Author Topic: Compact LT2400 Voltmeter  (Read 6684 times)

0 Members and 1 Guest are viewing this topic.

Offline hillflyerTopic starter

  • Contributor
  • Posts: 21
  • Country: de
Compact LT2400 Voltmeter
« on: September 18, 2016, 09:46:47 pm »
Hello everyone,
I am building a small 5 to 6 digit Voltmeter, based on the LT2400 and a Max6350 (DIP) reference.
I have run the reference for a few weeks (~1000 hrs) for aging and have added a  NTC tempco
compensation. This has reduced the 1ppm tempco to about 0.1ppm for temperatures from 15 to 25°C.
I am powering the LT2400 from the ref. and the reference from a LT1763.
So far no additional input circuitry, I can measure 0-5V DC only, which is fine for me.
The idea is to use this as sort of a portable voltage reference with the added 'bonus' of a high resolution
(and hopefully good accuracy) voltmeter.
It runs of 3x1000 mAh Lipo cells (good for 20 hours of operation.)
I have 2 questions:
Should I add a smoothing capacitor on the input pin?
So far it only has a 1k resistor as input protection.
I don't fully understand page 23 of the datasheet about dynamic input currents.
And what would you do about shielding and grounding around the LT2400 input?
Thanks for any hints!
Hannes

 

Offline EmmanuelFaure

  • Regular Contributor
  • *
  • Posts: 154
  • Country: 00
Re: Compact LT2400 Voltmeter
« Reply #1 on: September 18, 2016, 10:04:47 pm »
Wash your pcb :)
 
The following users thanked this post: TiN

Offline JS

  • Frequent Contributor
  • **
  • Posts: 947
  • Country: ar
Re: Compact LT2400 Voltmeter
« Reply #2 on: September 19, 2016, 01:00:42 am »
  I've built a 4 1/2 digit converter, going your route but one step at a time, I didn't saw the point starting a 5 1/2 without having the previous step... did missed the previous step, seems like grabbing a µC and a decent voltage reference and call it a day. I still had a few issues with it and never had the time to work on it or testing it further since I made it a few month ago.

  For what I understand in that page what's trying to say is the higher the input capacitance and the source resistance, the higher it takes to settle, and what error would that make. If I read correctly, it's a single cycle converter, so if you multiplex the input it would do a valid conversion for the next sample. If you add more capacitance, that cap will take some time to charge, and what error that introduces. It doesn't mention the time which makes things confusing, this is what I understand by the settling time mention on the beginning of the page, reading the last to bits of the page 22 makes it more clear I guess.

 Adding more capacitance would make the high frequency normal mode rejection greater, just filtering the noise, but it also would make the input of the converter take longer to settle. If you are interested that each sample is accurate you should minimize that settling time, otherwise the sample will be highly influenced by the signal before. I was making a mental exercise yesterday regarding of something similar, for this application (any measurement) longer time constants makes them have lower noise and could give better precision but if it takes too long to settle it becomes a problem. Having a smart timing for this would make a lot of sense, a fast settling when big changes occur, mid timing when smaller changes occur and long settling when the input remains in a small window. Having a µC makes sense to implement this in the digital domain, for that you should be aware of the aliasing which should be taken care before the conversion (in analog) and then just make what makes sense inside the µC.

JS
If I don't know how it works, I prefer not to turn it on.
 

Offline hillflyerTopic starter

  • Contributor
  • Posts: 21
  • Country: de
Re: Compact LT2400 Voltmeter
« Reply #3 on: September 19, 2016, 05:57:24 am »
oops, yes washing the ADC board is a good idea...
My intended measurements would be static or very slowly changing DC voltages.
I have about 2 uV noise, (std. dev.) which I tackle with some averaging.
« Last Edit: September 19, 2016, 09:41:14 am by hillflyer »
 

Offline JS

  • Frequent Contributor
  • **
  • Posts: 947
  • Country: ar
Re: Compact LT2400 Voltmeter
« Reply #4 on: September 19, 2016, 06:45:50 pm »
  I was talking about the noise in the source signal, not in the converter. If you have to measure a steady DC source but it's HF noise becomes a problem and you can wait a few seconds to get an acceptable reading you could add capacitance at the input of your converter, in any other case you probably shouldn't. Also you need to be careful with the leakage of such capacitance, since could introduce error when the source/series input resistance is high enough. With such a high sampling freq you probably will be displaying a few values per second. Having a 1ms time constant there won't be a problem, since in 20ms you'll be settled to do the next conversion. Then, you could ignore those few initial samples and start averaging after that. That way you won't have any correlation between samples. Maybe then you want an even steadier reading, you could add some correlation adding a 20s pole if you please but just using filteredSignal=(1-a)*filteredSignal+a*lastSample. I think 1/a is about the number of samples you want your filter to be working on.

