[SOLVED] Op Amp with open drain output + delay circuit.

Hello dudes!

I am wondering if the next circuit will work a I expected.







I want to design a signal conditioning stage with some opamps working like comparators. These opamps compare an incomming voltage with a reference voltage and they generate an "enable" signal to control a digital buffer. In the previous pic I have used NMOS trt's in order to generate an open drain output. I have simulated the behaviour of the circuit and it looks working as I expected. However i would like to ask someone in the forum if I could have any problem in the real life...Maybe I am skipping something.

Further, once the buffer is disabled, it is also important to control the enable of the buffer. In order to do that I have thought to use combination of RC network and opamp. I have also simulated the circuit and it looks works fine but I would like to know the opinion of the people here, because maybe they have better ideas...

In the simulations every opamp had a one impuse spaced between them 100us.



And the time adjustment circuit looks works like I expect. If the pulses are fast, the err signal goes low instantaneously. but it recovers its natural state after the delay introduced by the RC network...



Thanks in advance.

Hi,

a general recommendation: If you want to "compare" a signal then use a comparator.
There´s a good reson why they designed "opamps" and "comparators".
In the forum you will find a discussion going deeper in this regard.

***
To your circuit:
Why this high effort with the MOSFETs?

Why not replace the inputs of each of your "comparators"?
Then the ouput is inverted. Thus you don´t need the inverting character of the left side MOSFETS.
--> For push-pull outputs a simple diode will do instead of the mosfets.
--> If you use comparators with open_drain outputs then you dont need the left side mosfets at all. Just connect the outputs.

Klaus

Hi KlauST

Thanks for your quick answer.

i think that I had not explained myself very well. The LT1716 is a Rail-to-Rail comparator...So I have been using the righ part and it looks that it has also an open drain output...

http://cds.linear.com/docs/en/datasheet/1716fa.pdf

i have also followed your recommendations and I run new simulations. But As you can see in the next pictures, the pulses don't achieve 0V.....Any suggestion to reach the 0v?







Thanks again

Hi,

I just saw your headline - and didn´t open the datasheet
--> yes it´s a comparator.

it looks that it has also an open drain output...

Click to expand...

It "looks" or you "know"? this are different things.
Reading the datasheet gives "knowledge"

On the first page it says:
"The output stage includes a class “B” pull-up current
source, eliminating the need for an external resistive
pull-up and saving power. Output voltage swings to within
35mV of the negative supply and 55mV of the positive
supply, which makes the comparator a good choice for
low voltage single supply operation."

Additionally in the "electrical specification table" it says:
ISC + Output Source Current VOVERDRIVE = 5mV, VOUT = –14V l 70 105 μA

--> So every comparator actively drives output current when output is HIGH. --> It´s not a true "open drain" output.

But it is relatively low current, thus the LOW voltage should be less than 0.3V.

Please disconnect the negative feedback resistors. Comparators are not meant to operate with negative feedback.

Tell us what happens.

Klaus

hi KlauST,

I have already run the simulations again and it did not reach the 0V either...





Perhaps, I should use another comparator...

Greetings

KlausST said:

Hi,

I just saw your headline - and didn´t open the datasheet
--> yes it´s a comparator.


It "looks" or you "know"? this are different things.
Reading the datasheet gives "knowledge"

On the first page it says:
"The output stage includes a class “B” pull-up current
source, eliminating the need for an external resistive
pull-up and saving power. Output voltage swings to within
35mV of the negative supply and 55mV of the positive
supply, which makes the comparator a good choice for
low voltage single supply operation."

Additionally in the "electrical specification table" it says:
ISC + Output Source Current VOVERDRIVE = 5mV, VOUT = –14V l 70 105 μA

--> So every comparator actively drives output current when output is HIGH. --> It´s not a true "open drain" output.

But it is relatively low current, thus the LOW voltage should be less than 0.3V.

Please disconnect the negative feedback resistors. Comparators are not meant to operate with negative feedback.

Tell us what happens.

Klaus

Click to expand...

Hi,

I assume the datasheet is correct.
Possible issues are:
* the signals are too fast
* the overdrive is too low (datasheet says 5.5 ...10us @ 100mV overdrive)
* the pull up current is too high

Klaus

Hi KlauST,

you were right, the input signals were too fast. I have increased the pulse width from 5uS to 10uS and now it is able to reach the 0V. However 5us input signals are a requirement design....

so I think that this solution fits with everything:





Thanks.

KlausST said:

Hi,

I assume the datasheet is correct.
Possible issues are:
* the signals are too fast
* the overdrive is too low (datasheet says 5.5 ...10us @ 100mV overdrive)
* the pull up current is too high

Klaus

Click to expand...

You have already answered the question. You should definitely use a comparator with sufficient speed.

Using a comparator with specified 5 to 10 us response time for an input signal with similar pulse width is simply wrong.

Yes

I found this part from TI: TVL1701 and it works nice now:



ç

thanks so much for your help!

Greetings ;-)

FvM said:

You have already answered the question. You should definitely use a comparator with sufficient speed.

Using a comparator with specified 5 to 10 us response time for an input signal with similar pulse width is simply wrong.

Click to expand...

- - - Updated - - -

Yes

I found this part from TI: TVL1701 and it works nice now:



ç

thanks so much for your help!

Greetings ;-)

FvM said:

You have already answered the question. You should definitely use a comparator with sufficient speed.

Using a comparator with specified 5 to 10 us response time for an input signal with similar pulse width is simply wrong.

Click to expand...