| cowanrg |
i think this has been covered before, but i couldnt find it.
basically, when i turn off or unplug my bosoz, it keeps playing for up to 2 minutes or so. its slightly bothersome because it's like the power button is broken. is there a way to simply make it turn off when you cut power?
ive heard of people putting something across the + and - rails that bleeds them when it gets turned off or something? i guess i could also do some sort of relay at the output or whatever, but id rather not put anything in the signal path. |
|
|
| Blues |
| You can press stop on your CD player first :) |
|
|
| apassgear |
Add smaller bleader resistors over your PSU caps.
In any PSU you should add bleader resistors, so if you don't have them better add them. |
|
|
| cowanrg |
Blues,
thanks, but thats not how i make stuff :)
apassgear,
my PSU doesnt have bleeder resistors. its a by-the-book design of the original power supply. here is schematic and board:
http://web.vip.hr/pcb-design.vip/bls-pwr.html
ive never done a bleeder resistor before, but doesnt it go in between the + and - of the PSU caps? and if so, what value is good, and does it need to be higher powered? i have some 0.47ohm and 1ohm 5-watters laying around, would those work? |
|
|
| Blues |
| powered down, your circuit should be enough to bleed your ps caps if you want to work inside your chassis...as I've said in other threads I always keep my ps caps fully charged by cutting supply at the rails and just running my ps with LED as the standby load. Cutting off the rails at the circuit board will cut the music too if that's what you want...you're wasting energy in charging/discharging caps everytime you flip the power switch. |
|
|
| MikeW |
Get a power switch with three poles double throw. Wire the first for power and the other 2 opposite for the outputs.
I don't know if you want the AC that close to the outputs. :bigeyes: |
|
|
| cowanrg |
Blues,
the circuit bleeds down slowly unfortunately. it gradually gets quieter and quieter until it just fades out. its maybe a minute or so realistically.
i understand what you mean about having the power switch control the rails and wasting energy charging and discharging, but that isnt really an option right now.
plus, i would rather NOT have the PSU boards running 24/7. the preamp wont get turned on and off that often. i just want it to stop playing when its turned off.
MikeW,
getting a new power switch is NOT an option. that power switch thing took me 2 months to design and make. its not getting changed. it has more poles on the bottom that i could use, but i would rather not run AC and signal through the same switch.
and as far as the AC being close to the inputs, that was just a design flaw. it doesnt seem to hurt anything, i cant hear any noise on the output, it seems very quiet. i will either move that connecting board or build a steel shield around the AC inlet.
SO, back to the original problem at hand... there should be a very simple way to bleed off the PSU quickly when there isnt any power. what exactly will a bleed resistor accomplish? i would rather not rewire the whole damn thing with new switches, relays, etc...
can a bleeder resistor accomplish what i want, and if so, what value should i use? |
|
|
| GRollins |
I'm not a big fan of bleeder resistors for any number of reasons. The usual argument for bleeder resistors is safety. This will probably start another firestorm, but voltage is kinda like the "threat" that Iraq posed to the US pre-invasion...overblown. People are absurdly, irrationally scared of voltage. Below, say, 75 to 100VDC it doesn't even produce any sensation as long as you have dry hands. Does some current flow? Of course. But it's not a big deal. Above 100VDC (or 50VAC) I start taking things more seriously. Wet hands? I don't touch anything more than a 9V battery. (Sweaty car repair guys can get a tingle from a car battery, i.e. around 13V. Sweaty telephone repairmen cuss like the dickens when they touch 48V, but note that they're still alive, just tingly and mad.)
Okay, take the safety reasoning any way you like, it's no skin off my nose.
Onwards and upwards.
What? The preamp plays after you drop power and this is a problem? Gracious me, I'm offended by a piece of equipment that doesn't play after I turn it off. Tells me the power supply isn't big enough.
Okay, okay...if you want it to turn off, you want it to turn off.
(mumble, mumble...mutter, mutter...grumble...)
Possibilities:
--Use a relay at the signal output. This will have to have its own (pretty small) power supply. Less than a hundred uF, since you want to deplete the charge--via the relay coil--fairly quickly. The "it's in the signal path" objection doesn't have to be a problem. Don't switch the signal, switch ground. The signal goes straight through at all times--unswitched. When you want the output to mute, engage the relay to bring ground to the signal, not the other way around.
