Snubber or just a diode?

Discussion in 'General Electronics Chat' started by electronice123, Jan 28, 2014.

1. electronice123 Thread Starter Senior Member

Oct 10, 2008
304
0
Hello all,

I am getting ready to start working with my new wideband transformer.
Before I do though I want to be sure I protect the drive circuit.

Info on the coil:
Primary coil L=140uH, R=400mΩ Leakage L=5uH Applied voltage=+10V

I will be pulsing the coil in a range from 1kHz to 20kHz.
Current will be max 5 amps max at 1kHz.

My question is, can I just use a HV diode across the primary coil, or do I need a RCD snubber? I need the wave on the secondary side of the coil to be as square as possible.

2. #12 Expert

Nov 30, 2010
16,690
7,330
Probably a diode, but, drive it with what? and the output goes to what load?

3. electronice123 Thread Starter Senior Member

Oct 10, 2008
304
0
Driving it with a FET, using a square wave at the gate.

Secondary will be used to charge a cap.

4. #12 Expert

Nov 30, 2010
16,690
7,330
Yes. I think a diode will do to protect the fet. The diode will allow current to flow in the coil after the fet is off. If you are measuring energy, you will have to account for that.

5. crutschow Expert

Mar 14, 2008
13,498
3,374
So are you using the transformer in a flyback circuit to generate a high voltage on the output? If so then you don't want a diode as that will kill the voltage. Instead you could use a zener across the FET drain-source connection with a voltage rating perhaps 20% below the FET rating.

6. ErnieM AAC Fanatic!

Apr 24, 2011
7,436
1,626
No accurate comment can be made without a schematic.

7. vrainom Member

Sep 8, 2011
109
19
So it's a flyback converter, you need a snubber calculated by the frequency of operation and leakage inductance of the transformer, I guess you could try and calculate it for 10 khz and hope it does well all over the working band?

8. #12 Expert

Nov 30, 2010
16,690
7,330
This is why we get so cross with people that won't post a schematic.

9. electronice123 Thread Starter Senior Member

Oct 10, 2008
304
0
Sorry. I tried to post more but had problems logging in.

I don't have a schematic. But it is a very simple circuit.
Just a wideband xmfr driven by a fet. It's not a flyback.

I'm trying to teach myself to design a wideband xmfr which will give me a good square wave output.

The load on the secondary side is just a resistor.

I think I need a snubber rather than a diode since the diode will direct the current back through the coil and mess up the trailing edge.

I found the following calculator online:
http://www.daycounter.com/Calculators/Snubbers/Snubber-Design-Calculator.phtml

There seem to be a lot of posts on snubber design as well as a ton of info online. I think this site should have a sticky in snubber design. Most everything I read just leaves me confused, there should be an easier way. ..

Anyways, just want some suggestions on that online calculator. Is it any good?

Are there any other considerations I should be taking into account besides what's on that site?

Thanks for all the help.... and, Sorry for the trouble.

10. #12 Expert

Nov 30, 2010
16,690
7,330
I've used that calculator and it seemed valid for what I needed that day.

You can hand draw a schematic and scan it if you don't have MS Paint and some symbols.
I got my symbols from Bill Marsden.

File size:
19.3 KB
Views:
37
11. crutschow Expert

Mar 14, 2008
13,498
3,374
A transformer will not work well with a unidirectional pulse that your MOSFET driver generates since that signal has a large DC average component which may saturate the transformer. What is the duty-cycle of you pulses?

12. #12 Expert

Nov 30, 2010
16,690
7,330
True. A bipolar drive will also absorb the stuff a snubber would have to deal with.

13. electronice123 Thread Starter Senior Member

Oct 10, 2008
304
0
The duty cycle is 60%. The core I choose has a Bsat of 1.5T so I don't think saturation will be an issue. At least I hope not. I used an online software program to design the coil, as well as a lot of math on my own.