Recent ~$20 AD584 based precision voltage reference purchase experiences?

[eas]

These ~$20 AD584 based precision voltage references have been discussed here before, but I wondered about recent experiences with them. To start, I'll share my own.

I ordered one of these devices off ebay from 1984yht888, a seller mentioned in one of the previous threads. I'd recently purchased another item from 1984yht888, and was impressed that they had a tracking number to me in less than a day, with the item being delivered from Shenzen to Seattle in 6 days. This time, the tracking code was again provided quickly, but delivery took a still acceptable 11 days.

The item arrived packed in a padded envelope. Inside, it was further wrapped in multiple layers of bubble wrap. Inside that was a ziplock bag containing the device, which was wrapped in a small piece of fabric to protect the acrylic case.

The lithium ion battery was in place and charged enough to put the device through its paces. The unit I acquired had a calibration "certificate" on the bottom. The certificate identified the calibration device as a Agilent 34401A, the calibration date as "2015" and temp as 18C. This information, and the nominal voltages, was all computer printed. The measured voltages were written in pen.

I bought this to have a sanity check for my UniT139C DMM, and, not surprisingly given people's reports for both this voltage ref and the 139C, my readings were a good match for the readings on the calibration certificate, given the precision limits of the DMM.

Later, I had a chance to scrounge up a barrel connector that was compatible with the power socket on the unit, and used to it charge the device off USB. Given the low current drawn, and the quick ramp down, it appears that the battery was nearly fully charged when I received it.

None of what I've said so far really merits a post though, but there are a few things I thought worth noting.

In previous threads, some people reported that the battery had been punctured by through-hole leads on the bottom of the PCB. My unit had the battery fixed to the bottom of the case, and plenty of clearance between the PCB bottom and the top of the battery. I don't know if the battery is thinner, the enclosure taller, or the PCB mount higher.

On the other hand, the problem of banana plugs on test leads sticking through the bottom of the open sockets and contacting components on the PCB hasn't been addressed in any meaningful way. Given this, I doubt there has been any revision to the enclosure height or the PCB mounting position.

The main reason for my post though, is that I didn't really pay attention to what version of the AD584 was in the device I ordered until after I received it. As it turns out, the listing was for an AD584KH, and inspection of the marking on the can confirms that is what I received. Previously, these devices have had the somewhat higher-spec  AD584LH variant, and most of the ebay listings still claim to be for AD584LH. I also note that the Taobao that I think is the origin of all these devices, currently only lists the KH variant, though they have in the past offered both KH and LH versions.

So, if anyone else has ordered one of these recently, what did the listing claim? What did you actually receive, based on the markings on the can?

[JacquesBBB]

If you look for example to this ad

http://www.ebay.fr/itm/AD584LH-4-Channel-Reference-Module-Haute-Tension-de-Precision-de-4-Canaux-NEUF-/121394360069?pt=FR_YO_MaisonJardin_Bricolage_ElectroniqueComposants&hash=item1c43aaef05

it says AD584LH

But if you look to the associated picture, you can see KH on the chip.

[eas]

If you look for example to this ad

http://www.ebay.fr/itm/AD584LH-4-Channel-Reference-Module-Haute-Tension-de-Precision-de-4-Canaux-NEUF-/121394360069?pt=FR_YO_MaisonJardin_Bricolage_ElectroniqueComposants&hash=item1c43aaef05

it says AD584LH

But if you look to the associated picture, you can see KH on the chip.

Nicely spotted. Most of the other sellers I've seen are using photos off the original TaoBao listing, so the photos aren't much use in those cases.

[chrisalbertson]

Just to add one more recent data point to this topic.

I just bought one of these.  It is the type with the cut-out for a 15V battery and the set of four double jumped voltage selectors.   It came with a paper that listed measured voltages.  The paper had hand written numbers on it that I could read.  Most of the paper was in Chinese and my wife says she'd have to know "technical Chinese" to understand it.

The numbers were written by hand using ink.  It was not a photocopy.

The board has a metal can "L" version of the chip.   The chip looks as if it might be a used chip as there are minor scratches and the "shine" is gone from the metal.

The hand written numbers matched the output of my HP-3478A.   In fact, the 3478 rounded the last few digits correctly.   All except for the 2.5 volts where my HP-3478A can show more decimal places there was a disagreement between my DMM and the paper of a few hundred uV.

My HP-3478 is a recent eBay purchase and who knows when it was last calibrated.  But the fact that the DMM and this paper agree to about 5 digits tells me the paper is not fake.   

I powered the AD584 board with a low-cost adjustable linear power supply and swept the input voltage over the range allowed b the AD584 data sheet and within the limits of my HP-3478A the output voltage did not change

I paid a little under $8 with free shipping.  Well worth the place and seems to be exactly as the seller describes.

[not1xor1]

Just to add one more recent data point to this topic.

I just bought one of these.  It is the type with the cut-out for a 15V battery and the set of four double jumped voltage selectors.   It came with a paper that listed measured voltages.  The paper had hand written numbers on it that I could read.  Most of the paper was in Chinese and my wife says she'd have to know "technical Chinese" to understand it.

The numbers were written by hand using ink.  It was not a photocopy.

could you provide a link please? 

