I have a Phoenix Gold XS2500 which had 2 blown IRF540s in Rch + GND trace that goes to heatsink was melted open. After FETs replaced and trace repaired Amp now oscillates (on-off-on-off) when powered up. I've checked all diodes/transistors with a multimeter (in circuit), any questionable I compared readings with undamaged channel devices , no discolored resistors.I've discovered if I connect the positive probe of my multimeter in diode check mode (it applies ~ 2.8VDC) to pin 8 of SG3525A = pin 8 of LM339 the amp's SMPS runs generating +/- 37VDC on rails with no on-off oscillation , no hot components , but no output R or L nor any VDC offset R or L Ch. In this mode the amp is drawing ~ 850mA.The other thing I've noticed is the amp draws ~ 25mA with only B+ and B- connected.
When these particular amps blow outputs they take out gate resistors, and gate drivers, and the Diff input pairs also take a beating.
The on off pulsation is due to the safety circuitry being triggered and the amp is jumping from turn on to turn off protection over and over again. Its likely sensing a fault and triggering a shutdown to protect itself and your speakers.
These amps tend to be more of a complete rebuild of the damaged channel when they fail the outputs in my experience. 2SC2389 are your input pair, the rest are jelly bean MPSA-06 and 56 transistors up to the output 540 mosfets. The larger 2SA1964A plastic transistors rarely fail in my experience.
Also check the larger resistors near the outputs they tend to burn up in value. 200 ohm 2 watt and 4.7 ohm 1 watt, and 4.7 ohm 2 watt.
There are 4 each 1N5233b 6.0 volt zener diodes < D206,D207,D212,D213 > and 2 each 1N5242B 12.0 volt zener diodes <D208, an D209 >, all of these need to be checked for damage related issues. 100 numbered zener's are in left channel and 200 numbered zener's are the right channel, change the first number to 1 for the left channel of the numbers listed above.
Your bias control sets forward bias at the diff pair not at the outputs like a normal class AB amp would. If you trace it you will see it connects back at the input diff pair and not the outputs. Bias voltage is controlled back on the 12 volt side of the power supply from the safety circuitry section < Thermal, over voltage etc..>
Rails should be +&- 34 VDC and +&-15 VDC and a special +37 VDC to the front end 2SA1964A transistors attached to the 2SC2389 pair.
You can PM me if you like for more info and past experience with these......🙂
The on off pulsation is due to the safety circuitry being triggered and the amp is jumping from turn on to turn off protection over and over again. Its likely sensing a fault and triggering a shutdown to protect itself and your speakers.
These amps tend to be more of a complete rebuild of the damaged channel when they fail the outputs in my experience. 2SC2389 are your input pair, the rest are jelly bean MPSA-06 and 56 transistors up to the output 540 mosfets. The larger 2SA1964A plastic transistors rarely fail in my experience.
Also check the larger resistors near the outputs they tend to burn up in value. 200 ohm 2 watt and 4.7 ohm 1 watt, and 4.7 ohm 2 watt.
There are 4 each 1N5233b 6.0 volt zener diodes < D206,D207,D212,D213 > and 2 each 1N5242B 12.0 volt zener diodes <D208, an D209 >, all of these need to be checked for damage related issues. 100 numbered zener's are in left channel and 200 numbered zener's are the right channel, change the first number to 1 for the left channel of the numbers listed above.
Your bias control sets forward bias at the diff pair not at the outputs like a normal class AB amp would. If you trace it you will see it connects back at the input diff pair and not the outputs. Bias voltage is controlled back on the 12 volt side of the power supply from the safety circuitry section < Thermal, over voltage etc..>
Rails should be +&- 34 VDC and +&-15 VDC and a special +37 VDC to the front end 2SA1964A transistors attached to the 2SC2389 pair.
You can PM me if you like for more info and past experience with these......🙂
Last edited:
Yes, I have checked all transistors BJT & FET , all Diodes Signal & Zener , all Resistors 1W or greater , all checked in-circuit with multimeter and haven't found any bad parts. Because of the on-off oscillation it's not stable enough to make any voltage measurements, as I said I can verify the SMPS is generating +/- rail voltages. I have no schematic so I can't tell what the protection circuits are sensing or how they work.
Well U1 the LM339 is your safety circuitry . U1: A & D is the under volt remote and over volt 16 volt sense. And U1 B is your thermal circuitry. And U1 C would appear to be your mute control line output all feed off of the Sg3525 5.1 volt reference line .
there is no opto coupling from the secondary back to the primary so I have to ask what is your power supply feeding this amp rated at current wise? And are you measuring the DC current draw of the amp when it tries to turn on?
I have seen these amps draw a ton of inrush current when the channel was not repaired completely and it could be pulling your supply down and once its about 9.5 volts or less the supply will inhibit itself by way of U1:A 🙂
there is no opto coupling from the secondary back to the primary so I have to ask what is your power supply feeding this amp rated at current wise? And are you measuring the DC current draw of the amp when it tries to turn on?
I have seen these amps draw a ton of inrush current when the channel was not repaired completely and it could be pulling your supply down and once its about 9.5 volts or less the supply will inhibit itself by way of U1:A 🙂
It looks like you are correct, I was using a 6A supply and the voltage was dipping under 10V during on-off oscillation.I switched to a 30A supply and now no on-off but the current initially goes up to ~ 1.5A , hesitates and then rapidly goes to 8A at which time I pull the plug for fear of causing more damage. After this 10-15 SEC exercise BOTH sides have hot MOSFETS so I guess besides the first 2X bad 540s in the R Ch it looks like there is also a L Ch prob. I still cannot find any bad semis or resistors in either channel.
Sounds like somebody ran subs mono with this and ran both channels into submission. Whats going on is that both channels have damaged components and both channels are forward bias'ing out of control. If you remove both channels sets of outputs you should be able to test the amp without causing anymore significant failures to deal with. So please take all of the 540 outputs out and the the amp will not draw excessive current and you should be able to test all you like with power on...🙂
With all 540s removed , amp draws ~ 1.4A. Q108 , 103 , 106 all get hot as well as they're R Ch equivalents. Q108 bce = .2V,-26V,.75V...Q103/106 = 37V,.5V,37V. Q10/11 also get hot , emitters = +/- 14V
- Status
- Not open for further replies.