Once I got it in the right ballpark, I’d probably just quit fiddling with it so nothing gets broken. If I was using a pot, I’d probably try higher, but back off if I see it getting touchy. Just with the upgrades there are now, it’s going to sound 10x better, and be able to run both the internal and an external pair of speakers, with zero issue.
So leave it be with the PNP at 14.5 mA then. The other PNP is closer to 14 mA, be fine as well?
Thank you all so much! I bet she’s going to love it.
Dan
Thank you all so much! I bet she’s going to love it.
Dan
How did you finally end up mounting the diode stack to the heat sink? They do need to be at least loosely thermally coupled to the heat sink to prevent runaway when it’s turned up to 11.
It’s kind of hard to get a good photo, but I have them clamped in the same clamp the original diode used. I slipped them in and then squeezed the sides in and they’re held quite tight.
You think that will be sufficient?
Dan
One small concern I have is that on power up it has a fairly loud power on transient. If it’s been powered on after being powered off for say 2 minutes on power up is sounds like a short burst from a machine gun. Like 3 or 4 pops. If you cycle the power and it isn’t powered off too long then the woofer will throw, but it’s silent. I don’t know if it did this before hand. I dont remember it doing this, but at the same time one channel was out of commission so it’s not like I used it much.
Can’t tell if it’s and AC or DC transient. I measured the the speaker output in both AC and DC and I’m getting a reading on both, about 4-1/2 volts was peak seen.
Is this something that can be remedied or something to worry about as far as causing future issues?
One thing I thought about was to power it up with the speaker switches off, but found that if you power it and then turn the speakers on too quickly you get a loud pop. But if powered up and then wait 3-5 seconds you’re in the clear.
Dan
Can’t tell if it’s and AC or DC transient. I measured the the speaker output in both AC and DC and I’m getting a reading on both, about 4-1/2 volts was peak seen.
Is this something that can be remedied or something to worry about as far as causing future issues?
One thing I thought about was to power it up with the speaker switches off, but found that if you power it and then turn the speakers on too quickly you get a loud pop. But if powered up and then wait 3-5 seconds you’re in the clear.
Dan
Turn on/off transients are an issue with capacitor coupled amps. You might not have noticed as much with the tiny output capacitor. One thing that would have been affected by the changes is the time constant of the boot strap capacitor, when the VAS load was decreased. I would try reducing the cap by a factor of 10 and see if that helps with it.
A low frequency simulation is usually pretty accurate, even if transistor models are approximate. It’s dominated by the RC time constants and “a large amount of feedback”. Sim and play with it.
A low frequency simulation is usually pretty accurate, even if transistor models are approximate. It’s dominated by the RC time constants and “a large amount of feedback”. Sim and play with it.
Which one would be the bootstrap capacitor? Not C80? Here is the schematic for reference.
I have to say that the amplifier sounds excellent, I’ll have to see how the internal speakers while in the cabinet sound, but if she uses external she should definitely be pleased. Have it wired to a pair of Klipsch (to keep efficiency up) and the amp sounds great.
Dan
I have to say that the amplifier sounds excellent, I’ll have to see how the internal speakers while in the cabinet sound, but if she uses external she should definitely be pleased. Have it wired to a pair of Klipsch (to keep efficiency up) and the amp sounds great.
Dan
If somebody claimed "C82 appears to be a bootstrap capacitor" connected between nodes 119 and 120, I wouldn't argue with them
C82 is the boot strap capacitor. Worth a try, normally I like a “big” cap there, but it might be causing everything to come up out of sync. One can choose all the low frequency time constants to come up gently (therefore silently), and Spice can help with that. But on a low power circuit (in this case about 12, maybe 15 watts), and a cap in series with the speaker it won’t actually damage speakers. It’s not uncommon for these things to thump pretty good when the changer turns off at the end of the record.
