looking for power supply
- Thread starter al_p
- Start date
Help Support GroupDIY Audio Forum:
Al,
It's called ssltech power supply, and it works great.
http://beatbazar.com/guests/ssltech
Good luck!
It's called ssltech power supply, and it works great.
http://beatbazar.com/guests/ssltech
Good luck!
owel
Well-known member
An 18-0-18 AC transformer will give you 27Vdc after rectification.
Owel, I tried that last night, but the 27V is at the very bottom of what a 7824 needs to put out 24V, some datasheets say 28V, and it works but there's a lot of ripple coming through the regulator and the preamp hums like crazy. I've got a 22-0-22 on order, hopefully that will be better.
CJ
Well-known member
I have gotten away with 2 volts over the reg. output. There is a darlington pass transistor, so you need to overcome the dual e-b junctions, which is about 1.2 volts.
Did you amp hum with phantom applied? I had the same problem last saturday, turned out to be insufficient filtering on the phantom supply.
Just curious, does anybody know of a situation where phantom might be applied to just on side of a mic cable, like in an unbalanced mic input but using a condenser mic? If so, much more phantom filtering would be needed as you don't have the cancellation of the phantom ripple.
cj
Did you amp hum with phantom applied? I had the same problem last saturday, turned out to be insufficient filtering on the phantom supply.
Just curious, does anybody know of a situation where phantom might be applied to just on side of a mic cable, like in an unbalanced mic input but using a condenser mic? If so, much more phantom filtering would be needed as you don't have the cancellation of the phantom ripple.
cj
No, the phantom rail had about 8mv of ripple, the 24V rail had almost half a volt. I also tried the preamps with a Power One 24V supply I have lying around and they were clean. I'm guessing the 25-0-25 transformer I have was pushing the upper limit of what the regulators could do, and the 18-0-18 was just at the lower limit. Hopefully the 22-0-22 I have on order will be just right. Does this sound like Goldilocks?
I'm also acquiring a nice collection of toroid power transformers. I hope I can figure out something to do with them.
If I have a 24V single secondary transformer, could I split the output to feed both of the 24V inputs on the board? Will that lower the voltage or load down the transformer incorrectly?
Thanks
CJ
Well-known member
Wow, 1 volt ripple.
What is you capacitance value on the transformer side of the regulator?
After thee rectifier that is...
What is you capacitance value on the transformer side of the regulator?
After thee rectifier that is...
owel
Well-known member
Half a Volt ripple... that's unheard of... maybe it's a defective regulator?
27Volts will still regulate properly to 24Volts. 3 volts is still a big "margin". If the input voltage becomes less than 24Volts, then that's the only time you're going to get less than 24Volts.
For the hum, check your transformer's orientation... sometimes, even if you have a regulated PS, if the transformer is near your audio circuit, it will cause hum (via magnetic field induction). Move your transformer far away. Check that AC wires don't come near low-signal wires, or run parallel with the regulated DC output.
27Volts will still regulate properly to 24Volts. 3 volts is still a big "margin". If the input voltage becomes less than 24Volts, then that's the only time you're going to get less than 24Volts.
For the hum, check your transformer's orientation... sometimes, even if you have a regulated PS, if the transformer is near your audio circuit, it will cause hum (via magnetic field induction). Move your transformer far away. Check that AC wires don't come near low-signal wires, or run parallel with the regulated DC output.
[quote author="owel"]An 18-0-18 AC transformer will give you 27Vdc after rectification.[/quote]
How did you get 27VDC?
18Vrms x 1.414 = 25.5Vpeak, minus two diode drops of .6V each, = 24.3VDC... Too little!
The schematic in the link has a 20-0-20 transformer... that's about 27VDC after rectification/filtering.
Now, the dropout voltage spec for the Fairchild 7824 is 2V. In other words, you don't want to feed the regulator anything less than 26V, which can happen if the AC line voltage drops by like 1%...
The 22-0-22 tranformer will give you about 30V post-rectification. Better.
I suck at math, so excuse me if I ****** up... :?
Peace,
Al.
How did you get 27VDC?
18Vrms x 1.414 = 25.5Vpeak, minus two diode drops of .6V each, = 24.3VDC... Too little!
