Scullcom Hobby Electronics #8 - Precision Voltage Reference Modules

[@borayurt66]

For the last one, they actually wrote Red (Top), Black (Bottom). So, when you turn it around and keep the red connectors at top, left and right becomes apparent. Funny way to indicate input and output ports but nevertheless it should work. :D I was looking at these voltage reference kits from China for some time now, I wasn't sure of their accuracy, your video really helped to decide. Thank you for taking the time and doing this very informative review.

[@MyNameIsNobody57]

Great review. The second unit @17:20 was calibrated on the 49th day of November, 2013!!

[@scullcomhobbyelectronics1702]

Thanks. I see your point. Could be a typo or could they mean week 49 not sure. So I would just use November 2013. Regards, Louis

[@boggisthecat]

This is a useful review of these devices (I have one of the second type, with some minor differences). A couple of important things not noted: the ‘calibration data’ accuracy is dependent on ambient temperature and an adequate warm-up time; all commercial instruments are specified based on a Gaussian sampling scheme. With my unit around half a degree lower than the ambient temperature on the label, it matched our laboratory 3458A multimeter fairly well. (That multimeter is traceable to international standards, and the uncertainty is considerably lower than the manufacturer’s specifications.) The data does not give uncertainties, and it isn’t stated what instrument was used to measure ambient temperature. If you expect to require an accurate output down to the second to last digit then you’ll have to verify these units somehow - the last digit usability is unrealistic. If you only need an accurate output down to one millivolt then these should be fine to use as they are (indefinitely).

[@scullcomhobbyelectronics1702]

Thanks for your comments and info. Regards, Louis

[@ConsertandoTudo]

Great review, i'm building a full multimeter with arduino, that will be of great help, for many reasons, but most of all because i do not have any precision equipament to mensure resistors, and calibrate the voltage divider...

[@pomonabill220]

I know this is an older video, but I was a little concerned about the second reference with the DC boost converter. I would think that since this is supposed to be a precision reference, you would not want a noisy DC converter close to the reference. It seems like noise is not a problem though, unless the DVM is ignoring any switching noise. That is my only concern. Thank You for posting!

[@ivolol]

DVM is probably doing a decent amount of averaging, which mostly negates the noise. It only updates its reading a couple times a second. If you needed a continuous, stable voltage, then the noise from the switching might make its appearance.

[@freeelectron8261]

Very useful demo and explanation. Maybe the cheaper V refs could be improved by putting them in a temperature stabilized environment ? Arduino nano with temp probe and heater resistor in a small case.

[@michaelkaliski7651]

It does say red at the top and black at the bottom, then specifies left and right, if you read the back label properly.

[@Markus2801A]

Thank you very much for this very informative and detailed explanation! You included everything from the basics, input Voltage, schematic, etc. it is awesome how very well this review is done!

[@techobsessed1]

No reason to think the calibration is off on your Keithley. It has 24 hour accuracy of 15ppm of reading + 4ppm of range. That works out to uncertainty of 154uV for 10v, 79uV for 5v and 41.4uV at 2.5V. 1 year uncertainties are ~ roughly double. I didn't bother looking up the accuracy on the 34401A, but I expect its similar to the Keithley 2000. So, all your readings are within the combined uncertainties of the two meters For what it is worth, the two AD584 references you show originate from the same designer/maker. I think I even found gerbers for more involved version at some point. the We don't know how recently their 34401A was calibrated, but they tend to stay within their one year uncertainties for years, unless something has gone awry. You'd think though, checking dozens and dozens of Vrefs over years, they'd notice major drift.

[@scullcomhobbyelectronics1702]

Tech Obsessed Thanks for your comments.

[@divewizard]

there is no reason to believe accuracy of the "calibration certificate."

[@VJrobby]

Very interesting - thanks :) .... but some serious inflation going on now with No.3 ....... price up to $227 !! Also it seems No.2 are now using the the K version of the chip, but still a great bargain for most uses.

