Author Topic: Rigol measurement issue with capacitive detection module  (Read 1158 times)

0 Members and 1 Guest are viewing this topic.

Offline François

  • Contributor
  • Posts: 5
  • Country: fr
Rigol measurement issue with capacitive detection module
« on: June 05, 2016, 09:18:58 AM »
Hi everybody,

First of all, excuse my English i’m not a native speaker.

So i’m studying a circuit which uses the internal capacitive module of a PIC 16LF1823 to detect the presence of a finger on a pad, and then, lights up an LED, if the signal reaches 1V at the Vmax.

I wanted to probe the input of the PIC which is connected to the capacitive pad, as you can see on the schematic attached, to watch on the oscilloscope the signal depending on the finger’s touch.

The problem is the following : I’m using two oscilloscopes, a Voltcraft DSO-1062D (which is the same as the Hantek DSO 5062b) and a Rigol DS1054Z. When I probe the signal with the Voltcraft, everything is okay, if my finger is touching the capacitive pad, the LED lights up and the signal on the scope is correct as you can see on the picture attached.

However, with the Rigol in the same conditions, I get a signal completely attenuated (see the picture attached) and the LED doesn’t light up even if my finger is touching the pad.
If I remove the probe while i’m still touching the pad, then the LED lights up.

It’s like the Rigol interferes the capacitive module.

I switched the Voltcraft and Rigol probes and nothing changed. However, the input impedance of these two scopes are nearly the same : 1Mohm // 20pF for the Volcraft VS 1Mohm // 13pF for the Rigol.

Does anybody has the explanation of the Rigol issue ? Since both input impedances are the same, what could make the Rigol worst than the Volcraft for this measurement ?

Thanks in advance and have a good day/night.

François

PS : the pictures show the result with the probe in 1:1 mode. With the probe in 1:10 mode, it’s a bit better for the Rigol but it’s still bad.
 

Offline Earendil

  • Regular Contributor
  • *
  • Posts: 87
  • Country: hu
Re: Rigol measurement issue with capacitive detection module
« Reply #1 on: June 05, 2016, 10:05:10 AM »
Do you use AC or DC input coupling?
Btw. your picture shows a square/sine wave at 50Hz. I find this slightly suspect. Is there a reason why you use 50 Hz?
 

Online tautech

  • Super Contributor
  • ***
  • Posts: 9203
  • Country: nz
    • Taupaki Technologies Ltd.
Re: Rigol measurement issue with capacitive detection module
« Reply #2 on: June 05, 2016, 10:47:57 AM »
Do you use AC or DC input coupling?
Btw. your picture shows a square/sine wave at 50Hz. I find this slightly suspect. Is there a reason why you use 50 Hz?
DC coupling as is visible in the GUI.

When one has *lost their way* with a DSO for sanity's sake press Default and start again.
That sets the DSO up for most basic measurements but I do agree 50Hz seems weird.

As each DSO is *not* set the same (timebase) and not showing similar amplitude one might suspect the DSO is not set correctly for the probes attenuation settings. (10:1 or 1:1)(channel settings)
Probes and channel input settings *should* be set to 10:1 unless there's a very good reason to do otherwise.(eg. single digit mV ripple measurement)

Avid Rabid Hobbyist & NZ Siglent Distributor
 

Offline François

  • Contributor
  • Posts: 5
  • Country: fr
Re: Rigol measurement issue with capacitive detection module
« Reply #3 on: June 05, 2016, 08:50:45 PM »
Thanks for your answers. I have no idea why the signal frequency is 50Hz, it's not my circuit, it's one from a commercial device that I bought and disassembled.

Indeed, as tautech said, I use DC coupling.

But yes, I'm sure at 100% the DSO channel settings was set correctly for the probes attenuation. On these pictures both probes were set to 1:1 and on the DSO channel settings, it was also set to 1:1.

But as I said, even when I set my probe to 10:1 and change the DSO channel settings to 10:1, with the Voltcraft I still have the same good signal and with the Rigol it's slightly better but still not good compared to the Voltcraft.

Tomorrow, I'll make and post the same pictures with 10:1 attenuation, and you'll see it's still not good for the Rigol :-//
 

Online tautech

  • Super Contributor
  • ***
  • Posts: 9203
  • Country: nz
    • Taupaki Technologies Ltd.
Re: Rigol measurement issue with capacitive detection module
« Reply #4 on: June 05, 2016, 09:26:13 PM »
Thanks for your answers. I have no idea why the signal frequency is 50Hz, it's not my circuit, it's one from a commercial device that I bought and disassembled.

Indeed, as tautech said, I use DC coupling.

But yes, I'm sure at 100% the DSO channel settings was set correctly for the probes attenuation. On these pictures both probes were set to 1:1 and on the DSO channel settings, it was also set to 1:1.

But as I said, even when I set my probe to 10:1 and change the DSO channel settings to 10:1, with the Voltcraft I still have the same good signal and with the Rigol it's slightly better but still not good compared to the Voltcraft.

