Low Cost Voltage Reference Experiment

[sync]

I'm unable to find those for less than $2. Do you have a link? It should restrict to small quantity and include shipping because unfortunately I do have to pay that.

0.61$ at TME (Europe):
http://www.tme.eu/en/details/lm329dz/reference-voltage-sources-circuits/texas-instruments/lm329dznopb/#

1$ at Digikey.
http://www.digikey.com/product-detail/en/LM329DZ%2FNOPB/LM329DZ%2FNOPB-ND/9546

[paulie]

Thanks for the links but in small quantity and with shipping they don't seem to qualify as "pennies". Cheaper on Ebay at two for $4 and change but even they are just over the limit. Purchased in 100 or more these would be pennies each but unlikely I'll have need for that many.

At this time bandgap references look more usefu, at least for the DMM transfer standard purpose. Zeners require quite a bit of support circuitry (op amps, precision resistors, etc) but the bandgaps work by themselves. Some without even a bypass cap it seems. Also 2.5v is more useful for modern ADC reference as opposed to 7v from the zeners. I do plan to test zeners though in case there is future need for the better long term stability and lower noise.

[JohnnyBerg]

I have a lot references in stock. If you want a LM385, I can send you one. Send me a PM if you want one.

[Galaxyrise]

Looks like the "zener" in LTZ1000 is not available as a real diode either.

I thought pin 3 & 4 were the buried zener on a LTZ1000.  What did you find to the contrary?

[Miroslav]

While fully appreciating attempt to build something usefull for penniew or less, just wanted to add the "high end" solution that's for most cases still acceptable ($2 in quantities, up to $5 single unit orders).

Check REF102. No external componets required (most cases)

10V ±0.0025V OUTPUT (factory laser trimmed)
VERY LOW DRIFT: 2.5ppm/°C max
EXCELLENT STABILITY: 5ppm/1000hr typ
EXCELLENT LINE REGULATION: 1ppm/V max

hackaday has a nice article posted recently

[paulie]

Check REF102. No external componets required (most cases)

Yes, there is no doubt REF102 is a good part as evidenced by it's use in top performing commercial standards like D105 from Calibratory. Unlike some others mentioned here very good initial accuracy. This may be useful for laboratory standards and top end bench lab meters which need absolute 10v but not so much in my case where calibrating friends meters or building ADC circuits are the goal. Because many units will be required cost is more important in addition to temperature and time stability.

ps. 2.5v is also more friendly for my application because most modern ADC chips can't handle the higher voltage.

[paulie]

I thought pin 3 & 4 were the buried zener on a LTZ1000.  What did you find to the contrary?

I'm not an expert but a semiconductor product engineer customer of mine thinks the LTZ die picture below hints at many more junctions than a simple zener diode. In addition he says there are other parasitic elements that are not visible. Probably not a huge amount of extra circuits like in the LM399 diagram I posted but not a simple diode like one might be led to believe by use of the term "zener".

ps. I recall the reaction these comments created in the volt-nut thread. Probably chalked up to religious fervor sorta like in church mentioning archaeologic evidence Jesus might have been married or had a brother. LOL

[paulie]

I have a lot references in stock. If you want a LM385, I can send you one. Send me a PM if you want one.

I really appreciate your offer but fortunately janaf informed me he already included some along with the buried zener samples. I also ordered a bunch from polida2008 but they probably won't arrive for a month or so. Thanks again. I'm constantly amazed at the generosity of fellow EEvblog members.

ps. I'm am collecting another set of bandgap references but there's no rush because the next test run will be 7v zeners only. 2.5v devices will have to wait for next run after because my 10x expander don't have wide voltage range.

[paulie]

On request I've added another photo to 1st post with more details on the amplifier/multiplexer board. This design was actually for a customer who was investigating battery performance. They built over a hundred PCB versions but as usual I got to keep the prototype. One of two actually since I firmly believe in redundancy and backup in case things don't go so well.

