Hi Guys, I just finished my power supply for my bridged clones, but the thing is that I have around 45Vdc at the output, what would be best way to lower that to 35Vdc? I was thinking of the LM388K or the use of a high power zener....what you guys think? What would be the best way to go about this. Thank you in advance for your time and concideration.
There is a very simple method of uprating the power of a 1.5 amp regulator like standard 317 or low drop out 1086.
The method is stolen from a common circuit for 12v power supplies. In these supplies they use a 2n3055 NPN bipolar transistor. These are used because they are very tough and reliable.
Unregulated plus in goes to the collector of 2n3055 and also to in pin of reg. Voltage devider between out and ground, centre of which connected to adj. Out connected to base of 3055. Regulated output taken from emitter of 3055.
Voltage devider calculation is: Use 100ohm for upper resistor (R1), that is out to adj. Lower resistor use a 5k preset and adjust for output required.
Output from 3055 becomes output voltage of reg plus 0.6v.
In the actual 12v supply zenners are used and a small signal transistor to drive the 3055. My way is simpler I think, and allows complete adjustment. I also use a zobel on output of reg (2.7r and 0.1uF tant). There are many refinements to the reg, I have just given you the simple way.
You have to stand this circuit below ground for the - rail of your supply, in which the + above refers to ground, and the - to the - supply. The whole thing is floating.
The method is stolen from a common circuit for 12v power supplies. In these supplies they use a 2n3055 NPN bipolar transistor. These are used because they are very tough and reliable.
Unregulated plus in goes to the collector of 2n3055 and also to in pin of reg. Voltage devider between out and ground, centre of which connected to adj. Out connected to base of 3055. Regulated output taken from emitter of 3055.
Voltage devider calculation is: Use 100ohm for upper resistor (R1), that is out to adj. Lower resistor use a 5k preset and adjust for output required.
Output from 3055 becomes output voltage of reg plus 0.6v.
In the actual 12v supply zenners are used and a small signal transistor to drive the 3055. My way is simpler I think, and allows complete adjustment. I also use a zobel on output of reg (2.7r and 0.1uF tant). There are many refinements to the reg, I have just given you the simple way.
You have to stand this circuit below ground for the - rail of your supply, in which the + above refers to ground, and the - to the - supply. The whole thing is floating.
it's just a standard linear regulato with a series pass transistor...I think you can see it in the datasheet of a LM7812...go look for it on National's site...
www.national.com
www.national.com
janneman said:
not at all... explain why everyone insists on 1000uF filter caps. you get large ripple under load. As i see it you can use series impedance or an attenuator. the attenuator could be a better approximation as the fully loaded and unloaded voltages would be closer.
also, i'd use the series impedacne on the primaries, less current, no worries of matching.
in any case with any linear regulator, the voltage drop -- current product will determine heat dissipation. for 10V of drop and 1 amp of current you dissipate 10 watts. for 2 amps you'd drop 20 watts. now if your amp only draws 1amp from the secondaries it's feasible to use a linear regulator. i'd bet from the discription that this is a subwoofer amplifer. if so you will draw more then 1 amp from the secondaries, and things will begin to get hot.
theChris said:
Not " everyone insists on ...". And why do SOME insist on it? Well, this one guy reported that it improved his music, or so he thought. The rest, as they say, is history.
But yes, you can use secondary series resistors, but usually they are used to smooth the cap rippel, kind of an erzats choke. But if you want to use them to lower the secondary voltage I think you very quickly get a very hot resistor. I would, if feasible, wind some windings on the core (well insulated of course) and put that in opposite phase in series with the primary. I have used that often, always with complete succes. You need to experiment a bit to get the right # of turns. Start with 10 and see what you get. Then you know the voltage per turn. The rest is downhill.
