AD5270 Command Operation

Hi Akira-san,

Apologies for the delay.

My interpretation of the commands are the following:

1. 0x2400 = device is removed from shutdown mode

2. 0x1000 = software reset

3. 0x1C02 = allow update of wiper position through digital interface

4. 0x0400 = write zero-scale

From these commands, the expected behavior of the digipot output is at zero scale. Could you try removing the software reset and check if the part is responding? If the part responds, could you check your timing in the SPI lines as you may be violating a specification there.

Regards,

Mark

Hi Mark,

Thank you for your answer. Our customer tried it, but AD5270 was no responding. On the other hand, the customer tried to change the timing of the SPI.

The following is NG case;

And the following is OK case;

I can't find the timing specification about this area except t7 and t9. In my understanding, the NG case doesn't go against the specifications not only t7 and t9 but also others. Are there any constraints of timing about this area except Table.7 in the datasheet of AD5270?

Best regards,

Akira

Hi Akira-san,

Using the evaluation board, I tried the 4 commands in the same sequence and the digipot does not change the wiper position. When I removed the software reset command, everything works perfectly.

Looking at your scope shots, there should be no problem with this. How much delay did you add for the OK case?

Regards,

Mark

Hi Mark.

Thank you for your reply. Our customer removed the software reset command and sent the other commands by using NG case timing. But the wiper didn't work. And changing the timing to OK case, the wiper did work. In the OK case,

Sync edge set forward 40ns from SCLK edge. Do you have any idea about this phenomenon?

Thanks & Regards,

 Akira

Hi Akira-san,

The scope shots that you have previously shared are from a logic analyzer, is this accurate? If so, can you ask the customer for actual oscilloscope shots? I have not seen this phenomenon so i would like to understand it better.

Also, just for clarification, did you try the commands with Reset using the OK case?

Regards,

Mark

Hi Mark.

I got the oscilloscope shots. Please refer followings;

(1) NG case;   CH1=SCLK, CH2=SYNC

(1)' NG case (expanding) ; CH1=SCLK, CH2=SYNC

(2) OK case 1; CH1=SCLK, CH2=SYNC

(2)' OK case 1 (expanding); CH1=SCLK, CH2=SYNC

(3) OK case 2 ; CH1=SCLK, CH2=SYNC

(3)' OK case 2 (expanding); CH1=SCLK, CH2=SYNC

According to the customer's evaluation result, the SYNC edge needs to rise before the SCLK edge. Do you have any idea about the cause of these phenomenon? Are there any timing constraint between the rise edge of SYNC and the rise edge of SCLK? Or, is the changing of low to high of SYNC available only SCLK = low case?

>Also, just for clarification, did you try the commands with Reset using the OK case?

When the cusotmer send the command with Reset, the wiper didn't work.

Best regards,

Akira

Hi Akira-san,

We will look into the cases that you suggested. As I have mentioned earlier, I have not seen this phenomenon. We are now in discussion with other teams and we'll get back to you as soon as we have updates.

Regards,

Mark

Hi Akira-san,

I noticed that CLK is high when idle. This would mean that the clock polarity of the SPI transaction is CPOL = 1. Please set the CPOL to 0 (clock is idle low) and CPHA to 1 (data clocks in the falling edge) as indicated in the data sheet for proper operation.

Regards,

Mark

Hi Mark,

Thank you so much for your answer. As you said, the Clock was high when idle.

 By the way, does AD7520 support only mode 1 (CPOL=0, CPHA=1)? And can we change the CPHA mode of AD5270? If yes, how should we do?

Thanks & Regards,

Akira

Hi Akira,

As the DS states, the only SPI mode supported by the AD5270 is CPOL = 0, CPHA = 1, and there is no way to change the mode.

You can check http://www.analog.com/media/cn/technical-documentation/application-notes/AN-1248.pdf if you are considering to use external components to operate the device in a different mode.

Regards,

Miguel