Tomorrow, I'll make and post the same pictures with 10:1 attenuation, and you'll see it's still not good for the Rigol :-//
Ensure you set Defaults on both scopes before you do. There are several differences in the settings of each if you look closely.  ;)
Avid Rabid Hobbyist & NZ Siglent Distributor
 

Offline JPortici

  • Super Contributor
  • ***
  • Posts: 1674
  • Country: it
Re: Rigol measurement issue with capacitive detection module
« Reply #5 on: June 05, 2016, 10:51:14 PM »
If you read the datasheet, you'll find that the capacitive touch module works as such:
A programmable current source/sink connected to the selected pad
Voltage is buffered and passed into a comparator with programmable threshold.
Comparator switches the input connection from source to sink.
The output can be selected as timer 1 clock (or timer 0)

So, if probing didn't affect the measurement you should see a triangle waveform
Though i haven't tested it, this is a conclusion based from the info in the datasheet... but i don't see the point
« Last Edit: June 05, 2016, 10:54:19 PM by JPortici »
 

Offline Jasper

  • Contributor
  • Posts: 26
Re: Rigol measurement issue with capacitive detection module
« Reply #6 on: June 06, 2016, 02:02:12 AM »
Since it's a touch sensor, 50Hz noise makes sense to me — that's what a human is going to insert the most of into a circuit like that.
 

Offline François

  • Contributor
  • Posts: 5
  • Country: fr
Re: Rigol measurement issue with capacitive detection module
« Reply #7 on: June 06, 2016, 04:40:45 AM »
If you read the datasheet, you'll find that the capacitive touch module works as such:
A programmable current source/sink connected to the selected pad
Voltage is buffered and passed into a comparator with programmable threshold.
Comparator switches the input connection from source to sink.
The output can be selected as timer 1 clock (or timer 0)

So, if probing didn't affect the measurement you should see a triangle waveform
Though i haven't tested it, this is a conclusion based from the info in the datasheet... but i don't see the point

About the triangle waveform you're right but, even if we don't see a perfect triangle waveform on the Voltcraft scope, the capacitive detection of the module and the behavior of the LED when I'm touching the pad is the same, with or without the probe on the test point.

And that's not the case with the Rigol. When the Rigol's probe is on the test point, it affects the capacitive detection of the module and the LED doesn't light up. But why the Rigol and not the Voltcraft ? Which difference between them could cause this interference ?
 

Offline commongrounder

  • Frequent Contributor
  • **
  • Posts: 255
  • Country: us
Re: Rigol measurement issue with capacitive detection module
« Reply #8 on: June 06, 2016, 05:04:57 AM »
Please attach both scopes to the test point together, at the same time, with the probes/scopes in 10x mode and report back the readouts on each screen.  This should help narrow down whether the issue is with the Rigol scope/probe loading or a setting problem.  We assume here that you are attaching the probe ground reference leads to the same point in the circuit layout. :)
 

Offline François

  • Contributor
  • Posts: 5
  • Country: fr
Re: Rigol measurement issue with capacitive detection module
« Reply #9 on: June 06, 2016, 07:29:32 AM »
Please attach both scopes to the test point together, at the same time, with the probes/scopes in 10x mode and report back the readouts on each screen.  This should help narrow down whether the issue is with the Rigol scope/probe loading or a setting problem.  We assume here that you are attaching the probe ground reference leads to the same point in the circuit layout. :)

Yes, that's a good idea ! I'll do it tomorrow when I'll be in front of the two scopes (at my workplace) and I'll post the pictures ;)
 

Offline D3f1ant

  • Regular Contributor
  • *
  • Posts: 246
  • Country: nz
  • Doing as little as possible, but no less.
Re: Rigol measurement issue with capacitive detection module
« Reply #10 on: June 06, 2016, 12:30:19 PM »
Neither waveform looks correct.
With that particular cap sensor you don't need/want those resistors, connect the touch pad directly to the pic pin. Not sure why they would have added them, probably to reduce sensitivity or maybe ESD if touching bare copper?  You can measure the frequency on the pin directly with your scope with 10:1 probe, you should see sawtooth waveform, a few kHz from memory depending on what the source current is set to, from memory there is 2 or 3 levels. The probing will affect the sensor so don't expect to get more than a ballpark figure but is a good idea if you want to figure out what a suitable sampling period is for your code.
« Last Edit: June 06, 2016, 12:44:35 PM by D3f1ant »
 

Offline François

  • Contributor
  • Posts: 5
  • Country: fr
Re: Rigol measurement issue with capacitive detection module
« Reply #11 on: June 06, 2016, 06:51:45 PM »
Okay guys, I’ve finally found out what’s behind it. It’s simply an earth reference issue.

Let me explain: when I made the tests, I had the Rigol scope at my home, and the Voltcraft scope at my workplace.
It was the same circuit but which was tested in different places (because I don’t live at my workplace  ;D) with a different earth reference.

So the Rigol’s waveform picture was token at my home and the Voltcraft’s waveform picture was token at my workplace.
Today, I brought my Rigol to my workplace to make the additional test, and against all odds, both waveforms were exactly the same.

In addition, I’ve noted that, at my workplace, when I plug the alligator clip of the probe (which is actually the earth) to the ground of my circuit, the circuit is a bit more sensitive to my finger.

In conclusion, when measuring a capacitive pad, the result on the scope (even if the waveform isn't totally correct as some of you said) will depend on the connection to the earth at the measurement place, which is logic in fact, because the capacitive measurement is made between the pad and the earth (the finger is connected to the earth through your body).

Thanks everyone for your help!
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf

 

http://opalkelly.com/