DAC60501: Sequential Update Max Frequency a lot Lower than Expected.

Hi Bill,

Are you using the device in synchronous mode and sending LDAC commands, or in asynchronous mode?  Can you capture a write commands (or series of write commands) with a scope? 

Thanks,

Paul

Hi Paul,

I'm running in asynchronous mode.

Bill

It will take me a minute to capture the spi output on the scope.  Hold on...

This is not at 4Mhz, but rather at 20Mhz fyi

This is not at *40Mhz*... This is at 20Mhz

So to clarify, you are sending SPI with a 20MHz clock.  When you send a 24-bit frame every 6µs (166kSPS), you do not see an update. If you reduce the sample rate (but still use 20MHz SCLK) and send a 24-bit frame every 140µs (7kSPS), you do see an update?

That is to say that your SCLK is always 20MHz, but your sample rate is changing. Correct? Or are you reducing SCLK proportionally to get a slower sample rate?

Yes, this is correct.

I misstated 40MHz SCLK in the original post.  SCLK has not changed from 20MHz across all experiments.  SCLK has been 20Mhz the whole time.

I am reducing the sample rate by introducing a delay between sample updates.  When the delay >= 140us, updates work.  When the delay <= 130us, updates don't work.

Bill

Do you have any cap load on the output? Is there any analog change on the output? For example, if you try to create a square wave of 0x400 and 0xC00 codes, do you see any change?

No cap load on the output... beyond parasitics which I estimate to be in the 5pf range.

For testing, I have been using a ramp pattern... 0xFFF -> 0x000 in 0x001 steps.

I am not at my bench at the moment, so cannot try the square wave pattern for you.  Do you feel there would be significantly more diagnostic information in a square wave than in the ramp?

Opps... forgot about the 13pf from my scope probe, so cap load ~18pf.

So what is the analog behavior on the output? Does it just look like there is no output, and the voltage stays at 0V?  If you issue the same code repeatably , say 0x800, while updating every 6µs, do you see any analog change.

Thanks,

Paul

The analog output stays at 0.

Thanks,

BIll

Hi Bill, when you get a chance to do the square wave experiment, can you sweep the delay time between updates to see if there is some critical delay that causes the issue?

I do not see any issue with your SPI frame, and the device should support a higher sample rate than what you are seeing. 

Thanks,

Paul