I bought 4 of Njw0281/Njw0302 Power Amplifier Board 450W+450W Assembled Amp Board Dc 45V H1Z6 194031106746 | eBay kits to implement in a Home theater system.
They ran fine upon initial testing and I was actually surprised how good they sounded out of the box for a cheap kit. I first tried a 1kva 65-0-65 transformer hooked to them and I got 350w@8e and 550w@4e (resistive load 300Hz sinewave test). Emboldened by this, I did add bias adjustment pots. I wanted to modify them into either a bridge or a class-A low powered unit. Both experiments ran fine first time, but then I had a thermal failure. I noticed one OP device was running a lot hotter than the other 3 on the same rail. I figured once they hit thermal equilibrium, I'd be ok, but Thermal runaway had other ideas. The device that was running hotter than the other 3 shorted out soon.
Since then I have removed all OP devices and gain and Vbe matched them and re-installed, so far so good. I have lowered the bias to 40mA per device now and it runs warmish but stable. Heatsink stays below 50c and the OP devices between 45-53c on the printed side.
Now to the question: I can either run it with the 92v rails like before and get 350/8 (the speakers are for the most part resistive.. as there are no crossover parts in it (Bose 901 running surround duties). OR: as I recently brought to use another Transformer, this one also 1kva (40-0-40) giving me 55vdc rails, but running the kits in bridge mode. I have been testing them with 3 way test speakers in short bursts, and it plays very clean. Sound is warm, not brittle. In this mode I can get about 528 watts/8e at clipping. In this mode each channel drives about 264 watts (vs the 350watts on the single channel / High rail version). However, since its bridged, it is probably taxed more... so:
which verison should I choose for a high spl home theater system?
V1: Single channel 92vdc rails @ 8 ohms (350w)
V2: 2 channels bridged @ 55vdc @ 8 ohms (528w)
in terms of reliability for use with resistive 8e load?
They ran fine upon initial testing and I was actually surprised how good they sounded out of the box for a cheap kit. I first tried a 1kva 65-0-65 transformer hooked to them and I got 350w@8e and 550w@4e (resistive load 300Hz sinewave test). Emboldened by this, I did add bias adjustment pots. I wanted to modify them into either a bridge or a class-A low powered unit. Both experiments ran fine first time, but then I had a thermal failure. I noticed one OP device was running a lot hotter than the other 3 on the same rail. I figured once they hit thermal equilibrium, I'd be ok, but Thermal runaway had other ideas. The device that was running hotter than the other 3 shorted out soon.
Since then I have removed all OP devices and gain and Vbe matched them and re-installed, so far so good. I have lowered the bias to 40mA per device now and it runs warmish but stable. Heatsink stays below 50c and the OP devices between 45-53c on the printed side.
Now to the question: I can either run it with the 92v rails like before and get 350/8 (the speakers are for the most part resistive.. as there are no crossover parts in it (Bose 901 running surround duties). OR: as I recently brought to use another Transformer, this one also 1kva (40-0-40) giving me 55vdc rails, but running the kits in bridge mode. I have been testing them with 3 way test speakers in short bursts, and it plays very clean. Sound is warm, not brittle. In this mode I can get about 528 watts/8e at clipping. In this mode each channel drives about 264 watts (vs the 350watts on the single channel / High rail version). However, since its bridged, it is probably taxed more... so:
which verison should I choose for a high spl home theater system?
V1: Single channel 92vdc rails @ 8 ohms (350w)
V2: 2 channels bridged @ 55vdc @ 8 ohms (528w)
in terms of reliability for use with resistive 8e load?
Last edited:
Bridged should be more reliable in normal use with your matched devices and rebiased amps, because
of the much lower rail voltage, provided the power supply capacitors are good quality, high ripple types.
Are they 100V caps?
of the much lower rail voltage, provided the power supply capacitors are good quality, high ripple types.
Are they 100V caps?
Last edited:
chinese transistors are well known for poor specs.
smaller dies etc to save money.
i bought some irfp240/9240 from china
worked ok to start with but one glitch on the mains when i unplugged my soldering iron killed them.
so bought in some genuine parts and despite trying hard to glitch them they refuse to fail.
the chinese transistors must have had a much lower breakdown voltage.
smaller dies etc to save money.
i bought some irfp240/9240 from china
worked ok to start with but one glitch on the mains when i unplugged my soldering iron killed them.
so bought in some genuine parts and despite trying hard to glitch them they refuse to fail.
the chinese transistors must have had a much lower breakdown voltage.
They look very similar to the ON devices I know are genuine. The Vbe is similar but the Chinese ones have higher gain 120 vs 90 @ 640mA
I had that board running for a while (85V), played with the bias also, forgot what I had settled on for that but it was conservative. It was fine running at low volume for three months nearly constantly, then one day, while I was nearby, it smoked an output at low volume.
I suspected fake parts, and the lackluster sound didn’t make me want to run out and replace all of those outputs to test that theory.
I would probably lean towards using a lower voltage.
I suspected fake parts, and the lackluster sound didn’t make me want to run out and replace all of those outputs to test that theory.
I would probably lean towards using a lower voltage.
Last edited:
Mine sound a whole lot better after re-biasing and changing out the Electrolytic input cap to an MKP type. The amp is not cold and sterile now as it was stock.
The other issue I am finding is that the Thermal sense bjt they used is not a very high gain type, so the amp has a tendency of running into thermal runaway without a couple of mods. I am installing the thermal comp trannie on to one of the OP devices for less Thermal inertia, and also installing a couple of thermal switches that will activate fans.