Folded cascode !! yes , that is the only way to get super fast VFA slew.
Search member Roender's threads ..... He has a few tested successful
projects based on this topology.
PS - maybe not 360V/us ... but >200V/us , quite fast for a VFA.
Edit - (below) Roender clone "abomination" - very fast , and built successfully. replace Q1/Q18 with Jfet's , tweak degeneration R38/R37
to 22-33R .....
OS
Attachments
Last edited:
Not really , hard to get a close match between P/N channel devices.
Much less beta with the FET's .... my CFA needs 350Hfe BJT's to get
sufficient closed loop gain. Both input diamond and super-pairs are
mid grade BC550/560 BJT's.
Roenders design can use matched or dual N-channel fet's (or bjt).
OS
but can we add jfet at the input before BJT of the ND? without disturbing the BJTs at the input?
Last edited:
There are few doubts regarding
1. The Driver Vbe transistor does this transistor needs to be on same heatsink of predriver and driver transistors?
In that case we can have two stage input which might increase the loop gain by another few db by using jfet stage.
CFP ops vbe is (closely) coupled to the drivers.
You could of looked that up .... Lazy ?
OS
Generally the Vbe multiplier will be hooked to the heatsink since its sensing the bias just like that the driver vbe would be with the drivers.
it does not sense the bias.
The conventional Vbe multiplier senses the output device temperatures of an EF output stage.
It applies the temperature compensation that you have set up your Vbe multiplier to provide.
If you don't have an EF output stage then you need to find which devices need the most temperature compensation and move the Vbe multiplier to monitor the temperatures of that part of the output stage.
do you have two different temperature compensating mechanisms?
Last edited:
true it senses the temperature but eventually to set the bias.
The ckt has two vbe multipliers so the driver vbe needs to track the driver temperatures thats what I felt that the drivers and the driver vbe to be on the same heatsink. The output stage and the output vbe multiplier on another large heatsink. Correct me if its wrong.
No.
If you have one temp compensator, then it goes to monitor the worst case for variation of it's Vbe causing bias voltage changes. The two systems, EF or CFP, require the tempco to be in different locations.
If you have two tempcos, you can locate each in the separate locations of pre-driver, or driver, or output. But you have to decide which of those three need compensation and work out what tempco is required for each of the compensators.
If you have one temp compensator, then it goes to monitor the worst case for variation of it's Vbe causing bias voltage changes. The two systems, EF or CFP, require the tempco to be in different locations.
If you have two tempcos, you can locate each in the separate locations of pre-driver, or driver, or output. But you have to decide which of those three need compensation and work out what tempco is required for each of the compensators.
Please find the attached pic of the pcb positions where you can see the extreme right middle transistors are the vbe multipler transistors and the one on the left are the output of VAS, Predriver and Driver on the large heatsink.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.