Skip to main content ↑ ↓
Topic: Me too, My Sniffer may be Snuffed! No High Traces? (Read 3139 times) previous topic - next topic

Me too, My Sniffer may be Snuffed! No High Traces?

Hi Folks!

Got mine yesterday!  5/11  Took about 2 Weeks shipping to So. Calif, USA, not bad?

Got USB driver working on port 7 and changed speed in system to 115200.  Is it necessary to change PC port Baud?
Will it even connect if something is wrong here?
Seems fine?

Got it connected to pic counting on portb of a 16F873 running 5 volts on a proto board.
I have LEDs flashing on some of the slower bits.

Counting at about 1ms delay.

I click the ROKET, set up, and am using only 8 inputs, trigger on 0, and a 2K sample at 10Mhz
My ARM led is ON.  Both Arm and Trig are RED leds on my board.

Just see all low levels.  Nothing high on logic trace.

Tested same counter with Bus Pirate with Sump Client, works great.

 
A hint might be I don't see the time base at the top change with sample speed?
Its scale in in microseconds, so it is not seeing this much slower signal and setting its scale?

I have spent many hours, don't know if it is a dead Sniffer?

Am I putting the signal in the correct pins?

With USB connector on LEFT SIDE, I am using 2nd pin from BOTTOM as CH 0, then up from there.  Bottom Pin connected to my
pic counter ground.  Pins 2-9 for the input channels.

I see pictures on the sites with connectors connected to the top of this connector?

What is the display trace B0 below the ch7 mean?  It has continuous hi and low trace lines and says "00" in the middle?
Trigger?

Thanks for ANY help!


 
Mike
Crestline, Ca.

Re: Me too, My Sniffer may be Snuffed! No High Traces?

Reply #1 –
Hello MotoDog, Sorry for the late response, your post somehow slipped our attention.

From what you describe, your setup appears to be correct.

The buffered probe header on the right (with the USB connector pointing to the left) has channels 0-15 (from bottom to top) when using the Inside Numbering Scheme. In capture setup you should select/check channel groups 1 and 2. With GND at the bottom edge connected to GND on your MCU board and the channel pins counting up from the GND pin connected to the PORTB signals on your MCU board you should get your resulting captures on the channels up from 0.

The B0 line gives you the hex value of the byte for channels 0 to 7 - it's an alternative (byte) representation of the 8 channels of channel group 0.

I suggest you upgrade to the latest (stable) firmware v0.4, bitstream (r2.04) and Java client (included in the bitstream r2.04 package) releases which incorporates a self-test for the OLS. For this you will need a PIC programmer unless your OLS shipped with the bootloader installed. The firmware and bitstream update procedures are described in detail here.

Quite a few of the early problems have been resolved with firmware v0.4 and bitstream release 2.04 ... if the OLS passes the test you should be all set for sniffing on your PIC project.

/EDIT/ Just saw in an other thread that you and Ian finally established that your OLS may have a bad buffer chip ... did you contact Seeed?

Re: Me too, My Sniffer may be Snuffed! No High Traces?

Reply #2 –
Thanks,
I got another sniffer delivered yesterday as a replacement.    I sent the other back as suggested by Ian.
Thanks!

I also bought  some Seeed Sniffer cables.  A little confusing with black wire ground and white channel 0 and so on?
I guess they used what wire and clip colors they had available?

I am hooked onto my 8 bit Pic Generator board which fastest bit moves at 100Khz.

Set up with a few signals on Bank 0, I get no triggering when set up on Ch0, when I turn off triggering, I DO SEE signals on BANK 2, if I enable bank 2 and 3?  Nothing down on Bank0?

Since this is a new board?  I don't know what software versions are in it?  Could the PC client program get a self-test built in some day to do some version reading?

So, I wonder why do signals on Bank0 ?  The signals on Bank2 are the un-buffered channels correct?
I don't know if they are real, or artifacts, or feedthu?

I may try to update?  This board was sent a week ago, so I would assume it would be close to the latest?

Thanks
Mike

Re: Me too, My Sniffer may be Snuffed! No High Traces?

Reply #3 –
If you upgrade your OLS bistream to 2.04 and your pic to 0.4 a test mode is available. Ian made a topic with the fimware/bitstream here : http://dangerousprototypes.com/forum/in ... opic=580.0

You connect the unbuffered pins 1-to-1 to the buffered pins. I described the proces here: http://dangerousprototypes.com/forum/in ... 25#msg5225

I still havent found the time to solder a proper header and do the test again.. But the diagram 'Test_waveform.png' should be the result if everyhting is ok.

BTW assuming is never good ;) 'assumption is the mother of all fuckups' and 'assume, makes an ass out u and me' ;) To check which software is loaded issue the command: 'pump-loader -p:comx -status' (see http://dangerousprototypes.com/2010/05/ ... om-update/ )

Re: Me too, My Sniffer may be Snuffed! No High Traces?

