Played it yesterday for about 6 songs on Aux and then entire side of an LP. Had both top and bottom covers off so I had it sitting on its side. At that angle the major heat is not rising straight upward to the Voltage Amp board and the FETs. Played it most of the evening on mix of Aux iPod and LPs. No issues.
Today I put covers on, laid it upright but put my small fan w thermostat on top of case blowing downward. Again, it played for an hour or so w no issues. I suspect I may have a heat issue and maybe the fan is the best way to go regardless knowing how much heat this amp puts out.
Today I put covers on, laid it upright but put my small fan w thermostat on top of case blowing downward. Again, it played for an hour or so w no issues. I suspect I may have a heat issue and maybe the fan is the best way to go regardless knowing how much heat this amp puts out.
Its all sounding promising. The only real heat source would be from the output stage and as such it is very important to get the biasing correct so that it is at the nominal value when all the covers are on and its fully up to temperature. If that's OK then there should be no issues in normal use.
Mooly. Should I check DC offset at speakers when fan is running to see how biasing is doing w "cooler" temperature? I know I'm putting a band aid on the real issue but not knowing what is not liking the heat is a challenge. If it's the FETs, I'm hard pressed to find replacements. If it's the ICs that control the bias, those are $26 each from Pioneer. If it's something else, hmmm.
The fan(s) need to blow only on the heat sinks, NO components. The board mounted components need still air to stabilize, the moving air on the diff. transistors and the bias transistor will cause more problems.
Craig
Craig
Any normal variation of DC offset will not cause temperature problems. Even a massive offset of 0.25 volts would only cause around 30 milliamps to flow in an 8 ohm load.
The bias current (the current that flows in the output transistors) is the thing that could cause temperature increases if it were to high. Soldering a couple of wires across the appropriate emitter resistor and then bringing those wires outside the case to allow monitoring is one way of confirming all is well.
The bias current (the current that flows in the output transistors) is the thing that could cause temperature increases if it were to high. Soldering a couple of wires across the appropriate emitter resistor and then bringing those wires outside the case to allow monitoring is one way of confirming all is well.
Ah. I can't control the fan that much. It's too big (3.5"). I had it positioned over one sink and blowing down through there it was allowing air to circulate back up across the other. I doubt there was still air anywhere.
As for the output transistors, I know they are hot. 114F hot on heat sinks where they connect.
What I was suggesting was checking DC offsets to see if they are any difference with fan blowing vs without fan.
What I was suggesting was checking DC offsets to see if they are any difference with fan blowing vs without fan.
DC offset should change very little with temperature, but even if it did swing from say +0.25 volts to -0.25 volts as it warmed up, that change would not produce much change in how hot or cool the amp runs.
What produces heat is current flow in the output transistors. 114F is not really hot in the scheme of things.
What produces heat is current flow in the output transistors. 114F is not really hot in the scheme of things.
Ok. Next step is more tests without fan to see if things act up again.
Which transistors are the diff pair on this amp? Those are the one I replaced on the SA-8500 that you ( Mooly) helped me with a few months ago. It would cut out after extensive playing (4-8 hours) if you recall.
Which transistors are the diff pair on this amp? Those are the one I replaced on the SA-8500 that you ( Mooly) helped me with a few months ago. It would cut out after extensive playing (4-8 hours) if you recall.
Right
and they are in one package to give near perfect thermal balance. This is why I say your offset should change little with temperature.In my opinion, we haven’t found the problem yet.
From the beginning it is suspected that R115 (10 Ohm, 2W) became very warm.
With C39 and C41 in good condition, this indicate the presence of a substantial signal at a frequency above 100kHz at the amplifier output.
Since you don’t have an oscilloscope, I suggest to connect a small signal diode (like 1N4148) in series with a 0.1 uF capacitor between the amplifier output and the common (ground). This circuit will act as a detector to detect the presence of an oscillation. It should help for troubleshooting.
With no input signals, the DC voltage across the 0.1 uF capacitor should read 0 Vdc. It would require volts to make R115 warm, not just few mV.
The presence of the oscillation seems to be sporadic. Switching between the inputs, like moving from AUX to Phono, playing with the tone control may initiate the oscillation and create a significant voltage across the 0.1 uF. Probing the circuit, applying heat are other options to initiate the oscillation.
From the beginning it is suspected that R115 (10 Ohm, 2W) became very warm.
With C39 and C41 in good condition, this indicate the presence of a substantial signal at a frequency above 100kHz at the amplifier output.