--Use a "bleeder" resistor to do something useful, like run the LED. Not enough current draw? Drop even more by having lotsa lights. Make it look like some of the old '70s gear just after LEDs came out and people couldn't get enough of them. Use small lightbulbs instead of LEDs, they use a lot more current.
--The brute force method--use a relay, not to ground the output, but to drop the rails. Use a 10 ohm 5, 10, or 20W resistor so as not to have a dead short to ground. When the power goes off, the relay connects the rail via the resistor to ground. A 10 ohm resistor will make short work of all but the most robust power supplies. Note that the wattage of the resistor is not really very critical. Sure you're starting out at 50V or whatever, but a tenth of a second later you're only dropping 20V, and a tenth after that it's only 10V, and a tenth after that...you get the point. You can put a surprising amount of power through a resistor as a decaying pulse and it'll barely get warm to the touch. Just don't sit there and turn the poor thing off and on a thousand times in a row.
--Use the psychological approach. Use a DPDT switch. Use one pole to turn off the power. Use the other to switch off the LED instantly. If the LED doesn't glow, you know it's off even if the music is still playing.
--Take out some of the capacitance in your power supply. Don't blame me if it doesn't sound as good afterwards.
(...grumble, grumble...who'd want a power supply that ran out instantly, anyway?....mutter, snarl...harrumph!...confounded stereo doesn't even sound decent until it'll play for half an hour after you kill the power... assorted curmudgeonly noises...grumble...)
Grey |
|
|
| Blues |
| R=t/C; where: R=resistance in ohms, t=discharge time in seconds, C=capacitance in farad...increase resistor wattage if it smokes:) |
|
|
| cowanrg |
well, thanks GRollins.
i know not everyone on this forum follows my projects, so i remind everyone that this preamp isnt for me, its for my dad. and, i am a perfectionist. i would lose sleep (literally, i wouldnt be able to sleep) knowing that i spend a year building something and the off switch doesnt actually turn it off. the thing has to be perfect. the input switch switches inputs, the volume controls volume, and off should mean off.
oh, and none of this is for safety at all. i have exposed transistor casings, live heatsinks, etc... like all my projects, it has a case :) the case isnt live, so all is good.
out of the possible options, i of course only like a few. i wont be adding any more lights or LED's. plus, i like how they are wired right now. they turn on and off just how i want them too. plus, there are only two at a time, so that wouldnt be enough current draw to kill the PSU's.
i really dont want to use a relay at the output. the amp is silent when kicking on and off (well, its not exactly silent when going off, but from dead off to on there are no pops) and switching inputs whilst music is playing is dead silent too. i would really hate to add relays that might potentially introduce popping and clicking. once again, im really anal.
the psychological approach is a good one, but i see it as a band-aid, and im not for cutting corners. once again, i would lose sleep at night knowing that the led was really just for show and didnt indicate power :) its just not a good design.
now you say you dont like bleeders for reasons of safety? well, in your "brute force" approach, couldnt i omit the relay all togehter and just use a bleeder at the caps (there are only 4 relatively small ones) to drain them? no relays needed. i have a little DC motor that i use as a drain for PSU's when im playing with them. and this motor takes about 1/10 of a second to drain the rails completely. just a tiny tap with the wires and they are down to 1-3v. so, wouldnt a resistor accomplish this?
OR, do i have a problem with my preamp? shouldnt it drain quicker? hmm..... |
|
|
| cowanrg |
| quote: | Originally posted by Blues
R=t/C; where: R=resistance in ohms, t=discharge time in seconds, C=capacitance in farad...increase resistor wattage if it smokes:) |
oh ok. hum. so, if i have a 10,000uf cap, and i want to discharge it in 1 second, i need a 100ohm resistor. i wonder if that will smoke...
and i just connect it across the + and - terminals on the cap right? |
|
|
| Blues |
| W=V^2/R.....W=watt, V=volt, R=ohm |
|
|
| Blues |
| quote: | Originally posted by cowanrg
and i just connect it across the + and - terminals on the cap right? |
Yes. |
|
|
| cowanrg |
hmmmm.
by that formula i would need a 65-watt resistor... its 80v before regulation. so, 6400/100.
sigh. i was hoping to be done with this by this weekend. i guess not. |
|
|
| jh6you |
I would lower volume to 0, then, turn off the power . . . It could take total time much less than 2 min . . .
