TXS0108E: Open side waveform distorted

Steve Clynes

Part Number: TXS0108E
Other Parts Discussed in Thread: LSF0108, LSF0204, TXB0108

I am using the TXS010x devices to level shift between 2.5V and 3.3V and I am seeing some signal distortion.

Below is the level shifter schematic.

In the below waveform the bottom trace is the L side, which is being driven push-pull to 2.5V, the top trace is the H side and is floating. H is being used to monitor the L side but at 3.3V.

The H side does also go to the input of an op-amp but should be loading minimally.

We need this to be bi-directional since these signals are GPIO signals.

Is this expected?

over 6 years ago

0 Emrys Maier over 6 years ago

TI__Guru* 96591 points

Hi Steve,
We will look into this Monday.

0 Emrys Maier over 6 years ago

TI__Guru* 96591 points

Hey Steve,
I don't see the waveform you're referring to - can you re-upload the image?

0 Steve Clynes over 6 years ago in reply to Emrys Maier

TI__Mastermind 43875 points

Try again :)

0 Emrys Maier over 6 years ago in reply to Steve Clynes

TI__Guru* 96591 points

Hey Steve,
Thanks for the scope shot, that actually answers a lot.

It looks like the input is ramping up slower than the one shot fires (~30ns), which is causing the output to be pulled back down again. If you can increase the input edge rate, the problem should go away. If you can get a close-up scope shot of that edge you can probably see the issue there directly.

From the scope shot, I estimated the output charge time to be about 0.2us, which means (with our built in 10k resistor and assuming a perfect RC time constant) the output capacitance is about 20pF, which sounds right for the output + scope probe.

0 Steve Clynes over 6 years ago in reply to Emrys Maier

TI__Mastermind 43875 points

The input rise time (bottom trace) is about 20ns and the input fall time is about 15ns.

These particular signals are under my control so I might be able to speed them up, but I am using the TXS for other signals which are external hence I don't have real control over.

Can you recommend a replacement device family which is more likely to remain monotonic since the ringing is causing our analysis tools some issues?

0 Emrys Maier over 6 years ago in reply to Steve Clynes

TI__Guru* 96591 points

The LSF series of translators (LSF0204, LSF0108, etc) lacks the one-shot that's causing the problems, so that would be my first recommendation.

0 Emrys Maier over 6 years ago in reply to Emrys Maier

TI__Guru* 96591 points

Looking at this again, I would actually recommend the TXB series -- as long as the loading is light and the driver can handle ~3mA then it will have much better signal integrity than the LSF or the TXS.

TXB0108

0 Steve Clynes over 6 years ago in reply to Emrys Maier

TI__Mastermind 43875 points

OK, thanks.

I won't be able to replace all the level shifters in my system since some are used for I2C, but I will replace the ones that are not open drain and see how that goes.

BR,

Steve

Logic

Logic forum

TXS0108E: Open side waveform distorted