[@ElmerFuddGun]

Where to start! First one with the sliding switch: seeing that capacitor bodged on the back says a lot. No case - easy to damage if some pins come in contact with metal or tools. A case would cost a fraction of what the actual reference cost in the can, especially in China. Short power wires where a power jack would have been better. Second one with the nice case: it is the best of the lot. Scratching the part number off the control IC is lame as it isn't exactly complicated to select the voltage and turn on some LEDs. But using a switching DC to DC converter where you want as clean of a reference as possible isn't a good idea IMHO. Third one in black case: is a peace of crap. Wouldn't trust it and would be screaming at eBay that I want my money back! No labelling on the from. Made on a proto board!! Fine if the guy selling it says it is a one off that he made at home and says so BUT if it is claiming to be a professional grade (and at that price) it is a total scam that some might more likely expect from China than the USA. Serial number 36 so it looks like they've sold some before yours yet it is on a proto board! It doesn't show what the actual claimed voltage is just some made up accuracy. What is the accuracy of the resistors (including the pot) over time and temperature? Total crap. Thanks for the informative video though! I appreciate it. Might be time for me to design and build my own voltage reference! ;-)

[@scullcomhobbyelectronics1702]

+ElmerFuddGun Thanks for your comments. If you check my You Tube Channel you will see I have already designed and built a number of voltage reference sources which you may find of interest.

[@DeeegerD]

Just ordered one from eBay ($8 Cdn) - it's similar to the first but takes a 15 volt battery which costs more than the unit itself. Also has a barrel plug for power input so the battery is not necessary, although likely a cleaner input. This unit has the AD854L component which is the main reason I purchased it as I'm not in love with module layout. Should satisfy my requirements though.

[@scullcomhobbyelectronics1702]

+Digger D Thanks for the comments.

[@nickpelov]

The last one had text "+red (top) - black(bottom)". So you put the RED terminals on top and then the left is input. Chinese will rarely give you any manual, but even then it's not that obvious :)

[@waswestkan]

The Ober/November issue of Nuts and Volts has an article on voltage Ssandards. The author showed his home brew version that uses this IC. I though to my self I can build this. I can however when I went to order what parts that I didn't have is my salvage boxes, I decide to order a ready to unit. Because it was shown in the article I ordered the second unit scullcom reviewed here. The only complaint about it is unlike the machine printed “calibration” label shown her the label on mine had hand written values that are barely decipherable. Not sure how scullcom went about determining what pair of banana plugs where what, it was clear to me how to go about that, and seems like I would have been correct Regardless what one thinks of the construction, the third unit had the better design. Simple straightforward use. In the event one needs to calibrate lower range, most wont, it's a simple manner to create an outboard circuit to to that. On e caution though, the magazine article seems to indicate a load greater than 10ma place on the IC could damage it. Best to use high impedance meters. Unless they are VTOMs or FETVOMs, at these low voltage levels analog meter will not be high impudence. Being a DMM is no indication of high impedance. Going beyond the scope of this article,if your are working with electronic hardware anywhere, you should be using high impedance meter I hope this comment is informative. I will in tame homebrew this circuit. In the event I can purchase a Weston cell for the right price, I'll gut it it, to place the circuit inside that. Going for the far out factor of course.

[@PurityVendetta]

Thank you for a very interesting and informative video. Very useful.

[@jwrtiger]

As you mentioned part of the price of the unit is to cover the calibration. I was surprised how poorly constructed the last reference module was in your video.

[@AB1Vampire]

Thank you. I need an inexpensive standard to calibrate my meters.

[@audiotecnologia4084]

the control logic ic could be using a CD4017, Hugs from Brazil ....

[@scullcomhobbyelectronics1702]

Thanks for the comment.

[@elboa8]

The last one, not even a PCB, and no external labeling. What a scam!

[@dirtrider88]

27:56 how on earth did you come to that conclusion? would have been nice if you explained it.

[@yusdiy]

The third one was built point to point style for better tone and money transfer.

[@benadams6332]

Have the battery powered Chinese one, it works well and is a great deal with the included certificate.

[@mnovo]

I believe the 1N4148 is to lift the Vref output above the input noise coming into the circuit. Or maybe that is a lucky added feature...

[@scullcomhobbyelectronics1702]

+Mike Novo Thanks for your comments. I do not think that was their original intention but as you say its a lucky added feature!! I think they were more worried about reverse battery connection.

[@leonerduk]

Some good design insides and teardowns here, but I was hoping for more of a review of these units. What were your opinions on them, especially considering the prices? The middle of the three for example uses a simple cheap boost-mode SMPS to step up the 3.7V battery to what I'd imagine is around 15V to power the reference. What effect does that have on the output stability, for example? Does that switching noise appear on the output? Does it matter?