Since you don’t have an oscilloscope, I suggest to connect a small signal diode (like 1N4148) in series with a 0.1 uF capacitor between the amplifier output and the common (ground). This circuit will act as a detector to detect the presence of an oscillation. It should help for troubleshooting.
With no input signals, the DC voltage across the 0.1 uF capacitor should read 0 Vdc. It would require volts to make R115 warm, not just few mV.
The presence of the oscillation seems to be sporadic. Switching between the inputs, like moving from AUX to Phono, playing with the tone control may initiate the oscillation and create a significant voltage across the 0.1 uF. Probing the circuit, applying heat are other options to initiate the oscillation.
So I'm trying to get a hold of an oscilloscope. Though I I would t know where to start with using, hooking it up to look for oscillation. I'll do some searches. In the meantime it appears to have stabilized.
I played it on Aux for about an hour last night with covers on, no fan.
As for the signal diode and cap test, I'll need more help with exactly where to connect it, which way diode should be oriented.
I played it on Aux for about an hour last night with covers on, no fan.
As for the signal diode and cap test, I'll need more help with exactly where to connect it, which way diode should be oriented.
What DASutton is suggesting is this:
The diode (a small IN4148 type etc) would connect in series with a small cap of say 0.1uf. You should also include a series resistor of anywhere from 10 to 1000 ohms. So three components in series with the cap at one end.
Connect this chain across your speaker output terminals and have the cap connected to the speaker negative end. With the amp on you should see zero DC voltage across the cap. Any voltage indicates the presence of an AC component at the speaker output and this is then getting rectified by the diode and giving a voltage reading on the DVM. Diode polarity doesn't matter.
You can't beat a good scope though.
The diode (a small IN4148 type etc) would connect in series with a small cap of say 0.1uf. You should also include a series resistor of anywhere from 10 to 1000 ohms. So three components in series with the cap at one end.
Connect this chain across your speaker output terminals and have the cap connected to the speaker negative end. With the amp on you should see zero DC voltage across the cap. Any voltage indicates the presence of an AC component at the speaker output and this is then getting rectified by the diode and giving a voltage reading on the DVM. Diode polarity doesn't matter.
You can't beat a good scope though.
Ok. I'll see if I can locate the diodes locally this weekend. So with the diode/cap tester, if I'm seeing DC voltage, that indicates oscillation may be occurring? What level would be acceptable? Less than 10mV?
You should see essentially 0.00 volts DC. Any small signal type silicon diode or Shottky type will do (lol, and germanium is even better for this because of its low forward voltage).
Don't use a big cap, a small 0.1 to 0.47uf is fine. No bigger.
Don't use a big cap, a small 0.1 to 0.47uf is fine. No bigger.
I ordered extra .1 film caps when I replaced C39/C41. Are those ok?
Also, do I check voltage at different volumes of just at idle?
Also, do I check voltage at different volumes of just at idle?
Those caps are fine.
You check at idle, the theory being that any AC signal (such as oscillation) will, as long as it is more than the forward voltage of the rectifier cause a voltage to develop across the cap.
Its just the same idea as a power supply half wave rectifier.
If there were anything present but less than around 0.5/0.6 volts then the diode would not turn on and you wouldn't register it. Germanium would get you down to around 0.15 volts threshold.
You check at idle, the theory being that any AC signal (such as oscillation) will, as long as it is more than the forward voltage of the rectifier cause a voltage to develop across the cap.
Its just the same idea as a power supply half wave rectifier.
If there were anything present but less than around 0.5/0.6 volts then the diode would not turn on and you wouldn't register it. Germanium would get you down to around 0.15 volts threshold.
I found some N4148 diodes at the Shack. So, to be clear, I put in series:
10 ohm resistor --> diode --> .1uF cap
Create 2 of these and connect resistor to positive speaker out and cap to negative speaker out.
Measure idle voltage across each cap.
10 ohm resistor --> diode --> .1uF cap
Create 2 of these and connect resistor to positive speaker out and cap to negative speaker out.
Measure idle voltage across each cap.
Here is what I measured.
DC Offset before connecting diodes/caps.
Left: 7.4mV
Right: -4mV
Across Caps.
Left: 2.4mV
Right: -1.6mV
DC Offset before connecting diodes/caps.
Left: 7.4mV
Right: -4mV
Across Caps.
Left: 2.4mV
Right: -1.6mV
- Status
- Not open for further replies.