Forced Revision DMMCheck Plus

[rkeller]

Hi Everyone: I hope you all had a nice Christmas season and best wishes for a joy-filled and prosperous new year. 
I wanted to share a product update and also ask for feedback before we make our latest offering available.
Finding parts, as everyone knows, has gotten a little dicey at times.  Many of the thru-hole components we use on our DMMCheck Plus have risen in price by hundreds of percent or are not as readily available.  Even though, for example, we are only talking about resistors that used to be 0.23 cents U.S. going to $1.25, in masse, it still impacts pricing and the "do we continue doing this" equation.  Instead of getting discouraged or throwing in the towel, Doug and I have been working on an improved reference with more readily available surface mount components.  The benefit is, we can keep the price down and also go with higher tolerance parts.  Since we were redesigning, we implemented the often suggested (and thank you for all the great constructive feedback) idea of having an external power supply to keep the reference always on, and always ready without the 15-minute warm-up period.  As many of the metrology geeks (me a metrology geek wanna-be) already know, the longer you run a reference LT2021, the higher its stability.  With the forced redesign, you can keep it running and improving 24/7 or until your solar panels crack and your inverter fries... I digress.  We were able to add everything on board, and now instead of seven functions, there are 9 or 10 if you count the second reference frequency. DC volts, AC volts, DC current, AC current, Frequency 1, Freq. 1 duty cycle, Frequency 2, Freq. 2 duty cycle, Resistance, Inductance and Capacitance.
We are not making it active on the website till Jan 1, 2024.  It's visible as of five minutes ago, but the inventory is set to zero for now.  If you are interested in commenting please do so, I welcome your constructions.
You can see the offering here:
https://dmmcheckplus.com/
Attached is one of the beta units in pre-production.  The finished product and specs can be seen on the website.

[BILLPOD]

Thanks Rkeller.   I've been contemplating on getting one of these for quite a while, and looks like it's good that I did, as this new model has extra functions.  I'll be sending in my order in a week or two.

[CatalinaWOW]

I have sure enjoyed mine and am glad you are carrying on.  Don't know if I will buy another, but would happily recommend to anyone needing anything resembling this.

[bicycleguy]

Don't understand the logic of a 15Vdc power supply, especially with what looks like a 12Vdc connector.

I am really looking forward to all my low voltage stuff being USB-C and getting rid of all these weird non-standard connectors, cables and supplies.  Many USB-C power supplies will do 15V @ 2A, way more than you need and probably way more efficient and possibly cheaper.

[J-R]

FYI, direct link to product being discussed: https://dmmcheckplus.com/shop/ols/products/dmm-check-plus-fully-loaded-with-all-options-lc-board-enclosure-and-dual-frequency

Glad to see continued production and improvements.  It makes sense that SMT is far cheaper than through-hole, due to automation and the reduction of materials.

With regard to the new design, of course it's great that the components are underneath and protected and the connections are easier to reach.

[alm]

I am really looking forward to all my low voltage stuff being USB-C and getting rid of all these weird non-standard connectors, cables and supplies.  Many USB-C power supplies will do 15V @ 2A, way more than you need and probably way more efficient and possibly cheaper.

If they are using a linear regulator, then 12V might be tight for an output of 9V. Especially if there's some ripple on the power supply output.

For a reference like this I would be concerned with common mode and normal mode noise. So I'd ideally want a linear power supply. A phone charger or laptop would be the last thing I'd want to use. So I think an USB-C connector would be worse. Never mind the logic necessary to actually be compliant.

[Flux-Sucking Shunt]

Just got mine as a birthday gift middle of last year.  Very happy with it.  Now I know what to ask for this year!
All my meters (Fluke, including original 1983 77) are still spot-on.  And now I'm the cool kid at the local nerdfests. 

For a reference like this I would be concerned with common mode and normal mode noise. So I'd ideally want a linear power supply. A phone charger or laptop would be the last thing I'd want to use.

Pretty sure the supplied wallwart is a switcher, so internal filtering in it and in the DMMCheck must be fine:

[TERRA Operative]

Hmmm, this looks tempting...

Are the capacitance and inductance values measured and noted down like voltage and resistance values as part of the calibration?