I did something similar with a thermometer, 0.01ºC with consistent readings, well showing the trends, of course not that exactitude, but that precision was still usable, put your hand a bit closer and you'll see the numbers rise up just a little bit, but it would take quite a few seconds to end settling when a sudden change was presented.

JS
If I don't know how it works, I prefer not to turn it on.
 

Offline branadic

  • Super Contributor
  • ***
  • Posts: 2378
  • Country: de
  • Sounds like noise
Computers exist to solve problems that we wouldn't have without them. AI exists to answer questions, we wouldn't ask without it.
 

Offline hillflyerTopic starter

  • Contributor
  • Posts: 21
  • Country: de
Re: Compact LT2400 Voltmeter
« Reply #6 on: September 19, 2016, 09:09:42 pm »
@Emmanuel: I washed the board, very embarrassing...  ::)
Thank you, JS and branadic for your input!
@JS : I will try some measurement with vs without input smoothing cap. Oversampling helps a lot too.
@branadic: What you built is very very helpful and a lot along the lines to what I am doing . (Just much more refined!)
I have used 220 Ohm resistors in the SPI lines to the uC, maybe going up to 4.7k and 22k isolates the ADC more from the uC digital noise.
I need to study your design more.
I have included some serial output of measuring a 2.5V reference. So far both the meter and the ext ref are uncalibrated. :(

 

Offline EmmanuelFaure

  • Regular Contributor
  • *
  • Posts: 154
  • Country: 00
Re: Compact LT2400 Voltmeter
« Reply #7 on: September 20, 2016, 12:13:22 am »
With a very simple circuit I've already achieved 1.5µVrms input noise, as specified in the datasheet. Do you?

Advice : Plot your voltage graph with y axis in microvolts.
 

Offline hillflyerTopic starter

  • Contributor
  • Posts: 21
  • Country: de
Re: Compact LT2400 Voltmeter
« Reply #8 on: September 20, 2016, 08:57:09 am »
Emmanuel, how did you get to 1,5uV?
Currently I am at 3 uV , that is ADC and voltage source noise combined.
I just tested it with shorted input: now I am much closer to your figure.
(See graph of standard deviation in uV, time axis in seconds)
 

Offline rob77

  • Super Contributor
  • ***
  • Posts: 2085
  • Country: sk
Re: Compact LT2400 Voltmeter
« Reply #9 on: September 20, 2016, 09:46:24 am »
i would probably add an input buffer to increase input impedance and provide some level of protection to the ADC.

btw.. i'm working on a similar project but aiming only at 20bits usable output (hopefully :D ) with 4.096V reference.

here is my 1st version proof of concept board:
it has LTC2400 + LTC2050 buffer , MAX6241 reference and LM7705 to create a -230mV negative rail for the input buffer opamp.

 

Offline hillflyerTopic starter

  • Contributor
  • Posts: 21
  • Country: de
Re: Compact LT2400 Voltmeter
« Reply #10 on: September 20, 2016, 11:28:06 am »
That board looks very good!
As far as I can tell you have a resistor and 2 diodes for input protection,
and an RC Lowpass combination between OP and ADC. What are the values?
Or could you maybe post the schematic?
Do you worry about getting noise from the charge pump into the signal?
Are those inductors around the 78L05's?
« Last Edit: September 20, 2016, 12:09:42 pm by hillflyer »
 

Offline rob77

  • Super Contributor
  • ***
  • Posts: 2085
  • Country: sk
Re: Compact LT2400 Voltmeter
« Reply #11 on: September 20, 2016, 11:57:37 am »
That board looks very good!
As far as I can tell you have a resistor and 2 diodes for input protection,
and an RC Lowpass combination between OP and ADC. What are the values?
Or could you maybe post the schematic?
Do you worry about getting noise from the charge pump into the signal?

