My home audio system is basically a two-way mono HiFi cabinet situated under the television. The woofer and tweeter each have their own channel of amplification from a stereo power amplifier and input switching is done through a passive switch box.
The crossover/mixer is a simple first-order affair presently implemented with passive line-level components. This works very well for sources with their own volume control, but other components (I'm looking at you, PS3) require an external adjustment. To that end, I would like to add a dual-channel buffer followed by a stereo volume control pot after the passive crossover/mixer components, but I think there are some issues with my buffer design that need to be ironed out.
Previous implementations of a similar buffer have worked perfectly with two exceptions: lightning strikes and Bluetooth adapters. With the buffer in place, electrical storms would, without fail, cause the system to buzz loudly until power to the buffer had been reset. More troubling, however, was the interaction of the buffer with a Bluetooth audio receiver for streaming music from mobile devices. It was VERY finicky, and required constant toggling of the power to get a hum-free connection. In both cases, switching to a passive front end completely eradicated these issues.
My buffer was a garden variety TL082 circuit driven from a +/- 17vdc supply (+/- 12.6v power transformer > bridge rectifier > 10uf electrolytics > .01uf ceramics near power pins). I suspect I could significantly lower the supply voltage since it is running near the capacity of the chip, but I'm not sure if there is some other filtering issue I'm missing.
Thanks in advance for any ideas,
Nick
The crossover/mixer is a simple first-order affair presently implemented with passive line-level components. This works very well for sources with their own volume control, but other components (I'm looking at you, PS3) require an external adjustment. To that end, I would like to add a dual-channel buffer followed by a stereo volume control pot after the passive crossover/mixer components, but I think there are some issues with my buffer design that need to be ironed out.
Previous implementations of a similar buffer have worked perfectly with two exceptions: lightning strikes and Bluetooth adapters. With the buffer in place, electrical storms would, without fail, cause the system to buzz loudly until power to the buffer had been reset. More troubling, however, was the interaction of the buffer with a Bluetooth audio receiver for streaming music from mobile devices. It was VERY finicky, and required constant toggling of the power to get a hum-free connection. In both cases, switching to a passive front end completely eradicated these issues.
My buffer was a garden variety TL082 circuit driven from a +/- 17vdc supply (+/- 12.6v power transformer > bridge rectifier > 10uf electrolytics > .01uf ceramics near power pins). I suspect I could significantly lower the supply voltage since it is running near the capacity of the chip, but I'm not sure if there is some other filtering issue I'm missing.
Thanks in advance for any ideas,
Nick
Just quickly reading your first post and it seems you are not using any voltage stabilisation at all. A 12 volt transformer could easily output more than that under light load and high mains conditions and exceed the IC rating.
These should explain it all. Put your own numbers in and do the maths (not math
)
http://www.diyaudio.com/forums/chip-amps/166100-preamp-chip-amp-same-power-supply.html#post2674522
http://www.diyaudio.com/forums/analog-line-level/167495-simple-stereo-gain-stage-2.html#post2200615
These should explain it all. Put your own numbers in and do the maths (not math
)http://www.diyaudio.com/forums/chip-amps/166100-preamp-chip-amp-same-power-supply.html#post2674522
http://www.diyaudio.com/forums/analog-line-level/167495-simple-stereo-gain-stage-2.html#post2200615
This response is a long-overdue update, but I would like to thank everyone who offered suggestions to my initial problem. Shortly after posting, a different amplifier was substituted in my system, necessitating some changes to the overall layout. I now have three sources (cable box, iPad, and CD player) connected to a passive switch box. The output of the switch box goes to a passive RC network that sums and filters the signal before hitting the input of the amplifier - one channel of which controls the tweeter and one the woofer. Each source has its own volume control (the amplifier has no volume control) and the system works flawlessly.
Almost.
As stated in my original post, the PS3 has no provision for controlling the master volume of the analog audio output, which is a major hassle. I can't readily add a passive volume control to the PS3 because it would mess with the RC network filter response. To that end I built the buffered volume control pictured below (hopefully):
The PS3 output goes a stereo potentiometer, and each channel has its own op-amp buffer. Based on the advice in previous threads, I have incorporated the following suggestions:
-reduced power supply voltage from +/-18 volts to +/- 9 volts
-increased filter capacitors from 10uF to 100uF
-added a series resistor on the output (though it is part of the RC network downstream, maybe fifteen feet of cable away)
When working properly, the circuit has negligible hum and noise and sounds very acceptable. Occasionally, however, a very loud and objectionable power supply hum is introduced into the mix. As you can see from the photos, I don't have a tremendous amount of room to work, so I am trying to determine the best course of action, given the good design practices I have already implemented.
The regulation scheme described at Mooly's link looks promising, but it would be difficult to implement on my board unless it is acceptable to have the zener diode and current-limiting resistor acting directly on the unfiltered signal coming directly off of the rectifier.
