This is a circuit i came up with this morning, i have just finished building it and it turns out it works well above expectation.
I'm running it from 2x35VDC and the fets(IRFP450) is not even getting warm.
Schematic: http://i.imgur.com/giy22.png
I'm running it from 2x35VDC and the fets(IRFP450) is not even getting warm.
Schematic: http://i.imgur.com/giy22.png
It's a child in growth pains, i havent dealt with these triangular wave oscillators in a looong time so im mostly getting sine waves out, maybe LM4562 isent a particularly good choice, even though im only asking it to oscillate at around 100kHz.
You can try triangle oscillator from JBL 200kHz with single OP (1/2 TL072).
Attachments
That, or rework thed values of the resistors in the existing circuit, i lean toward the latter as i doubt i have any 5.6V zeners, i have 7.5V, 10V and i think 12 and 15V.
That jbl circuit kinda reminds me of the circuit of a relaxation oscillator.
That jbl circuit kinda reminds me of the circuit of a relaxation oscillator.
I never got that far, i had it runnig on 2x37V when a ground wire came loose and landed straight onto the gate for the high side fet = *POOF* and after that the IR2110 and both IRFP450's were dead.
My plan was to use a 50A 600V IGBT module for the output stage, i alrready tried this an d on 2x38VDC and 500kHz, the module with no heatsing only got slightly warm, the audio part dident worth though as i tried it with transformer coupled gate drive.
I'll resume the work on this this weekend, and i will socket the IR2110 so it can be easily replaced when needed.
I'll resume the work on this this weekend, and i will socket the IR2110 so it can be easily replaced when needed.
The old igbts were slow, the modern ones are almost as fast as mosfets.
As i said i ran a modern fairchild transfer molded igbt halfbridge module(http://www.fairchildsemi.com/ds/FP/FP7G50US60.pdf) at 500kHz with 2x38VDC and it only got slightly warm just laying on the table, and that with lousy transformer gate drive and no heatsink.
As i said i ran a modern fairchild transfer molded igbt halfbridge module(http://www.fairchildsemi.com/ds/FP/FP7G50US60.pdf) at 500kHz with 2x38VDC and it only got slightly warm just laying on the table, and that with lousy transformer gate drive and no heatsink.
This Fairchild IGBT just proved itself, a short piece of thin wire lietrally melted at near max power output, was cranking to near 99% duty cycle when i looked over and saw smoke.
Despite this abuse, the IGBT module was only ever so slightly warm.
Despite this abuse, the IGBT module was only ever so slightly warm.
A new video: Feedbackless class d amp pounding it - YouTube
Rail voltages are now 2x60V, but for some reason 30-35V is the max i can get out, even when its clipping and hitting the rails.
The fets remain barely warm to the touch.
Rail voltages are now 2x60V, but for some reason 30-35V is the max i can get out, even when its clipping and hitting the rails.
The fets remain barely warm to the touch.
Good evening.
You can send me the complete layout of this amplifier?
Thank you.
luizcpimenta@yahoo.com.br
luizcpimenta@gmail.com
- Status
- Not open for further replies.