Shop OBEX P1 Docs P2 Docs Learn Events
Electrical Gurus Needed :) ---- Measuring The Millivolt With High Accuracy - Page 2 — Parallax Forums

Electrical Gurus Needed :) ---- Measuring The Millivolt With High Accuracy

2456

Comments

  • Seems like we are trying to re-invent the wheel. If possible incorporate a high end meter into your application. In the high meter you have a custom precision networks on the front end, guard circuit, filtering stage, amplication stage, intergration stage, and display stage. Fluke, Keysight, Kiethley etc. have already engineered the solution for you. Unless you intend on making high end D.C. Meters for sale.
    Look at some of old HP and Fluke dmm service manuals. You'll find that fluke made their own proprietary integrator IC's.
  • Another question is:

    Do you need one magic measuring device, or lots of them?

  • Well someone mentioned environment at last. What sort of noise are you going to have to contend with, how will you filter it out? Are your sample times long enough to average the noise out in an analog fashion or will you just do some digital smoothing on the snapshots you capture assuming you can capture enough samples that give you what you need when you add in the processing time to fold, spindle and mutilate said samples. If you have a MYTi.com account, this link may help you a bit...

    http://www.ti.com/analog-circuit/bonnie-baker-ebook.html

    Good luck




  • The other problem is that your requirements are a "secret" so you can't tell us every little detail where any one detail might be a game changer.

    Alright!!!.... At the risk of gaining some unwanted competition... Without divulging too much of my concept... As I am almost certain that there must already be similar devices in existence... I will share some information :)

    Basically, my goal is to develop a device for tracking down parasitic current draws within a 12VDC automotive electrical system and perhaps a 24VDC electrical system for other equipment. To find these parasitic current draws, I will be measuring the voltage drop across the various fuses of these electrical systems, in milli-volts. The readings for this milli-volt meter (at least for a 12V system) will be limited to the following:

    0.0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0, 1.1, 1.2, 1.3, 1.4, 1.5, 1.6, 1.7, 1.8, 1.9, 2.0, 2.1, 2.2, 2.3, 2.4, 2.5, 2.6, 2.7, 2.8, 2.9, 3.0, 3.1, 3.2, 3.3, 3.4, 3.5, 3.6, 3.7, 3.8, 3.9, 4.0, 4.1, 4.2, 4.3, 4.4, 4.5, 4.6, 4.7, 4.8, 4.9, 5.0, 5.1, 5.2, 5.3, 5.4, 5.5, 5.6, 5.7, 5.8, 5.9, 6.0, 6.1, 6.2, 6.3, 6.4, 6.5, 6.6, 6.7, 6.8, 6.9, 7.0, 7.1, 7.2, 7.3, 7.4, 7.5, 7.6, 7.7, 7.8, 7.9, 8.0, 8.1, 8.2, 8.3, 8.4, 8.5, 8.6, 8.7, 8.8, 8.9, 9.0, 9.1, 9.2, 9.3, 9.4, 9.5, 9.6, 9.7, 9.8, 9.9, 10.0
  • ErNaErNa Posts: 1,738
    My impression is, you like to measure voltage drop on a high side shunt
  • Seems like we are trying to re-invent the wheel. If possible incorporate a high end meter into your application. In the high meter you have a custom precision networks on the front end, guard circuit, filtering stage, amplication stage, intergration stage, and display stage. Fluke, Keysight, Kiethley etc. have already engineered the solution for you.

    I am sure that many of you will be thinking the same way...

    However, since I have divulged the intended application, I contend that there is an easier method of finding a parasitic current draw than by using a common multi-meter. Working under dashboards with a multi-meter can be some real back breaking work and my goal is to make it easier.
  • Heater.Heater. Posts: 21,230
    edited 2018-05-04 12:46
    Details matter. Turns out you don't want to measure 0 to 24V with a 100uV resolution but only 0 to 10mV with a resolution of 0.1mV.

