Problems with ADIS16265 EVAL Board ...

When you refer to the evaluation board, which part number are you referring to?  I suspect you are referring to the ADIS16265/PCBZ, but it would be nice to know for sure.  How are you connecting this board to your processor?  While we are glad that the other SPI device is communicating will in your system but, by itself, that does not assure successful communication with other devices.  Have you tried reading the PROD_ID register, while proving all of the SPI signals with a scope?  This allows you to check for the proper bit coding, clock polarity, clock phase, timing and signal integrity.  There are more details on this, for another production that has a similar interface protocol, in the following discussion.

http://ez.analog.com/message/44238#44238

I hope that helps and look forward to helping with any follow-up questions.

Best,

NevadaMark

Dear NevadaMark,

I will check all these points and will come back with results.

Thanks for the fast support, up to now

Best regards

Ritchie

Hello,

I check the voltage on the break board (top of the connector) and get 4.95 V.

I also checked the Voltage on the board, SMD pads of the connector J1. Same value.

The board give back a stable value of dez 16265 for the product ID, but the diagnose bit is toggling sometimes the bit 0 (low voltage).

Up to now, only two commands are running on the SPI bus. (Product ID 0x56 and Diag_stat 0x3C).

Could still not found the reason. I there a special point on the board, where I have to check the voltage.

Connectors are fine so far.

Best regards

Ritchie

Since you are able to read the PROD_ID register on a regular basis, I would focus my scope probing on the SPI signals, and the VDD pin.  You might be able to catch a "bad read," while also monitoring the VDD pin.  This is strange.  I am starting to wonder if there is a poor solder joint or another potential connection issue.

Best,

NevadaMark

Hi,

I have check the signal state. All signals are fine (rise and fall of voltage is o.k.).

Only on Pin 4of the Board (DOUT), the signal stage is not correct.

I has steps in between 3.3V and 5.0 Volt. See attached picture. I also tried this with the internal pull up's of the controller and a external pull up (6.8k). Always the same. If no pullup is configured, the signal is only going up to 3.3V as high level).

Edit2:

I read in the documentation, that the signals are 5V compatible, but I will add a level converter (74HCT125) for this pin to get the full 5V. A normal Pull up will not help in my configuration so far.

Edit1:

Just in addition, do I have to mount a condenser for C1 / C2 on the board, or is this a option ?

Best regards

Ritchie

1. Answer

You can see only the signal level, which is up to 3.3V for High level. My controller (ATMEGA644) is running with 5V and I am not 100% sure, If the SPI Interface will running proper with such a high level.

Yes, 5V is only reach by tristate level.

2. Answer

Yes

3.Answer

Pull Up installed.

4. Answer

Nothing installed up to now

Chip 74HCT125 is on the way (already order) and I will see, what will happen, when the voltage level is proper 5V.

Best regards

Ritchie

1. What is the attached picture showing us? I can't see the transitions and wonder if we are losing important information (noise, ringing, etc)
2. You do not need to run 5V on your logic. 
3. If the DOUT is moving to +5V, it is because the ADIS16265 has put it in a tri-state (very high output impedance) and it is floating there due to external influences.
4. You do not need C1 or C2 in place.

Helpful screen shots (I hope are not confusing)

The SPI signals have some noise but I hope you can make it out as I listed them below.

1) Both 16 bit signals.

2) The Din register = 5600

3) The Dout register = 3f89

Thank you, !

Ritchie: This type of signal representation is more typical of our troubleshooting approaches.  platform is not running in full duplex mode, but it does allow you to see the DIN = 0x5600 cycle, followed by the DOUT = 0x3F89 (16,265) cycle .  If you are able to write DIN = 0x5600 on every cycle, that would allow you to focus on one communication cycle, where DIN = 0x5600 and DOUT = 0x3F89, on every cycle (except the first one).

Best,

NevadaMark

Hi NevadaMark,

my circuit to get a proper 5V Signal does not work. I only know a circuit for level shifting with a 3.3V and 5V as supply voltage.

Do you have any good circuit for this chip to get a proper 5V Signal, if you have only a 5V supply.

Best regards

Ritchie