how long is your amp been operational?? please share your schematic with the modified inverter
Thanks stewin. I do not have the schematic. I do routing with eagle manually. This is a habit . Inverter is like aud600 with 74hc00 instead of 74hc04. It is 2 months that I test it 4 hours work with high power with 4 ohm resistive load without problem. But irfp250n needs a big heatsink and it stay cool. With small heatsink temperature. Increases rapidly + 70 °. I use aluminum 1cm thick / 6cm / 10cm for 2 irfp250n and its cool
Thanks stewin. I do not have the schematic. I do routing with eagle manually. This is a habit . Inverter is like aud600 with 74hc00 instead of 74hc04. It is 2 months that I test it 4 hours work with high power with 4 ohm resistive load without problem. But irfp250n needs a big heatsink and it stay cool. With small heatsink temperature. Increases rapidly + 70 °. I use aluminum 1cm thick / 6cm / 10cm for 2 irfp250n and its cool
D
Deleted member 148505
have you ever tried to remove the resistive 4ohm load when the amp is operational at high spl??
If you're experiencing problems when speaker is disconnected at high volume. Try to put antiparallel diodes (>= 4 amps rated, ultrafast) after the output filter. The amp module also needs very good overcurrent protection circuit. So that when speaker is removed while playing loud, the amp will survive.
Or you can put a dummy 150 ohms (30W) load that is permanently connected to the output.
Use dummy load with lowest resistance that will make the amp stable, but high enough such that when amp is driven at full volume without speakers attached, the dummy load will dissipate low heat as possible.
Attachments
Last edited by a moderator:
If you're experiencing problems when speaker is disconnected at high volume. Try to put antiparallel diodes (>= 4 amps rated, ultrafast) after the output filter. The amp module also needs very good overcurrent protection circuit. So that when speaker is removed while playing loud, the amp will survive.
Or you can put a dummy 150 ohms (30W) load that is permanently connected to the output.
Use dummy load with lowest resistance that will make the amp stable, but high enough such that when amp is driven at full volume without speakers attached, the dummy load will dissipate low heat as possible.
thanks lester
have you ever tried to remove the resistive 4ohm load when the amp is operational at high spl??
mr stewin. without load when i increase volume ocp is activate ! it is high current problem when lc filter resonnate? anti parallel diode is it a sollution? i see them a standart rectifier diode at output with iraudamp 7 schematic
mr stewin. without load when i increase volume ocp is activate ! it is high current problem when lc filter resonnate? anti parallel diode is it a sollution? i see them a standart rectifier diode at output with iraudamp 7 schematic
If you're experiencing problems when speaker is disconnected at high volume.......................... So that when speaker is removed while playing loud, the amp will survive.
Nonsense
If you make a normal , stable feedback in your amp - nothing of that kind may happen .
Last edited:
D
Deleted member 148505
erro no d900
Good morning, Merry Christmas to all, with the irfp4768 can use 120 + 120 volts, in the scheme of the D900 hybrid, it would not be wrong to connect the base of q1, because it should have an LC network to oscillate. And happy new year to all Can publish the archie in ltspice
I have something faster if you like.
I make this as I am not happy with the output of two legs of resistor under LTP
ltspice file is attached in the PDF also.
Good morning, Merry Christmas to all, with the irfp4768 can use 120 + 120 volts, in the scheme of the D900 hybrid, it would not be wrong to connect the base of q1, because it should have an LC network to oscillate. And happy new year to all Can publish the archie in ltspice
Jlester,
These compromized solutions help somewhat at the price of higher loss, lower transmittion bandwidth, but the correct solution is much better.
You can set high resonant freq, but it will increase loss and output ripple. You can set input bandwidth limit, but it is a loss in quality, and still doesn't solve the problem, only eases. A good UcD amp can have significantly higher bandwidh than its output filter, and no resonance can be observed with any practical load.
In case of UcD a zobel network is also beneficial, but not the way as in a pre-filter fed back amp. In case of UcD switching freq is affected somewhat by load impedance, especially at very high modulation index, and this dependance is lower with zobel. But no resonance at the filter freq.
These compromized solutions help somewhat at the price of higher loss, lower transmittion bandwidth, but the correct solution is much better.
You can set high resonant freq, but it will increase loss and output ripple. You can set input bandwidth limit, but it is a loss in quality, and still doesn't solve the problem, only eases. A good UcD amp can have significantly higher bandwidh than its output filter, and no resonance can be observed with any practical load.
In case of UcD a zobel network is also beneficial, but not the way as in a pre-filter fed back amp. In case of UcD switching freq is affected somewhat by load impedance, especially at very high modulation index, and this dependance is lower with zobel. But no resonance at the filter freq.
More than 10 years ago I designed a ClassD amp with resonating unloaded filter. The manufacturer/distributor refused it.
why ? all self oscilate class d have this problem. and many class ab also!! congratulation pafi for your hard work
By the way , the UcD is more unstable vs the prefilter feedback amps. They need carefully adjusted feedback .The amps based on IR2092 configuration have no such issues in ALL conditions !!!A good UcD amp can ....
I believe that there is a mistake in Q4 of this scheme because I think that the base should not be linked to the gnd, someone can give a help because it is not able to make it work ..
An externally hosted image should be here but it was not working when we last tested it.
why ? all self oscilate class d have this problem. and many class ab also!! congratulation pafi for your hard work
This problem? What do you mean? Resonance of output filter without load? This is independent from the oscillation method. This depends on feedback topology (pre- or post-filter).
And why is this unacceptable? Because it activates protection as you saw. And this can happen not only at completely disconnected speaker, but also with a woofer with low pass filter (very high impedance at high freq). Also because zobel can burn out by a sustained resonance.
Last edited:
- Home
- Amplifiers
- Class D
- UCD 25 watts to 1200 watts using 2 mosfets