Lowest voltage reading ?

Ask any questions relating to the PicoScope hardware or kit contents here.
If you have any questions prior to purchasing the kit post them here.
User avatar
Robski
Advanced User
Advanced User
Posts: 598
Joined: Tue Sep 14, 2010 6:36 pm
Location: Yorkshire

Lowest voltage reading ?

Post by Robski »

What's the lowest voltage reading the 4423 will display ?

I have been trying to measure voltage drop across a component, I am looking at sub 20mV (square wave), all the scope wanted to do was draw a flat line.
I was using CHA & CHC each side of the component with a math channel of A-C & C-A, I did away with the scope & used my DMM & got what I expected voltage-wise.Why wouldn't the scope display the same ???????? do I have to adjust sample rate & vertical resolution or is it just not up to the job ?

EDITED typo on scope model.
Last edited by Robski on Fri Sep 04, 2015 3:39 pm, edited 1 time in total.

Steve Smith
Pico Staff Member
Pico Staff Member
Posts: 1578
Joined: Sun Aug 25, 2013 7:22 am

Re: Lowest voltage reading ?

Post by Steve Smith »

Hello Robski and thank you for the post.

Could you post your psdata file of the square waveform as we have a number of options at our disposal for voltage measurements sub 20 mV?

Here I could place some annotations that will help with the various explanations.

Could you confirm the scope model too as I assume you have a 4423? (sorry if this is not the case)

Either way, the specification of the scopes (12 bit, resolution enhance, sample rate, scaling, filtering etc.) are more than capable of such low voltage measurements, complemented by various techniques we can adopt depending upon the signal under test. (Voltage of the square waveform for example)

Take care…….Steve

User avatar
Robski
Advanced User
Advanced User
Posts: 598
Joined: Tue Sep 14, 2010 6:36 pm
Location: Yorkshire

Re: Lowest voltage reading ?

Post by Robski »

Steve- yes automotive 4000 series (typo).

No capture as I didn't save because it wasn't what I wanted.

PS. I did vertically zoom X30 iirc on a ±50mV scale.

User avatar
Fat Freddy
TwoWaves
TwoWaves
Posts: 335
Joined: Fri May 04, 2012 4:52 am

Re: Lowest voltage reading ?

Post by Fat Freddy »

It shouldn't have any problems at all. That thermocouple probe I made range maxed out at 50mV @ 1250°C. The original covered 5mV 100°C. AFAIR no zoom or anything. Probably used the 50mv range for the probe.

HTH
FF

User avatar
Robski
Advanced User
Advanced User
Posts: 598
Joined: Tue Sep 14, 2010 6:36 pm
Location: Yorkshire

Re: Lowest voltage reading ?

Post by Robski »

so why did the DMM give me the expected Hi/Lo figures I expected & scope a flat line & no movement ?
I just thought another hour of my life wasted when I could have just used the DMM .

I'll try again when I get another spare minute .........

User avatar
Fat Freddy
TwoWaves
TwoWaves
Posts: 335
Joined: Fri May 04, 2012 4:52 am

Re: Lowest voltage reading ?

Post by Fat Freddy »

TBH I can't answer that. Also my measurement was not using two channels so not a direct comparison although theoretically it shouldn't make a difference. But I remember Firanocars (?) having a problem measuring low millivolts on a thermocouple.

If I get the opportunity/time/remember (the odds are not stacked in your favour :wink: ) I'll have a crack myself.

What sort of voltage difference were you looking for, 10 - 20mV?

User avatar
Robski
Advanced User
Advanced User
Posts: 598
Joined: Tue Sep 14, 2010 6:36 pm
Location: Yorkshire

Re: Lowest voltage reading ?

Post by Robski »

FF,

PM your email addy I can't find it in my inbox & I'll drop you an email :wink:

Steve Smith
Pico Staff Member
Pico Staff Member
Posts: 1578
Joined: Sun Aug 25, 2013 7:22 am

Re: Lowest voltage reading ?