The schematic in the link has a 20-0-20 transformer... that's about 27VDC after rectification/filtering.
Now, the dropout voltage spec for the Fairchild 7824 is 2V. In other words, you don't want to feed the regulator anything less than 26V, which can happen if the AC line voltage drops by like 1%...
The 22-0-22 tranformer will give you about 30V post-rectification. Better.
I suck at math, so excuse me if I ****** up... :?
Peace,
Al.
owel
Well-known member
>How did you get 27VDC?
I measured it with a tester!
My 18-0-18 rated transformer probably gives more than 18-0-18 VAC.
I measured it with a tester!
My 18-0-18 rated transformer probably gives more than 18-0-18 VAC.
Al, I think you're describing my situation. The power supply is in a different box from the preamps, so the hum is not a proximity issue. I also looked around the internet and found people saying the 7824 needed to see 27 or 28 volts to not create hum. I'm not actually sure what the mechanism is - is the regulator passing ripple or creating hum itself?
owel
Well-known member
Here's a thought... what if you wire the 24-0-24 secondaries in series, that gives you 48VAC, rectify and regulate for 48VDC phantom power using an LM317HVT (take note, this is the high voltage version of the 317).
Then get that regulated 48VDC phantom power and tap into it and regulate to 24VDC with a 7824 (or another LM317T).
Then get that regulated 48VDC phantom power and tap into it and regulate to 24VDC with a 7824 (or another LM317T).
You might get 27VDC out of a lightly-loaded 18VAC transformer. Won't take a lot of load to sag it.
Under some load, you might have 1V of average ripple. But what really matters is the bottom of the ripple. Common DC meters read the average DC, the ripple zig-zags both sides of that. Not sure how you are measuring your ripple, or how sharp your zig-zag is... this all depends on many things. But as a wild guess, you could have 3V peak-peak ripple. Or in other words +/-1.5V around 27V. And 27VDC+/-1.5VDC gives 25.5V at the bottom of the ripple. That is just too low for most 24V regulators.
Anyway, it is crazy to try to shave things too fine. You use a regulator because the raw supply varies. One day the wall outlet is 122VAC, another day it may be 112VAC. Or the bass-amp hits a power chord and your house line sags to 105VAC for an instant. Or you get a different transformer where the 18VAC isn't really 18VAC. Or one day you have one 1mA mike, and another day you have a dozen 8mA mikes.
The idea of the regulator is you give it too much, and it throws away the excess. (as heat... that metal tab is on there for a reason. Bolt some metal to it, and don't let it run hot enough to burn your finger.)
Don't be cheap!!! Another few volts of transformer is not expensive. Being a tenth-volt short sorta ruins the whole point of a regulator.
There is a simple rule of thumb:
Volts AC equals Regulated Volts DC.
Let's see how that works.
We want regulated 24VDC. Per the above, we buy a 24VAC transformer. If it is small, then at rated line input and at no-load, it might make 28VAC. But we are going to load it, so we don't care (yet) about no-load voltage.
Rated voltage ========= 24VAC
Low line voltage -10% ==== 21.6VAC
Times 1.414 ========== 30.5V peak
Minus 2 rectifiers -1.5V === 29VDC
3V peak-peak ripple ======== 27.5V at bottom of ripple
We see that VAC=VDC is barely enough on a bad day. Sure, it will be plenty most days, but you have to design pesimistically. If you don't, Murphy will bite you, usually in the middle of a great gig.
I said we don't care about no-load, but you have to check the regulator's maximum input for another kind of bad day. Small transformers will rise 20% above rating at no-load. So a "24VAC" winding is really 28.8V at no-load. Then say half of Florida goes off-line again, and the rest of the US grid bounces up 5%. You now have 30.24V. 30.24V*1.414= 42.7V peak. At light load, the rectifier drop may be very small. Clearly you can have over 40VDC into the regulator, which isn't rated for it.
So to get 24VDC with standard regulators, you "have" to use 24VAC windings, don't load them so hard they sag all the way to 24VAC, don't undersize the capacitor, and it would be wise to have some permanent load (like an LED) on the raw DC to keep it from floating too high if the load becomes disconnected.