Regards
jH |
|
|
| Blues |
| try 5W and see if it can take the heat:hot: |
|
|
| Blues |
| hitting Source stop button is still the best option...simple and quick. |
|
|
| GRollins |
A bleeder resistor robs current that could otherwise be used by the circuit. Unless you choose to switch it in and out with a relay or an unused pole on the power switch, it's going to be there all the time; a nasty little leech sucking power out of your circuit.
Keep in mind that the circuits you're comparing this to have pathetic, puny power supplies that collapse quickly because they were anemic to begin with. You're trying to provide your father with a perfect circuit...by making it less perfect. If you're going to saddle the poor power supply with a bleeder, a far more elegant solution would be to simply remove part of the capacitance that's in there. The power supply will collapse more quickly and you can put the caps on the shelf to use in another project later on. I'm not sure I see the logic in putting together a beefy, high end power supply, only to hobble it down to the level of a mid-fi unit. Most, if not all, high end circuits run on for varying periods of time after they're turned off. The ones that don't invariably use relays to drop the outputs. A good power supply is the mark of a good piece of gear.
A weak argument can be made that a bleeder resistor serves as a crude regulator of sorts, but there are far better ways to accomplish the same thing. Using one just to drop a power supply after the circuit it turned off is...oh well, it's your/your father's circuit. Do as you wish.
Sigh.
I'll head on to bed and leave you folks alone...g'night.
Grey |
|
|
| lgreen |
I like the LED idea. My stock BOSOZ, which actually uses double the capacitance (oops, guess its not stock after all), has a nice quick fade when turned off. Image of BOSOZ (More).
Heck you could put a few on the inside or shine them down for a bottom glow, or....
Hey, nice gears! |
|
|
| cowanrg |
GRollins,
you just convinced me NOT to use bleeder resistors.
after thinking about it, i will do a small xformer with 12v secondaries that will power a couple relays. they will be connected in between the PSU and the preamp. when power is applied, the relays will latch and put power to the preamp. when power is turned off, they will unlatch and disconnect the preamp from power. BUT, the BOSOZ does have 1000uf caps at its power inputs, so hopefully that wont operate it for long. ill just have to see.
Blues,
when i build something, i dont like compromises. i want a complete anyone to be able to use it. i dont want someone to say "how come when i turn it off, it doesnt stop playing music". in my mind, thats just a poor design.
to all:
here is something that is bothering me though... is my preamp working adequately. im re-reading the spec sheets from passdiy.com and it appears i should be getting a 10W draw from each channel. so, 10W being delivered by 4000uf of capacitance (two 1000uf caps on PSU and two 1000uf caps on the preamp itself) . seems like its not drawing nearly enough current... i guess ill have to measure that myself. from the specs it looks like it draws 80mA for each FET, so 0.16A for each channel. wow, i guess an 800va transformer really is WAY overkill :) i didnt realize that. |
|
|
| Blues |
As a matter of fact, a PS that can keep music playing for 30 seconds or more is not a poor design. It's more than solid. The caps in both your pre and power amp is what keeps the music playing at power off and not just your pre-amp's caps. BTW, caps are better off fully charged all the time.
I suggest you stop worrying about it and just keep in mind to power off your Source first. |
|
|
| cowanrg |
well, looking at the formulas, it doesnt look like im drawing the full power of the preamp.
i still need to get in and measure it to make sure, but here is the math...
given the previous two formulas, we get:
W = v^2/(t/C)
we know W, V, and C, we are trying to find the time it takes to discharge the caps, assuming all 2 1000uf caps, two per rail.
so, 10 = 3600/(t/0.002)
10 = 7.2t
t = 1.39 seconds.
so, sound lasting more than about 2 seconds doesnt seem like it should be happening... maybe i need to check the amp first before i start doing anything with relays :)
Blues,
it aint happening man. i WILL fix it and it will be how i want it. its just me. i dont half-*** anything. you may not see it as such, but i do. i dont want my dad to have to worry about turning off pressing stop, swithing inputs, turning down the volume first, or turning off the amp, etc... i want it to work exactly how it should. there is a reason i put an on/off switch. if i WANTED it to work like this, i wouldnt have bothered for 2 months making an on/off switch that i liked.
when i build something, i dont build it to be just ok, or average. i try to build it not only the best that i can, but i try to keep the standard that it should be one of the best of the bunch. and i feel that when this is done, it will be right up there with the best BOSOZ's ever made.
its the reason i dont finish my projects very quickly and often run into many problems. but, when im done, i have no regrets and am always happy with the turnout. |
|
|
| jh6you |
800VA Transformer for BOSOZ . . . ??!!??!!??!!
install 8000CC engine on the bike . . . :eek: . . .
