Hi,

I have a transducer driving circuit composed of a push-pull network of N & P channel FETs, driven by a microprocessor with a 40KHz TTL output through 6 inverters in parallel. The circuit was successfully switching 48 volts to the transducer until I wanted more voltage. I replaced the 48 volt source with an EMCO high voltage unit and biased it to about 140 volts. I upgraded my P & N channel FETs to a higher rating. Problem is now the FETs won't switch the new voltage source through. Does anyone know what the problem might be? Do I need to increase the power of the driving TTL signal? What minimum conditions need to be met for FET switching that maybe I am not meeting?

http://www.emcohighvoltage.com/

Thanks

 

I was surprised no one posted. Was it what I said? Do FETs offend you guys? Haha...Anyhow, for anyone looking at this post later replacing the P channel of the push-pull created a working network. Works fine now.

 

Nobody answered because you didn't attach your schematic, and didn't say which Mosfets you used.

__________________
Uncle $crooge

 

Yes please post a schematic...

but keep this in mind, you must overcome the threshold voltage Vgs in order for the FET to conduct. So, first thing is to make sure that is happening. FETs that switch very fast may also draw significant gate current (due to Cdg charging/discharging) so don't assume the gate never draws more than pA. If driver circuit output impedance is too high, it may not be able to properly switch the FET. The driving impedance may also include your power supply. A calculation can determine roughly what drive current you'll need to provide on the gate without droping all your drive voltage (thus failing turn-on)

These are the two most common problems I see when FETs do not switch properly (other than wiring errors)

 

Plus you don't want to exceed the maximum gate voltage of 20V and use ferrite beads or small resistor in series with the gates to prevent RF oscillation.

__________________
I also post at the following sites:
http://www.stop-microsoft.org http://www.heated-debates.com
Screen name: Aloone_Jonez

 

Thanks for the reply, I had a case of the newbies.

Here is the schematic: http://216.77.188.54/coDataImages/p/...PullDriver.JPG

Here are the datasheets:

FET Driver: http://rocky.digikey.com/WebLib/Telc...C426,27,28.pdf

N Channel FET: http://www.alliedelec.com/Images/Pro...s/273-1178.pdf

P Channel FET:
http://www.alliedelec.com/Images/Pro...s/273-1263.pdf

The voltage source I am switching has about 1ma of current (the EMCO converter). A few times I was able to switch a 9v battery with nice 40KHz square waves but not the 100 or so volts from the DC converter. Is the converter output power too small?

Thanks gk

 

The Mosfets are spec'd with 10V gate to source. Your N-channel gets only 5V and your P-channel gets only 4.3V.
If your driver had a 10V power supply then the gate voltage would be much better but the driver doesn't provide dead-time so both Mosfets would be turned on for a moment during switching and conduct a huge current from their supply.

__________________
Uncle $crooge

 

TC428 has a voltage doubling application to do just that. I gave it a try with a 9v voltage regulator (hence 18v of gate voltage) and it worked (nice 40KHz square wave). It switched the 9v test battery but it still won't switch the EMCO unit (100volts). Do you think that that getting a higher amped EMCO unit would help?

National semi is sending some bridge drivers to address the MOSFET issue you mentioned.

 

When both your Mosfets conduct during switching they try to conduct many thousands of times the tiny amount of current from the EMCO supply. Then the supply gets completely drained before the switching has completed.

__________________
Uncle $crooge

 

Simple, add a small inductor in series with the 100V supply to stop the transients.

__________________
I also post at the following sites:
http://www.stop-microsoft.org http://www.heated-debates.com
Screen name: Aloone_Jonez