Tweet
Our Guitar Dept sent this older Eden WT 800 Hybrid Bass Guitar Amp over to the shop for service. This one has the PM 400 Power Amp modules, only having 2 pairs of output xstrs.
All I had from the Guitar Dept notes was ‘inconsistent output’ to go by. Plugged it in, connected it and found the Right Ch was grossly distorting. No DC on the outputs, thankfully. The Right Ch would work find under NO LOAD, but with any load, the positive half of the output would collapse. The Right Ch module has the Bridge Mono signal soldered to the power supply PCB, so that had to be unsoldered.
I also found there was substantial noise on both channels that appeared to be contact noise in the wiring. I could bang on the power supply PCB Assy and generate noise in both channels. I powered down, then after discharging the supply caps, I went thru the Power Supply PCB reseating every connection to it, as well as reseating all of the connections to the Front Panel PCB Assy as well as both amp modules. Checked again, powered up and reconnected to speaker to see if that contact noise had ceased, and it had, but still had the positive half clipping on the Right Ch module.
Just to check, I swapped power amp modules to see if the problem remained with the chassis wiring, or moved with the module. It moved with the module, so that was what I expected. When I removed the Right Ch module, the bridge input resistor and it’s wire fell off, so now had that to restore.
I removed the Right Ch module for inspection. Checking semiconductors, everything was measuring nominal, and since I did see full output under no load, that pointed towards there being either components that are breaking down under load, or solder joints breaking down under load. Initial inspection didn’t reveal any obvious solder joint issues.
I tried replacing the 1N4734 5.6V 1W Zeners (D50 & D3 on the WT800C schematic), along with C10 and C11 (10,000uF/10V caps—I used 12,000uF/16V caps). That made no change. Removed & re-soldered R24 & R25 6.8k/2W resistors, with no difference. Replaced Q10 TIP 31C and Q11 TIP32C with no change. Replaced Q7 2N5415 PNP and Q8 2N3440 NPN in the Voltage Gain Stage with no change.
I didn’t really see any bad solder joints, but, I re-soldered the 0.47 ohm emitter resistors, the output network, the output header connector to the harness and one of the emitters of the output stage. Cleaned up the flux, and put this Right Ch back into the chassis. It now drives load current.
NOW, I have to find just WHERE to re-connect the Bridge Mono resistor that fell off. I don’t clearly see where that is connected. Working thru the schematic, I found the circuit path, and in trying to get the wire-connected 27k resistor soldered into place, the wire end lead broke off, so had to fetch a new 27k resistor. Got that, prep’d the leads & lead wire, soldered that into place. Glued it down with RTV, and a tie wrap around the wire leaving the PCB assy to the source location on the power supply board.
I verified both channels working. So I set it up into Bridge Mono, initially fixed Sine Wave drive to verify all was working, driving load resistor bank. Then switched over to Burst Pink Noise, and connected the 4 ohm Ampeg Test Speaker, and ran up the level to yield 105dB SPL, let it run and went to take a leak. Came back, all still working fine. I was about to sit down, when I then heard very distorted output. Turned the level down, switched out of Bridge Mono, and checked Left Ch. That’s working fine. Plugged output/analyzer into Right Ch, and it’s once again severely clipping. So, whatever WAS working fine and conducting current to the load isn’t any longer. RATS!
Now I’m back to trying to find out what has failed this time. Do I have output level without load? YES. Full output until I load it. Verified this with the scope probe, that it IS the upper half of the waveform that has dropped out again. WHAT is ceasing to conduct? I re-soldered the High Current connections this last time…the emitter resistors, output inductor/resistor, Header terminal.
After further re-soldering, and lifting the power supply board up for inspection, I tried again, and still…under load, I loose positive half drive current! I’m puzzled. I’m not yet ready to believe I have power xstrs causing this.
This time, I replaced R19 100 ohms on the cathode of Q7 2N5415 PNP, R26 100 ohm emitter resistor of Q10 TIP31 Driver, along with the 22pF around Q7 and the 470pF around Q10 Driver. Also re-soldered the two NPN Power Xstrs Base, Emitter and Collector leads.
I now have the amp once again passing load current, so I have it in Bridge Mono driving the 4 ohm test speaker at 107dB SPL level (nominal) with Burst Pink Noise, and pulling up to 4.7Amps AC Mains. Letting it run to see if it continues, as I did before. Started this burn-in process @ 1:40PM. At 2:05PM, I lost signal. Got up to have a look. Both heat sinks are extremely hot, so I suspect the thermal switch(s) have opened. I see idle current of 0.58A/46W being drawn. The internal fan is running, so I’ll see how long it takes for it to turn back on…assuming this has hit thermal-cycle shut-down.
At 2:20PM, I have signal again. I turned down the drive signal for 100dB SPL, pulling around 2A Peak. Letting it run at this level. So, that was thermal cycle cut-out, as I suspected. Using the Averaging Function of the Magtrol Power Analyzer, it’s drawing an average of 1.36A/112W using the 3 Sec On/2 Sec Off burst cycle. Output level shows around 8V RMS into the 4 ohm speaker for this nominal 100dB SPL level. I wonder if it will thermal cycle again at this level? NO….still running at 3:20PM.
I don’t recall having one of the WT800 amps fail in this mode before. I’ve lost power transistors, voltage gain stage xstrs, drivers and associated resistors. This one took a while to find just what was cutting out. Still uncertain that it WAS solder joints, but suspect that. When you look at the way the power amp modules are built, relying solely on the solder joints to hold the PCB into place, I’m not surprised. But usually, you see the tell-tale signs of joint fractures. Not this time.