Regards
jH |
|
|
| jleaman |
cowanrg,
Got any pictures ? I admire your work and would love to see what ya got so far..
j' |
|
|
| cowanrg |
| quote: | Originally posted by jh6you
800VA Transformer for BOSOZ . . . ??!!??!!??!!
install 8000CC engine on the bike . . . :eek: . . .
Regards
jH |
when the project is done, i will have a FULL write-up on my site including all the reasons why i decided on an 800va xformer :) i get this a LOT. |
|
|
| Coulomb |
Hey Cow man, not to milk this discussion any more as it seems to have an unusually long life as it is, but here is my two cents.
Every power supply I ever built and continue to build is either equipped with bleeder resistors or load resistor on a relay. For dual rail applications always use two bleeders one from each rail to common ground. for 80V rails 2.2K to 3.9K @ 5W should be sufficient to drain the caps below thier ability to deliver power in a few seconds. Unloaded it can take several minutes to drain the caps, but with a load in parallel with the bleeders the caps lose thier ability to present sufficient current to the load in a hurry.
The other process I use is a DC relay holding the rails closed while power is applied, as soon as the AC is disengaged the relay shunts the load to a 10R 25Watt metal housed resistor clamped to the casing. I use this on my really big Amps.
My two cents
Regards
Anthony |
|
|
| cowanrg |
| quote: | Originally posted by Coulomb
Hey Cow man, not to milk this discussion any more as it seems to have an unusually long life as it is, but here is my two cents.
Every power supply I ever built and continue to build is either equipped with bleeder resistors or load resistor on a relay. For dual rail applications always use two bleeders one from each rail to common ground. for 80V rails 2.2K to 3.9K @ 5W should be sufficient to drain the caps below thier ability to deliver power in a few seconds. Unloaded it can take several minutes to drain the caps, but with a load in parallel with the bleeders the caps lose thier ability to present sufficient current to the load in a hurry.
The other process I use is a DC relay holding the rails closed while power is applied, as soon as the AC is disengaged the relay shunts the load to a 10R 25Watt metal housed resistor clamped to the casing. I use this on my really big Amps.
My two cents
Regards
Anthony |
thanks for the advice and the pun.
yeah, the thread has gone unusually long, but thats because i havent found a satisfactory answer yet :) i think the relays will be the ticket, but that requires another xformerer (dont have much room left) and some more work to make it look nice.
but im still concerned if the amp is drawing what it should. the PSU seems awfully small to sustain the preamp for a minute. a couple seconds i might be able to handle.
here is an idea... would it be possible to just boost the bias up a bit? i know on an amp that could be possible. i have MORE than ample heatsinking. so, if i could boost the bias up to where the PSU would only last a second or so, that could even better the sound? hmmmm. |
|
|
| Coulomb |
BTW, the spring on the idler arm is to keep tension off of a power switch at the back there? Can't make out the switch from the Photo.
Oh, and I am terribly dissapointed to see you cheaped out on a plain Bridge rectifier instead of a more exotic expatial Diode Rectifier with snubbers.
Regards
Anthony |
|
|
| Coulomb |
Well if you used an AC Coil relay you would not need another transformer, besides I doubt you could fit another 800VA transformer in that case for a low power relay!!
Anthony |
|
|
| cowanrg |
| quote: | Originally posted by Coulomb
BTW, the spring on the idler arm is to keep tension off of a power switch at the back there? Can't make out the switch from the Photo.
Oh, and I am terribly dissapointed to see you cheaped out on a plain Bridge rectifier instead of a more exotic expatial Diode Rectifier with snubbers.
Regards
Anthony |
well, the spring is to keep the rod pushed away from the power switch, yes. the power switch itself has a spring, but not enough to sustain the whole assembly.
sorry about the bridge rectifier. it was the original design. as you notice, i always use off-the-shelf designs and PCB's. at least i upgraded the 1N4005's to MUR1520's! im not even sure what a expatial diode rectifier with snubbers is. i think i know that snubbers are small value caps across the diodes? i just used the basic design with decent parts is all. sorry! i did use south-american bubinga wood though, not cherry... |
|
|
| Coulomb |
| quote: | Originally posted by cowanrg
well, the spring is to keep the rod pushed away from the power switch, yes. the power switch itself has a spring, but not enough to sustain the whole assembly.
sorry about the bridge rectifier. it was the original design. as you notice, i always use off-the-shelf designs and PCB's. at least i upgraded the 1N4005's to MUR1520's! im not even sure what a expatial diode rectifier with snubbers is. i think i know that snubbers are small value caps across the diodes? i just used the basic design with decent parts is all. sorry! i did use south-american bubinga wood though, not cherry... |
HEXFRED's are expatial Diodes, not to be confused with MUR devices which are simply fast or ultra fast devices.
Regards
Anthony |
|
|
| Coulomb |
I like the front panel, you must have mounted the LEDs and trimmed the ends off, sanded them over to blend into the wood and then varnished over them. Looks good
Anthony |
|
|
| cowanrg |
| quote: | Originally posted by Coulomb
HEXFRED's are expatial Diodes, not to be confused with MUR devices which are simply fast or ultra fast devices.
Regards
Anthony |
ah ok. i learned something then. oh well, there is always next time :rolleyes:
oh, duh. forgot about an AC coil relay. so something like this:
http://www.allelectronics.com/cgi-b...120L&type=store
connected to the + and - rails of each PSU and optionally connected to a bleeder resistor of some sort. i guess i would still have to worry about the caps on the main board. so, i should make the relay not necessarily disconnect the power, but connect a bleeder resistor when its off, thus discharging all the power. |
|
|
| cowanrg |
| quote: | Originally posted by Coulomb
I like the front panel, you must have mounted the LEDs and trimmed the ends off, sanded them over to blend into the wood and then varnished over them. Looks good
Anthony |
nope. simpler than that.
sandwiched the wood between two other pieces of wood (for a perfect hole) and drilled a "just less than" 1/8" hole. then, took 1/8" plexiglass rod and carefully hammered it into the holes, then sanded flush. its perfectly smooth, even before it was lacquered. then, there are pockets in the back that are milled out to house the LED's. since its just a see-through rod, i can acutally position the LED's so they dont directly shine though, but rather reflect through. so, you get a bright LED, but its only really visible looking straight at it. i like this better because at night i dont want my LED's to light up the room. and if you stare straight at those blue buggers it can be BRIGHT. these are very subtle. but as you see from the pics, you can still see them even with flash from a camera. |
|
|
| Coulomb |
| quote: | Originally posted by cowanrg
ah ok. i learned something then. oh well, there is always next time :rolleyes:
oh, duh. forgot about an AC coil relay. so something like this:
http://www.allelectronics.com/cgi-b...120L&type=store
connected to the + and - rails of each PSU and optionally connected to a bleeder resistor of some sort. i guess i would still have to worry about the caps on the main board. so, i should make the relay not necessarily disconnect the power, but connect a bleeder resistor when its off, thus discharging all the power. |
Yes quite, the rails or output from the power supplies would be the common contacts and at rest would connect to the shunt resistor through the (N/C at rest) contacts. The (N/O at rest) contacts is where the preamp power would be supplied from when the coil is enegizedl
Edit:
Add a 10,000pf cap across the AC contacts to the coil for noise suppression. |
|
|
| cowanrg |
ok, im satisfied with that.
i can mount it on the front wall of the metal plate the isolates the PSU from the preamp. so around a 3k, 5w resistor? ill give that a go tomorrow, but first i will measure the current draw on the amp to make sure all is going well. i need to wire up my volume control too. i finally mounted the front panel (all those front panel pics were just hot-glued togehter).
if i want it to "power down" quicker, can i just get a 10W resistor of a lower value?
well, time for bed now. i probably wont get back to this until the weekend because ill be working hard until then. unfortunately i didnt make near the progress i wanted to today.
thanks everyone. if it wasnt for you guys, i wouldnt know what im doing, and i wouldnt care about it because i wouldnt have anyone to say "nice work". i really do appreciate it. one of these days i will get back to my idea about a design contest. but i need to work on my entries first :) |
|
|
| Coulomb |
As the bleeder is not in the circuit generating heat all the time I would suggest somthing like 25R to 100R at about 20Watts.
Ceramic epoxied to the frame or an aluminum can type.
Nite' |
|
|
| GRollins |
I'm still unclear on how you arrived at the conclusion that running after the switch is cut is somehow bad...but this is DIY and we're all allowed our eccentricities. Still, I think you might want to review the circuits of existing designs that you're comparing this to. I think that you'll find that what I said earlier is true: High end designs run after cutoff or use relays to drop the signal (meaning they're still running--you just can't hear it). Ah, well.
You're on a "bandersnatch hunt," also known as "looking for the nonesuch." That's your right. Just try not to be dissappointed when you discover that there's no magic way to accomplish what you're trying to do. A number of things have been proposed in this thread, but they're all just bandaids. I still recommend doing what the high end designers do.
Technically, 800VA is way overkill, but I happen to like using oversized transformers. Secondary impedance is lower, which means that the circuit doesn't have to snarf so hard on the rail to get current.
...However...
Note that the transformer has absolutely nothing to do with the amount of time the circuit runs after it's turned off.
Boosting the bias on the circuit suits me just peachy. Keep in mind two things:
--voltage and current ratings for the devices
--rated heat dissipation
Don't be surprised if increasing the bias still leaves the music running longer than you might like.
If I can come up with another suggestion, I may drop back in on this thread, but right now I'm all out of bandersnatches and my shipment of nonesuches has been delayed once more.
Good luck.
Remember to have fun.
Grey |
|
|
| Coulomb |
| quote: | Originally posted by GRollins
I'm still unclear on how you arrived at the conclusion that running after the switch is cut is somehow bad...but this is DIY and we're all allowed our eccentricities. Still, I think you might want to review the circuits of existing designs that you're comparing this to. I think that you'll find that what I said earlier is true: High end designs run after cutoff or use relays to drop the signal (meaning they're still running--you just can't hear it). Ah, well.
You're on a "bandersnatch hunt," also known as "looking for the nonesuch." That's your right. Just try not to be dissappointed when you discover that there's no magic way to accomplish what you're trying to do. A number of things have been proposed in this thread, but they're all just bandaids. I still recommend doing what the high end designers do.
Technically, 800VA is way overkill, but I happen to like using oversized transformers. Secondary impedance is lower, which means that the circuit doesn't have to snarf so hard on the rail to get current.
...However...
Note that the transformer has absolutely nothing to do with the amount of time the circuit runs after it's turned off.
Boosting the bias on the circuit suits me just peachy. Keep in mind two things:
--voltage and current ratings for the devices
--rated heat dissipation
Don't be surprised if increasing the bias still leaves the music running longer than you might like.
If I can come up with another suggestion, I may drop back in on this thread, but right now I'm all out of bandersnatches and my shipment of nonesuches has been delayed once more.
Good luck.
Remember to have fun.
Grey |
Hey Grey, Tube Preamp still on the burner?
Regards
Anthony |
|
|
| steenoe |
Well, let me suggest a bleeder resistor of say, 2k's, it doesnt even have to be high wattage! 1 watt is fine. That would take care of your problem! I dont think your relative will ever notice its precense ;) One from each rail, to GND.
Steen. |
|
|
| cowanrg |
steenoe,
ill give it a try.
GRollins,
i understand what you are saying. but to me, its just not ALL about the sound. its about the functionality and aesthetics too. if the on/off switch doesnt do what you would traditionally think it should, then its a failed design in my opinion.
the relays/bleeder circuit is my other option, but as anal as i am about all this stuff, i would have to find the perfect place for it, as well as redesigning some wiring layout, etc... so, for me, its a bigger undertaking. but, it could solve the issue all together.
however, i have some ~2k resistors at home, ill give it a try. even a second or two of music is fine. but it plays for at the least, a minute or more. it just feels like its broken as long as it keeps playing. |
|
|
|