One can also go on the warpath for small coupling and bypass caps further up the chain that could be going bad and causing preamp stages to come up too quickly, too, resulting in a pop being sent to the power stages. Which it faithfully amplifies. But I’d normally leave it alone unless there’s a telltale hum or wonky frequency response.
One can also go on the warpath for small coupling and bypass caps further up the chain that could be going bad and causing preamp stages to come up too quickly, too, resulting in a pop being sent to the power stages. Which it faithfully amplifies. But I’d normally leave it alone unless there’s a telltale hum or wonky frequency response.
lol, that was my second guess.
It seems like much of this, just like loudspeaker design, is a bunch of trade offs. I was doing some reading last night and read exactly what you had said about cap coupled amplifiers having a turn on/off thump. There doesn’t seem to be much of a turn off. I’m guessing if I went back to the original 470 µF that the thump would be reduced greatly.
You say that you like having larger values for the bootstrap. If I were to reduce it by a factor of 10, that would take it down to 22 µF. With this in any way affect what I have done downstream with the bias? Would I have to change the value of that resistor again? I’m wondering if it’s just better to keep the 220 µF in there. I can tell her to just turn on the speakers a few seconds after powering up the amplifier or just deal with the noise. It’s certainly not loud enough to damage the speakers as you say. Do you think that’s a good assessment? Or would you think the negative effects of reducing that capacitor value are minimal and the circuit would benefit more? I’ve heard of spice modeling, read a little about it, but have zero experience. Just another thing I should familiarize myself with.
Dan
The original 470 uF output caps would obviously help with the serverity of turn on thumps, but also cut bass. Noticeably. Which is why you do a bigger one. Going to 22 uF on the boot strap would take it’s time constant back where it was originally, reducing the contribution to what it would have been originally. It might not even be dominating things. The corner frequency at 12 Hz is probably low enough, although personally I prefer single digits there. It’s just a simple experiment one can try.
The real solution is a speaker relay - that’s why everything these days has one. It’s debateable if one is effective in the event of direct-coupled amp failure but it does prevent this. If there is a speaker switch, it works too. If there is an output stage failure the cap will always work to keep DC out of the speaker. A lot less likely to fail now, though. Those old Ge’s don’t live forever, and a VAS running at 2 watts can unsolder itself off the board.
The real solution is a speaker relay - that’s why everything these days has one. It’s debateable if one is effective in the event of direct-coupled amp failure but it does prevent this. If there is a speaker switch, it works too. If there is an output stage failure the cap will always work to keep DC out of the speaker. A lot less likely to fail now, though. Those old Ge’s don’t live forever, and a VAS running at 2 watts can unsolder itself off the board.
Ok, not to play dumb, the time constant has to do with the low frequency cutoff, just as the size of the coupling cap does. So they both play a roll? So going down to 22uF would likely take the frequency cutoff back to what it was before I increased the coupling cap size? I’ll leave it be, if it were mine I’d prefer the low frequency extension. Even though the internal speakers won’t need it it was mentioned that external speakers might be hooked up.
I thought of adding a small relay circuit, one of the generics out there. Probably not worth it unless she can’t remember to turn the speakers off or if the thump really bothers her.
Dan
The low frequency roll off due to the boot strap is inside the feedback loop and doesn’t directly affect the -3dB point like the coupling cap does. But the amount of feedback is reduced at low frequency. This increases low frequency distortion near and below its corner frequency. Changing it may or may not even affect the thump. It’s just one thing that was inadvertently changed so it cannot be ruled out as the cause until explored.
If this piece was originally a “price point” unit, it would tend to lack refinements of more expensive ones. It may just be par for the course. There is no way for the output stage to draw “stupid” amounts of current during these evens, and the amount of power produced won’t blow a woofer voice coil, so it is “safe”.
If this piece was originally a “price point” unit, it would tend to lack refinements of more expensive ones. It may just be par for the course. There is no way for the output stage to draw “stupid” amounts of current during these evens, and the amount of power produced won’t blow a woofer voice coil, so it is “safe”.