So today's repair is a Yamaha P4500 power amp, which went into thermal protect when driven hard. Fine once it cooled. It appears that the fans aren't kicking in.
Fans work when connected to a DC supply, but not turning in normal use, getting 0V when tested.
They're variable speed fans, so there will be a temp sensor somewhere inside. Does anyone know where?
I'd rather not resort to plan B, which is to wire a permanent 24V supply to the fans.
Of course, it could be that it's all working fine now - the fans maybe just needed cleaning to free them up - but since I don't have a dummy load or tolerant neighbours, I can't really test it under load at the moment....
Full Version: Yamaha P4500 repair
QUOTE (bruce @ 17 Jun 2009, 7:44 PM) 
but since I don't have a dummy load or tolerant neighbours, I can't really test it under load at the moment....
Do you have a 1kW electric fire heater? The element is often about the right resistance for amp testing!
Sorry, I don't know where the thermistor is....
Circuit diagram obtained.
Dummy load improvised from a few parcans
It looks like there are 2 thermistors on each of the main heatsinks, one of which triggers at 60 degrees and turns the fans on, and one which triggers at 90 and turns on the protection circuit. The "fan" circuit is not coming on, the "prot" is.
Looks like maybe there's a fault in the 60 circuit. 2 thermistors, 3 transistors, a couple of zeners and resistors. And, of course, it's in the most inaccessible place...
Dummy load improvised from a few parcans
It looks like there are 2 thermistors on each of the main heatsinks, one of which triggers at 60 degrees and turns the fans on, and one which triggers at 90 and turns on the protection circuit. The "fan" circuit is not coming on, the "prot" is.
Looks like maybe there's a fault in the 60 circuit. 2 thermistors, 3 transistors, a couple of zeners and resistors. And, of course, it's in the most inaccessible place...
I can't vouch for the Yam specifically but many amps use the temperature signal to not only control the fan speed but also the very same signal to keep the bias stable.
Job done. It's frustrating how you sometimes work through a job in a structured methodical order, then eventually come across the fault completely by accident....
I found a pdf of the circuit diagram.
Fan circuit is completely independent of everything else - a couple of thermistors in series on the 2 main heatsinks, that and a fixed resistor forming a potential divider between the 0V and -27V rail. That then feeds a simple circuit (3 transistors, 2 Zeners, a few resistors and capacitors) which drives the fans. As the heatsink gets hotter, the fans go faster.
Stage 1. Test the fans. Both fine.
Stage 2. Locate the thermistors, stick a meter across them, apply gentle heat, see what happens. Voltages change as expected, so next port of call is the fan circuitry. Which is on the most inaccessible board - the same one as the main PSU smoothing caps. So remove the IO/preamp board to get access to it.
Stage 3. Measure voltages at input and output of this circuit. Input changes, output doesn't. We're getting closer to the fault. But the board needs to come out to test the components. But before doing that, let's measure the voltages around the final A1837 transistor. Stick the voltmeter between
0V and one of the pins, and the fan comes on. The initial daft thoughts of "could the miniscule current passed by the meter be switching the transistor" were immediately replaced by "don't be daft, Bruce, it's just a dry joint".
So, one dab of solder, and all is well. But getting to it meant dismantling the whole thing!
I found a pdf of the circuit diagram.
Fan circuit is completely independent of everything else - a couple of thermistors in series on the 2 main heatsinks, that and a fixed resistor forming a potential divider between the 0V and -27V rail. That then feeds a simple circuit (3 transistors, 2 Zeners, a few resistors and capacitors) which drives the fans. As the heatsink gets hotter, the fans go faster.
Stage 1. Test the fans. Both fine.
Stage 2. Locate the thermistors, stick a meter across them, apply gentle heat, see what happens. Voltages change as expected, so next port of call is the fan circuitry. Which is on the most inaccessible board - the same one as the main PSU smoothing caps. So remove the IO/preamp board to get access to it.
Stage 3. Measure voltages at input and output of this circuit. Input changes, output doesn't. We're getting closer to the fault. But the board needs to come out to test the components. But before doing that, let's measure the voltages around the final A1837 transistor. Stick the voltmeter between
0V and one of the pins, and the fan comes on. The initial daft thoughts of "could the miniscule current passed by the meter be switching the transistor" were immediately replaced by "don't be daft, Bruce, it's just a dry joint".
So, one dab of solder, and all is well. But getting to it meant dismantling the whole thing!
Well done Bruce, you certainly know you're stuff!
So a dry solder joint after all that?
Who'd have thought!
Personally, I would have spent an hour with a fine tipped iron dabbing the whole thing, if it didn't work after that, maybe a few taps with a lump hammer....
At least I know where to bring my faulty kit now.
So a dry solder joint after all that?
Who'd have thought!
Personally, I would have spent an hour with a fine tipped iron dabbing the whole thing, if it didn't work after that, maybe a few taps with a lump hammer....
At least I know where to bring my faulty kit now.
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.