Help selecting a 1A BJT for 200KHz switching?

[Lightium]

I need help selecting a BJT that's rated for around 1 amp collector current and can handle 200KHz switching. VCE should be greater then 20V.

I ask because as I increase the frequency on my oscillator section the driver sections output current reduces.

 

[Lightium]

Ok, here's my latest circuit while I read your message.
View attachment 144773

 

[Lightium]

OK, I followed your suggestions from post #20. I am getting pulses on HV_out of 3KV and 1.5mA at 43KHz.

 

[Nigel Goodwin]

OK, I followed your suggestions from post #20. I am getting pulses on HV_out of 3KV and 1.5mA at 43KHz.

Why have you left R5 in place, and randomly changed R7 to a ludicrous 1 ohm, and R10 to 500 ohm?.

 

[Lightium]

ok something weird is going give me a minute.

 

[Lightium]

I posted the right picture, but it posted the previous one. Here's what I am doing now.
Are the 1.4A spikes ok for Q1 and Q2?

 

[Nigel Goodwin]

What's happened to the 680 ohm resistor to the base of Q5 - it's essential to have a resistor there to limit the base current to sensible values.

 

[Roff]

Do you realize that you don't have a spice model for Q5? I'm guessing you got that error message, and ignored it, so LTSpice uses a VERY basic model. This model has a beta of 100 at any collector current, and it has nothing to reduce switching speed; no capacitance, no storage time, no ft. Real high power bipolar transistors generally have very slow switching speeds. In effect, the generic model is a current-controlled switch. It does have a base-emitter diode, but that only serves to slightly reduce the base current.
I could not find a spice model for Q5 (FLJ9620). You could use a different transistor. In your simulation, the collector voltage is only reaching 50V, but I suspect it will go higher if you ever get it to work...
You can't damage a transistor in a simulation.
EDIT: Have you considered a MOSFET? Seems to me that would be a better choice.

 

[Lightium]

Do you realize that you don't have a spice model for Q5? I'm guessing you got that error message, and ignored it, so LTSpice uses a VERY basic model. This model has a beta of 100 at any collector current, and it has nothing to reduce switching speed; no capacitance, no storage time, no ft. Real high power bipolar transistors generally have very slow switching speeds. In effect, the generic model is a current-controlled switch. It does have a base-emitter diode, but that only serves to slightly reduce the base current.
I could not find a spice model for Q5 (FLJ9620). You could use a different transistor. In your simulation, the collector voltage is only reaching 50V, but I suspect it will go higher if you ever get it to work...
You can't damage a transistor in a simulation.
EDIT: Have you considered a MOSFET? Seems to me that would be a better choice.

https://www.onsemi.com/download/models/lib/fjl6920.lib

 

[Roff]

Well, I feel a little stupid. That transistor works a lot better than I thought it would.
You could have saved my bruised ego if you had included that spice model when you posted the .asc file.

 

[Lightium]

I am still learning about MOSFET's. I have got a MOSFET circuit I am working on, but that's another subject.
Don't let it sink you down. I was just being helpful.
If you find a use for the transistor, let us know.

 

[Tony Stewart]

serv

200KHz is only a very low frequency, any transistor should be perfectly fine there.

It is rise time that matters not the switching rate.
FET's will have much lower Ron but much higher Coss. This is why C of coils are critical.

Can you verify RLC values of Inductors are they coupled? M =?

 

[Roff]

Your latest simulation in post #18 said you were getting "20KV 5mA 13KHz", but I ran the sim and got 1.3kV@323 kHz.
What's up with that?
I have made some mods, and currently have 9.5 kv p-p @20 kHz, but the pulse width is only about 2.2us.
Don't know if this is an improvement or not.

 

[Lightium]

No, that was a bad circuit and I didn't say that.

 

[Lightium]

Try this one. Supply voltage is down to 18VDC(12VDC is target). It output's 24KV (on target) and 12mA (20mA is target) at 17KHz.

 

[Roff]

No, that was a bad circuit and I didn't say that.

You didn't say "20KV 5mA 13KHz", but it was a comment on the schematic that you posted in #18.
Please post MJE18004 spice model. Without it, your sim won't run.

 

[Roff]

Deleted.

 

[Lightium]

.MODEL MJE18004 NPN IS=1e-17 BF=28.9081 NF=0.85 VAF=30.1386 IKF=5.56749 ISE=5.41307e-12 NE=3.42903 BR=2.8908 NR=0.782418 VAR=3.18374 IKR=2.47573 ISC=4.31878e-13 NC=3.99474 RB=1.01858 IRB=0.100033 RBM=0.100017 RE=0.000655569 RC=0.193973 XTB=0.119644 XTI=1.00078 EG=1.10976 CJE=2.59175e-09 VJE=0.4 MJE=0.345984 TF=1e-08 XTF=0.05 VTF=0.1 ITF=620.447 CJC=1.81952e-10 VJC=0.8 MJC=0.536669 XCJC=0.1 FC=0.8 CJS=0 VJS=0.75 MJS=0.5 TR=1.29051e-07 PTF=0 KF=0 AF=1 Vceo=450 Icrating=5 mfg=On_Semi

 

[Lightium]

I couldn't keep up with the circuit text with all the changes I was making.

 

[Roff]

Why does your power supply have 1.2 ohms series resistance?
Are the values of R8 and R9 set in stone, or can they change?
With 24kv, you can only get 12 ma through 2 Meg.

 

[Lightium]

R9 is a HV wire. R8 is a stand by for an air gap.