Hello, I am new to electrical engineering, but I am competent enough to understand a lot of the basic things. I am trying to fix a Bogen C100 that I acquired, and have attached the schematic provided by the manufacturer. When I got it 2 of the resistors had catastrophic failures and at first I didn't know why, and don't fully understand still. They are R70 and R77 both 22ohms. I replaced those with the same type and value, and when I turned in on, the BJT, G17, burned up. I believe R74 also was overstressed because part of it charred, but it didn't fail completely. It's value is also unlabeled on the schematic and I pointed it out. I got another BJT, a 2SD718 NPN BJT. I replaced that, but I didn't replace R74. I connected everything back, but a connector to another part of the circuit, wasn't plugged in. It was a 3 pin connect that I believe connected to the front potentiometer volume knows and it's board. There is only one actual plug connector, everything else is tab connectors on the main board. From what I remember that should mean that any other boards connect through that boards connection as well. When I turned it on that time, i saw sparks on what looked like the main board near the ac main side, but cant be 100% certain. I checked everything and nothing seem bad, burnt, or otherwise hurt. I noticed connectors between GND and COM as well as between LINK and 4ohm. I was wondering is something was shorted, so I checked for continuity on those connected terminals and all the others to include the 25V, 16ohm, 70V, and right end GRD. They all had continuity with each other, but I don't know if that is supposed to be like that or not. I ended up takeing off the connectors between them and turn it on again, I believe I plugged the 3 pin connector back in. Well the new BJT catastrophically failed and burned a hole on the right side, which may be the emmiter. The resistor that was unlabeled R74 also seemed worse. I measured the resistance of it before and it measured 219ohms in the circuit, but now value starts below that and climbs. I can't read the value because the middle 2 rings are charred off and or lost their color. I don't have a full picture of the internal components right now but but will be able to attach one later. I am attaching everything else I have and an image I found showing those back connectors.
So my questions are as follows;
1. What is the value of R74 supposed to be?
2. What could I have caused to make components fail?
3. Why is the device and components not working correctly.
4. Can I safely test components to make sure they are good and if so how? (If desoldering is needed that's is fine)
So my questions are as follows;
1. What is the value of R74 supposed to be?
2. What could I have caused to make components fail?
3. Why is the device and components not working correctly.
4. Can I safely test components to make sure they are good and if so how? (If desoldering is needed that's is fine)
Attachments
When a directly coupled amplifier fails, there is a cascade of failures that ripples back through the circuit destroying several devices. Any attempt at repairs by replacing one of the failed parts simply repeats the failure and sometimes worse. Whenever an amplifier repair is attempted, the initial power up must be done under controlled condition so that overlooked problems can be detected before more new parts are destroyed. The classic device is a power limiting device such as a series incandescent lamp in the AC power and perhaps the use of a Variac transformer to bring up the voltage slowly. Initial power up is done with no load connected.
1. 220 Ohms is a reasonable value for R74.
2. The circuit includes an overload detection and protection, but it may not be effective, especially if C35 and C36 have dried out. Failure is usually due to an output short, but this amp uses an output transformer which is a hazard to such an amp.
3. It's clear that the output transistors have failed. This results in a current surge that shorts some transistors and opens others. This is a classic failure.
4. Transistor shorts can be detected in circuit but ultimately you have to remove and test the bias VBE multiplier Q7, the protection circuit D1, Q13, Q15, D4?, C35, C36, drivers Q14, Q16, outputs Q17, Q18, Q19, Q20, resistors R74, R81, R82, R83, R84.
Once you have tested and replaced all these devices, you may attempt a power up using a power limiting device, and no load. When you get zero DC on the output ("4 Ohm" no link) and then connect a speaker to test the sound. A signal generator and oscilloscope is a better way to test the "sound" if you have them. Transistors can be partially tested as two diodes using an ohmmeter, however this is not a totally conclusive test.
1. 220 Ohms is a reasonable value for R74.
