Tweet
This came in from our rental inventory, marked No Output. Amp was set for Bridge Mono. There is always a momentary pause of dread seeing these, knowing how much ‘fun’ they are to disembowel as well as the fragility of solder pads & traces on the power amp PCB.
I pulled the top cover for the initial inspection. Peered down into the cavity where the 0.47 ohm/5W Source resistors live to see if there’s shrapnel present. All looked ok. Looked at the front side of the N-Channel MosFET’s, not seeing any burnt gate resistors. I checked all ten of the IRFP240’s for shorts as well as seeing if I could turn them on and off with my Fluke 8060A. All ok, and were behaving like working parts. With the amp fully assembled in the rear, access to the main PCB and the P-Ch MosFET’s is denied without further effort, so I just hooked up two of my long EZ-Hook probes between GND and each channel’s output buss (clipping onto accessible Source Resistors ahead of the output relays).
Plugged the mains cord to my variac/power analyzer, switched the power switch on, then slowly began raising the AC mains, watching the two DMM channels on the 4-Ch HP3467A Logging DMM. The outputs of both channels began going negative quickly, while no abnormal AC Mains current was being drawn. I powered down, removed the probes & mains cord, and began disassembly to extract the power amp PCB assembly.
I got it out of the chassis, this one only having US Primary, and not the split primary. Wire terminals had already been labeled as to their mating PCB connections from previous servicing. I moved the chassis and rear panel PCB assemblies aside to have a look at the P-Ch MosFET’s and the rest of the front end of the power amp channels. As both channels yielded the same results, this suggested power supply issues. Still, I went thru all of the basic checks, verifying the P-Ch MosFET’s & their behavior, the source & gate resistors, then checked main power supply bridge and it’s solder joints, known to be a source of radial fractures around their leads. Those were fine, as were all the joints on the main supply caps, relay terminals, rear panel jacks and the like.
Checked all of the front end semiconductors, all measuring ok. I removed the power supply PCB assembly from the rear heat sink for inspection & checking diodes & connections, all looked ok, measured ok.
With this amp assembly out of the chassis, I next connected my Tek PS503A Dual Tracking Supply to the LV AC input header, so I could check the IC regulators as well as the IC’s on the main PCB. I got +/-15V with normal tolerances on the regulators, found near-0VDC on the quad op amps, and high voltages of both polarities on IC3, which controls the two relays as well as the fan circuits, so that seemed ok. IC1 had both of it’s outputs latched negative. Looking at the schematic, that op amp is getting feedback from the power amp output thru it’s non-inverting input, accounting for the phase shift in the voltage gain circuit that follows it…driving the MosFET power stages, so without the HV present, these op amps would latch up like that. So, that made sense.
I moved another power supply to the bench…my HP 6227B, which is dual tracking, though only goes to +/- 25VDC @ 2A. Still, should be adequate to see LV behavior. With it connected, I switched on the Tek PS503A for the +/- 15V supplies & relays, and began raising the HP supply. Output again going negative. I replaced IC1, being the controlling gain stage for the two channels, but got the same results. With it’s outputs while the HP supply wasn’t up, now latching positive. Still seemed reasonable. When I moved my other DMM probes to monitor the +/-15V outputs of the IC regulators, this time the relays clicked in when I applied power from the TEK PS503A, and raising the HP supply, the outputs weren’t going negative. Huh?
I also noticed the terminals on the HP supply where my joining wire req’d for Tracking Mode connected the (-) terminal of the Plus supply to the (+) terminal of the Neg supply wasn’t fully tight, I gave them a twist to tighten them back up, and the two supplies suddenly went low. HUH? I had the Plus supply’s Current level turned part-way up, while the Neg supply’s Current level turned all the way up, it being the tracking supply, and use the Plus supply’s Voltage knob to adjust both supplies, normal operation for it. Turning it back up, both meters began to rise, then both went to near-zero. I disconnected their outputs, checking the supply unloaded. Same behavior in Dual Tracking. Switch to Independent, to find the Plus supply (master in Tracking mode) still latches, while the Neg supply works fine. No idea what happened, but I just killed the power supply somehow.
That left me with having to power this board using extender cables from the chassis power xfmr. So, set that up, putting the top cover partially on and insulating material under the power amp PCB assembly and re-attached my probes, now using the variac to control everything, and no longer using the TEK PS503A LV bipolar supply. And, now the amp is powering up, no longer going negative as it had been doing. Does that mean the problem WAS IC1? I guess. I switched the supply on and off, with the AC mains set for 120VAC, still in this open configuration.
