DS91M040: Enable time from receiver to driver

Alessandro

Do you have a scope capture showing the toggling of DE/RE, the input signal, and the output signal?

Thanks

David

David,

in the scope capture you can see the non inverted output signal A (violet) and the inverted output signal B (green).

The toggling of DE/RE is exactly and only when B starts to toggle and the input signal is equals to A.

I hope this explanation is enough for you, otherwise I restore the set-up test in laboratory and take a new scope picture.

Alessandro

This is the scope capture with all signals. 

Alessandro

What is connected at the A and B? It looks like when DE/RE is driven from low to high, B tries to drive high, but something is pulling it down for ~3.7us and then release it.

Is it possible to disconnect what is connected at the A and B and connect 50ohm to GND on A and B and repeat the test?

Thanks

David

Hi Alessandro,

Below is a simulation of the behavior of and A and B after the driver is enabled generated with the IBIS model:

The delay is only a couple of nanoseconds. There may be something on your B lane that is pulling it down.

Regards,

I.K. 

In the measures carried out to date, there were 100 ohms between A and GND and between B and GND. Moreover, in practice we have 2 DS91M040 that communicate with each other and therefore the load is nothing other than a DS91M040 always set in reception (with others 100Ohm A-GND and B-GND).

This morning I repeated the measures disconnecting the cable and with 50ohm between  A and GND and B and GND: nothing is changed.

It is very strange since after this transition, from receiver to transmitter, the driver works perfectly. It seems that for a few microseconds the driver struggles to drive the load.

Alessandro

We will see if we can duplicate this issue on our EVM and get back to you. 

But with cable already disconnected and only 50ohm load on output A and B, DS91M040 should not have issue driving this load.

Can you please check VDD? Are you seeing a VDD voltage drop when you switching from input to output mode? 

DS91M040 also has a thermal GND pad, can you please check to see if you have a solid connection between the thermal GND pad and the PCB board ground?

Thanks

David

David,

first of all, thank you for your fast support. I checked VDD and it is always constant to 3.3V.  The DS91M040 Thermal pad is correctly connected to ground.

Best regards

Alessandro

Alessandro

Thanks for the update, please give us couple days to see if we can duplicate this in our lab.

Thanks

David

Alessandro,

Sorry for the delay. The EVM we ordered finally arrived today, and we'll be able to look at this in the lab early next week.

Regards,

I.K.

Alessandro,

Here's what we see in the lab: 

It only takes about 3.7ns from the time DE is pulled high before the outputs begin transmitting valid data. 

I believe the behavior you're seeing is due to the fact that you are not inputting any data until microseconds after DE is pulled high. There is a weak internal pulldown on DE but it does not inhibit the enable time. 

Regards,

I.K.