    That reduces you accuracy requirement from 0.004% to 1%.

    That changes everything the situation from impossible to mundane!

    Use an op-amp to amplify that up to 1 or 3 volts or whatever and a cheap 8 bit ADC.

    Create a differential amplifier, with inputs taken from each end of your fuse.

    https://www.electronics-tutorials.ws/opamp/opamp_5.html

    You'll need to be sure that 24v is divided down to something within the opamps power supply range.

  • Details matter. Turns out you don't want to measure 0 to 24V with a 100uV resolution but only 0 to 10mV with a resolution of 0.1mV.

    That reduces you accuracy requirement from 0.004% to 1%.

    Yea, yesterday was a tough day for me, so I had a wee bit too much rum, and my thought process was surely skewed. :) Sorry.
  • To repeat - " If possible incorporate a high end meter into your application. In the high meter you have a custom precision networks on the front end, guard circuit, filtering stage, amplification stage, integration stage, and display stage. Fluke, Keysight, Kiethley etc. have already engineered the solution for you." Work with one of the high end DVMs and engineer / modify / add op amps / sensors / whatever, to the probe end to suit your needs. This is done all the time, with excellent results. Many high end meters have i2c, RS 232 data ports you can use. Use the Prop to add intelligence to your readings, data capture, time stamps, store the data, fancy display, timing & sequencing of the data capture to enhance the end user's requirements.
  • RaymanRayman Posts: 13,797
    I've used this chip to measure big and small voltages: Max1270

    Very nice chip.
  • JonnyMac mentioned the INA219 chip that is designed for just this thing and is easy to use. Why wouldn't you use this? I would.
  • kwinnkwinn Posts: 8,697
    JonnyMac mentioned the INA219 chip that is designed for just this thing and is easy to use. Why wouldn't you use this? I would.

    +1 - Simple reliable and inexpensive approach.
  • The INA219 is easy to use. I've put it into two industrial designs and it's worked well. Being and I2C part it can share the EEPROM bus pins.
    JonnyMac mentioned the INA219 chip that is designed for just this thing and is easy to use. Why wouldn't you use this? I would.
    +1 - Simple reliable and inexpensive approach.

    Okay so three votes for the INA219 chip.

    Adafruit sells an INA219 breakout board for $9.95. Here is a link: https://www.adafruit.com/product/904

    I will pick one up and see how it compares with the Fluke 112 meter.

    If this works out the way I want it to, then that will be a quick and easy solution the measuring part of the device, which will allow me to devote plenty of time for the UI and coding.

    As a side note, the INA219 is a available in two different versions, as stated in the TI datasheet:
    The INA219 is available in two grades: A and B. The B grade version has higher accuracy and higher precision specifications.

    From the photo on the product page, they used the B grade version.
  • Tony

    Yea, but the INA219 part# is speaking to me, because I live in a 219 area code :)

    Could it be a good omen??? LOL I intend to find out.
  • max72max72 Posts: 1,155
    I would suggest to check the free low level measurement handbook.
    It is a great source of informations and inspiration..
    Massimo

    http://download.tek.com/document/LowLevelHandbook_7Ed.pdf
  • They also make amp with built-in adc with a i2c output. (but all are 0.5mm pin pitch)
    https://www.mouser.com/Semiconductors/Amplifier-ICs/Current-Sense-Amplifiers/_/N-6o48gZ1yzvvqx?Keyword=i2c&Ns=Pricing|0


    Then you would need a 8channel analog mux for the low side of each shunt/fuse if you want to messure 8 fuses but only using one sense amp-adc.



  • I've attached my INA219 object -- you may find it helpful. That said, I've structured it to read millivolts, so you'll need to write another method.
  • Thanks Massimo, that appears to be a very informative document.
  • idbruceidbruce Posts: 6,197
    edited 2018-05-04 17:15
    Thanks Jon

    It looks well thought out and written.