You can go through all those calculations every time. But some factors are guesses. And you are limited by available transformers. So for regulated DC over about 10VDC (up to hundreds of volts), it is usually expedient to just buy VAC=VDC.
Under 10V, rectifier and regulator fixed-voltage losses bite you badly enough to hurt, and you have to buy more VAC.
Under some load, you might have 1V of average ripple. But what really matters is the bottom of the ripple. Common DC meters read the average DC, the ripple zig-zags both sides of that. Not sure how you are measuring your ripple, or how sharp your zig-zag is... this all depends on many things. But as a wild guess, you could have 3V peak-peak ripple. Or in other words +/-1.5V around 27V. And 27VDC+/-1.5VDC gives 25.5V at the bottom of the ripple. That is just too low for most 24V regulators.
Anyway, it is crazy to try to shave things too fine. You use a regulator because the raw supply varies. One day the wall outlet is 122VAC, another day it may be 112VAC. Or the bass-amp hits a power chord and your house line sags to 105VAC for an instant. Or you get a different transformer where the 18VAC isn't really 18VAC. Or one day you have one 1mA mike, and another day you have a dozen 8mA mikes.
The idea of the regulator is you give it too much, and it throws away the excess. (as heat... that metal tab is on there for a reason. Bolt some metal to it, and don't let it run hot enough to burn your finger.)
Don't be cheap!!! Another few volts of transformer is not expensive. Being a tenth-volt short sorta ruins the whole point of a regulator.
There is a simple rule of thumb:
Volts AC equals Regulated Volts DC.
Let's see how that works.
We want regulated 24VDC. Per the above, we buy a 24VAC transformer. If it is small, then at rated line input and at no-load, it might make 28VAC. But we are going to load it, so we don't care (yet) about no-load voltage.
Rated voltage ========= 24VAC
Low line voltage -10% ==== 21.6VAC
Times 1.414 ========== 30.5V peak
Minus 2 rectifiers -1.5V === 29VDC
3V peak-peak ripple ======== 27.5V at bottom of ripple
We see that VAC=VDC is barely enough on a bad day. Sure, it will be plenty most days, but you have to design pesimistically. If you don't, Murphy will bite you, usually in the middle of a great gig.
I said we don't care about no-load, but you have to check the regulator's maximum input for another kind of bad day. Small transformers will rise 20% above rating at no-load. So a "24VAC" winding is really 28.8V at no-load. Then say half of Florida goes off-line again, and the rest of the US grid bounces up 5%. You now have 30.24V. 30.24V*1.414= 42.7V peak. At light load, the rectifier drop may be very small. Clearly you can have over 40VDC into the regulator, which isn't rated for it.
So to get 24VDC with standard regulators, you "have" to use 24VAC windings, don't load them so hard they sag all the way to 24VAC, don't undersize the capacitor, and it would be wise to have some permanent load (like an LED) on the raw DC to keep it from floating too high if the load becomes disconnected.
You can go through all those calculations every time. But some factors are guesses. And you are limited by available transformers. So for regulated DC over about 10VDC (up to hundreds of volts), it is usually expedient to just buy VAC=VDC.
Under 10V, rectifier and regulator fixed-voltage losses bite you badly enough to hurt, and you have to buy more VAC.
JLM Audio
Well-known member
PRR has it exactly right. Also remember the regulators have thermal protection and without a heat sink can only handle about 100 to 150mA Max depending on the voltage dropped across them. I have seen noise get through a regulator where the top of the ripple just pushes the wattage dropped across a regulator too high at its peaks. This starts the regulator into shut down just enough to let the 50 or 60Hz to the output. Basically bolt the regulators (with insulator kits) to the side of the case or a heat sink so they can handle most of their 1A (78 or 79series) or 1.5Amp (LM317, LM337 series) rating. Most regulators run better when they are not running to their thermal limits.
JLM AC/DC power supply product plug below.
Adjustable +/-2 to +/-37volts and 48v
http://www.jlmaudio.com/JLM%20Power%20Supply.htm
Joe
http://www.jlmaudio.com
JLM AC/DC power supply product plug below.
Adjustable +/-2 to +/-37volts and 48v
http://www.jlmaudio.com/JLM%20Power%20Supply.htm
Joe
http://www.jlmaudio.com
