Digital I/O

cancel
Showing results for 
Search instead for 
Did you mean: 

Impedance / Termination Problems

I'm using a 6551 card to drive a nand flash memory and my program seems to be ok but i got some troubles with the data that my flash memory sends to me.

If i make consecutively 2 reads of a same flash memory, I have several differences between the two binaries created. 

Clock of the system is 10MHz. 

 

Indeed, i check the 6551 outputs and it's like (and even worst) the right figure of this picture :

 

 

 

I test one type of termination (parallel) but no better results.

 

Is there a solution ?

 

Thanks for helping me. 

Message Edité par xela75 le 03-27-2009 11:01 AM
Message Edité par xela75 le 03-27-2009 11:01 AM
Message Edité par xela75 le 03-27-2009 11:02 AM
0 Kudos
Message 1 of 4
(3,467 Views)

Hello xela75,

 

It looks like you have already found the developer zone article on  Proper Termination for High-Speed Digital I/O Applications. You mention that you are using the NI 6551.  If you look inside the NI Digital Waveform Generator-Analyzer Help under Devices >> NI 655x >> Hardware Architecture and click on Channel Electronics you will see a diagram of the a generation/acquisition channel.  This diagram shows that the generation channel has a source impedance of 50 Ohm.  For most applications this means that if you have a 50 Ohm characteristic impedance cables going from the 6551 to your nand flash memory (DUT) you should not need any termination at your DUT (note:  the SCH-C68-68-D4 cable on the 6551 product page is a 50 Ohm cable).  

 

You mention that you tried placing parallel termination at your DUT, but this had no effect.  This suggests that you have having reflections in other locations along your cable path.  I would verify that your connections and insure that you are providing a 50ohm cable path from the 6551 to your nand memory   If all of your cable connections are good, I would verify where you are probing the waveform.  To understand what ringing/reflections your nand memory is seeing you will need to probe as close as possible (perhaps on the pins of the memory) to measure the input waveform.  

 

Regards,

Jesse O. | National Instruments R&D
0 Kudos
Message 2 of 4
(3,454 Views)

Looks like cabling/interconnection capacitance is too high.

 

-AK2DM

~~~~~~~~~~~~~~~~~~~~~~~~~~
"It’s the questions that drive us.”
~~~~~~~~~~~~~~~~~~~~~~~~~~
0 Kudos
Message 3 of 4
(3,443 Views)

Ok thank you for helping me !

 

I'm going to check every connections because it's what i understood : I don't need to make specific terminations in my case.

I confirm you I have a SCH-C68-68-D4 cable.

Message Edité par xela75 le 04-01-2009 09:07 AM
0 Kudos
Message 4 of 4
(3,414 Views)