2. The circuit includes an overload detection and protection, but it may not be effective, especially if C35 and C36 have dried out. Failure is usually due to an output short, but this amp uses an output transformer which is a hazard to such an amp.
3. It's clear that the output transistors have failed. This results in a current surge that shorts some transistors and opens others. This is a classic failure.
4. Transistor shorts can be detected in circuit but ultimately you have to remove and test the bias VBE multiplier Q7, the protection circuit D1, Q13, Q15, D4?, C35, C36, drivers Q14, Q16, outputs Q17, Q18, Q19, Q20, resistors R74, R81, R82, R83, R84.
Once you have tested and replaced all these devices, you may attempt a power up using a power limiting device, and no load. When you get zero DC on the output ("4 Ohm" no link) and then connect a speaker to test the sound. A signal generator and oscilloscope is a better way to test the "sound" if you have them. Transistors can be partially tested as two diodes using an ohmmeter, however this is not a totally conclusive test.
Yes, about 60W or 100W.
Transistors before and including G3, G4, G5, G6, G8, G9 are unlikely to have failed because the current in these devices is limited by resistors etc. But all the outputs, drivers and the bias are suspect. Normally transistor designations are "Qn", but you seem to have figured that out.
Using an Ohmmeter, a working transistor appears to be 2 diodes, base-emitter and base-collector. Be sure there is no conduction from emitter to collector. As I said before, this is a basic test and is not definitive, but it will detect 95% of transistor failures. Some meters have a transistor/diode test, which will detect 98% of failures. The outputs are matched pairs, which should be replaced together even if one survived.
Transistors before and including G3, G4, G5, G6, G8, G9 are unlikely to have failed because the current in these devices is limited by resistors etc. But all the outputs, drivers and the bias are suspect. Normally transistor designations are "Qn", but you seem to have figured that out.
Using an Ohmmeter, a working transistor appears to be 2 diodes, base-emitter and base-collector. Be sure there is no conduction from emitter to collector. As I said before, this is a basic test and is not definitive, but it will detect 95% of transistor failures. Some meters have a transistor/diode test, which will detect 98% of failures. The outputs are matched pairs, which should be replaced together even if one survived.
Last edited:
Oh OK I think I understand most of what you said. So what is the resistance supposed to be between the Base and Emitter/Collector, or is it more just looking that current flows through them in one way? Is it only supposed to flow from base to emitter/Collector? Current is not supposed to flow from emitter to collector correct? Is current supposed to be able to flow from Collector to emmiter?
Voltage drop across B-E or C-E should be 0.55 to 0.7 v on silicon transistors. You'll see this number or 100x that on the DVM display when using diode scale. Current should not flow from C-E unless B-E is biased to .6 v. Current flow is in direction of arrow on diagram. Plus goes to tail of arrow.
DVM diode test is at 2 v or less. Transistors can stand that voltage and fail at 10, 20, 30, 50 v whatever the actual in circuit voltage is. I have found that testing current flow at 24 v predicts bad transistors that passed the DVM diode test. I put milliamps scale of dvm series 47k resistor series 24 v supply series junction of transistor. If backwards voltage causes 24/47000 microamps, the transistor is bad. Forwards is plus to C on npn, minus to C on pnp.
Really easier to replace the transistors that fail the DVM diode test, then power up with speaker disconnected and 60 w tungsten bulb series the AC line. Bulb lights up, you still have a problem. I then connect negative of DVM to analog ground (before the transformer) with alligator clip lead, probe around looking for stupid voltages like b-e 0, c-e 0, voltages across capacitors 0, voltages across resistors higher than the wattage of the resistor would allow (ie burnt up). Diodes can blow, short, or open too. According to the schematic you posted ground on the terminal block is the analog ground.
Do not use 2 hands to check voltages, voltage >24 across your heart can stop it. Do not wear jewelry on hands, wrists, or neck, high current at 1 v can burn your flesh to charcoal. Wear safety glasses, parts can explode.
