Indeed wg_ski, the regulator transistors would also absorb part of the heat dissipation. So given 50v supply, it looks like 35v would be a reasonable mid point to spread the heat. While i'm at it might as well regulate eaxh channel separately. Looks like I have a use for 12 transistors after all.
Reading the thread through, it seems in class A, the heatsinks are a limiting factor. It was suggested 30v rails could heat them nicely, with the power that transformer could supply. With such ideas repeated by different users, can this amp afford to drop a third of it's power into the heatsink through regulation. My immediate thought is if chokes would present cooling problems. Could there be a way to combine the two methods, where the chokes do most of the work. After which, could come regulators. Although I'm feeling regulators are a part I don't usually see.
Or is this amp to be viewed as the regs on the lesser heatsink, and amps on the better. That was about the spread in the original design. If amp count on each is considered.
Or is this amp to be viewed as the regs on the lesser heatsink, and amps on the better. That was about the spread in the original design. If amp count on each is considered.
My experience with faulty av amps is that the power transformers often have a 2nd set of taps on the main winding at half the rail voltage.
Eg the onkyo nx 609 had 42-21-0-21-42.
Eg the onkyo nx 609 had 42-21-0-21-42.
Read this thread https://www.diyaudio.com/community/...ut-adjustable-stabilized-power-supply.225737/ about a good voltage regulator. I have an unused pcb board if you're interested.
Cheers
Throw the guts out, keep the chassis and power supply (consider the choke input power supply mod suggested by others) and decide which DIY/Audio store Uncle Nelson amp will work in there.
My 2 cents, ymmv, dyor etc etc
Regulators after chokes is a good idea. Choke supplies are partially regulated, but not fully. The problem with linear regulators is dissipation/heat, which using choke input supply will solve. With choke supplies, you don’t even need that “10,000 uF per 4 ohm load” rule of thumb. But do use caps rated for high ripple current. What’s in there may be at risk of early failure used in a high average demand setting.
Yes, they have a critical current to maintain 180 degree conduction angle and “become” regulated. Below that, however, they are unregulated supplies — with poorer regulation than capacitor input. Class A operation has the advantage of providing a high minimum current draw. If you draw less, the voltage rises up to the full peak voltage (at no load, it’s pretty linear in between). The linear regulator can chop this off. Dissipation is still reduced substantially - the VI product on the regulator won’t ever get crazy high, averaging about 1/3 of what it would be with a cap input filter. The smaller heat sink would have no trouble keeping it cool.
With this technique, one could make a push pull design that stays in class A for 8 ohm operation and moves into AB when higher currents are demanded. Doubling power at 4 ohms, and doubling again at 2. Use 4 output pairs per channel, and find a nice 5-ppm-capable front end to put on it….. Presto - instant Mini-Krell.
One wouldn’t need 2SC5200’s for the regulator pass elements, but they wouldn’t hurt either. I’d use it as an opportunity to use up old low voltage high current TO-3’s I have lying around.
Yes, they have a critical current to maintain 180 degree conduction angle and “become” regulated. Below that, however, they are unregulated supplies — with poorer regulation than capacitor input. Class A operation has the advantage of providing a high minimum current draw. If you draw less, the voltage rises up to the full peak voltage (at no load, it’s pretty linear in between). The linear regulator can chop this off. Dissipation is still reduced substantially - the VI product on the regulator won’t ever get crazy high, averaging about 1/3 of what it would be with a cap input filter. The smaller heat sink would have no trouble keeping it cool.
With this technique, one could make a push pull design that stays in class A for 8 ohm operation and moves into AB when higher currents are demanded. Doubling power at 4 ohms, and doubling again at 2. Use 4 output pairs per channel, and find a nice 5-ppm-capable front end to put on it….. Presto - instant Mini-Krell.
One wouldn’t need 2SC5200’s for the regulator pass elements, but they wouldn’t hurt either. I’d use it as an opportunity to use up old low voltage high current TO-3’s I have lying around.
If you want to use an output capacitor, maybe you could use the transformer's centre tap as positive (or negative)?
Given I do not want an output cap, and am stuck with the +-50V-55V supply, there appears to be a concensus:
-Drop the voltage (+- 25%) with chokes
-Each channel has its regulators to drop voltage (and eliminate ripple) a bit (+-10%), on the smaller heatsink. Using power transistors (4 per channel) from 2 decommissioned channels.
-Use 8 transistors per channel for the amps.
-Trial and error adjustments to the regulator voltages trying to spread the power fairly between them monitoring heatsink temperatures..
-Drop the voltage (+- 25%) with chokes
-Each channel has its regulators to drop voltage (and eliminate ripple) a bit (+-10%), on the smaller heatsink. Using power transistors (4 per channel) from 2 decommissioned channels.
-Use 8 transistors per channel for the amps.
-Trial and error adjustments to the regulator voltages trying to spread the power fairly between them monitoring heatsink temperatures..
Normally secondary is over the primary, so sometimes is possible to remove turns from secondary winding without tearing apart the core.
Is it possible in your case?
Is it possible in your case?
You are correct that the secondary is outside, unfortunately it is center tap and one half covers entirely the other...
I have a 676 in as spares, and that range does much the same.
In the UK we are falling over 240v 55-0-55 transformers. We use 110v power tools on site, with the center taps our earth. Reducing shock voltage. We actually call them site transformers. 750VA is a baby. 1500VA-3000VA typical. Just shy of 80vdc though. Not a lot of cheap boards like that.
https://www.google.com/search?clien...OTc5MzM2NjI3NA==,pvt:hg,pvo:3&oshop=apv&pvs=0