Hy lorylaci
.I am still on test board.
Diferent ground setups and diferent feeback wire placements give diferent distorsion paterns.I made a ground place with aluminium foil.Now i use t106-2 thanks to savu.I also use irf540z.Heat has decreased alot..and noise by a small margin..
Even if i wanted to do a pcb and give it to someone else to etch it(printer still broken) i have no ideea how to design it corectly since all my eforts have been invested into test board manipulations with no result..
First i will give a diagram of curent ground setup...
Last edited:
I found a t106-2 Amidon core worked very well. I made the mistake of using power inductors start with and they got seriously hot 120 degrees C !
Making your own class d pcb is fun, keep connections as short as possible and decouple really close to devices. Keep outputs away from inputs.
Heareable
..so >5-10 %.I do not have distorsion analizer.Load impedance 2 ohm.I will pun 100 nf capacitors on bridge.Thanks.
Hy lorylaci
I am still on test board.
Diferent ground setups and diferent feeback wire placements give diferent distorsion paterns.I made a ground place with aluminium foil.Now i use t106-2 thanks to savu.I also use irf540z.Heat has decreased alot..and noise by a small margin..
Even if i wanted to do a pcb and give it to someone else to etch it(printer still broken) i have no ideea how to design it corectly since all my eforts have been invested into test board manipulations with no result..
First i will give a diagram of curent ground setup...
FORGET THE TEST BOARD!
As i said previously, you most of your problems will simply dissappear when you switch to PCB layout!
Nigel said it short and right: keep connections, especially critical connections short. These incluede FET et driving path, FET dupply path from nearest cap.
Start developing a PCB layout, if you search this tread, there are lot of tries, some are bad but some are quite good.
This amp normally is able to kill the 50Hz hum, due to supply rejection of the a feeadback. The hum is because of long ground paths and the test board. Most of the noise and distortion is also becuase of test board.
And at the end: use one big ground plane, as Bruno reccomended for UcD.
After reading over 100 pages of the topic have not come to a conclusion ...
Many say they are bad, others say it is very good
Unfortunately I do not have time to "try to mount, keep making changes, measuring, simulating, etc. .." so much that I read these 100 pages today, a Sunday ..
After all, if I mount the original circuit, the original board, as the toroid correct, it works well?
Has anyone actually got to 1000 W of power?
Someone uses it continually? at parties for example: 5,6,7 hours of continuous use at full power?
thanks
Many say they are bad, others say it is very good
Unfortunately I do not have time to "try to mount, keep making changes, measuring, simulating, etc. .." so much that I read these 100 pages today, a Sunday ..
After all, if I mount the original circuit, the original board, as the toroid correct, it works well?
Has anyone actually got to 1000 W of power?
Someone uses it continually? at parties for example: 5,6,7 hours of continuous use at full power?
thanks
The original schematic with the given given IRFP fets is not capable of 1kW quality, continnous power. But with modifications, with improved layout, and better fets, it is sure capable of about 1kW into 2 Ohm with proper supply.
I tested my modified schematic and layout with 4 Ohm load at +-70V supply and worked pretty well. I am working on improveing this schematic and layout, had figured out some guidelines for further improvement...
It is a good schematic and layout to start with and develop your plan. But if you are lasy and just want to rebuild one guaranteed succes then you should pay some first!
I tested my modified schematic and layout with 4 Ohm load at +-70V supply and worked pretty well. I am working on improveing this schematic and layout, had figured out some guidelines for further improvement...
It is a good schematic and layout to start with and develop your plan. But if you are lasy and just want to rebuild one guaranteed succes then you should pay some first![/QUOTE]
can the amp be used in bridge mode with the irfp 250 will it be more powerfull?
It is a good schematic and layout to start with and develop your plan. But if you are lasy and just want to rebuild one guaranteed succes then you should pay some first![/QUOTE]
can the amp be used in bridge mode with the irfp 250 will it be more powerfull?
Those IRFP250Ns already has a hard job in this schematic.
In full bridge at 4 Ohm there is a great chance of overheating for them, with excessive use. Okay, it takes music signal, but a there is no safety limit. Take attention also that Rdson increases with temperature.
If you want that high power, buy some modern FETs. IRFB4227 is one of the bests on the market, if compare the price with the power you can get with it, its nothing. I buy componentes from at least 4 shops here in Hungary. (my favourite shop is 2 min walk from my apartment, but of course they do not have good FETs, but the shop which has good FETs is expensive and shipping...)
In full bridge at 4 Ohm there is a great chance of overheating for them, with excessive use. Okay, it takes music signal, but a there is no safety limit. Take attention also that Rdson increases with temperature.
If you want that high power, buy some modern FETs. IRFB4227 is one of the bests on the market, if compare the price with the power you can get with it, its nothing. I buy componentes from at least 4 shops here in Hungary. (my favourite shop is 2 min walk from my apartment, but of course they do not have good FETs, but the shop which has good FETs is expensive and shipping...)
Here is my board...first version.Please give feedback.
Thanks alot.
Thanks alot.
An externally hosted image should be here but it was not working when we last tested it.
Traces are short except lm311 to level shifter..a longer path to lm311 is worse in my opinion(like original board) since analog path is more sensitive to noise..
Mosfets are under the board and output inductor can be moved if needed(i will do this in testing phase) but that will add aditional switching traces wich act as an antenna.
Long traces and jumpers to voltage regulators are not a problem because of local decoupling...
Mosfets are under the board and output inductor can be moved if needed(i will do this in testing phase) but that will add aditional switching traces wich act as an antenna.
Long traces and jumpers to voltage regulators are not a problem because of local decoupling...