I fix amps successfully without a scope. Sometimes you need a music source like a FM radio earphone jack, and an AC voltage detector like an ANALOG voltmeter with a pointer. Cheap DVM give random AC voltage numbers on music, and the $160 fluke RMS meter ignores RF oscillation above 8000 hz. 20 VAC scale is useful, 2 v AC scale is even better ahead of the VAS transistors Q6 Q9. A set of Pamona grabbers for probing helps you blow up fewer parts with your hands.
A cheap speaker (I use junkyard salvage ones) across the output while probing can help you find bad solder joints. POP means you just pressed one. You can use the transformer winding to prevent DC on speaker (25 v to common). But on amps without a speaker transformer you have to put ~2000 uf capacitors negative to negative series the speaker to prevent damage from DC.
DVM diode test is at 2 v or less. Transistors can stand that voltage and fail at 10, 20, 30, 50 v whatever the actual in circuit voltage is. I have found that testing current flow at 24 v predicts bad transistors that passed the DVM diode test. I put milliamps scale of dvm series 47k resistor series 24 v supply series junction of transistor. If backwards voltage causes 24/47000 microamps, the transistor is bad. Forwards is plus to C on npn, minus to C on pnp.
Really easier to replace the transistors that fail the DVM diode test, then power up with speaker disconnected and 60 w tungsten bulb series the AC line. Bulb lights up, you still have a problem. I then connect negative of DVM to analog ground (before the transformer) with alligator clip lead, probe around looking for stupid voltages like b-e 0, c-e 0, voltages across capacitors 0, voltages across resistors higher than the wattage of the resistor would allow (ie burnt up). Diodes can blow, short, or open too. According to the schematic you posted ground on the terminal block is the analog ground.
Do not use 2 hands to check voltages, voltage >24 across your heart can stop it. Do not wear jewelry on hands, wrists, or neck, high current at 1 v can burn your flesh to charcoal. Wear safety glasses, parts can explode.
I fix amps successfully without a scope. Sometimes you need a music source like a FM radio earphone jack, and an AC voltage detector like an ANALOG voltmeter with a pointer. Cheap DVM give random AC voltage numbers on music, and the $160 fluke RMS meter ignores RF oscillation above 8000 hz. 20 VAC scale is useful, 2 v AC scale is even better ahead of the VAS transistors Q6 Q9. A set of Pamona grabbers for probing helps you blow up fewer parts with your hands.
A cheap speaker (I use junkyard salvage ones) across the output while probing can help you find bad solder joints. POP means you just pressed one. You can use the transformer winding to prevent DC on speaker (25 v to common). But on amps without a speaker transformer you have to put ~2000 uf capacitors negative to negative series the speaker to prevent damage from DC.
Last edited:
Ok sorry but after the first sentence or two you lost me. By DVM, do you mean digital voltage meter? Also by diode scale, do you mean the diode mode on the meter? After that I am confused by your wording, I don't know if you are talking about specific things or are not spacing and wording things right. I am wondering if you are talking about testing things in series or whatnot. Also does the bulb need to be connected after the amp or before when testing, or does it even matter ?
1. https://en.wikipedia.org/wiki/DVM answer 3
2. yes
3. light bulb is in series with AC input to amp. Mine is in a grounded steel case with a power cord, AC socket for load, and a circuit breaker. The top is mesh so I can see if light bulb is lit or not. Light bulb stays in series with AC cord until it is out, DC voltage with no signal on direct (4 ohm) output is <100 mv, and amp plays music at low volume into trash speaker. Then you can remove light bulb. Purpose of light bulb is to prevent faults from burning up additional components, while you probe with DVM for suitable DC voltages through the circuit. Light bulb also tends to limit tendency of faulty components to blow up and bounce bits off the ceiling or your safety glasses.
After all those signs of good repair, with no signal check voltage across emitter resistors (R81,82,83,84) to ensure no output transistor has idle current in excess of 25 ma (0.12 mv/0.47 ohms). This amp has no adjustment for idle current, but if different output transistors causes a problem, changes to R57 or R60 would be used to adjust the voltage across Q7.