    Odds are that I will most likely be programming in C and port Adafruit's code, but like you said, I may find it helpful to have additional information and to see how things are done.

    However, on second thought, your object will be of great help, because it will allow me to test the chip, before doing any porting of C.

    Thank you for your kindness and generosity.
  • RaymanRayman Posts: 13,797
    I see I should have read further into this thread before posting!

    Sounds like you have a plan, but I think what you want is called a "current sense amplifier"...

    Found this note from TI:
    http://www.ti.com/lit/sg/slyb194c/slyb194c.pdf


  • GenetixGenetix Posts: 1,740
    edited 2018-05-04 20:47
    idbruce,

    According to the Pareto principle 80% is USUALLY caused by the top 2 or 3 things but you never know when uncle Murphy is going to pay you a visit.
    If you have a lot of little problems then it's precision meter time, otherwise just look for big changes.
    Say a small value just gives a b-e-e-p, well then if there is a whine, chirp, or a screech then that's the culprit.
  • Another good choice is the ICL7135 4-1/2 digit DMM IC. It already has a built in integrator and it interfaces easily with a PIC. Just add a resistor network like a Caddock.
  • idbruceidbruce Posts: 6,197
    edited 2018-05-04 22:35
    Genetix

    Everything has changed over the last several decades, as it pertains to finding parasitic draws in an automobile.

    When I first started testing the suspected vehicle, it appeared to have a 5A draw. As it turns out, with today's automobiles, you must wait for everything, to go into sleep mode, and then, it is still not as simple as it once was, where you could simply disconnect the negative terminal and insert your ammeter in series between the negative post and terminal. Nope, can't do that anymore, because it will wake the systems back up and you will be drawing massive current again. Instead, you must attach a booster pack or spare battery to the terminals, then remove the negative terminal, then connect your ammeter in series with the negative terminal, and then finally, disconnect the booster battery, just to get an accurate reading for total current draw. However, if you have a clamp on dc meter, it is a breeze, just wait for the system to sleep, and clamp it over the negative terminal.

    Anyhow, what initially appeared to be a 5A current draw, when done properly according to today's standards, resulted in 42mA, which is perfectly acceptable according to the automobiles today.

    So I took the battery to the auto parts store for a load test, which resulted in them telling me the battery was still good. I came to the conclusion that either the battery was in fact bad or I am having an intermittent problem with a sticking relay somewhere. To eliminate guesswork and prevent myself from going in circles, I bought a new battery yesterday. As of today, everything seems to be okay, but time will tell if it is an intermittent sticking relay. One thing is for sure, I will certainly be lugging the booster pack along for the ride for a while, just to make sure I do not get stranded.
  • idbruce wrote:
    To eliminate guesswork and prevent myself from going in circles, I bought a new battery yesterday. As of today, everything seems to be okay, ...

    Maybe all you really needed was a wire-brush battery post and terminal cleaner. :)

    -Phil
  • jmgjmg Posts: 15,140
    edited 2018-05-04 23:13
    idbruce wrote: »
    Yea, but the INA219 part# is speaking to me, because I live in a 219 area code :)

    There is a series of those parts, and the INA233 has a higher voltage rating and tighter drift specs. (Likely similar SW drivers)
    The higher voltage rating will matter for 24V systems.

    Microchip also have something, IIRC from SMSC
    Addit : PC1720, 40V, but not as many bits as the TI part, & in DFN10, so not casual use friendly, but also has a lower price.
    idbruce wrote: »
    ..
    Basically, my goal is to develop a device for tracking down parasitic current draws within a 12VDC automotive electrical system and perhaps a 24VDC electrical system for other equipment. To find these parasitic current draws, I will be measuring the voltage drop across the various fuses of these electrical systems, in milli-volts. The readings for this milli-volt meter (at least for a 12V system) will be limited to the following:

    0.0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0, 1.1, 1.2, 1.3, 1.4, 1.5, 1.6, 1.7, 1.8, 1.9, 2.0, 2.1, 2.2, 2.3, 2.4, 2.5, 2.6, 2.7, 2.8, 2.9, 3.0, 3.1, 3.2, 3.3, 3.4, 3.5, 3.6, 3.7, 3.8, 3.9, 4.0, 4.1, 4.2, 4.3, 4.4, 4.5, 4.6, 4.7, 4.8, 4.9, 5.0, 5.1, 5.2, 5.3, 5.4, 5.5, 5.6, 5.7, 5.8, 5.9, 6.0, 6.1, 6.2, 6.3, 6.4, 6.5, 6.6, 6.7, 6.8, 6.9, 7.0, 7.1, 7.2, 7.3, 7.4, 7.5, 7.6, 7.7, 7.8, 7.9, 8.0, 8.1, 8.2, 8.3, 8.4, 8.5, 8.6, 8.7, 8.8, 8.9, 9.0, 9.1, 9.2, 9.3, 9.4, 9.5, 9.6, 9.7, 9.8, 9.9, 10.0

    Why choose an arbitrary 0.1mv LSB, do you have only a 2 digit display ?
    When you say 'measuring the voltage drop across the various fuses' is that by fuse replacement, or do you try to squeeze probes alongside the existing fuse ?

    How many channels of these do you need ?
  • idbruceidbruce Posts: 6,197
    edited 2018-05-05 00:54
    @Phil
    Maybe all you really needed was a wire-brush battery post and terminal cleaner.

    Nice try :) I had already cleaned the posts and installed new terminals, so that problem had already been ruled out.

    @jmg
    There is a series of those parts, and the INA233 has a higher voltage rating and tighter drift specs. (Likely similar SW drivers)
    The higher voltage rating will matter for 24V systems.

    Thanks for that information.
    Why choose an arbitrary 0.1mv LSB

    Without divulging too much of my idea, I did not choose the 0.1 mV LSB arbitrarily.
    do you have only a 2 digit display ?

    Actually I need 3 digits... Don't forget the 10.0 mV reading. However, I might not use a display. I have not decided yet.
    When you say 'measuring the voltage drop across the various fuses' is that by fuse replacement

    No... The fuses are measured in the circuit. The milli-volt voltage drop as measured across the in circuit fuse, correlates with a specific milli-amp current draw on that circuit. The fuses are not removed, as you would with measuring current with an ammeter.

    Basically, if there is no parasitic current draw, the reading will be 0.0 mV, but if there is a parasitic current draw, then there will be a mV reading across the fuse.
    do you try to squeeze probes alongside the existing fuse ?

    In some cases, that will be necessary, but the Mini ATM fuses and the Standard ATC fuses have little nibs on the top that you can probe for readings and continuity.
    How many channels of these do you need ?

    I need only one channel.
  • idbruceidbruce Posts: 6,197
    edited 2018-05-05 01:28
    How many channels of these do you need ?
    I need only one channel.

    Please allow me to correct myself. I would need two channels to enable differential readings. This would eliminate the need to have the probes in a specific orientation.
  • jmgjmg Posts: 15,140
    idbruce wrote: »
    Basically, if there is no parasitic current draw, the reading will be 0.0 mV, but if there is a parasitic current draw, then there will be a mV reading across the fuse.

    Depends on what you call Parasitic.
    If a large fuse drops 10mV at rated current (eg 25A) , that 0.1mV LSB is 250mA - rather higher than most consider 'Parasitic'


    idbruce wrote: »
    Please allow me to correct myself. I would need two channels to enable differential readings. This would eliminate the need to have the probes in a specific orientation.

    That means you need differential current sense, and you |absolute| that elsewhere.
    The INA233 series are differential, so will work just fine.
  • The INA219 breakout board includes a sense resistor. To take a reading across a fuse, you will want to remove the fuse first to get an accurate reading.

    -Phil
Sign In or Register to comment.