I have for repair a P-128 mono block power amp.
This is the version with two pairs of MOSFETs (2SK225 and 2SJ81)
I measured about 160 Watt before clipping into 8 ohm load.
During the bench test, I noticed a 0R22 black resistor getting really hot.
This is R36, and I am very suprised with this resistor in serie with the output.
It we have 36V RMS into 8ohms at max power (160W), this means about 36 x 0,22 / 8 = ~1V RMS across this resistor, equals to more than 4 Watts in R36.
This gets worse with 4 ohms loads:
240 Watts is 31V RMS on 4 ohms, so the voltage is 1,7V on R36, and the power is now about 13 Watts in R36 at max power !
The resistor itself is a wirewound, but it seems it will not dissipate more than a couple of Watts.
I am puzzled: why add a resistor at this place ?
- it cannot withstand the power, but this could be arranged with a bigger part
- it adds to the amp's output impedance, reducing the damping factor
Any idea ?
This is the version with two pairs of MOSFETs (2SK225 and 2SJ81)
I measured about 160 Watt before clipping into 8 ohm load.
During the bench test, I noticed a 0R22 black resistor getting really hot.
This is R36, and I am very suprised with this resistor in serie with the output.
It we have 36V RMS into 8ohms at max power (160W), this means about 36 x 0,22 / 8 = ~1V RMS across this resistor, equals to more than 4 Watts in R36.
This gets worse with 4 ohms loads:
240 Watts is 31V RMS on 4 ohms, so the voltage is 1,7V on R36, and the power is now about 13 Watts in R36 at max power !
The resistor itself is a wirewound, but it seems it will not dissipate more than a couple of Watts.
I am puzzled: why add a resistor at this place ?
- it cannot withstand the power, but this could be arranged with a bigger part
- it adds to the amp's output impedance, reducing the damping factor
Any idea ?
I don't know if you (Andrew) have already fixed it, but to me it seems like overvoltage when connecting the input for the time only the core is connecting. An input clamp would fix that to some degree, depending on how aggressively you design the input clamp. I added input clamps to a preamp I designed especially for this purpose.
As for avoiding offsetting by a blown fuse, that can be entirely avoided by having a separate front end supply. If the whole amp runs off one supply I don't find it strange it goes into offset. One caveat though, since during a fault condition the front-end keeps working while OPS fuses have blown, you'll have to deal with limiting VAS/drive current as to not blow up the front-end.
As for avoiding offsetting by a blown fuse, that can be entirely avoided by having a separate front end supply. If the whole amp runs off one supply I don't find it strange it goes into offset. One caveat though, since during a fault condition the front-end keeps working while OPS fuses have blown, you'll have to deal with limiting VAS/drive current as to not blow up the front-end.
Last edited:
This low value resistor, 0,2 Ohm, is usually used enhance stability, specially with capacitive loads and in substitution of the usual Zobel network, a inductor and resistor in parallel. Strangely in the case of the P-128 we have both, the resistor and the Zobel.
Some British constructors use this small resistor at the output. Naim come to mind...
http://www.acoustica.org.uk/t/naim/poweramp_pix/NAP250 schematic.jpg
Obviously it adds to the amp's output impedance, reducing the damping factor, because it is outside the overall feedback loop.
About the low dissipation part used, the designer must had thought that the maximal RMS output power will not be used for prolonged periods of time, and as a sinusoidal signal is much more challenging than a musical signal, in power dissipation terms, it would be hardly noticed in "normal" operation...
- Status
- Not open for further replies.