[@scullcomhobbyelectronics1702]

+Paul “LeoNerd” Evans Thanks for your comments. The first two where both well built and there accuracy was as per the calibration reading which were sent with the units. The last one in the black box came from a guy in the USA and although the build quality was poor this unit was the most accurate as a reference standard. I think the key to accuracy is they type if voltage reference IC used.

[@leonerduk]

+Scullcom Hobby Electronics Interesting. So you don't find that the switching PSU in the battery-based unit upsets the output at all? That was the one part of its design that put me off the idea

[@scullcomhobbyelectronics1702]

+Paul “LeoNerd” Evans No there was not a problem with the unit with the switching PSU it did not effect the output. That particular unit worked very well. In fact it was the best build of the 3 I reviewed and was good value for the money.

[@leonerduk]

+Scullcom Hobby Electronics Ah I see. I was mostly looking for advice on whether I should get one, so it sounds like the answer may be "yes". Thanks

[@waswestkan]

For for less than $20 US including shipping I wasn't expecting a $200 construction and packaging quality. I have the clear plastic enclosure model here on my bench. I first learned of this IC in a nuts and volts magazine article. I was about to chase the parts to construct my own.Howe ever I notice the Author's artarticalcal was base on duplicating what the product that in a phot within the article I purchased. The author didn't say anything, one way or another about the commercial item I hav think meters with less than 10 Meg ohm impedance, may load the IC to degree, where it can't give a decent measure

[@BeetleJuice1980]

some sellers say that their units are calibrated with an agilent. do you think that the unit 1 and 2 can be calibrated , cause based on your schematics , they cant (unit 3 can be calibrated of course).

[@scullcomhobbyelectronics1702]

The first two rely on the accuracy of the I.C. As that particular I.C. provides 4 reference voltages. What the sellers does is check the accuracy of the reference voltages and then prints the actual voltage measured with a precision calibrated meter on a test sheet attached to the unit. In the video I do point this out.

[@BeetleJuice1980]

You are correct! Some sellers dont provide the written notes though.maybe we can trust your seller to be sure. Thank you!

[@zx8401ztv]

The middle one does look much better made, and very ideal for checking cheap multimeters for basic accuracy, they have put more work into the product. the last and most expensive one just looked rather diy and dispite its accuracy i would give it a wide birth. I would trust your keithley more than those references, keithley is quality design :-)

[@scullcomhobbyelectronics1702]

zx8401ztv Yes I was surprised at the quality of the most expensive one. Soldering was not to good!!

[@zx8401ztv]

Scullcom Hobby Electronics Hope you dont mind comments on your earlier videos, some people get annoyed with revisiting them again.

[@scullcomhobbyelectronics1702]

zx8401ztv No problems.

[@SirMo]

Great review, thank you! Wow that last unit, what a bunch of snake oil!

[@scullcomhobbyelectronics1702]

Yes agree the construction and soldering leaves a lot to be desired!

[@reggiebacci]

I always wonder - when talking about a few 100000ths of a volt - surely things like test lead and meter input resistances come into play? I know these things are normally negligable but at these tolerances they must start to have an effect.

[@scullcomhobbyelectronics1702]

+reggiebacci Thanks for your comments. Yes you are correct. In professional equipment the resistance of leads are usually cancelled out with additional circuitry.

[@michaeliv6064]

+reggiebacci For this particular case, not really, resistance of the leads is probably 0.1Ohm and impedance of the Keithley is somewhere around 10MOhm, meaning that the voltage drop across the leads is 0.000`000`1V - Under the Keithley's capabilities.

[@vernonbosshard9317]

What would you use one for?

[@tablatronix]

What kind of equipment can references be salvaged from? Any commercial or consumer electronics use them ?

[@scullcomhobbyelectronics1702]

You will find precision reference voltage devices in precision bench multimeters. You will also find the used with equipment that uses DAC or ADC devices.

[@BellA-mf1ti]

which way is right and wiich is left ? that's a classic desgn.

[@warplanner8852]

Accuracy or quality of test leads? That surely will affect/explain the difference between the vendor's readings and your DMM.

[@ConsertandoTudo]

also the temperature diference, from Scullcom's lab and factory tested temperature... but this leads, damn, any noise from power transformer outside, on street, may induce that error.

[@GeorgeGraves]

I have the middle one (clear case) - the PBC is really interesting. Why did they layout the can with holes around it like they did?

[@scullcomhobbyelectronics1702]

They look like isolation slots used to guard against interference or leakage from other parts of the PCB. The reference IC is very close to the DC to DC boost converter components so that may be a reason. The other point to make is that the PCB is a fairly tight fit in the plastic case so it could also help with air release when pushing the PCB in to the case!

