.  On my older Agilent 34401A's it reads 10.000,69 and 10.000,16.  The 34465A using stats, 10V range, and 700+ readings at 10PLC, std.dev 0.000,004.

Note the consistent approx 160uV difference between the two references across all four meters, so I don't think that is due to Front/rear inputs or meters.  Essentially it appears that one of the 34401A's is out of 1-year spec, but the other are all likely good.  My total uncertainty with these measurements is probably too high to be sure, but it gives me some confidence for those three other meters.  That one with a reading much higher is my only SG manufactured one (approx 2005), all the others are MY manufactured (approx 2014, 2007, 2001).

This was just a quick check I did this morning.  I am just using common banana plug leads, daisy chained between meters, not low EMF cabling although I have bought some to (eventually) set up some longer term logging.  The weather yesterday and today have been hotter than usual, but all of the meters have been powered on constantly for months.  I need to check my notes if I can find them, but I seem to recall that these readings match quite well with those from about 6 months ago.

Both references were bought in December 2024, and have been powered on almost continuously since then.  The CV ADR1399 was one of the early batch, and it didn't get a label of its initial 10V reading (only Vz).  I wasn't too worried as I had planned to leave it to age and eventually get a reading on a local 7.5+d calibrated meter.  The Emeco LM399 10V output was labelled as 9.99962 using a DMM7510, but it likely has drifted somewhat in nearly 2 years (no idea how much burn in they do).

BTW, if you are passing through Sydney at some point in your travels I'd be happy to do a comparison measurement for you.

[enut11]

Hi @Kean
Interesting comparison of your Vrefs. It actually reminds me of when we had an active Aussie Cal Club (now dormant) and we shipped our Vrefs around the country for comparative tests. We had members in most Aussie states. Those were the days...

We are currently in QLD and don't expect to be in NSW for some months. I could post the LM399 Vref if that is ok with you. It is expected to arrive here in about 2 weeks.

See Reply #26 for a peek inside. There is some interesting additional circuitry on The PCB
https://www.eevblog.com/forum/metrology/mini-metrology-kit-for-travelling/

[Kean]

Thanks enut11, I hadn't seen that Tony Albus video.  Nor had I even opened it up to look inside (a major fail there).

Also, silly me didn't think to compare against my PDVS2mini.  I've got that hooked up this evening, and after a little while to warm up and settle the reading is 9.999,899 on the 34465A.  I'll add that to my spreadsheet.

I'd be very happy to take some comparison measurements if you want to send me your unit.  I might even order myself another Emeco LM399 unit with current cal for comparison, as I was thinking about modding one to add a battery backup.

I've just cleared some shelf space and fired up the old Prema 6047 to see how its 10V range looks.  That lost its cal when Dave had it, so reverted to factory - but with only a few minutes warm up it reads 10.000,113 off the PVDSmini which isn't bad (80s integration time!).

[Kean]

I got my extra Emeco YMC-05 LM399H reference in the post today.  I am at home now and only have a couple of basic 4.5 digit meters to hand.  I let it warm up for a few hours while I had dinner and a nap.

It is labelled as measured with a Hioki DM7275 (not DMM7510 this time) at 25C 60% RH, but not dated - so I don't know how fresh that reading is.  And of course we have no uncertainty info, but I believe the DM7275 should be within 20ppm when in spec.

The 10V output was labelled as 9.98177, and my basic home DMMs read either 9.972 (Owon SPM6053) or 9.987 (Aneng AN8008).
The 7V output was labelled as 7.03219, and again I get 7.025 and 7.036 - one lower and one higher.

I'll be in the office for a fair bit of the long weekend, so I'll see how it look across the 6.5d meters and compare with the other two older references (& PDVS2mini!).

[Kean]

As I mentioned, the newly received LM399 reference from Emeco on AliExpress was labelled with a measured voltage of 9.98177V.

Let us assume for now that the labelled voltage is spot on...