BTW rather than matching number of output transistor and drivers, I would use a modern replacement. If 1" wide, MJL1302/3281 or MJL21193/21194 should be in stock somewhere.
2. yes
3. light bulb is in series with AC input to amp. Mine is in a grounded steel case with a power cord, AC socket for load, and a circuit breaker. The top is mesh so I can see if light bulb is lit or not. Light bulb stays in series with AC cord until it is out, DC voltage with no signal on direct (4 ohm) output is <100 mv, and amp plays music at low volume into trash speaker. Then you can remove light bulb. Purpose of light bulb is to prevent faults from burning up additional components, while you probe with DVM for suitable DC voltages through the circuit. Light bulb also tends to limit tendency of faulty components to blow up and bounce bits off the ceiling or your safety glasses.
After all those signs of good repair, with no signal check voltage across emitter resistors (R81,82,83,84) to ensure no output transistor has idle current in excess of 25 ma (0.12 mv/0.47 ohms). This amp has no adjustment for idle current, but if different output transistors causes a problem, changes to R57 or R60 would be used to adjust the voltage across Q7.
BTW rather than matching number of output transistor and drivers, I would use a modern replacement. If 1" wide, MJL1302/3281 or MJL21193/21194 should be in stock somewhere.
If R74 has fried, it is because Q14 and/or Q16 are shorted. They provide the excess current path which causes this. Outputs can be blown too, but only blown outputs wont usually take out the charge- suckout resistor since the fault current does not flow through it.
In case of a driver or output failure, you’re supposed to replace the whole kit and kaboodle, and at least TEST the vbe multiplier. You can get away with just replacing which side is blown if only one is - but you do risk leaving overstressed devices in there. In any case, if one fails in a parallel bank, you must replace the whole bank so that there is CONSISTENCY.
B817/D1047 is a modern output pair, but the bean counters have decided not to keep them when higher-running, higher rated, similarly priced alternatives are available. The 140 volt line can always be replaced with 230 volt versions, so they got dropped. They are obsolete, so the usual caveats surrounding sourcing them apply. If you can get them and trust them, great. Personally I wouldn’t try. C5200N/A1943N (TO-3P-N not L case version), the Fairchild equivalents (C5242/A1962), or any of the “NJW” or “MJW” pairs from ON will work in their place. The “MJL’s” may not fit, and cost more anyway. Even the 4 MHz types are fine in this application. I might be inclined to use C5200N driver and MJW21194 outputs.
In case of a driver or output failure, you’re supposed to replace the whole kit and kaboodle, and at least TEST the vbe multiplier. You can get away with just replacing which side is blown if only one is - but you do risk leaving overstressed devices in there. In any case, if one fails in a parallel bank, you must replace the whole bank so that there is CONSISTENCY.
B817/D1047 is a modern output pair, but the bean counters have decided not to keep them when higher-running, higher rated, similarly priced alternatives are available. The 140 volt line can always be replaced with 230 volt versions, so they got dropped. They are obsolete, so the usual caveats surrounding sourcing them apply. If you can get them and trust them, great. Personally I wouldn’t try. C5200N/A1943N (TO-3P-N not L case version), the Fairchild equivalents (C5242/A1962), or any of the “NJW” or “MJW” pairs from ON will work in their place. The “MJL’s” may not fit, and cost more anyway. Even the 4 MHz types are fine in this application. I might be inclined to use C5200N driver and MJW21194 outputs.