thanks ;)

yes you're correct about the protection and lowpass, the values are more or less eye-balled but generally should be ok : the input protection resistor is 1k and the RC lowpass is 1k2 + 1uF (not planning to measure anything fast ;) ).

regarding the noise form the charge pump: the worst case PSRR of the LTC2050 is 115dB but typycal 130dB , the ripple of the -0.23V rail is 5mV max as per datasheet. so the influence on the signal will be literally nonexistent.

the  board is powered from +12V (the reference needs at least 8V) and each component has it's own 5V linear regulator (so the charge pump can't influence the buffer or ADC through the supply rail).

but as mentioned is the 1st version, basically a proof of concept. it works, but haven't had the time to do some intensive testing yet. but definitely planning to build another one (cheaper one) with MCP3550 ADC (22bit delta-sigma) and AD8628 buffer opamp.
 

Offline Alex Nikitin

  • Super Contributor
  • ***
  • Posts: 1149
  • Country: gb
  • Femtoampnut and Tapehead.
    • A.N.T. Audio
Re: Compact LT2400 Voltmeter
« Reply #12 on: September 20, 2016, 12:21:55 pm »
regarding the noise form the charge pump: the worst case PSRR of the LTC2050 is 115dB but typycal 130dB , the ripple of the -0.23V rail is 5mV max as per datasheet. so the influence on the signal will be literally nonexistent.

Not quite correct. The ripple frequency from the charge pump is ~90kHz, and -PSRR of LTC2050 at that frequency is 65-70dB only .

Cheers

Alex
 

Offline hillflyerTopic starter

  • Contributor
  • Posts: 21
  • Country: de
Re: Compact LT2400 Voltmeter
« Reply #13 on: September 20, 2016, 12:58:08 pm »
I have noticed several ppm change in the Max6350 (simlar to your Max6241) output when there is a variation in the supply voltage.
That's why I put in an extra regulator in front of my ref. I used a LT1763 (20uV noise on output), because the 78 series ae quite noisy by comparison.
 

Offline rob77

  • Super Contributor
  • ***
  • Posts: 2085
  • Country: sk
Re: Compact LT2400 Voltmeter
« Reply #14 on: September 20, 2016, 02:25:02 pm »
regarding the noise form the charge pump: the worst case PSRR of the LTC2050 is 115dB but typycal 130dB , the ripple of the -0.23V rail is 5mV max as per datasheet. so the influence on the signal will be literally nonexistent.

Not quite correct. The ripple frequency from the charge pump is ~90kHz, and -PSRR of LTC2050 at that frequency is 65-70dB only .

Cheers

Alex

even 65dB PSRR is enough , 5mV atenuated by 65dB is 2,8uV , so still acceptable for 20bit.
furthermore there is a ferrite bead + capacitor on that rail , so the ripple will be less than 5mV at the opamp.
« Last Edit: September 20, 2016, 02:27:20 pm by rob77 »
 

Offline Andreas

  • Super Contributor
  • ***
  • Posts: 3221
  • Country: de
Re: Compact LT2400 Voltmeter
« Reply #15 on: September 20, 2016, 07:48:29 pm »

I am building a small 5 to 6 digit Voltmeter, based on the LT2400 and a Max6350 (DIP) reference.
I have run the reference for a few weeks (~1000 hrs) for aging and have added a  NTC tempco
compensation. This has reduced the 1ppm tempco to about 0.1ppm for temperatures from 15 to 25°C.
...
So far no additional input circuitry, I can measure 0-5V DC only, which is fine for me.

Should I add a smoothing capacitor on the input pin?
So far it only has a 1k resistor as input protection.


you can easily extend the voltage range to 10V with a LTC1043 2:1 divider with good accuracy.

see also first picture here:
https://www.eevblog.com/forum/metrology/t-c-hysteresis-measurements-on-brand-new-lt1027dcls8-5-voltage-reference/msg968038/#msg968038

Is the tempco compensation on the trim input of the reference or do you calculate out the T.C. mathematically?

My MAX6350 samples are drifting around 10 ppm/year without decreasing the drift over time unlike other (hermetically) references.