I shut it down, let the supplies drain off (they go down quickly), so put the amp back together, and went to power it up again, still monitoring the outputs ahead of the relays. Outputs go negative again. Sigh………. I connected my other DMM channels across the outputs of the two IC regulators, disconnected the supply cable to the preamp, and now, powering up, the outputs no longer go negative and the relays click in. Something squirrely on the power supply PCB assy. I removed the rear panel PCB’s to gain full access to the mounting screws of that power supply PCB, removed it, and unsoldered the two IC regulators. Installed the new regulators, now using greased mica to mount them. While soldering them into place, I noticed I had solder joint fracture rings around the LV supply header! How did I miss that? De-soldered and re-soldered those, put the supply assembly back onto the heat sink, re-connected the output cable to the preamp board, and for the moment, left the other rear panel PCB ‘s out. Went to power up the amp again, and now everything powers up fine. Patched signal into the power amp inputs, and had output from both channels.
Powered down, finished putting the rear panel PCB assemblies back into place, turned the amp around facing the front, re-connected the output monitoring cables just in case, and powered back up. Outputs still less than 10mV of offset. Connected signal to the input, with the output being monitored on the analyzer/scope. Turned up the gain & master, now finding the output level is low. Checked the power amps again, patching oscillator into the Power Amp inputs, and got full output, so their gain structure is nominal. Dead-patched the power amp inputs, and connected the analyzer/scope to the Preamp Output, signal applied to the input. Output level still low. Tubes? Removed the hold-down clamp. Moved the analyzer/scope to the Send Jack, ahead of the Preamp Output. Checked the output level, noting its’ reading. Swapped the input tube, now have increased 6dB gain. Swapped that tube with the second tube, output level went down 6dB. Swapped that second tube with another, gain restored. Swapped the output tube, no difference, so restored it. Moved the analyzer/scope cable back to the preamp output, and level much better.
Changed to the shop bass, and connected test speaker, while monitoring the amp output on the analyzer/scope. Sounds fine, everything works. Switched back to Bridge Mono, working fine. Put the amp back together. Sure took long enough! And, it cost me damage to my HP supply in the process! RATS! Still don’t know if the IC regulators were bad…..may have just been those solder joints on the three-pin LV header!
I pulled the top cover for the initial inspection. Peered down into the cavity where the 0.47 ohm/5W Source resistors live to see if there’s shrapnel present. All looked ok. Looked at the front side of the N-Channel MosFET’s, not seeing any burnt gate resistors. I checked all ten of the IRFP240’s for shorts as well as seeing if I could turn them on and off with my Fluke 8060A. All ok, and were behaving like working parts. With the amp fully assembled in the rear, access to the main PCB and the P-Ch MosFET’s is denied without further effort, so I just hooked up two of my long EZ-Hook probes between GND and each channel’s output buss (clipping onto accessible Source Resistors ahead of the output relays).
Plugged the mains cord to my variac/power analyzer, switched the power switch on, then slowly began raising the AC mains, watching the two DMM channels on the 4-Ch HP3467A Logging DMM. The outputs of both channels began going negative quickly, while no abnormal AC Mains current was being drawn. I powered down, removed the probes & mains cord, and began disassembly to extract the power amp PCB assembly.
I got it out of the chassis, this one only having US Primary, and not the split primary. Wire terminals had already been labeled as to their mating PCB connections from previous servicing. I moved the chassis and rear panel PCB assemblies aside to have a look at the P-Ch MosFET’s and the rest of the front end of the power amp channels. As both channels yielded the same results, this suggested power supply issues. Still, I went thru all of the basic checks, verifying the P-Ch MosFET’s & their behavior, the source & gate resistors, then checked main power supply bridge and it’s solder joints, known to be a source of radial fractures around their leads. Those were fine, as were all the joints on the main supply caps, relay terminals, rear panel jacks and the like.
Checked all of the front end semiconductors, all measuring ok. I removed the power supply PCB assembly from the rear heat sink for inspection & checking diodes & connections, all looked ok, measured ok.