Post by Steve Smith »

Hello and thank you for the posts.

I think it would be helpful here to briefly explain 12 bit resolution as this will help all when looking at low voltage measurements.

12 bit resolution allows for 4096 measurement levels or steps across the full voltage range.

For example.
A voltage scale of +- 50 mV = 100 mV full voltage range (100 mV = 0.1 V)
Divide 0.1 V by 4096 gives us 0.000024 (24) microvolt measurement steps across the full voltage range.
This is more than ample to measure volt drop values sub 20 mV.

Each of the vertical measurement steps/levels from the base of the screen to the top of the screen equates to 24 microvolts (24 microvolt vertical resolution on 100 mV range)
12 BIT RESOLUTION.jpg
In the case of Robski’s square waveform, let’s assume we have a 12 V square wave, our voltage range will be +- 20 V. (40 V full voltage range)
40 divided by 4096 = 0.0097 (9.7) millivolt measurement steps across the full voltage range.

This will still reveal sub 20 mV, volt drop values but could now be getting challenging if we have excessive noise on the measurement signal.

Low pass filtering and scaling factors remains an option here to reveal the signal or voltage value of interest.

Please remember there will always be “noise” of some form on our signals and of course, some of these measurement steps will be consumed by noise.

Keep the number of samples on screen to a minimum of 1 MS as “detail is everything” when looking at such small signals.

1 MS is a good bench mark for a “one size fits all” sample control setting for all automotive signals. 2MS is better still for the ultimate horizontal resolution but be aware, files sizes will increase given the amount of detail we have captured on screen along with the number of waveforms in your buffer.

Assuming we have a 12 V square waveform we could always AC couple the signal removing the DC component of the signal.

This will reveal the AC component present on the signal and display the signal about the 0 V point on our scale.

Because the AC signal will generally be small, we can revert back to our +- 50 mV scale and take advantage of the 24 micro volt measurement steps.

Another option would be the resolution enhancement feature of PicoScope, where the best possible explanation can be found here at

https://www.picotech.com/library/oscill ... nhancement

In effect we can increase the vertical resolution, (Measurement steps) in 0.5 bit increments.

For example if we jump to 14 bit resolution when looking at the 40 V full voltage range we then have
40 divided by 16384 = 0.0024 (2.4 mV) millivolt measurement steps across the full voltage range.
14 BIT RESOLUTION.jpg

There is however a trade-off for this clever feature, improved vertical resolution comes at the expense of bandwidth/frequency response. (See the video in the link above)

I am not sure of the frequency of the square waveform under test, but once again with PicoScope we can use a combination of resolution enhancement, sample control/sample rate adjustment (to off-set the effects of reduced frequency response), filtering, scaling and maths to find exactly what we are looking for regardless of voltage level.

I hope this helps and if you can post a captured signal that would be great as we can see just how the application of the above features would help to obtain the volt drop values required.

Take care…….Steve

User avatar
Robski
Advanced User
Advanced User
Posts: 598
Joined: Tue Sep 14, 2010 6:36 pm
Location: Yorkshire

Re: Lowest voltage reading ?

Post by Robski »

Steve,

I was on a ±20V scale & 1MS sample rate, AC coupling is not an option I want DC 0-20mV (for the moment) & its the 'differential' I want to see between 2X channels hence a math channel, as stated with my DMM across i got the expected values measuring the difference between the black lead & red lead.
So in short I expect the scope using 2X channels & a math to do the same job but show me a square wave with a min & max amplitude.

Steve Smith
Pico Staff Member
Pico Staff Member
Posts: 1578
Joined: Sun Aug 25, 2013 7:22 am

Re: Lowest voltage reading ?

Post by Steve Smith »

Thanks for the feedback Robski.

I would like to simulate this test and post the results.

I can see how we are trying to achieve the Min/Max voltage drop value of a 12 V square waveform (using 2 x channels and maths)

The scope should produce these results without compromise.

Can you tell me what component you were testing, this way we should produce realistic and relevant results?

Take care......Steve

Post Reply