The smothing cap depends a bit on your buffer amplifier on the input of the LTC2400.
I am using generally 820R and around 680-820 pF (for LTC1050) to obtain minimum noise and without compromizing linearity.

with best regards

Andreas


 

Offline Andreas

  • Super Contributor
  • ***
  • Posts: 3221
  • Country: de
Re: Compact LT2400 Voltmeter
« Reply #16 on: September 20, 2016, 08:04:23 pm »

regarding the noise form the charge pump: the worst case PSRR of the LTC2050 is 115dB but typycal 130dB , the ripple of the -0.23V rail is 5mV max as per datasheet. so the influence on the signal will be literally nonexistent.


Hello,

I am also thinking about to save my negative battery supply. So this cirquit is very interesting.

What counts is what you see on the output of the buffer: have you measured this?
Is there a influence on 1/f or broadband noise if you exchange the charge pump by a battery?
Is there a drift in the ADC measurement value.

With best regards

Andreas
 

Offline hillflyerTopic starter

  • Contributor
  • Posts: 21
  • Country: de
Re: Compact LT2400 Voltmeter
« Reply #17 on: September 22, 2016, 06:03:51 am »
@Andreas,
the TC trim is via the Trim pin on the Max6350. It is a linearized NTC resistor network.
I am quite pleased that it reduced the TC by factor ~ 10.
The Max6350 is 2001 Vintage, it was mostly powered off, but I have preaged it ~1000 hours at 50°C powered on.
That was 2 years ago. Since it lives in my basement workshop, I get quite repeatable results.
Thank you for pointing out your LT2400 setup to me. I had missed that till now.
Good hint with the LT1043 divider, I had shied away from trying sth with fancy resistive dividers.
I'll post a pic of my compensation circuit.
 

Offline hillflyerTopic starter

  • Contributor
  • Posts: 21
  • Country: de
Re: Compact LT2400 Voltmeter
« Reply #18 on: September 22, 2016, 06:35:59 am »
Here is the Max6350 trim circuit.
It needed 1,7mV tc compensation @ trim pin for 10°-30°C temp difference.
 

Offline hillflyerTopic starter

  • Contributor
  • Posts: 21
  • Country: de
Re: Compact LT2400 Voltmeter
« Reply #19 on: September 22, 2016, 01:27:57 pm »
Is the 10ppm/year drift for the Max6350 based on a device that is powered the whole time?
At what temperature?
Hannes
 

Offline Andreas

  • Super Contributor
  • ***
  • Posts: 3221
  • Country: de
Re: Compact LT2400 Voltmeter
« Reply #20 on: September 23, 2016, 06:43:46 pm »
Hello,

yes the 2 MAX6350 are powered 24/7
Temperature is near room temperature (+ some degrees for self heating within a styrofoam box).
During measurement the references are heated up to 50 deg C for 1 hr per day to get constant conditions.

X-Axis: day from start measurement
Y-Axis: drift in ppm measured with ADC13
(ADC13 drift against LTZ#2 is also given for reference)

with best regards

Andreas
 
The following users thanked this post: EmmanuelFaure

Offline hillflyerTopic starter

  • Contributor
  • Posts: 21
  • Country: de
Re: Compact LT2400 Voltmeter
« Reply #21 on: September 24, 2016, 07:42:17 pm »
Do I understand your graph correctly: This measurement has been running for >4 YEARS?
Wow!
I don't understand why you cycle them daily from room temp to 50°C and back.
Wouldn't 25 to 30°C max be enough for repeatable temp conditions?
Why didn't you just keep them at 30°C all the time?
Questions, questions...
« Last Edit: September 24, 2016, 08:27:05 pm by hillflyer »
 

Offline Andreas

  • Super Contributor
  • ***
  • Posts: 3221
  • Country: de
Re: Compact LT2400 Voltmeter
« Reply #22 on: September 24, 2016, 09:08:16 pm »
Hello,

yes it is running since a few years.
But from time to time I earn a reference for use in a ADC from the box.
30 degrees is not a stable temperature in my lab.
Room temperature can change from 18 deg C in winter to 32 deg C in summer.
And self heating is also some deg C.
So I decided to heat up to 50 deg C during measurement.

The temperature controller needs the PC for calculating the complex control algorithm with 2 temperature sensors.
And the PC is also needed for other automated measurements like ageing of LTZ1000 or LM399 references against each other ...

With best regards

Andreas
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf