I think it's broken

[KeithRussell]

I have an AD2 that I use a lot and today the Channel 1 started showing an input voltage of -21 with nothing connected. Channel 2 and all other functions work great.

I restarted the computer, updated the software, and power cycled the AD2. No change.

I calibrated the waveform generator and then the scope, but scope calibration failed saying that Channel 1 is showing too high on the input when grounded. I tried reverting to the factory calibration, but the symptoms remained the same.

So, is it truly broken? If so, is there a way to repair it?

[attila]

Hi @KeithRussell

The scope inputs of AD2 are rated to +/-50V.

On your device probably the ADG612 analog switch got damaged:
https://reference.digilentinc.com/reference/instrumentation/analog-discovery-2/reference-manual#scope_input_divider_and_gain_selection

[KeithRussell]

Agreed. I'm checking it out now. I didn't know if it was the mux or the buffer, but on your recommendation, I will concentrate on the mux.

I'll let you know.

[KeithRussell]

You were correct, attila, it was the mux. After confirming with measurements, I cut pins 10 and 15 to be double-sure it wasn't the buffer. 

I'm ordering anther ADG612 now.

After going over all my measurements and checking everything I did, I am confident that I did not damage the mux. I don't have anything over 9 volts on the bench, and I did not have a mis-matched ground problem anywhere. It just died of its own accord.

[KeithRussell]

I ordered an ADG612 on Monday from Mouser and it came yesterday. I put in the new part, ran the calibration, and everything works great. 

Thanks for the help, Attila.

[attila]

Hi @KeithRussell

Good to hear you managed to fix it.

[Camarillo]

Hello.

 

Mr Attila AKA "Technical Forum Moderator".

 

I think i have the same problem as Mr Russell.

 

My AD2 will not go down from +20 volts. I was doing some measurements with a transformer and the BNC kit for the AD2 and after some measurements, it looks like I also broke my Channel 2 input. The probes were x10 in the attenuation department.

 

Will the solution suggested work also for me?     I have almost the same issue but  is not -21, is +21.

 

The device works well measuring frequencies, the attached image shows the generator connected to channel 2, measuring 10 MHz.

 

Scope and WFG 10 MHz.pdf

Edited February 18, 2020 by Camarillo
There. Had to convert the image to PDF for it to be uploaded.

[Camarillo]

Sorry, forgot to add the MHz part.

 

 

Edit: can´t upload the other picture.

 

 

EDIT: I was able to upload the picture but as a PDF.

Edited February 18, 2020 by Camarillo
Was now able to upload a file now.

[attila]

Hi @Camarillo

Probably some high voltage transient input or accidentally having the probe at 1x with high voltage damaged the ADG612.

Before replacing the ADG612 you could verify the VREF_SC2 and VOFF_SC2 voltages.
https://reference.digilentinc.com/reference/instrumentation/analog-discovery-2/reference-manual#scope_reference_and_offset
- the VREF should be about 2V
- the VOFF with:
   0V offset about 2V
   +2.5V offset with 500mV/div range about 3V, like the screenshot below
   --2.5V offset with 500mV/div range about 1V

Some further troubleshooting posts:

 

[Camarillo]

Hello Mr Attila.

 

I think I did the measurements but not sure if I did them in the right manner.

 

My setup:

-OSC1  ---->  GND

+OSC1 ----> to test points

 

Did all measurements with respect to GND of the AD2. In the last measurement I remembered that it also could have been differential measurements.  Did all the differential measurements and the result is the same with the other two with respect to GND of the AD2.

 

I had to modify the +2.5 offset with 0.5 scale because the measurement will not show it self (higher than 0 volts I guess).

 

 

 

This was the VREF_SC2, with respect to GND it was close to 2 volts.

 

 

 

This one is the NETC410_1 with respect to GND. It was 1.1947v

 

 

 

This one is the diferential measurement, just took it after I´ve finished...

 

 

 

This one is the VOFF with 0 volts offset. It was almost 2 volts (1.9969)

 

 

 

 

This one is NET409_1 with 0 volts Offset respect to GND. It was 0.5855 volts.

 

 

 

 

This one is VOFF with -2.5 volts offset and 0.5 volts/div.   It measures 1.99 volts.

 

 

 

 

 

This one is NETC409_1 with -2.5 volts offset with 0.5 volts/div.   It measures 0.5786 volts.

 

 

 

 

 

 

[Camarillo]

I´m sorry with all those uploads. Don´t know what I did but apparently I uploaded all the files again.

 

 

The last two measurements:

 

 

This one is the +2.5 Offset at VOFF.  I had to modify the range from 0.5 to 1 volt/div.  It measures 2 volts.

 

 

 

This one is the NET409_1 with +2.5 volts offset.  I left the same 1 volt/div as the other one. It measures 0.5959 volts.

 

 

 

What do you think?  Could the MUX be the problem?

 

 

Thank you so very much.

[attila]

Hi @Camarillo

I meant measuring the VOFF_SC2 relative to the device GND with external DMM.
Having the scope channel 2 range 500mV/div or lower, changing the offset between -2.5V and 2.5V should change the VOFF_SC2 between ~1V and ~3V
Above 500mV/div for offset between -25V and +25V change VOFF between ~1V and ~3V.

[Camarillo]

On 2/20/2020 at 1:47 AM, attila said:

Hi @Camarillo

I meant measuring the VOFF_SC2 relative to the device GND with external DMM.
Having the scope channel 2 range 500mV/div or lower, changing the offset between -2.5V and 2.5V should change the VOFF_SC2 between ~1V and ~3V
Above 500mV/div for offset between -25V and +25V change VOFF between ~1V and ~3V.

