Hi there. I want to amplify multiple PWM signals generated by a 3.3V MCU. The system specifications are:

1. Input: PWM frequencies up to 1Mhz and fixed duty cycle 50%.
2. Output: 0-24V (don't care if inverted) and sink/source current ~10mA or more. Also trise,tfall must be reasonable small <50ns.
3. ICs with multiple channels should be used in order to save space.

I have searched for anything that would do the job (buffers, line drivers, mosfet drivers etc) but nothing matches the specs needed.
The best choice till now is SN7407 but it has open collector output and not push-pull. Also CD4041UB could be a good choice but it support only up to 18V. Can someone please suggest a specific IC or at least what type of ICs to search? Thanks in advance

 

Welcome to AAC!

What you want is a level shifter. If you can't find one for the voltage you want, look for a comparator with push pull output.

 

Thank you for your response! I have also looked for level shifters (nothing above 18V) and for comparators (nothing above 16V with push-pull output). Any other suggestions?

 

If your requirements exceed the specs of any available parts, you'll need to roll your own. Your sub 50ns rise/fall time will be problematic.

PWM with a fixed duty cycle doesn't make much sense to me.

 

Yes I know that fixed duty cycle seems strange but this is how it is. I mentioned it cause I thought it could make the solution simpler. Thank you for your answers, I will keep searching for a solution. However, I would appreciate it if someone else could suggest a different solution.

 

If the duty-cycle is fixed at 50% then it's not PWM, it's just a square-wave signal.

You might look at MOSFET gate drivers.

 

Ok thanks for your response. I found IX4427 mosfet driver. I will test it and post the results.

 

LM393(dual) and LM339(quad) Comparators work on +36 Volts. Why can't you use those?

http://www.ti.com/lit/ds/symlink/lm393-n.pdf
http://www.ti.com/lit/ds/symlink/lm2901.pdf

 

Correct me if I am wrong but this has open drain outputs and the max output high level current is only 1μA.

 

Correct me if I am wrong but this has open drain outputs and the max output high level current is only 1μA.

The outputs are open collector. The output high current is controlled by the size of the pullup resistor and the load impedance, not by the device itself. So use the comparator output to control either a high-side or a low-side switch depending on your load.

 

From datasheet "The open-drain output allows the user to configure...". But it doesn't matter if it is open drain or open collector because I cannot use pull up resistors. I want push/pull outputs. If I wanted to use pull up resistors I could just use SN7407, as I mentioned in my first post. Thanks for your response.

 

Some octal push/pull drivers over here to 28V -

https://para.maximintegrated.com/en/results.mvp?fam=mosfet_drivers&793=10 to 36&tree=interface

As an aside how many PWM channels do you have to generate ?

Regards, Dana.

 

Thank you very much for your response. Unfortunately this chip doesn't meet the speed requirements. For push-pull mode the typucal t-rise is 500ns and t-fall is 50ns. The best solution would be something like the TC4427, but capable for higher supply voltage (only 18V) and more inputs-outputs for less cost/space. The only IC that seems to meet the requirements is IX4427, but it has only 2 channels.

 

That's of course into 100 pF from spec sheet, timing that is.

Do you need to switch the load to the 24 VDC or is that the load supply buss,
and all you need to do is to get enough current thru load ? If latter use a zener
I series with output to drop output V driver sees ?

Just a thought.

Regards, Dana.

 

No, I need just a push/pull 0-24V ~10mA <50ns output. I could do this by many means but the problems come out when someone wants to do it for multiple signals in a small board. I think the zener would fix a problem that does not apply in this situation. Regarding the trise/tfall of the driver you suggest, I don't think that smaller capacitance would make huge difference. My opinion is based on a similar simulation using the TC4427, where decreasing capacitance didn't make much difference to the trise/tfall.

 

Ok thanks for your response. I found IX4427 mosfet driver. I will test it and post the results.

Just in case this information might be useful for someone else, I have successfully used IX4427 for my project. The result is shown in the attached picture.