WT800C schematic-1.pdf
WT800A POWER SUPPLY SCEMATIC.pdf
PM-400 Pwr Amp Sch-1.pdf
All I had from the Guitar Dept notes was ‘inconsistent output’ to go by. Plugged it in, connected it and found the Right Ch was grossly distorting. No DC on the outputs, thankfully. The Right Ch would work find under NO LOAD, but with any load, the positive half of the output would collapse. The Right Ch module has the Bridge Mono signal soldered to the power supply PCB, so that had to be unsoldered.
I also found there was substantial noise on both channels that appeared to be contact noise in the wiring. I could bang on the power supply PCB Assy and generate noise in both channels. I powered down, then after discharging the supply caps, I went thru the Power Supply PCB reseating every connection to it, as well as reseating all of the connections to the Front Panel PCB Assy as well as both amp modules. Checked again, powered up and reconnected to speaker to see if that contact noise had ceased, and it had, but still had the positive half clipping on the Right Ch module.
Just to check, I swapped power amp modules to see if the problem remained with the chassis wiring, or moved with the module. It moved with the module, so that was what I expected. When I removed the Right Ch module, the bridge input resistor and it’s wire fell off, so now had that to restore.
I removed the Right Ch module for inspection. Checking semiconductors, everything was measuring nominal, and since I did see full output under no load, that pointed towards there being either components that are breaking down under load, or solder joints breaking down under load. Initial inspection didn’t reveal any obvious solder joint issues.
I tried replacing the 1N4734 5.6V 1W Zeners (D50 & D3 on the WT800C schematic), along with C10 and C11 (10,000uF/10V caps—I used 12,000uF/16V caps). That made no change. Removed & re-soldered R24 & R25 6.8k/2W resistors, with no difference. Replaced Q10 TIP 31C and Q11 TIP32C with no change. Replaced Q7 2N5415 PNP and Q8 2N3440 NPN in the Voltage Gain Stage with no change.
I didn’t really see any bad solder joints, but, I re-soldered the 0.47 ohm emitter resistors, the output network, the output header connector to the harness and one of the emitters of the output stage. Cleaned up the flux, and put this Right Ch back into the chassis. It now drives load current.
NOW, I have to find just WHERE to re-connect the Bridge Mono resistor that fell off. I don’t clearly see where that is connected. Working thru the schematic, I found the circuit path, and in trying to get the wire-connected 27k resistor soldered into place, the wire end lead broke off, so had to fetch a new 27k resistor. Got that, prep’d the leads & lead wire, soldered that into place. Glued it down with RTV, and a tie wrap around the wire leaving the PCB assy to the source location on the power supply board.
I verified both channels working. So I set it up into Bridge Mono, initially fixed Sine Wave drive to verify all was working, driving load resistor bank. Then switched over to Burst Pink Noise, and connected the 4 ohm Ampeg Test Speaker, and ran up the level to yield 105dB SPL, let it run and went to take a leak. Came back, all still working fine. I was about to sit down, when I then heard very distorted output. Turned the level down, switched out of Bridge Mono, and checked Left Ch. That’s working fine. Plugged output/analyzer into Right Ch, and it’s once again severely clipping. So, whatever WAS working fine and conducting current to the load isn’t any longer. RATS!
Now I’m back to trying to find out what has failed this time. Do I have output level without load? YES. Full output until I load it. Verified this with the scope probe, that it IS the upper half of the waveform that has dropped out again. WHAT is ceasing to conduct? I re-soldered the High Current connections this last time…the emitter resistors, output inductor/resistor, Header terminal.
After further re-soldering, and lifting the power supply board up for inspection, I tried again, and still…under load, I loose positive half drive current! I’m puzzled. I’m not yet ready to believe I have power xstrs causing this.
This time, I replaced R19 100 ohms on the cathode of Q7 2N5415 PNP, R26 100 ohm emitter resistor of Q10 TIP31 Driver, along with the 22pF around Q7 and the 470pF around Q10 Driver. Also re-soldered the two NPN Power Xstrs Base, Emitter and Collector leads.
I now have the amp once again passing load current, so I have it in Bridge Mono driving the 4 ohm test speaker at 107dB SPL level (nominal) with Burst Pink Noise, and pulling up to 4.7Amps AC Mains. Letting it run to see if it continues, as I did before. Started this burn-in process @ 1:40PM. At 2:05PM, I lost signal. Got up to have a look. Both heat sinks are extremely hot, so I suspect the thermal switch(s) have opened. I see idle current of 0.58A/46W being drawn. The internal fan is running, so I’ll see how long it takes for it to turn back on…assuming this has hit thermal-cycle shut-down.
At 2:20PM, I have signal again. I turned down the drive signal for 100dB SPL, pulling around 2A Peak. Letting it run at this level. So, that was thermal cycle cut-out, as I suspected. Using the Averaging Function of the Magtrol Power Analyzer, it’s drawing an average of 1.36A/112W using the 3 Sec On/2 Sec Off burst cycle. Output level shows around 8V RMS into the 4 ohm speaker for this nominal 100dB SPL level. I wonder if it will thermal cycle again at this level? NO….still running at 3:20PM.
I don’t recall having one of the WT800 amps fail in this mode before. I’ve lost power transistors, voltage gain stage xstrs, drivers and associated resistors. This one took a while to find just what was cutting out. Still uncertain that it WAS solder joints, but suspect that. When you look at the way the power amp modules are built, relying solely on the solder joints to hold the PCB into place, I’m not surprised. But usually, you see the tell-tale signs of joint fractures. Not this time.
WT800C schematic-1.pdf
WT800A POWER SUPPLY SCEMATIC.pdf
PM-400 Pwr Amp Sch-1.pdf
Attached Files
Comment