Oh.... sorry, did not understad that.

 

Did the measurements. They are within range:

VREF2V = 0V

VOFF_SC2 with 0V offset = 1.94V,

VOFF_SC2 with 2.5V offset = 2.99V

VOFF_SC2 with -2.5V offset = 1.01V

 

All measurementes respecto to the device´s ground and with 500mV/Div for channel 2.

Edited February 21, 2020 by Camarillo
Definded the voltages.

[attila]

Hi @Camarillo

According your measurements the offset DAC is working.

I almost forgot that odd reading or output offset could be due to wrong calibration.
Since the offset voltage is applied this it likely good, but please perform a "Load factory" and see the oscilloscope readings again to make sure the calibration is not culprit.

Otherwise the 21V reading is likely due to damaged ADG612.
The one for channel 2 is on the right side, the 16pin IC above the trimmers.

 

[Camarillo]

Thanks for the reply.

 

Did the factory reset but the measurementes were the same. Although something happended that I did not notice: if I start channel 2 with the 500mV/Div, if i change it to 5V/Div, the signal from channel 2 goes to -3V and after a few seconds, it goes to +21 (with nothing conected to the probes).

 

Did a calibration after loading the factory data. The waveform channel, both 1 and 2 passed. But the oscilloscope did not, and it showed the behaviour described above (-2.7 volts to +21 volts.

 

Do you think it might be the MUX ADG612?

 

Thanks.

 

 

 

Edited February 24, 2020 by Camarillo
Deleted the same image. Uploaded it twice.

[attila]

Hi @Camarillo

It is most probably the C2 ADG612 damaged, the first IC in the input path.

The 2.7v is the maximum input reading with high gain, <= 500mV/div

Sorry for the late reply, I'm off work this week.

[Camarillo]

On 2/27/2020 at 1:30 AM, attila said:

Hi @Camarillo

It is most probably the C2 ADG612 damaged, the first IC in the input path. 

Hello. 

Thanks for the info. 

 

Just one question, and this might be very transparent for everyone but me:

Which of the two ADG612 is the channel 2?

 

I have all the intention to do that repair in this week. Good thing I have the AD1 for testing. 

[attila]

Hi @Camarillo

See the following picture:

[Camarillo]

At the time of your response, I was looking at this other of your responses: 

 

Was looking for those resistors IDs in the reference document to determine which one was Channel 2.

 

 

Thanks, I´ll now proceed to operate ?

 

[Camarillo]

Hi again......

 

I think I broke it for sure now.

 

 

This image is with the wave generator #1 with its output at 5 volts with no offset.

 

 

 

After resoldering the ADG612, Channel two was able to mirror the information in channel one.

 

 

 

Here are both channels at the same time.

 

 

 

It´s strange because Channel one was not failing. Maybe it´s because my soldering skills with SMD is not that good. I just soldered the IC.

Maybe I broke it for sure now.

 

 

Did a device calibration setup and the first  two steps generate -5.05 volts. (it says demo mode but the measurements were made before disconnecting the device)

 

 

 

Right now I don´t have access to my AD1. Next week I´ll have access to it and will use it as a reference to measure analog signals.

 

Can you provide some insight?

 

 

Thanks.

 

 

[attila]

Hi @Camarillo

You have fixed the oscilloscope, but it looks like the AWG output is damaged too

You can verify the following rails, when the device is connected and opened with WaveForms, between the no load component pads and device ground.
Bottom left edge of the board:

 

In the following post you have further troubleshooting:

 

[attila]

Hi @Giovanni

First of all, connect to the device with WF application in order the analog circuitry to be enabled.

The VREF1V2_AWG and VREF1V_AWG is generated by the IC42. You should also verify the 3V3 input of this.
These are buffered by IC13 as VREF_AWG1/2.
https://reference.digilentinc.com/reference/instrumentation/analog-discovery-2/reference-manual#awg_reference_and_offset 

You can find the respective layout sections in the following post:
https://forum.digilentinc.com/topic/4714-analog-discovery-waveform-generator-failure/?do=findComment&comment=19153

 

[Camarillo]

Hello again. Almost a year and a half later.

 

After reparing the oscilloscope (with your help, of course), I did use it for some other low level measurements and the digital side as well.

 

Had to do some other measurements and was surprised by something:

On 2/19/2020 at 12:38 AM, attila said:

Hi @Camarillo

Probably some high voltage transient input or accidentally having the probe at 1x with high voltage damaged the ADG612.

 

One of my probes has a faulty switch for the x1 and x10 range. I started using a bench oscilloscope and that is where i found out.

This is odd since they were bought with the BNC PCB for the AD2.

 

 

 

 

[Camarillo]

On 2/19/2020 at 12:38 AM, attila said:

Hi @Camarillo

Probably some high voltage transient input or accidentally having the probe at 1x with high voltage damaged the ADG612.

It appears I can´t post more pictures. I can´t delete the ones in the post above (don´t know how).

 

Well, just to your point, it probably was that faulty switch that caused all of this.

[Camarillo]

Two more things:

 

1. I will now just measure +-40 volts. Can I use the differential measurement for that (Channel 1 and 2).
2. Will try to revive the waveform generator in the near future.

 

I think #1 is a given since here:

 

 

and here:

https://reference.digilentinc.com/test-and-measurement/analog-discovery-2/reference-manual?redirect=1#scope_input_divider_and_gain_selection

 

states that the maximum input can be +-50 volts using differential measurement.

 

 

Thanks in advance... and sorry for the super late reply.

 

 

[attila]

Hi @Camarillo

Yes.

The scope inputs are rated/protected up to +-50V relative to device ground.
The range is 50V, with inputs between +-50V and offset -25V..+25V