So I would need a transformer higher in voltage than 24-0-24 Vac? I would want maybe 6-7A (if not greater) current capability from the transformer so that I can possibly power a USB port for charging my phone or iPod while it is playing.

To do the USB port option would require a 5 volt regulator capable of at least 1A of current and I would install a 1A fuse after the regulator then install a 5.1V Zener between the output of the fuse and ground. That way if the regulator shorts and applies full rail voltage it will cause the zener to conduct and blow the fuse thusfore saving my device from over voltage.

Uh, I didn't say that, especially since I didn't know.

My feeling is that if the kit supplier specs a 48VCT 2 A transformer then it will probably work and the current loading math was just an exercise which, after correcting my boo-boos, indicates it can supply the current (as I would hope, since they spec'd it). The other comment was just saying there's 'more to it' if one were designing from scratch.

35 Watt into 8 ohms is the 'worst case' voltage at 16.7V peak. The amplifier topology seems to be bog standard so, as a wild swag, say it needs 3 V bias headroom. That means, under all conditions (low line AC, etc.), the power rail should be at least 20V. Fortunately, 8 Ohms is also the 'lower current' of the two loads and with the 'high current' 4 Ohms peak its only 11.8 V, so the power rails need only maintain 15 V in that case. That means a 5 V 'sag' under maximum load would be perfectly fine.

In theory one might think more voltage would allow more power but the schematic shows they've wisely included an instantaneous current limiter so it's debatable whether it would, or be enough to matter as another 9 Watts would be an imperceptible 1 dB increase and even if we magically turned it into a 50 Watt amplifier that's only 1.6 dB more. It just doesn't matter, except for 'specmanship'.

And, of course, there's my, by now, long standing contention that in any 'real world' use it's never going to be putting out anything close to 35 Watt continuous so this is mainly an exercise to 'pass' an artificial FTC 'continuous Watt' specification that requires line power (for U.S. equipment) during test "shall be" 120 VAC anyway (so line voltage 'sag' is moot).

On the other hand, there's nothing to keep us from making it 'beefier' than 'necessary'.

They also spec ambient at 25C ("at least," so you know it will be 25C) and, frankly, I'd be more concerned with heat because I think it's worse to have the thing burn up on a hot summer day than discover it only does 30 Watt during an AC line sag.
I don't know where all those extra amps are going to go.
IMO you'd be better off picking up a cheap USB power adapter off Ebay because regulating '35 V' down to 5 V 1A is going to burn off 30 Watt and that's a heck of a lot of waste heat just to power 5 Watt worth of USB devices.

Could be, but would be better using a regulator as I would be taking up one less power outlet.

You're kidding, right? Put a couple of AC service outlets on the amp, one switched and the other not. Or make them the same. Then you'll gain an outlet.

I could do that

While cleaning up I found a defunct 7.1 channel Onkyo home theatre receiver. It has this transformer in it



The power supply uses two 8,200 uF 63 volt caps and more than likely produces both a positive and negative voltage rail. If so will that transformer work?

I measured and at 120 Vac I get 39.5-0-39.5 Vac at probably a few amps of current. Good thing is it has other lower voltage outputs which I suppose could be used to power a USB port or two.

The other secondaries are

10.3 Vac
20.62-0-20.62 Vac
5.78 Vac

If necessary I can use one of those other windings to buck the primary to reduce the ac voltage going to the + & - power supply.

I went ahead and robbed the two filter caps and two bridge rectifiers before finally putting the chassis outside to be thrown away.

Not sure how they had both bridge rectifiers connected to provide the + and - voltages though. never seen that type of connection before. I will more than likely use only one rectifier or parallel them both.

It will be a heavy power supply, but it won't cost me anything to build it I even saved the power cord with strain relief from the receiver and I am sure I have a good SPST switch somewhere.

Will try to build all in one box if possible.

Given I ordered the speaker protection kit as well I can use one of those secondaries to produce the 12 Vdc necessary for the circuit to work.

The input to the amp will be a pair of 1/4" switched jacks with the switch tied to ground to eliminate any buzzing when I plug in the audio cables.

What was 'unusual' about the two bridges?