I had it warm up for a few hours last night, and then left it gathering statistics for some hours.  The result with about 30,000 samples at 10PLC on my 34465A is 9.981778V (+0.8ppm) with sd of 3uV.  There is some noise, with the span at 50uV.

Although my 34465A hasn't been calibrated since I bought it in 2015, I now feel pretty confident that it has stayed well within spec.

My 34401A meters read 9.98168V (-9ppm), 9.98182V (+5ppm), and 9.98236V (+59ppm).  As mentioned previously, I knew that 3rd one was out of cal.  All the above was using rear terminals.

The Prema 6047 reading is 9.981925V (+16ppm) at 10s integration time.  I'm actually quite surprised by it being in spec.  The spec is a bit looser than the other meters, even though it has an extra digit.

One notable thing is that when the Prema was connected it pulled all the other meter readings down by a few uV.  Thankfully I didn't have it connected overnight with the rest.  With just the Prema connected, the reading settled at 9.981940 (+17ppm) at 80s integration time.  I knew there might be some effect of load with the meters in parallel.

Because I did the above overnight tests with multiple meters in parallel, I'm am now running with just the 34465A connected and set to High Z and front terminals.  It currently says 9.981798V (+2.8ppm) with over 500 samples.  I'm interested to see how the stats compare if I leave it alone for a while.

Obviously, I don't know the uncertainty of the reference measurement, so all these single digit ppm variances are questionable.  As are both my metrology experience, and my sanity.

I also picked up some of the Emeco resistor standards.  Two wire 100R and 10k.  I'll take some measurements of them another day.

Edit to add: I left the 34465A running as the only meter connected to the new ref, and with about 220k samples at 10PLC Hi-Z, the avg was 9.981797V (+2.7ppm) with sd of 4uV and span of 30uV.

[tatel]

I got my extra Emeco YMC-05 LM399H reference in the post today.  I am at home now and only have a couple of basic 4.5 digit meters to hand.  I let it warm up for a few hours while I had dinner and a nap.

It is labelled as measured with a Hioki DM7275 (not DMM7510 this time) at 25C 60% RH, but not dated - so I don't know how fresh that reading is.  And of course we have no uncertainty info, but I believe the DM7275 should be within 20ppm when in spec.

The 10V output was labelled as 9.98177, and my basic home DMMs read either 9.972 (Owon SPM6053) or 9.987 (Aneng AN8008).
The 7V output was labelled as 7.03219, and again I get 7.025 and 7.036 - one lower and one higher.

I'll be in the office for a fair bit of the long weekend, so I'll see how it look across the 6.5d meters and compare with the other two older references (& PDVS2mini!).

Links to AE LMM399 calibrated references seem to be all dead, could you kindly give some pointer?

[Kean]

Links to AE LMM399 calibrated references seem to be all dead, could you kindly give some pointer?

An AliExpress search on "emeco lm399" should bring it up. On other sites like ebay you can try the search terms "YMC-05 LM399".

This is the one I bought and feel reasonably comfortable to recommend: https://www.aliexpress.com/item/1005007409159788.html

I have several other TE items from the AE Emeco store and they are pretty good.  There are resellers, but usually more expensive which is why I'd recommend sticking to Emeco.  There are also copies of the Emeco products that look identical externally but have different PCBs (without Emeco name on them) and inferior components.  Not sure if this LM399 ref has been copied, but their IGBT tester has been - as seen on Big Clive and noticed the differences to mine.

One example likely reseller on AE is this one, and I post it not as a recommendation (more expensive anyway) but because they got a detailed positive review that reflects my experience: https://www.aliexpress.com/item/1005009509473001.html

I will note that my previous reference from Emeco bought in Dec 2024 has appeared to drift, although I cannot find my old records so I am not sure how much (grrr!).  Anyway, I suspect they don't use aged or reclaimed parts.

PS. I just checked, and Tony Albus linked in his video review mentioned earlier to the same Emeco one I linked above.  He also got measurements very close to those on the label.