[@GeorgeGraves]

Scullcom Hobby Electronics Yea, I was surprised with a boost circuit in there. You'd think that would generate some noise, even if it was was on a separate, but near by PCB. An interesting follow up video might be to look at the output on the scope in AC couple mode, with powering the other two by battery or very smooth linear supply. Thanks for the great video! Thumbs up and subscribed!

[@techobsessed1]

Scullcom Hobby Electronics My guess is thermal isolation.

[@scullcomhobbyelectronics1702]

Maybe for thermal isolation but nothing in the module gets hot!

[@twjonckheere]

GeorgeGraves Maybe for stress isolation? The changing temperature of the reference won't stress the board as much and cause possible variances. I remember Dave (EEVBlog) mentioning that for one of his recent tear downs.

[@kaybhee6]

noted u need to stick in clean 15 vdc as vcc

[@Ziplock9000]

Does anyone have a link to the first one on UK eBay please and a comparable resistance reference board too please.

[@scullcomhobbyelectronics1702]

Try the following eBay link: www.ebay.co.uk/itm/Precision-Voltage-Reference-Module-4-Channel-AD584kH-For-Desk-top-Meter-Probes-/141731489866?hash=item20ffdae84a:g:zWEAAOSw3ydVuZsD

[@Ziplock9000]

Thanks mate

[@-yeme-]

if you search ebay for "AD584" and arrange for lowest price there are a couple of even lower cost variants. they come with fake test result sheets but the outputs are within spec for the chips they use and are perfectly accurate enough for most hobbyists

[@pulesjet]

That first PCB was well within the tolerances of the IC.

[@scullcomhobbyelectronics1702]

Yes it did work OK within tolerance.

[@Yuēhàn24]

The battery was marked Lithium-ion.

[@michaeliv6064]

The LT1021C is not nearly as good as the REF102C -- It's actually 20ppm/C max vs 2.5ppm/c ( search "LT1021C-10/LT1021D-10" in the datasheet ). Also its long term drift is 15ppm/1000hr vs 5ppm/1000hr. Line / load regulation is also worse. Sad that they would go for a $3 cheaper core part on an item that costs $60. Also the 6ppm claimed accuracy is questionable since their meter is only 8ppm accurate.

[@scullcomhobbyelectronics1702]

+Michael Iv Yes agree. I was surprised my self to see the LT1021C chip in the unit since I was expecting a REF102C as the seller advertised the module. Clearly, the seller did it to reduce his costs but at the same time did not adviser the buyer that he had changed it.

[@arthurharrison1345]

Odd that the polarity protection diode in the third unit is in the negative path.

[@scullcomhobbyelectronics1702]

It makes no difference. You can add a reverse supply protection diode in either path.

[@arthurharrison1345]

True, but it seems it ruins some of the versatility of the device, should it be used for something other than a meter calibrator. For example, if it were to be used as a voltage reference in a system, where the negative point on the regulator would usually be at ground potential. However, in such an instance, you could simply bridge the two black posts.

[@scullcomhobbyelectronics1702]

Good point. I think the supplier of this unit only intended this to be used for meter calibration check.

[@1pcfred]

That last reference was pretty sketchy being built on ring board. I mean break down and get boards made up for crying out loud.

[@scullcomhobbyelectronics1702]

Yes agree the last one was poorly built for the price.

[@1pcfred]

+Scullcom Hobby Electronics I have point to point wired up some circuits. Not only can it be unattractive, but it also tends to be a lot of work compared to assembling a printed circuit board. Perfboard certainly has its place, but definitely that place is not in a commercial product.

[@Yuēhàn24]

can someone please tell me what equipment needs a voltage to be within .00001 volt?

[@scullcomhobbyelectronics1702]

Usually required for checking the calibration/accuracy of 6½ digit bench multimeters. Regards, Louis

[@felixcat4346]

Funny conclusion: "it was done with Fluke calibration equipment, so I would imagine....". Boot, boot why?

[@muppetpaster]

And keep room temp in mind

[@saadaldossri5348]

Thanks

[@muppetpaster]

First one available for 6 euro

[@hglrr]

Too long an complicated video for me. Any conclusion? Is any of these 3 items worth it? Any good? Can we trust any of them? If you can answer that, please add it in your comments below video. Thanks!

[@pulesjet]

No Audio ? Fix Please.