[Flux-Sucking Shunt]

Assuming the procedure hasn't changed, yes, all values are measured and noted during calibration:

[rkeller]

I am really looking forward to all my low voltage stuff being USB-C and getting rid of all these weird non-standard connectors, cables and supplies.  Many USB-C power supplies will do 15V @ 2A, way more than you need and probably way more efficient and possibly cheaper.

If they are using a linear regulator, then 12V might be tight for an output of 9V. Especially if there's some ripple on the power supply output.

For a reference like this I would be concerned with common mode and normal mode noise. So I'd ideally want a linear power supply. A phone charger or laptop would be the last thing I'd want to use. So I think an USB-C connector would be worse. Never mind the logic necessary to actually be compliant.

Just unplug the reference and you are running pure DC.  The power supply provides aging and instant on and ready, eliminating the warm-up time.

[J-R]

As a test I powered my original DMMCheck Plus with a function gen and threw all kinds of evil at it aaand....nothing really.  Worst I saw was about 5uV on my K2010 depending on the evilness of the waveform on top of the 9V offset.  This isn't much in comparison to the maybe 25uV that I believe the reference is out currently (about 9 months since last cal) and which still far exceeds the spec of 0.007%.

So my opinion is the choice of AC adapter doesn't matter.

[rkeller]

FYI, direct link to product being discussed: https://dmmcheckplus.com/shop/ols/products/dmm-check-plus-fully-loaded-with-all-options-lc-board-enclosure-and-dual-frequency

Glad to see continued production and improvements.  It makes sense that SMT is far cheaper than through-hole, due to automation and the reduction of materials.

With regard to the new design, of course it's great that the components are underneath and protected and the connections are easier to reach.

We provide a 15VDC external supply that is compatible with domestic and international AC line voltages and frequencies.  As an alternative, we have connectors available for customers that prefer to provide their own 15VDC supply, I'll put a link to the connector on our web page.

[rkeller]

Hmmm, this looks tempting...

Are the capacitance and inductance values measured and noted down like voltage and resistance values as part of the calibration?

YES, Those values are noted on the as measured Calibration Certificate.  Our LCR test equipment is at Custom Cal at the moment so (expected return 1/26/2024) so we are two weeks out on orders until it returns.  The new Calibration Certificate will be posted here as soon as it arrives: https://dmmcheckplus.com/technical-information

[rkeller]

Hi J-R, you shouldn't notice any difference between shore power (plugged in) and battery power.  The incoming 15Vdc power supply voltage is filtered and regulated on board as well.  As far as we have measured there is no observable change in the reference outputs.  : )

[KungFuJosh]

Is the voltage standard only ~5V? I'd love to see a standard that covers multiple ranges including mV.

[J-R]

Is the voltage standard only ~5V? I'd love to see a standard that covers multiple ranges including mV.

C'mon man! The voltage info is in the specs!

There are at least a couple ways to tease out some additional lower AC & DC voltages from the DMMCheck Plus.

Easiest to set up and calculate is to use the current source combined with the 100 Ohm resistor to generate a precise voltage near 100mV (calculate the exact value from the calibration sheet).  This can allow you to verify the mV range on a lot of DMMs.  The current source tops out at around 500mV, so you can't do this with the 1k Ohm resistor, though.

Less than ideal but can be done is to use all four precision resistors with the 5V reference to make a voltage divider.  This drags the 5V down a little bit so you will want to measure it or your calculations will be off.  This yields individual values of approximately: 4.5mV, 45mV, 450mV, 4.5V.  When calculating the voltage drop across each individual resistor, be sure to include the DMM resistance since it will be in parallel, but only with that one resistor at a time (unless you have four DMMs!).

Some helpful calculators:
https://3roam.com/total-resistance-calculator-parallel/
https://3roam.com/voltage-divider-calculator-with-4-resistors/

Ultimately, if you want to take things very far you will need to purchase some additional equipment.  In my opinion the easiest solution is to just have a better & calibrated DMM rather than trying to have every imaginable reference.

[KungFuJosh]

Is the voltage standard only ~5V? I'd love to see a standard that covers multiple ranges including mV.

C'mon man! The voltage info is in the specs!

Ultimately, if you want to take things very far you will need to purchase some additional equipment.  In my opinion the easiest solution is to just have a better & calibrated DMM rather than trying to have every imaginable reference.

Cute unsolicited opinions. However, the point of a good reference is to verify your DMM's calibration. I have two calibrated bench DMMs that don't agree in the low mV range.

