LMX2594: Problems with readback not working

Jason Matusiak

Part Number: LMX2594
Other Parts Discussed in Thread: USB2ANY

I have an LMX2594 dev board I am trying to talk to but having issues. I //think// I am programming it correctly, but I can't be sure because the readback doesn't seem to be responding properly.

I am using a Silicon Labs CP2130 and a Linux.

Looking on a scope I see the CSB go low and I see the request start via a readwrite mechanism. What is odd is the first MOSI byte is the address I want to read followed by the two dummy bytes to keep the clock going. I would then expect 2 bytes of the registers value back during those two dummy bytes. What I see is that the TI part starts responding on that very first clock cycle with the value of register 0, and the third byte is the register address I had sent.

It doesn't seem to be waiting for the register address at all since it starts blasting data on the first clock pulse.

What am I missing here?

over 5 years ago

0 Jason Matusiak over 5 years ago

Prodigy 140 points

I attached a picture of the scope to give you an idea what I mean.

(Actually it looks like I don't have an attach option. Maybe I haven't posted enough to be allowed it. I'll supply a link)

0 Jason Matusiak over 5 years ago in reply to Jason Matusiak

Prodigy 140 points

OK, I see that the r/w bit wasn't set in the address portion, so I made the mod. It is doing the exact same thing though. Why would it ignore the bit?.... I attached an updated picture.

0 Jason Matusiak over 5 years ago in reply to Jason Matusiak

Prodigy 140 points

OK, I got it to work, but I don't understand it. I pulled out the TI SPI board and monitored its interaction with the LMX3594 and found something interesting, the clock idles at level high!!!!! I made the change with my CP2130 and it works now.

Why does the clock need to idle high when the data sheet clearly slows it idling low?

0 Vibhu Vanjari over 5 years ago in reply to Jason Matusiak

TI__Genius 11900 points

Hello Jason,

If you are referring to "Figure 1. Serial Data Input Timing Diagram", the important thing to note is "The CSB must be held low for data to be clocked. Device will ignore clock pulses if CSB is held high." The diagram shows the CSB pin going low, referring to a potential next register write after a time period of holding the signal high, "tEWH".

Thanks,

Vibhu

0 Jason Matusiak over 5 years ago in reply to Vibhu Vanjari

Prodigy 140 points

No, that isn't what I mean. The problem was that the clock seems to need to be idling high, but the document says otherwise, Here is an image of what I mean:

0 Vibhu Vanjari over 5 years ago in reply to Jason Matusiak

TI__Genius 11900 points

Hello Jason,

I see, you were referring to SCK signal. Is that the only change that was needed? That is, you were idling SCK low, it wasn't working, while the reference pro board idles SCK high? When you made this change you were able to communicate?

Thanks,

Vibhu

0 Jason Matusiak over 5 years ago in reply to Vibhu Vanjari

Prodigy 140 points

That is exactly right.

0 Vibhu Vanjari over 5 years ago in reply to Jason Matusiak

TI__Genius 11900 points

Hello Jason,

The SCK can be idle high or low, this is not the root cause of the issue you were seeing. The reference pro board / USB2ANY is implemented with idle high.

Can you tell me what the blue and red channels are indicating?

How is your MUXOUT pin connected on your board?

Thanks,

Vibhu

Clock & timing

Clock & timing forum

LMX2594: Problems with readback not working