F5 turbo problems

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi

I build the F5 turbo with low rails +/- 15V for driving a low impedance ribbon speaker.
Everything looks fine bandwidth 750Khz -3dB and I get about 35A from the output in this configuration.
I use 4 sets mosfets from IR the IRFP240 and IRFP9240 and 0,235 Ohm source resistors.
The fronted is the same as the F5 with 100 Ohm parallel in the feedback loop.
Now the problem if I feed the amplifier 100Khz and on the output the voltage rise above 20Vtt then the output collapse and the fets drawing lots of current about 10A each.
It doesn't mind if I load the output with a resistor or not.
What could be the problem.
I tried capacitors on the feedback resistors to reduce the bandwidth to 100Khz use input filter and 1N from drain to gate.
Nothing helps.

Rob
 

Attachments

  • F5 turbo.png
    F5 turbo.png
    22.7 KB · Views: 1,127
Hi,

Unability of the front end to supply enough current to properly swing the output mosfet's gate capacitance at such high freq ? this leading to what class D people call shoot through, ie N and P channel mosfets conducting too much at the same time, ie shorting V+ to V-

Regards,
--
stéphane
 
Hi,

Unability of the front end to supply enough current to properly swing the output mosfet's gate capacitance at such high freq ? this leading to what class D people call shoot through, ie N and P channel mosfets conducting too much at the same time, ie shorting V+ to V-

Regards,
--
stéphane
Hi stéphane

Should I leave it like that and pay no attention to it or can it cause problems.

Rob
 
The one and only
Joined 2001
Paid Member
I have tried these various and sundry, but I know I didn't run your
circuit up too 100 KHz at power.

You want to try higher Source resistance because it might give us a clue.

If it's not related to that, then in might simply be inadequate drive,
requiring parallel input devices as outlined.

It is possible that the IR parts are at fault, the N channel IR part has
about 50% more Ciss and the P channel 50% more than that. In
addition, the P channel device has what is referred to as the "IR P
distortion" in which the transconductance shelves.

Even so, are you planning on operating this as an audio amplifier?

:cool:
 
The one and only
Joined 2001
Paid Member
I am struck by the possibility that the cross-current conduction you are
experiencing is the result of the Jfets operating beyond their bias to charge
the capacitance of the output devices so that there is a net increase in their
average current.

There is something you can try for this.

Connect the Source pins of the two Jfets together and increase the values
of the feedback resistors 100 ohms and 1Kohms. This will increase the
bias current and it will also put more local degeneration on the Jfets. From
there you can parallel Jfets for more current.

I think you still need to try raising the values of the Source resistors of
the output devices.

:cool:
 
Gosh maybe I am off base here if so Please forgive my Ignorance,If 20khz is the limit for some humans hearing,mine is 16.5khz,Why are you at 100khz anyway ,anything over 20khz is out of audible hearing,try cutting the freq bandwith down to 20khz or If you think you can hear it 22khz and see what it does,I agree with the posts,you can't have them both on at once,And it will burn up,
It's just my 2 cents,goodluck to you.
Regards,
NS
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.