Reply #4 –
Once you have upgraded the firmware to v04, v05, v06 (all three work with bitstream v2.04)
and the bitstream to v2.04 the header/channel assignment is as follows:

Where you will see the signals depends on the Numbering Scheme you select:

1. If you select Numbering Scheme Inside and check Channel Groups 0, 1, 2 and 3 then
    - channel groups 0 and 1 are on the buffered header (right side with the USB connector pointing to the left) - channel group 0: 0-7, channel group 1: 8-15
    - channel groups 2 and 3 are on the unbuffered header (bottom - the header should not have been populated when you received the board) - channel group 2: 16-23, channel group 3: 24-31 (inside numbers! ;)
    - you can sample max 2k (max. possible Recording Size)!
    if you select Numbering Scheme Inside and check Channel Groups 0 and 1 only then
    - channel groups 0 and 1 are on the buffered header (right side with the USB connector pointing to the left) - channel group 0: 0-7, channel group 1: 8-15 (inside numbers! ;)
    - you can sample max 4k (max. possible Recording Size)!

2. If you select Numbering Scheme Outside and check Channel Groups 0, 1, 2 and 3 then
    - channel groups 0 and 1 are on the unbuffered header (bottom side with the USB connector pointing to the left) - channel group 0: 0-7, channel group 1: 8-15
    - channel groups 2 and 3 are on the buffered header (right side) - channel group 2: 16-23, channel group 3: 24-31 (outside numbers! ;)
    - you can sample max 2k (max. possible Recording Size)!
    if you select Numbering Scheme Outside and check Channel Groups 0 and 1 only then
    - channel groups 0 and 1 are on the unbuffered header (bottom side with the USB connector pointing to the left) - channel group 0: 0-7, channel group 1: 8-15 (outside numbers! ;)
    - you can sample max 4k (max. possible Recording Size)!

3. If you select Test Mode then you must connect pin 0/16 on the unbuffered header (bottom) to pin 16/0 and so on on the buffered header (right) with fly wires or a flat ribbon cable (ubh = unbuffered header/bottom, bh = buffered header/right side):

    ubh 0/16 --> bh 16/0, ubh 1/17 --> bh 17/1, ubh 2/18 --> bh 18/2 ... ubh 15/31 --> bh 31/15

    There is NO need to connect ubh GND -->  bh GND ... AND make sure that you will NOT connect any of the channels of the unbufferd header to GND on the bufferd header!!!!

Re: Me too, My Sniffer may be Snuffed! No High Traces?

Reply #5 –
Why you can't you connect the gnd to an unbuffered pin? AFAIK there is just one gnd for the whole board. The unbuffered pins max input level is 3v3 ( the minimum I guess is 0v or GND :) )

Re: Me too, My Sniffer may be Snuffed! No High Traces?

Reply #6 –
In test mode the unbuffered header is the output! Connecting an output directly to GND is potentially destructive to the FPGA - at least to the I/O buffer of the grounded output signal.

Re: Me too, My Sniffer may be Snuffed! No High Traces?

Reply #7 –
Ha ok!

Didn't see you meant it testmode, but you are totally right. Good point!

Re: Me too, My Sniffer may be Snuffed! No High Traces?

Reply #8 –
Thanks for the extensive reply.  I knew nothing about a new Client.  I no nothing about Java and am running the exe file in XP.  Will someone eventually make a upgraded exe program?


I went to the Bus Pirate location and re-installed USB.  System reports "CDC RS-232 Emulator Demo".

Same driver I have seen before from Microchip demo?

The run java batch file gives errors.  "Unsupported Class" ......... Exception in Main?

I am running the old exe Client, it has problems connecting to the USB.
I push the Reset button a few times on the Sniffer and Re open the exe file until I can get the sniffer to sample?
So I am able to Sniff, especially if I leave it going.
My first sniffer had a solder or connection problem around the buffer?
This one seems OK, even though the Trig LED jumped off.

Thanks
Mike

Re: Me too, My Sniffer may be Snuffed! No High Traces?

Reply #9 –
This post was split to keep things focused on single, not necessarily directly related issues:

1. This thread is/was about your suspicion that your Sniffer might be snuffed ... by what you reported it seems the hardware is working fine.

2. Software realted issues/questions were split, and moved to this new thread: Upgrade, what files to use?  - the questions were answered in detail, too.

3. Your question regarding where to find the "USB driver" has been answered in your original thread "USB Logic Sniffer CDC-232 file Location?"

Re: Me too, My Sniffer may be Snuffed! No High Traces?

Reply #10 –
[quote author="MotoDog"]
The run java batch file gives errors.  "Unsupported Class" ......... Exception in Main?

I am running the old exe Client, it has problems connecting to the USB.
I push the Reset button a few times on the Sniffer and Re open the exe file until I can get the sniffer to sample?
So I am able to Sniff, especially if I leave it going.
My first sniffer had a solder or connection problem around the buffer?
This one seems OK, even though the Trig LED jumped off.
[/quote]

There is no .exe (yet). The java error is because you don't have the rxtx library installed in your default java.

Copy the new analyzer.jar into the directory where the analyzer.exe (where BFP_Logic_Analyzer_1.0.zip is extracted) is and overwrite the old one. Also delete the exe (to prevent starting the wrong client). start a command box and issue the following command:

Code: [Select]
D:BFP_Logic_Analyzer_1.0>jre1.6.0_14-b08binjava -jar analyzer.jar