Playing games like you suggest, and using other components doesn't result in a calibrated/reliable reference, and is useless in this case.

Before I asked the question, I looked at the page OP linked, and there was literally no specs for voltage listed. In general, the information on the product page is inadequate.

The only voltage listing was on the cal sheet another user posted in this thread for a previous device version.

Have a pleasant day.

[alm]

Cute unsolicited opinions. However, the point of a good reference is to verify your DMM's calibration. I have two calibrated bench DMMs that don't agree in the low mV range.

To fully verify calibration of a DMM, you need a multi-function calibrator that covers all ranges and functions (from mV to 1000 V). This device offers a few useful voltage, currents, resistances etc to spot check a DMM. There are other devices (that would be off-topic for this thread) that offer adjustable voltage sources that may be more suitable for your particular task of adjusting the mV ranges.

Playing games like you suggest, and using other components doesn't result in a calibrated/reliable reference, and is useless in this case.

Actually most voltage references offer just one proper output and are used together with other ratio devices to cover other ranges. With good metrology practices, like a Hamon divider, you can transfer between ranges (e.g. 5 V and 500 mV) with a low uncertainty.

Before I asked the question, I looked at the page OP linked, and there was literally no specs for voltage listed. In general, the information on the product page is inadequate.

Try clicking on the "learn more" button near the top that links to this these specs.

[KungFuJosh]

Actually most voltage references offer just one proper output and are used together with other ratio devices to cover other ranges. With good metrology practices, like a Hamon divider, you can transfer between ranges (e.g. 5 V and 500 mV) with a low uncertainty.

Yes, but the important factor here is that whatever is in the circuit is permanent, and calibrated. Without the specs on paper, there's nothing to trust.

[Martin72]

This device offers a few useful voltage, currents, resistances etc to spot check a DMM.

And that's exactly why I bought the checker, which already has its purpose in its name.

[J-R]

There's no good reason to attack me.  My ribbing was clearly in jest and in my opinion high post count users shouldn't expect the gentle redirect that new members get.  There are plenty of video, written and forum reviews of the DMMCheck Plus as well as the Technical Information page that can provide tons of details about the product.  I can understand if you landed on the Shop pages there is little information provided, but a single click at the top of the page or a search engine would have solved that.  Couldn't you have just apologized for being in a hurry or whatever, and moved on?

The "games" I was playing with my DMMCheck Plus are not useless and I spent a lot of time running the calculations and actually testing it all with half a dozen DMMs before posting.  100mV is a very common DMM mV range and the 1mA current source with the 100 Ohm resistor provided solid results and in my tests were within 2uV based off of two calibrated 6.5 digit DMMs.  The voltage divider as stated is less than ideal, but everything can be calculated for and again the numbers were excellent or I wouldn't have posted the idea.  I wouldn't have any concerns doing this with a 50,000 count DMM, which is going to be a common market for the DMMCheck Plus, but it also provided some decent info for my 6.5 and 7.5 digit DMMs.

To dig a little deeper into my results, we can consider that Russ uses a calibrated 8.5 digit DMM but provides quite a bit less on the calibration sheet.  In my case, I can introduce some very basic uncertainty to the calculations and add or subtract from the LSD on the calibration sheet and get 100.170mV for the high, 100.159mV for the exact calculated value and 100.148mV for the low.  In my case I got 100.160mV on my 34461A and 100.161mV on my 34401A which is a really great result (both using high impedance mode).

[KungFuJosh]

Please don't take it personally, I meant specifically useless for my needs, not intended as an attack on you. Sorry if I wasn't clear enough there, I wasn't quite awake yet when I replied.

I understand that we each have different needs and use cases, and that was my point. For my needs, it has to be a specific device calibrated in the low mV range. I'm annoyed that I have 2 calibrated bench DMMs, and in the mV range they're far off enough that a 3rd calibrated device only agrees with one of them. The problem is, I was trying to see if the 3rd device was in spec or not, and one DMM says yes, the other says no.

The DMM that disagreed is going to be sent out for calibration, but I still need a calibrated device to test different voltage ranges, primarily low mV.

[CatalinaWOW]

Actually most voltage references offer just one proper output and are used together with other ratio devices to cover other ranges. With good metrology practices, like a Hamon divider, you can transfer between ranges (e.g. 5 V and 500 mV) with a low uncertainty.

