Problem with assembled PSN-ADC24 board

[Dannie .Hidayat]

Thank you Larry for providing all information you put in github to make PSN-ADC24 board.  I ordered them many many years ago, and they were very useful for our application creating low power low cost seismic data logger.  

I have a need to make a few of them, so based on the gerber files in the github I ordered the PCB from one of the suggested PCB maker. I got the components online locally as well as online from China and US. Some components are no longer available (e.g. KV3901 diode), but I found similar diode with similar specs.  I found a needed frequency counter to tune in variable capacitor to get the exact frequency.  Flashing the bootloader onto the chip to the microchip DSPic30F3014 was a challenged for a few months, I thought similar to U2 USB Programmer would work (i.e. Pickit3).  With Pickit3 a lot of errors occurred,  I suspect one of the reason was that a couple I ordered were not original.  Then I ordered the suggested programmer and finally it worked.

However, the PSN-ADC24 board I assembled is not working yet.  I tested using Earthworm.  The board LED blinking fast before the program started, then blinking slower after the program started.  The PsnAdSend module showed errors (see below).  It seems like no data received through the board.  In addition, after several cycles, Time Reference never get locked.  I still have the old working PSN-ADC24 and with it, the errors did not occurred, data received with increasing numbers of samples, and Time Reference locked.

 I am wondering if anyone have some thought of what causing them?

Regards,

Dannie

-----------------------------------

Webtronics 16-Bit Digitizer - PsnAdSend Version: 2.7

 

Program Start Time (UTC): 09/25/2021 04:20:03

Last Packet Time (UTC): ???

Channels: 4   Samples Received: 0

Time Reference: Not Locked

09/25/21 04:20:40 No Data Timeout!

-----------------------------------

Program Start Time (UTC): 09/25/2021 04:20:03

Last Packet Time (UTC): ???

Channels: 4   Samples Received: 0

Time Reference: Not Locked

Messages:

09/25/21 04:20:30 Config Error: Num Converters

09/25/21 04:20:31 Config Error: Num Converters

09/25/21 04:20:32 Config Error: Num Converters

09/25/21 04:20:33 Config Error: Num Converters

09/25/21 04:20:34 Config Error: Num Converters

09/25/21 04:20:35 Config Error: Num Converters

09/25/21 04:20:36 Config Error: Num Converters

09/25/21 04:20:37 Config Error: Num Converters

09/25/21 04:20:38 Config Error: Num Converters

09/25/21 04:20:39 Config Error: Num Converters

09/25/21 04:21:18 No Data Timeout!

…

 

[chris]

Hi Dannie,

    I presume that you have an Oscilloscope ? Just a few thoughts - which may be 

helpful. Are all the 'chips' on your board correctly oriented ?  

    I presume that the crystal is oscillating and that you have set the sample rate ? How 

many channels are you trying to use ?

    Are there any signals for the channel selection ? Is a DC signal on one input being 

transferred to the ADC input ?

    Are there 'convert' timing signals coming into the ADC ?   

    With an input voltage on the ADC, is there any output ?

    Have you checked with a multi-meter for dry solder joints ?

    Regards, 

    Chris Chapman

--
You received this message because you are subscribed to the Google Groups "Public Seismic Network Mailing List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to psnlist+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/psnlist/24586ecf-471a-4c11-abf0-cf2790a6077cn%40googlegroups.com.

[Larry Cochrane]

Hi Dannie,

Sounds like you are making some progress. The first thing I would do is test the board using WinSDR. This may help you
debug the problem. My guess is the CPU is not getting data from all 4 ADC chips. That would explain the Config Error:
Num Converters error. If you request 4 channels and the CPU on the ADC board only sees 2 or 3 ADC chips you will get
this error. Try setting the number of channels to 1 and see if you get any data out of the board.

Regards,
-Larry

> --
> You received this message because you are subscribed to the Google Groups "Public Seismic Network Mailing List" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to psnlist+u...@googlegroups.com

> <mailto:psnlist+u...@googlegroups.com>.

> To view this discussion on the web visit
> https://groups.google.com/d/msgid/psnlist/24586ecf-471a-4c11-abf0-cf2790a6077cn%40googlegroups.com

> <https://groups.google.com/d/msgid/psnlist/24586ecf-471a-4c11-abf0-cf2790a6077cn%40googlegroups.com?utm_medium=email&utm_source=footer>.

[Dannie .Hidayat]

Hi Chris and Larry,

Thank you for your inputs to test on the problem.

I installed WinSDR as you suggested to test the board.  Looking at the logs (see part of them below), I only have 2 good channels!  So I set the channel number from initially 4 to 2 and keep the sampling rate (50Hz), and it seems working for the two channels.  I changed the sampling rate to 100Hz and it seems to work as well.

I did not mentioned earlier that after uploading the .hex files to the CPU,I tested on the older (working) board and all channels are working.  I am thinking, it is not the CPU problem, it might be the ADC chips.  

I do not have oscilloscope to test the board.  I checked and very sure the orientation of the ADC chips are correct.  To be honest, I probably not doing a good job soldering these SMD chips.  I burned (I think) one of the chip when I tried soldering using the hot air gun.  Then I change to normal, temperature controlled and small tip soldering iron.  After many trials of soldering, desoldering, then I thought it was OK. The followed up question is, if there are 2 bad chips, which one are bad?  I thought ADC chip U3 and U4 were OK and I messed up in U1 (which I replaced later).  It seems U1 and U2 are the OK ones.

Thank you.

Regards,

Dannie

WinSDR.log

...