Jan Didden
don't use secondary series imedance. i guess you could, but you'll be pulling more current through it.
in any case, if you choose any form of linear regualtion, be it on the primary for lower loss, on the secondary after the diodes for lower ripple, or a transistor/IC, or a zener diode, realize that it'll get hot. if you're trying to get a lot of power from the amp, like 100 watts, you'll be dropping probably 20 volts across the entire secondary. with about 3.5 amps being passed. this means 70 watts lost in regualtion!
realistically this could be less (as music isn't a test tone) or more (as the ICs aren't 100% efficienct). in any case it's a big loss IMO and if your expecting some dinky TO-220 heatsink the size of a coin to cut it, well, nope. and same goes for a small TO-3 case.
for resistance on the secondary, well, this'll have a large loss too, and of the same magnitude.
for resistance on the primary, this loss will be less.
the idea with windings on the transformer seems to have merit as well. likely has the least loss, but also may not be possible.
in any case, if you choose any form of linear regualtion, be it on the primary for lower loss, on the secondary after the diodes for lower ripple, or a transistor/IC, or a zener diode, realize that it'll get hot. if you're trying to get a lot of power from the amp, like 100 watts, you'll be dropping probably 20 volts across the entire secondary. with about 3.5 amps being passed. this means 70 watts lost in regualtion!
realistically this could be less (as music isn't a test tone) or more (as the ICs aren't 100% efficienct). in any case it's a big loss IMO and if your expecting some dinky TO-220 heatsink the size of a coin to cut it, well, nope. and same goes for a small TO-3 case.
for resistance on the secondary, well, this'll have a large loss too, and of the same magnitude.
for resistance on the primary, this loss will be less.
the idea with windings on the transformer seems to have merit as well. likely has the least loss, but also may not be possible.
I personally would do half the voltage... so.. +-22 volts... that'll work, won't it??? parallel the transformer secondaries.... or use one winding per amplifier..?? yes? no??
Like this circuit... it will give the total transformer voltage per rail.. right??
Ignore the 3 transformers in that diagram... and the voltages... just the bridge configuration..
Like this circuit... it will give the total transformer voltage per rail.. right??
Ignore the 3 transformers in that diagram... and the voltages... just the bridge configuration..
Just for the record. I'd go with Paul's regulator idea. Those parts would be less expensive that a resistor sized to dissipate the required wattage. You do need the heat sink of course but you would for the resistor too.
Oh, by the way a resistor on the primary of the transformer would dissipate the same amount of power for a given voltage reduction @ X current of the secondary so no help there.
My question is why would anyone want to degrade the load regulation of the power supply that much? If you ar talking about a class A amp it would be OK because the amp requires nearly constant current, but with a class AB or B amp the power supply voltage would be all over the place with load demand. Not a good idea.
Later BZ
30 years an engineer
Oh, by the way a resistor on the primary of the transformer would dissipate the same amount of power for a given voltage reduction @ X current of the secondary so no help there.
My question is why would anyone want to degrade the load regulation of the power supply that much? If you ar talking about a class A amp it would be OK because the amp requires nearly constant current, but with a class AB or B amp the power supply voltage would be all over the place with load demand. Not a good idea.
Later BZ
30 years an engineer
"Oh, by the way a resistor on the primary of the transformer would dissipate the same amount of power for a given voltage reduction @ X current of the secondary so no help there. "
-nope. look you have a 40V and need 30V, well, for comparison...
ok so thats a 10V drop needed in the secondary. right?
ok so the ratio is 1:3 for the primary to the secondary. so a 3.3V drop on the primary translates to a 10V drop on the secondary. further, there will be less current flowing in the primary. so far there is a smaller voltage drop and less current. less loss, but the supply will rise when there is no music. i'm not overly concerned.
the load does need to be somewhat constant for good ripple rejection. this can be done with the attenuator idea, which will increase loss by an amount. and as i mention, people use 1000uF caps for these amps, i mean supply ripple doens't seem to enter in at all! otherwise you pretty much reduce efficiency by a third
-nope. look you have a 40V and need 30V, well, for comparison...
ok so thats a 10V drop needed in the secondary. right?
ok so the ratio is 1:3 for the primary to the secondary. so a 3.3V drop on the primary translates to a 10V drop on the secondary. further, there will be less current flowing in the primary. so far there is a smaller voltage drop and less current. less loss, but the supply will rise when there is no music. i'm not overly concerned.
the load does need to be somewhat constant for good ripple rejection. this can be done with the attenuator idea, which will increase loss by an amount. and as i mention, people use 1000uF caps for these amps, i mean supply ripple doens't seem to enter in at all! otherwise you pretty much reduce efficiency by a third
theChris said:
I'm afraid that you have it backwards. To drop 10V dc from each rail that is approx. 7-8VAC after rectification losses. to drop 8VAC from each secondary you would need a 24VAC drop in primary voltage. You actually get slightly better efficiency burning the extra power on the secondary as the primary carries slightly more VA (magnetising current etc) for a given secondary VA output.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.