[Marco]

I found a scan someone did of the LM723 temperature controlled reference, it wasn't as smart as suggested ... it actually used a second transistor array for the voltage reference, the LM723 was a preregulator and oven controller.

This seems wasteful to me ... the suggestion of using the internal LM723 transistors as sensor and heater for the internal reference seems better, although the way the transistors are connected makes it a bit non trivial.

[JohnnyBerg]

I wouldn't use a 723 as a reference. I just looked at the specs, even a 2 cent TL431 is better

[SeanB]

I wouldn't use a 723 as a reference. I just looked at the specs, even a 2 cent TL431 is better

It was pretty good though when it came out. The competition was a zener diode, potential divider and pass transistor. Still is a good device though if used correctly.

[JohnnyBerg]

It was pretty good though when it came out. The competition was a zener diode, potential divider and pass transistor.

I remember using it as a kid, however, that was decades ago.

Still is a good device though if used correctly.

hmm ..

[Marco]

I wouldn't use a 723 as a reference. I just looked at the specs, even a 2 cent TL431 is better

There are however no other cheap references with devices on the die you can use for temperature measurement AFAICS.

[JohnnyBerg]

I wouldn't use a 723 as a reference. I just looked at the specs, even a 2 cent TL431 is better

There are however no other cheap references with devices on the die you can use for temperature measurement AFAICS.

True!
But besides tempco, long term stability also sucks!

It is very difficult to rebuild a Deux Chevaux to a Ferrari. It always will be a .. Deux Chevaux

[rdl]

The 723 voltage regulator is a classic. Designed by Bob
Widlar and first introduced in 1967, it is a flexible, easy-to-
use regulator with excellent performance.

Of course, they are describing it as a voltage regulator, not as a voltage reference.

[EJE]

Hello, Paulie
Looking for discussions regarding voltage references (for DMM, for Power supplies, etc), I stumbled in this old thread and found your series of posts quite interesting, but had some difficulties to follow your experimental findings or to fully get your test method and steps you chose to do (but for the discussions held here this would be secondary, compared to your findings).

From what I understood:
(1) You measured several lower cost Voltage References, and of different types in an oven, after a given warmup time, using a kind of multiplexed ADC - as you pictured in that post

By "low cost" I mean pennies not dollars. "Voltage reference" refers to any device that might function as a transfer standard for calibrating 5 or 6 digit voltmeters or serve as reference for DIY meter or PSU with similar resolution.
….

(2) For several pictured devices, grouped as:
   (a) IC references

Low cost bandgap devices. Most 20 cents or less each, 2 cents for the TL431.

   (b) Zener+Diode assemblies

Back to back penny zeners. Less favorable voltages than the 2.5v bandgaps but potentially more stable. …

   (c) LED and Transistor samples

LEDs and bipolar transistors. Not as promising but certainly inexpensive.

For them, you typically used that “fixed current” through a given voltage:(as ~5V) and a known resistor to measure their “output” voltage.

(3) There are data handwritten about the voltages you found, but
(4) I could not identify the influence of Temperature (like Temp. coeff.) you might have identified and measured.
(5) I also didn’t find if you measured/rechecked them for mid/long term stability at a given test condition (~ current & temperature).

After such long time, you may have tabulated these findings, or somehow summarized them on an attachment, but I couldn’t find. So, if you have done so, could please point them where I could study your findings?

Thank you,
Euthymios

P.S.
I’m an Mech.Engineer professionally and passionate about Electronics for more than 40 years, studying topics as a “serious hobbyist”.
Sorry for the “inadequate indented” format, as this is an initial post (and done on an iPad, no mouse here); I had difficulties to do a nicer formatting and compromise this post.

[magic]

FYI Paulie hasn't been active here for many years.

https://www.eevblog.com/forum/profile/?u=91790

[iMo]

@Euthymios: But there are answers, sure, better to ask in the Metrology section, there are people playing with various kinds of references..

[EJE]

Thank you for the info & tip how to check that.
It is a pity if Paulie’s data might be now inaccessible.