THS4141: Slew rate issue on Pspice for TI model

Hi Téo,

the slew rate specification is only valid for the test condition. See figure 6 of datasheet

Kai

Hello Téo,

   For the output voltage and rise time for a square wave, you would need a large signal bandwidth of 0.35/15ns = 23MHz. This translates to an amplifier with a slew rate of 2 * pi * 23*10^6 * 15V = 2261V/us. We can help suggest another amplifier that would better suit your requirements. What is the maximum amount of rise time your design can handle, and would you be able to share schematic/sim file?

Thank you,
Sima

Hello Sima,

Thanks for this complete answer, as we want a precision of ~10 ns in the end, we want the rise time to be as fast as possible to match this requirements.
Please find attached the schematic/ simulation files

Tx Math Modification.zip

Best regards

Thank You 
Téo

Also, i don't understand how you get this 23MHz.

It seemed to me that the minimum slew rate was given by:  Slew rate = 2 π f V where f is my signal frequency (200kHz)  and V the maximum voltage of my signal.

So in this case it gives me : Slew rate = 18.8 V/us 
But of course, in this case the slew rate isn't high enough to match my 10 ns criteria.

But what we want is 7V/10ns → 700V/us 
Isn't that enough ?
Please not that we are looking for a Q1 Qualified component here

Best regards
Téo

Hi Teo,

can you post your schematic as a picture here, please?

Kai

Hi kai,
Yes of course, here it is.

Téo

Hi Teo,

I get this result with TINA-TI:

teo_ths4141.TSC

Looks like the slew rate is 264.5V/µs

Kai

Alright.

Is it something wrong with the PSPICE model of the THS4141 then ? What makes it so different from what you have on TINA ?
Even like this, i guess this is not enough for us.
We would like to have the rising edge in about 10ns.
Do you have a Q1 qualified reference that can match those criteria ? 

Thanks for your help

Téo

Hi Teo,

I imported the same PSPICE model into TINA-TI. So it should be ok?

What load is connected to the output of your fully differential amplifier?

Kai

Well

I have the same schematic on PSPICE for Ti, with no load like you.
But I have a ~200ns Rising Edge for each output

Téo

Hello Téo,

   The slew rate by taking the signal frequency and the equation 2piFV defines a sine wave rather than a pulse/square wave. We would have to use the rise time in this case and use the equation 0.35/rise time (this is a derived equation). 

   Thank you for the Pspice folder, and thank you Kai for Tina-TI version. It does look like PSpice model is giving around 200ns rise time while Tina-TI version gives around 25ns from 0 to 100 percent. I will need to look into this and get back to you. 

   For alternative parts, this is our list of high power supply FDAs. THS4501 would work in terms of slew rate, but unfortunately has a smaller output voltage range compared to THS4141. THS4141 is not automotive qualitied, here is the link of our automotive qualified FDAs. Unfortunately, neither of these FDAs are high power design of 15V or greater. 

   Another option would be to use two single ended amplifiers rather than FDA if space and distortion permit. 

Thank you,
Sima

Alright ! 
Thanks for this. 

What component will you advise if we go for the two single ended amplifiers option ? I have been looking for components that can provide what we need but didn't find the right one yet.

If we go for a not Q1 solution first for test what component would you advise then ? Maybe the choice is easier like this 

Best regards

Téo

Hello Téo,

   Here is a filtered list that would work for high power supplies + high slew rate. Unfortunately, none of these are automotive qualified. 

   I used THS3092 as an example below for fast output rise time + differential configuration:  

   THS3092_fastpulse.TSC

Thanks!
Sima