Adding a series output resistor would be another potential solution, but unless it is acceptable to include it within the feedback loop (I've read conflicting accounts), it would also alter the RC network response. As stated above, there is a series resistor between the buffer output and the amplifier input (part of the RC network), but I suspect its effect may be negated by the impedance of the lengthy cables in between.
I guess I'm looking for the most elegant solution since I am trying to avoid compromising my layout, which I am quite pleased with. Any additional advice would be much appreciated.
Sincerely,
Nick
Almost.
As stated in my original post, the PS3 has no provision for controlling the master volume of the analog audio output, which is a major hassle. I can't readily add a passive volume control to the PS3 because it would mess with the RC network filter response. To that end I built the buffered volume control pictured below (hopefully):
The PS3 output goes a stereo potentiometer, and each channel has its own op-amp buffer. Based on the advice in previous threads, I have incorporated the following suggestions:
-reduced power supply voltage from +/-18 volts to +/- 9 volts
-increased filter capacitors from 10uF to 100uF
-added a series resistor on the output (though it is part of the RC network downstream, maybe fifteen feet of cable away)
When working properly, the circuit has negligible hum and noise and sounds very acceptable. Occasionally, however, a very loud and objectionable power supply hum is introduced into the mix. As you can see from the photos, I don't have a tremendous amount of room to work, so I am trying to determine the best course of action, given the good design practices I have already implemented.
The regulation scheme described at Mooly's link looks promising, but it would be difficult to implement on my board unless it is acceptable to have the zener diode and current-limiting resistor acting directly on the unfiltered signal coming directly off of the rectifier.
Adding a series output resistor would be another potential solution, but unless it is acceptable to include it within the feedback loop (I've read conflicting accounts), it would also alter the RC network response. As stated above, there is a series resistor between the buffer output and the amplifier input (part of the RC network), but I suspect its effect may be negated by the impedance of the lengthy cables in between.
I guess I'm looking for the most elegant solution since I am trying to avoid compromising my layout, which I am quite pleased with. Any additional advice would be much appreciated.
Sincerely,
Nick
Thanks for the update...
Is this hum a pure deep tone or a rasping harsher buzz ?
Whatever the cause, I don't think it is down to having no proper regulation. The ripple voltage on the rails will be small with just the opamp loading and is constant, not intermitent. And the opamp rejects low frequency noise on the rails by a huge amount anyway.
Only two things come to mid. One, that the problem is physical... something touching something it shouldn't or a break in a lead of connection somewhere. The other is some hazy thoughts on whether it could be a grounding issue but thats hard to see given that it looks a stand alone unit with no mains earth.
You could temporarily add a single larger cap of say 1000uf across the rails (from pos rail to neg rail) and then when the hum occurs flick the power switch off. Tha cap will hold enough charge to run it for a few seconds and you can see if the problem is still there.
Is this hum a pure deep tone or a rasping harsher buzz ?
Whatever the cause, I don't think it is down to having no proper regulation. The ripple voltage on the rails will be small with just the opamp loading and is constant, not intermitent. And the opamp rejects low frequency noise on the rails by a huge amount anyway.
Only two things come to mid. One, that the problem is physical... something touching something it shouldn't or a break in a lead of connection somewhere. The other is some hazy thoughts on whether it could be a grounding issue but thats hard to see given that it looks a stand alone unit with no mains earth.
You could temporarily add a single larger cap of say 1000uf across the rails (from pos rail to neg rail) and then when the hum occurs flick the power switch off. Tha cap will hold enough charge to run it for a few seconds and you can see if the problem is still there.
Thank you for the prompt reply! It is reassuring to hear that regulation isn't likely the cause, because that would probably be the biggest headache to implement. I would say that the noise is predominantly a low fundamental with some harmonic raspiness.
One additional clue I forgot to mention is that the last time it happened I was able to fix it by selecting a different input on the switch box and then re-selecting the PS3.
One additional clue I forgot to mention is that the last time it happened I was able to fix it by selecting a different input on the switch box and then re-selecting the PS3.
Update - the preamp was malfunctioning when I got home from work, so I took the opportunity to run some diagnostics. To clarify, the hum is all fundamental - presumably 120 Hz. I was unable to alleviate it by toggling inputs on the switch box. I opened the case without turning it off or disconnecting any cables and, while I can't account for the cause, I was able to measure about 130 mV one output terminal (the other was less than five mV, with no DC on either input). I was unable to measure any appreciable AC on either output during no-signal conditions, so I assume the DC was upsetting the amplifier's stability and causing hum. Since the two outputs mix to mono downstream, unplugging the offending output eliminated the hum. I played around plugging and unplugging the outputs in different orders and I was actually able to shift the high DC component to the opposite output, quite by surprise. Eventually, I was able to get them to both measure less than five mV, but I don't trust they'll stay that way. I swapped the op-amp in the event it was defective. Fingers crossed!
-Nick
-Nick
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.