I think I understood a lot of that lol, I'm still new and don't understand most of how amp circuits work. I have a few simple questions, and I have 2 D718 NPN BJTs and 2 B688 PNP BJTs. So a good D718 NPN should have 0.6 V with positive on Base, and negitive on either Collector or emitter correct? With the B688 PNP it should be 0.6 V with Negitive on base , and positive on either emitter or collector correct ? These should not allow anything else correct ? Also should the Collector pin have continuity with the metal backing? With that said should the metal backing and in turn the Collector be grounded via the metal shell of the unit? For context they are screwed in to the back that has a heat sink. They have thermal paste on them, but they also had little plastic Rectangular sheets between them and the metal. I thought this might have been a mistake as I didn't know that any of the pins were in continuity with the metal backing, and only thought it was a heat sink/spreader surface. Also indianajo you didn't really answer 2 of my questions and a lot of what you said was not specific and confusing.
Those types are even smaller than the D1047 I’ve seen in this amp, just as obsolete, and even more likely to show up as fakes when you buy them. I’d be using the Fairchild pair (C5242 and PNP) because I trust the source. That and I have a drawer full of each.
Correct on the polaritires. If you short metal tab to chassis (omit the insulator) you short the power supply directly to ground. This will blow a fuse, and may or may not be fatal to the transistor. It’s a direct connection and doesn’t go through the device or bond wires.
Correct on the polaritires. If you short metal tab to chassis (omit the insulator) you short the power supply directly to ground. This will blow a fuse, and may or may not be fatal to the transistor. It’s a direct connection and doesn’t go through the device or bond wires.
can only speak english. If english is not understood read reference material like wikipedia or dictionary.
Plastic sheets insulate output transistor collectors from the heat sink. Collector can be wired though a pin. If sheets are clear (mica) they get brittle after 25 years and should be replaced. Modern insulators are silicon rubber, require no heat sink compound and do not get old.
R77 is base driver resistor to Q16 lower driver. Usually this blows when output transitor failure takes out Q16. O.T. could have been replaced by previous owner. Short of ZD1 or C33 could do it. R74 is not labeled but if it is the one next to D2 between bases of Q17 & Q19, then yes, output transistor failure or driver failure either one could burn it up.
When output transistors short extremely high current slips out the base line and blows up a lot of parts forward of that. Output transistors failure usually caused by wiring fault or shorted speaker coil or part. Screw terminal P.A. amps do not suffer the wiring faults that bar band phone jack amps incur all the time.
IC1 increases voltage of microphones, Q2 increases voltage of aux input (line level like CD player etc). IC2 are bass & treble controls. Q3 increases voltage of whatever signal is selected. Q4 is the negative feedback from the output that linearizes all these exponential gain transistors. Q5 is constant current source to provide steady current to Q3 & Q4. Q6 & Q9 are VAS provide final voltage gain to what comes out the end. Q7 is voltage spreader to make sure bases of drivers Q14 & Q16 are about 1.2 to 2.4 volts apart. This establishes O.T. idle bias current. Output transistors Q17,18,19,20 have to be on at all times to avoid the nasty crossover distortion of output transistors switching on and off. About 25 ma at no signal, idle, running DC through the output transistors, measured across emitter resistors R81,82,83,84. Drivers Q14 Q16 provide current amplification to the voltage from VAS. Emitter followers Q17,18,19,20 provide more current amplification to match speaker current requirement.
Last edited:
Should any components be grounded to the system, other that the obvious, such as transformers and such. Really I'm specifically asking about these two components screwed into the back of the case. I learned that the transistors beside them were likely shorted to ground when they weren't supposed to, so I'm making sure neither of the other two have the same issue. I believe one is only plastic though.
Attachments
The thing in the yellow circle may be a heat sensor transistor. I believe in that metal loop connected directly to the back of the case is a black plastic TO-92 case transistor. So the plastic of the case insulates the electrical parts. The metal loop is supposed to conduct heat to the transistor. the screw holds the metal loop to the case.
No heat temperature sense transistor in the schematic, so this could be R67 a thermistor. R67 has it's own insulating case. Looks like a thermal shutdown of signal to Q5 & Q8.
No heat temperature sense transistor in the schematic, so this could be R67 a thermistor. R67 has it's own insulating case. Looks like a thermal shutdown of signal to Q5 & Q8.