With this amp assembly out of the chassis, I next connected my Tek PS503A Dual Tracking Supply to the LV AC input header, so I could check the IC regulators as well as the IC’s on the main PCB. I got +/-15V with normal tolerances on the regulators, found near-0VDC on the quad op amps, and high voltages of both polarities on IC3, which controls the two relays as well as the fan circuits, so that seemed ok. IC1 had both of it’s outputs latched negative. Looking at the schematic, that op amp is getting feedback from the power amp output thru it’s non-inverting input, accounting for the phase shift in the voltage gain circuit that follows it…driving the MosFET power stages, so without the HV present, these op amps would latch up like that. So, that made sense.
I moved another power supply to the bench…my HP 6227B, which is dual tracking, though only goes to +/- 25VDC @ 2A. Still, should be adequate to see LV behavior. With it connected, I switched on the Tek PS503A for the +/- 15V supplies & relays, and began raising the HP supply. Output again going negative. I replaced IC1, being the controlling gain stage for the two channels, but got the same results. With it’s outputs while the HP supply wasn’t up, now latching positive. Still seemed reasonable. When I moved my other DMM probes to monitor the +/-15V outputs of the IC regulators, this time the relays clicked in when I applied power from the TEK PS503A, and raising the HP supply, the outputs weren’t going negative. Huh?
I also noticed the terminals on the HP supply where my joining wire req’d for Tracking Mode connected the (-) terminal of the Plus supply to the (+) terminal of the Neg supply wasn’t fully tight, I gave them a twist to tighten them back up, and the two supplies suddenly went low. HUH? I had the Plus supply’s Current level turned part-way up, while the Neg supply’s Current level turned all the way up, it being the tracking supply, and use the Plus supply’s Voltage knob to adjust both supplies, normal operation for it. Turning it back up, both meters began to rise, then both went to near-zero. I disconnected their outputs, checking the supply unloaded. Same behavior in Dual Tracking. Switch to Independent, to find the Plus supply (master in Tracking mode) still latches, while the Neg supply works fine. No idea what happened, but I just killed the power supply somehow.
That left me with having to power this board using extender cables from the chassis power xfmr. So, set that up, putting the top cover partially on and insulating material under the power amp PCB assembly and re-attached my probes, now using the variac to control everything, and no longer using the TEK PS503A LV bipolar supply. And, now the amp is powering up, no longer going negative as it had been doing. Does that mean the problem WAS IC1? I guess. I switched the supply on and off, with the AC mains set for 120VAC, still in this open configuration.
I shut it down, let the supplies drain off (they go down quickly), so put the amp back together, and went to power it up again, still monitoring the outputs ahead of the relays. Outputs go negative again. Sigh………. I connected my other DMM channels across the outputs of the two IC regulators, disconnected the supply cable to the preamp, and now, powering up, the outputs no longer go negative and the relays click in. Something squirrely on the power supply PCB assy. I removed the rear panel PCB’s to gain full access to the mounting screws of that power supply PCB, removed it, and unsoldered the two IC regulators. Installed the new regulators, now using greased mica to mount them. While soldering them into place, I noticed I had solder joint fracture rings around the LV supply header! How did I miss that? De-soldered and re-soldered those, put the supply assembly back onto the heat sink, re-connected the output cable to the preamp board, and for the moment, left the other rear panel PCB ‘s out. Went to power up the amp again, and now everything powers up fine. Patched signal into the power amp inputs, and had output from both channels.
Powered down, finished putting the rear panel PCB assemblies back into place, turned the amp around facing the front, re-connected the output monitoring cables just in case, and powered back up. Outputs still less than 10mV of offset. Connected signal to the input, with the output being monitored on the analyzer/scope. Turned up the gain & master, now finding the output level is low. Checked the power amps again, patching oscillator into the Power Amp inputs, and got full output, so their gain structure is nominal. Dead-patched the power amp inputs, and connected the analyzer/scope to the Preamp Output, signal applied to the input. Output level still low. Tubes? Removed the hold-down clamp. Moved the analyzer/scope to the Send Jack, ahead of the Preamp Output. Checked the output level, noting its’ reading. Swapped the input tube, now have increased 6dB gain. Swapped that tube with the second tube, output level went down 6dB. Swapped that second tube with another, gain restored. Swapped the output tube, no difference, so restored it. Moved the analyzer/scope cable back to the preamp output, and level much better.
Changed to the shop bass, and connected test speaker, while monitoring the amp output on the analyzer/scope. Sounds fine, everything works. Switched back to Bridge Mono, working fine. Put the amp back together. Sure took long enough! And, it cost me damage to my HP supply in the process! RATS! Still don’t know if the IC regulators were bad…..may have just been those solder joints on the three-pin LV header!
Comment