Yes, but the important factor here is that whatever is in the circuit is permanent, and calibrated. Without the specs on paper, there's nothing to trust.

It needs to be traceable, but that doesn't mean that each and every part used be calibrated.  You can match resistors to arbitrary accuracy using bridge type measurements (It may take a lot of bootstrapping, but it can be done).  Match large enough resistors and you can bring the load within the specified capability of your voltage source and you can eventually get a low voltage reference with accuracy numbers that you know. 

This process is long and tedious, requiring great attention to detail, and side tests to assure that nothing is happening that is unpredicted or controlled.  Buying a calibration box saves most or all of this labor.  But a box that meets a wide range of voltages (mV to kV for example) is very, very pricey.  The DMM check is actually quite inexpensive for what it does.   It would be nice if it did everything, but it meets a need at a price point. 

Perhaps you would like do develop one that meets your specs and sell them at a comparable price.  I am sure you will get substantial sales, particularly after the market determines that your device performs as claimed.  Whether you will make money is much more questionable.

[KungFuJosh]

It needs to be traceable, but that doesn't mean that each and every part used be calibrated.  You can match resistors to arbitrary accuracy using bridge type measurements (It may take a lot of bootstrapping, but it can be done).  Match large enough resistors and you can bring the load within the specified capability of your voltage source and you can eventually get a low voltage reference with accuracy numbers that you know. 

This process is long and tedious, requiring great attention to detail, and side tests to assure that nothing is happening that is unpredicted or controlled.  Buying a calibration box saves most or all of this labor.  But a box that meets a wide range of voltages (mV to kV for example) is very, very pricey.  The DMM check is actually quite inexpensive for what it does.   It would be nice if it did everything, but it meets a need at a price point. 

Not each part calibrated, the values need to be calibrated. Say for example it does 2.5V, 5V, 7.5V, 10V like a lot of standards do. Each point is calibrated as a (hopefully) trustworthy value. Now add in 1V, 500mV, 250mV, 100mV 25mV or something like that with calibrated values, that's what I need.

I'm well aware I could mess around and set other arbitrary values. The issue is that I would have to trust my meters to tell the truth, and I don't own a calibrated 8.5 digit DMM. Since I don't trust my DMMs like that, it's impossible to work with a reference that's not externally calibrated in the value ranges that I need.

Perhaps you would like do develop one that meets your specs and sell them at a comparable price.  I am sure you will get substantial sales, particularly after the market determines that your device performs as claimed.  Whether you will make money is much more questionable.

I'm working on a version of the Elektor reference in another thread. My issue is that it's quite expensive (to me/for what it is), and once complete, will need to be sent out for external calibration anyway. I would be super happy if there was already a device under or around $200 that included the ranges I need and calibration certification.

I'd love to develop my own at some point, but the development alone is expensive, then I either need to suffer from TEAS and get some higher end calibrated DMMs, or send them out for 3rd party cal. Who knows. It could happen. 😉

Thanks,
Josh

[alm]

Not each part calibrated, the values need to be calibrated. Say for example it does 2.5V, 5V, 7.5V, 10V like a lot of standards do. Each point is calibrated as a (hopefully) trustworthy value. Now add in 1V, 500mV, 250mV, 100mV 25mV or something like that with calibrated values, that's what I need.

The main value in the product like the DMMCheck Plus in my opinion is the external calibration. From there you can transfer to other values yourself. If you follow the traceability chain for the lab, then probably at the top of their equipment is something like a Fluke 732B DCV standard that gives 10V (I'm not sure anyone ever uses the other outputs). All other DC voltages are derived from there using ratio techniques like the Hamon divider I linked to above. As CatalinaWOW mentioned, such a divider can be built without any external calibration using bridge techniques where the only requirement is short term stability of the components involved. In exchange for a second mortgage, Fluke will also sell you a box (autocal multi-function calibrator) that uses the same techniques to generate voltages from mV to kV from a 10V standard (and with a resistance standard also resistance and current).

Of course it is fair to say you don't want to do that work. But then I'd expect to spend a lot more than for the DMMCheck Plus. Generally in calibration you pay per point that you want measured, so if they would have to measure a few dozen points, the calibration would become a lot more expensive. The PVDS2mini (well above €200 and not currently available) might be the closest to what you want, though it may not meet your traceability requirements and the specifications like INL appear to be not very tight.