[not1xor1]

OK
I can provide some more fresh informations about these sort of cheap references, at least about the one I received a few days ago:
http://www.ebay.it/itm/111293133692

the item I got has no brand and gets charged via a mini-USB port

measurement data on the back is handwritten

despite being advised as based on an AD584KH, it has a better AD584LH produced in 1991

I charged the battery and then left the reference and my multimeters on for a couple of hours

then I got the following measurements results:

nominal value --	label value --	5Ds rounded --	Fluke 8842A --	Brymen 867S --	UNI-T 61E
2.5	2.49843	2.4984	2.4985+	2.4989	2.495
5	4.99999	5.0000	5.0000	5.0007	4.966+
7.5	7.49863	7.4986	7.4985+	7.499	7.496
10	9.99976	9.9998	9.9997+	10.001	9.998

I added "+" to mean the value is unstable e.g. varies between 9.9997 and 9.9998

The fluke 8842A is about 25 years old, I got it from ebay and have no information about its calibration history, but since the results are quite close to the values reported on the reference label I guess it is still within the specified accuracy of 30ppm as it is quite unlikely that a fabricated label provides such close results.

Of course I cannot guarantee that other references sold by the above mentioned seller provide similar results.
As you know YMMV...

[technogeeky]

Can everyone be sure (or better yet, go back and edit) to include date information for the sticker? Have there been any 2017 dated stickers in the wild?

[P90]

Can everyone be sure (or better yet, go back and edit) to include date information for the sticker? Have there been any 2017 dated stickers in the wild?

 

I bought the $20 kkmoon version from Aamazon and it had the ad584kh version, and handwritten measurements on the back of the unit, and they seemed plausible since they were within a couple of milivolts of what I measured on my Fluke.
I also bought a version with the yellow circuit board that takes the 15v battery,from eeBay, and it has a ad584lh, and includes a printed paper with measured voltages, but it's fake because I've seen youtube videos with of same unit and exact same measurements and information on the included paper. So, I would conclude that on the $3-8 voltage references, they don't bother making the measurements.

[Muxr]

I bought a AD584KH version from Alice on ebay 3 years ago (this is the actual listing):

http://www.ebay.com/itm/AD584KH-4-Channel-High-Precision-Voltage-Reference-Module-2-5V-7-5V-5V-10V-/311849349264?hash=item489bab7090:g:p1kAAOSwtZJY9MS-

I'd much rather either supply the voltage from a battery or from my own linear power supply than use the LiPo battery with a Boost converter like on the KKMon version. I usually use my E3611A which has clean power.

You can never have enough references is how I look at it.. it's just another data point. In my experience measuring this unit with a myriad of DMMs I have (6 1/2, 5 1/2 bench, and various handhelds) over the last 3 years, it's been surprisingly drift free, at least at the resolution I am checking it at.. I don't have a in cal 8 1/2 digit meter or anything 30ppm or better way to tell. My office/lab space is both temperature and humidity controlled so I measure at the same temperature/humidity.

For the money it's a useful reference, especially if you're a hobbyist and you're just looking to keep track of your DMMs drifts. I would not rely on it for absolute reference for calibration however, as others have mentioned I am not sure the numbers that come with these are actual values. Also they don't list the temperature and humidity at which it was measured and the packaging used can upset the reference in shipment. But once you have it, like I said it's a pretty good data point and a DMM sanity check.

I also have the DMM Check plus, which also has a 1mA constant current source and and a few precision resistors that I also use to cross check my DMMs. I am currently in process of building a few of my own references as well.

[P90]

mine said 20° on the sticker.

[not1xor1]

You can never have enough references is how I look at it.. it's just another data point. In my experience measuring this unit with a myriad of DMMs I have (6 1/2, 5 1/2 bench, and various handhelds) over the last 3 years, it's been surprisingly drift free, at least at the resolution I am checking it at.. I don't have a in cal 8 1/2 digit meter or anything 30ppm or better way to tell. My office/lab space is both temperature and humidity controlled so I measure at the same temperature/humidity.

For the money it's a useful reference, especially if you're a hobbyist and you're just looking to keep track of your DMMs drifts. I would not rely on it for absolute reference for calibration however, as others have mentioned I am not sure the numbers that come with these are actual values. Also they don't list the temperature and humidity at which it was measured and the packaging used can upset the reference in shipment. But once you have it, like I said it's a pretty good data point and a DMM sanity check.

I also have the DMM Check plus, which also has a 1mA constant current source and and a few precision resistors that I also use to cross check my DMMs. I am currently in process of building a few of my own references as well.

regarding long term drift that should be quite low as the references are usually old savaged parts (mine is dated '91)
of course we do not know anything about real usage history so we can just guess 

regarding drift caused by humidity it should be very low or just none as the parts are in metal case

so there is just the temperature drift which may vary quite a lot among the different parts and labs where they are used...
in any case, even in the worst case, the handwritten label references should be OK to check most 3/4 digits multimeters

but you can always get a rough idea of temperature stability by putting the reference in a polyester foam box with an incandescent lamp (not so powerful to melt the box  ) to warm it to 10-15°C more than ambient temperature for a few hours and then measure the voltage delta...

I'll do that with my unit  and see how much I've been lucky with my LH IC

[P90]

I bought another one of these on ebay just to check out. It comes with the same printed "calibration" paper as the others seen on YouTube. The numbers are all the same, they just print the same BS paper!
second picture is from YouTube video of someone with same paper!   

[technogeeky]

I bought another one of these on ebay just to check out. It comes with the same printed "calibration" paper as the others seen on YouTube. The numbers are all the same, they just print the same BS paper!
second picture is from YouTube video of someone with same paper!   

That's a shame. We should compile a database of these devices or whatever, and warn people that the calibration on those is fake.

[P90]

I bought another one of these on ebay just to check out. It comes with the same printed "calibration" paper as the others seen on YouTube. The numbers are all the same, they just print the same BS paper!
second picture is from YouTube video of someone with same paper!   

That's a shame. We should compile a database of these devices or whatever, and warn people that the calibration on those is fake.

at this point I would hazard to guess that if the measurements are printed, they're most likely fake, but having said that, just because they're scribbled on a piece of paper doesn't necessarily make it legit.