DAC82001: There are strange jumps in the output signal of DAC

Hi SS, 

Are you able to isolate which codes these jumps are happening at? It's a bit hard to tell from your figure. That would help us confirm if the jump is happening at transition code like you pointed out, or some other cause. It looks repeated at consistent intervals which could be a result of a large switching transition, or it could mean there is some repeated bug in your code when these jumps are happening. 

Best,

Katlynne Jones

Hi, Katlynne

Thanks advance for your help.

In our observation, when we set following DAC datas, the output have unexpected jump:

0x801F<->0x8020,

0x805F<->0x8060,

0x809F<->0x80A0,

0x80DF<->0x80E0

when from low to high, the output is unexpected drop,

when from high to low, the output is unexpected rise.  

BR,

Steven

Hi SS,

As these are major carry codes, I think we should confirm if we have some digital issue here.  My concern is that the SPI timing is being violated here and maybe we have some data getting shifted on the wrong clock edge.  Can you please capture an oscilloscope image of the the SPI frame showing the edge relationships?

Hi, Paul

Many thanks for your information.

You means we should capture a snapshot by oscilloscope to check the timing between SPI SCK and MOSI when the issue happened? 

Actually, we had previously suspected an issue with SPI by reducing its clock from 32MHz to 3.2MHz, but the problem still persists.

and we checked spi timing by oscilloscope and it seems okay.    

BR,

Steven

Hi, Paul

These abnormal outputs are very regular, appearing at these points every time. Guess may not a timing issue?

BR,

Steven

Hi SS, 

Yes, capture the oscilloscope shot of a SPI write when the issue happens (the new code that sees the jump). Even with reducing the SPI frequency, there could still be some timing violation, and repeatability of the issue wouldn't necessarily rule out a timing issue. 

Best,

Katlynne Jones

Hi, Katlynne

Sorry for late reponse. We tried to do some compastion these days, but still observed this strange jump in some channels.

Following is the snapshot of 0x801F<->0x8020, seems clk is okay??

BR,

Steven

Hi Steven, 

Thanks for sharing the plots. The SPI frame looks ok to me. You are using a slow SPI clock, but I wouldn't think that's the cause of this issue. And there's nothing that would indicate data being shifted on the wrong edge. And the data looks correct on the scope, so no issue in your code sending the wrong command.

Just to double check, were you monitoring the output during the sequence and checking if this is where the jump occurs? I can't tell if this is what you are showing on CH1 of the scope. 

Best,

Katlynne Jones

Hi, Katlynne

Thanks for your response.

Yes, we monitoring the output during the sequence and the jump occurs. CH1 of the scope is SYNC of DAC82001.

BR,

Steven

Hi Steven, 

Thanks for confirming. In that case it does seem like the jumps are due to the major carry transitions of the DAC. 

Did you have luck with the compensation you tried? How were you trying to compensate? 

Best,

Katlynne Jones

Hi, Katlynne

Yes, we tried compensation, very simply, just skip the abnormal point and all following DAC points plus/minus same skip value.

Currently the compensation works well.

But we still want to known if this is output characteristics of DAC to ensure our program is okay for all scenario.

And we are curiosity why ony the switch between  b'011111 and b'100000 (See following observed abnormal list points)have this observation, but others carry transitions don't have this obervation.

0x801F<->0x8020,   

0x805F<->0x8060,

0x809F<->0x80A0,

0x80DF<->0x80E0

BR,

Steven

Hi, Katlynne 

Great, the information is very important to us to exclude our program's issue.

Many thanks for your help. 

BR,

Steven