I'm going to be using a pair of these amps on my doors and they need a little tweaking.
First, the filters are outta here. I'll only keep the pre-amp and substitute a low noise opamp for IC12. IC17 is the inverting buffer used for bridging but I should keep that to avoid having to rewire the speaker connection. I guess I could just re-label them and remove IC17 as well since I don't need bridging.
Second, I'll need a little help with the bias circuit. I need to make it adjustable. They used a thermistor to regulate bias but didn't mount it to the heatsink. It's just sitting on the pcb like a fish out of water.
Third, I need to shorten the turn-on delay. On a cold start it's about 6 seconds. It looks to me that C11 and R4 are responsible for the delay. Reducing C11 to 100uf should speed things up.
Perry - Would you have a moment to go over this?
First, the filters are outta here. I'll only keep the pre-amp and substitute a low noise opamp for IC12. IC17 is the inverting buffer used for bridging but I should keep that to avoid having to rewire the speaker connection. I guess I could just re-label them and remove IC17 as well since I don't need bridging.
Second, I'll need a little help with the bias circuit. I need to make it adjustable. They used a thermistor to regulate bias but didn't mount it to the heatsink. It's just sitting on the pcb like a fish out of water.
Third, I need to shorten the turn-on delay. On a cold start it's about 6 seconds. It looks to me that C11 and R4 are responsible for the delay. Reducing C11 to 100uf should speed things up.
Perry - Would you have a moment to go over this?
Attachments
It looks like the colors for the LED are wrong also.
To make the bias adjustable, you could substitute a 3.3k pot for R168/268. It's safer to have the pot on the low side of the bias transistors. If the pot opens (dirty), the bias will go low, instead of high.
I'd adjust the delay with C11. R4 changes the thermal shutdown. You can't go too low with C11 because, if I'm not mistaken, if it's too low, it will trigger the thermal protection when the amp powers up. This problem showed up on some amps that had defective C11 or broken solder connections.
To make the bias adjustable, you could substitute a 3.3k pot for R168/268. It's safer to have the pot on the low side of the bias transistors. If the pot opens (dirty), the bias will go low, instead of high.
I'd adjust the delay with C11. R4 changes the thermal shutdown. You can't go too low with C11 because, if I'm not mistaken, if it's too low, it will trigger the thermal protection when the amp powers up. This problem showed up on some amps that had defective C11 or broken solder connections.
Yes it seems the LED is labeled backwards. How do they build these things with so many errors?
In that case I might go with a 1k resistor and 2k pot to limit the bias range. At the moment I can not measure a voltage across the emitter resistors at all. Sounds ok for a class B amp. 🙂
I think I'd be fine going to 100uf. It's not that the delay is a bother but more then once a long power-on delay has tricked me into thinking an amp was dead.
Perry - At the end of the line in the filter stage, IC12 has the feedback resistor (R104) bypassed (C105). What's the reason for this? Only high frequency attenuation? What frequency?
In that case I might go with a 1k resistor and 2k pot to limit the bias range. At the moment I can not measure a voltage across the emitter resistors at all. Sounds ok for a class B amp. 🙂
I think I'd be fine going to 100uf. It's not that the delay is a bother but more then once a long power-on delay has tricked me into thinking an amp was dead.
Perry - At the end of the line in the filter stage, IC12 has the feedback resistor (R104) bypassed (C105). What's the reason for this? Only high frequency attenuation? What frequency?
You can find errors on most schematic diagrams. I don't think they build from this type of diagram.
Class B amps can sound perfectly fine.
After you adjust the bias to where you want it, you will need to heat it up the thermal shutdown to make sure that the bias holds.
I don't know what frequency it is but it's likely well outside of the audible range. It's likely used to ensure that the circuit with gain doesn't oscillate.
Class B amps can sound perfectly fine.
After you adjust the bias to where you want it, you will need to heat it up the thermal shutdown to make sure that the bias holds.
I don't know what frequency it is but it's likely well outside of the audible range. It's likely used to ensure that the circuit with gain doesn't oscillate.
Yup that's exactly it. I looked into opamp designs and see it's commonly used for opamp stability.
Last night I bypassed the filter and fed the audio to IC12 pin 3 and 5. It's much more transparent sounding this way.
There's so many mods needed I may just go back to one of my old amp designs and grab a couple of ebay powersupplies for it. It was clean, stable, and acoustically invisible.
Last night I bypassed the filter and fed the audio to IC12 pin 3 and 5. It's much more transparent sounding this way.
There's so many mods needed I may just go back to one of my old amp designs and grab a couple of ebay powersupplies for it. It was clean, stable, and acoustically invisible.
- Status
- Not open for further replies.