09/26/21 20:36:24 ADC: Config Error: Num Converters

09/26/21 20:36:25 ADC: Config Error: Num Converters

09/26/21 20:36:26 ADC: Config Error: Num Converters

09/26/21 20:36:26 No Data from A/D Board - Time Diff: 60

09/26/21 20:36:26 Resetting WinSDR

09/26/21 20:36:34 ADC: Config Error: Num Converters

09/26/21 20:36:34 DLL: Reset Board - NoUsbDataError=0 UsbDevice=0

09/26/21 20:36:34 DLL: Reset Board - Waiting 5 seconds

09/26/21 20:36:34 DLL: ADC Board Info: Channels:2 GoodFlags:0x03 NeedsRst:0

09/26/21 20:36:35 DLL: Sending Configuration Information Channels:2

09/26/21 20:36:35 ADC: Restart Flag

09/26/21 20:36:35 DLL: Data before configuration sent - Resetting Board

09/26/21 20:36:35 DLL: Reset Board - NoUsbDataError=0 UsbDevice=0

09/26/21 20:36:36 DLL: Reset Board - Waiting 5 seconds

09/26/21 20:36:45 DLL: ADC Board Info: Channels:2 GoodFlags:0x03 NeedsRst:0

09/26/21 20:36:45 DLL: Sending Configuration Information Channels:2

09/26/21 20:36:47 ADC: Restart Flag

09/26/21 20:36:49 DLL: ADC Board Firmware Version: 2.5

09/26/21 20:37:18 DLL: GPSRef: First Reset - Time Diff:-0.115

09/26/21 20:37:49 Test Header Error: Creating New Record File

09/26/21 20:41:50 DLL: GPSRef: Sts:1 Lck:1 Sats:10 LckTm:00:04 Vco:53% VcoChg:00:00/00:02 PPSDif:-2

09/26/21 20:46:50 DLL: GPSRef: Sts:1 Lck:1 Sats:10 LckTm:00:09 Vco:57% VcoChg:00:00/00:01 PPSDif:-2

...

[Larry Cochrane]

Hi Dannie,

This line in the log file tells you what ADC chips are working.

09/26/21 20:36:45 DLL: ADC Board Info: Channels:2 GoodFlags:0x03 NeedsRst:0

The GoodFlags is a bitmap where 0x01 = channel 1, 0x02 = channel 2 etc. So on your board the CPU sees ADC chips on
channel 1 and 2 but 3 and 4 are not working.

Regards,
-Larry

> <https://groups.google.com/d/msgid/psnlist/24586ecf-471a-4c11-abf0-cf2790a6077cn%40googlegroups.com?utm_medium=email&utm_source=footer
> <https://groups.google.com/d/msgid/psnlist/24586ecf-471a-4c11-abf0-cf2790a6077cn%40googlegroups.com?utm_medium=email&utm_source=footer>>.

>
>
> --
> You received this message because you are subscribed to the Google Groups "Public Seismic Network Mailing List" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to psnlist+u...@googlegroups.com
> <mailto:psnlist+u...@googlegroups.com>.
> To view this discussion on the web visit

> https://groups.google.com/d/msgid/psnlist/4684f9b3-269c-4e0f-938c-3a8a5dce3acdn%40googlegroups.com
> <https://groups.google.com/d/msgid/psnlist/4684f9b3-269c-4e0f-938c-3a8a5dce3acdn%40googlegroups.com?utm_medium=email&utm_source=footer>.

[Dannie .Hidayat]

Hi Larry,

Thank you for your help and info.  After knowing ADC chip 3 and 4 not shown in the the logs I gave another try to repair the soldering of ADC chips.  When it was powered and connected to WinSDR I got " Channels:24 GoodFlags:0x0f"! I was able to see signals on all 4 channels, testing giving input voltage to the input terminal block, I was able to see the voltage reading on WinSDR.

Another question: the board input can accept voltage +/- 2.5V max.  I would like to connect a sensor that give maximum +/-10V max.  In the board information you mentioned about changing resistors and capacitors.  For the case of channel 4 (see the attached photo), I would need to change the jumper wires R6 and R15 to 35.2K, R3 and R16 change to 2.2K, and capacitor C9 and R4 to 0.1uF, right?

Thank you again.

Regards,

Dannie

[Larry Cochrane]

Hi Dannie,

Congratulations on getting your board up and running!

On 10/1/2021 9:26 AM, Dannie .Hidayat wrote:
> Another question: the board input can accept voltage +/- 2.5V max.

Yes, but it's centered around a 2.5V DC offset not ground referenced. So the input range for maximum ADC counts is 0 to
5v with the ADC gain set to 1. You do not want to apply a negative voltage to the inputs without the +/-10 Volt input
configuration.

>I would like to connect a sensor that give maximum
> +/-10V max.  In the board information you mentioned about changing resistors and capacitors.  For the case of channel 4
> (see the attached photo), I would need to change the jumper wires R6 and R15 to 35.2K, R3 and R16 change to 2.2K, and
> capacitor C9 and R4 to 0.1uF, right?

Yes, but R6 and R15 should be 33.0K not 35.2k. Using 33k resistors will make the input exactly 10V. In this
configuration the inputs will be ground referenced with an input range of +-10V. To use the +/-10V input option you must
set the ADC IC's gain to 2 or more.

Regards,
-Larry

[Dannie .Hidayat]

Hi Larry,

Thank you for the explanation and correction to the resistor size for the modified input configuration.  I tried it on one of the channels.  It seems to work.  Comparing with the +/- 2.5V input configuration (gain1), the +/-10V input configuration (gain 8) resulted amplitude ratio about the expected input ranges.

Thanks again.

Regards,

Dannie