Greetings. Attempted to repair my XLS-402 which had very low output on one channel. In a moment of brain fade I attempted to clip my scope ground to one of the driver transistor (MJE15032/033) heatsinks, thinking it would be 0V but instead got a wakeup call with a bang and a flash. I was aware the output transistor heatsinks were live I just stupidly didn't test the smaller heatsinks first before looking for a convenient scope ground. Replaced all the MJE but no output at all in either channel now. There is 95VDC on the output heatsinks but nothing on the driver sinks, I assume there was before given the fireworks from shorting it out. Schematic shows these are all connected together, any thoughts out there on why this might be the case?
Cheers
Cheers
The schematic I have shows driver Q10 to be TIP50/47 and Q12 to be MJE5731. I will take your word you have found the drivers, production changes often substitute different transistors and MJE15032/33 are popular as drivers.
You could have burned a land on a PC board with a low impedance earth ground to HT1. You will have to inspect for such damage. If you find it, you will have to drill small holes in the PC boards near where the burn is. Then push solid core wire of suitable gage through the holes bend over and solder to the stubs of the remaining lands. You may have to scrape coating off the board in the area you are going to solder. If you are not lucky, the PCB is multilayer and such a drill hole may damage other internal traces, or short the HT1 to the internal trace. I use #40 or #54 drills in a pin vise I spin with my fingers, not a power drill.
Use safety glasses when soldering/desoldering. Also when probing with power on, parts can explode. Unplug the amp from wall plug and discharge rail caps and AC input caps with a resistor and clip leads before any work with tools or hands. Measure all metal at <1v to safety ground before you touch it with hands or tools. Use only one hand at a time, >24 v from one hand to the other can stop your heart. No jewelry on hands, wrists, or around the neck when working power amps.
I would definitely stop use of a scope that has such a hard connection on the probe to safety ground. Teks I have used had a bar from analog ground to safety ground which can be removed. I mostly fix amps with a DVM which is floating. If there is an AC problem, I probe around with 20 vac, 2 vac or 100 vac scale of a Siimpson 266 analog meter, which is also floating. I put a .047 uf 200 v cap series the negative probe on the analog meter to make sure I do not get AC readings on DC sources.
It is usual when testing power amps to put a 60 w tungsten bulb series the AC input to limit the damage when one slips when probing. I am not sure a switcher supply like XLS402 will spin up with such a limiter. My tungsten bulb is in a grounded steel box with circuit breaker, and mesh where I can see if the bulb is lit up or not. Those screw terminals on edison bulb bases, stranded wire can pop off them at any time. They are designed for 12 or 14 gage solid wire bent in a hook under the screw.
You could have burned a land on a PC board with a low impedance earth ground to HT1. You will have to inspect for such damage. If you find it, you will have to drill small holes in the PC boards near where the burn is. Then push solid core wire of suitable gage through the holes bend over and solder to the stubs of the remaining lands. You may have to scrape coating off the board in the area you are going to solder. If you are not lucky, the PCB is multilayer and such a drill hole may damage other internal traces, or short the HT1 to the internal trace. I use #40 or #54 drills in a pin vise I spin with my fingers, not a power drill.
Use safety glasses when soldering/desoldering. Also when probing with power on, parts can explode. Unplug the amp from wall plug and discharge rail caps and AC input caps with a resistor and clip leads before any work with tools or hands. Measure all metal at <1v to safety ground before you touch it with hands or tools. Use only one hand at a time, >24 v from one hand to the other can stop your heart. No jewelry on hands, wrists, or around the neck when working power amps.
I would definitely stop use of a scope that has such a hard connection on the probe to safety ground. Teks I have used had a bar from analog ground to safety ground which can be removed. I mostly fix amps with a DVM which is floating. If there is an AC problem, I probe around with 20 vac, 2 vac or 100 vac scale of a Siimpson 266 analog meter, which is also floating. I put a .047 uf 200 v cap series the negative probe on the analog meter to make sure I do not get AC readings on DC sources.
It is usual when testing power amps to put a 60 w tungsten bulb series the AC input to limit the damage when one slips when probing. I am not sure a switcher supply like XLS402 will spin up with such a limiter. My tungsten bulb is in a grounded steel box with circuit breaker, and mesh where I can see if the bulb is lit up or not. Those screw terminals on edison bulb bases, stranded wire can pop off them at any time. They are designed for 12 or 14 gage solid wire bent in a hook under the screw.
Last edited:
Many thanks for your reply. IU have that same schematic however it doesn't match my model (there must be different variations of 402's?), there is a different one I have found after some digging with the matching transistor p/n (attached). They don't show the HT part of the power supply on there which is frustrating but I assume it's the same as what's on the other schematic.
Understand the electrical safety issues, I work with 135kV regularly and somehow it's not as scary as working on power amps 😆
Thanks for tip re scope too.
Understand the electrical safety issues, I work with 135kV regularly and somehow it's not as scary as working on power amps 😆
Thanks for tip re scope too.
Should finish the story, nothing worse than finding a thread with no resolution.
Turns out after closer study of the schematic, the heatsinks on the driver transistors are not at HT potential (~95VDC). Sinks are connected to the collector, where the emitter is sitting at HT rail. Obviously grounding the collector caused full current through the transistor and blew it out. Luckily nothing else got got.
Replaced all the MJE15032/33 drivers, still no audio on ch2. Original post said both channels out but this was an issue with test speakers, ch1 ok. Long story short, the mode switch for dual/bridge had failed, preventing ch2 signal from reaching output stages. Hard wired into dual mode and all is good, not a big fan of bridge mode anyways. Switch not shown on schematic (grrrr) so it was a visual trace with all the boards out. Ch2 does a couple of laps from the back to the front of the main board just to provide the mode switching function.
Turns out after closer study of the schematic, the heatsinks on the driver transistors are not at HT potential (~95VDC). Sinks are connected to the collector, where the emitter is sitting at HT rail. Obviously grounding the collector caused full current through the transistor and blew it out. Luckily nothing else got got.
Replaced all the MJE15032/33 drivers, still no audio on ch2. Original post said both channels out but this was an issue with test speakers, ch1 ok. Long story short, the mode switch for dual/bridge had failed, preventing ch2 signal from reaching output stages. Hard wired into dual mode and all is good, not a big fan of bridge mode anyways. Switch not shown on schematic (grrrr) so it was a visual trace with all the boards out. Ch2 does a couple of laps from the back to the front of the main board just to provide the mode switching function.
This is a unique circuit - with the drivers configured as CFPs. Provides the current gain of EF3 with the temperature coefficient of EF2. Drivers themselves were subject to shoot-through at high frequency with the original TIP50/MJE5731 pair, hence the change to the faster and more appropriate MJE15032/3. Bet you’re not the first person to assume the tab was at the wrong potential.