TLV3502-Q1: TLV3502

Robin,

The TLV3202 has 1.2mV of typical hysteresis. If you think your signal is clean (and fast) enough such that it won't falsely trigger within 1.2mV of your respective trip points, then you don't need to design an external hysteresis network.

For the channel monitoring the voltage level of your signal, what is your voltage threshold you want to set? What are the values of all your signal nodes (vd1sns, voutp1, vln_smpl)?

For the channel monitoring ZCD, the inverting terminal should be tied to ground.

Regards,
Jonny

Jonny

Thanks a lot for the review.

Pl see the 2311.fast_detector_tlv3502.pdfattached correction of the ZCD  hook up with amplitudes of the other inputs. I will look into the TLV3202 meanwhile.

Jonny

I did not give you all the answers in the previous "reply".

I did a quick simulation using lT1011...just because it is so very handy! So that you can see the input comp levels.

I am going to see if TINA-TI has models to do the same.

Clearly, there is some difference: see what is required to do ZCD  in this comparator: I had to hook up the NI to GND ref.

Would TLV3502/ 3202  function as you indicated: for ZCD, I need to hook up the INV to gnd ref?

fast_detector_tlv3502sims.pdf

Jonny

I found TLV349 in TINA-TI model.

It is quite different in compare mode.

In fact, connecting INV to GND with a capacitor as +ve feedback looks to be a better circuit.

Will this be true for TLV3502 or 3202 as well?

If so, I will choose this configuration for sure.

thnx

robin

Hi Robin,

in any case I would add a hysteresis. A fast comparator is always profiting from it. And the TLV3502 is very fast! I would also add a small cap from the -input to GND, if the bandwidth requirements allow this. This minimizes stray coupling from the output to the -input. Also follow the layout guidelines given in section 10.1 of datasheet.

Kai