Sight unseen I'd 'guess' one is for the 39.5 V output power rails with the other for a bipolar preamp supply, and who knows what else, using the other +- winding. Another winding probably for logic power and who knows what for the 4'th, maybe 'lighting'.

Odds are the Onkyo did more power than what you're building so it ought to handle the current, but the voltage is high.

Both bridge rectifiers were connected In the circuit with one tied into the cap for the positive rail and one for the cap for the negative rail. I could understand that arrangement if the transformer was non center tapped, but it is center tapped. I am supposing that the rectifiers are connected in parallel maybe for more current handling since both were mounted to a heatsink.

If the voltage is indeed too high I can use a secondary winding to buck the voltage some.

Concerning the power supply I will build the amps then the power supply and bring the voltage up with a variac until I get + and - 35 Vdc then measure the ac input and subtract it from 120 (highest the ac line voltage gets) which will determine how much I need to buck the primary voltage. I will also test under load to see how much the dc voltage drops if at all. That way I will know the exact voltage to set the power supply to produce.

I'm looking for the voltage to stay close to +35 -35 Vdc dropping to maybe a minimum of +30 -30 Vdc under full music power at 4 ohms. Good thing is the speakers that will be used with the amp can more than likely only handle 20 watts and are 8 ohms. Provided the amp does put out 35 watts at 8 ohms there can be more leeway in the power supply voltage.

I am still considering using that car audio preamp/crossover/eq to step up the audio level to the amp. It has two selectable inputs and also a stereo subwoofer output with its own frequency and level control which will make it easy to integrate a couple small subs should the need arise, but the small vintage Olson speakers I am using sound good as is.

I temporarily connected the power supply and got more dc voltage than I will need. I used a secondary winding to lower the primary voltage and got these results.

120 vac in 89.3 vac out 30.7 Vac bucking voltage +41.6 Vdc -41.6 Vdc unloaded. The other two secondary voltages were lower as well.

Problem is how do I power that car audio preamp? Would more than likely draw too much current to power it from one of the 15 volt supplies on the amp since it uses a simple resistor and zener diode arrangement. I suppose I could take one of the secondaries and use a voltage doubler on it or use a separate transformer to power a 12 volt regulator for the car audio preamp and a 5 volt regulator for the USB charging feature I plan to integrate. Pretty sure I have one of those Radio Shack 6.3-0-6.3 Vac transformers sitting around somewhere. I have a DPDT switch which could be used in conjunction with a SPST switch to allow for a master switch that will turn on AC to the whole unit and use the DPDT to cut the dc voltage ( to the preamp and also the AC to the amplifier power transformer. That way I can still power up just the USB port for charging purposes and not need one of those small switch mode USB chargers so notorious for RFI.

Power supply schematic.



I'll be installing one of these USB connectors on the chassis http://www.usbfirewire.com/Parts/rr-211300-10.html

I got one of the two amps built today.

The manual says to break the collector connection of TR10 and using a milliammeter adjust VR1 until the current is 40 mA. I damaged my DMM's current function a good while ago so I needed some other way to measure the current.

What I thought about doing is using a DPDT switch and a couple 1 ohm resistors (one for each amp) then using three banana jacks. Each 1 ohm resistor would be in series with it's respective TR10 collector lead and the +35 volt supply. The DPDT switch would simply short out the 1 ohm resistor. The voltage developed across the resistor would be equal to the milliamps of current through the 1 ohm resistor.

An update on the project. I was out and about and found an ammo box in decent shape.

Only thing I will need to do after drilling all the necessary holes is to clean up the few spots of rust, give the whole cabinet a good scuffing, use some Dupli-Color Rust Fix then finish with some Eastwood chassis black satin spray paint.



I will make a change to the power supply schematic since the preamp will not be mounted in the case. I will simply switch the 12 volt line instead of switching power to the remote wire so I only have to run two wires from the box.




I have a question. Since both amps will be in an enclosed case should I install a small fan? If I need to I can just drill a few holes under where the heatsinks will be to permit airflow over the heatsinks.

I will more than likely use another 7812 regulator for the fan, but depending on the fan I use I might decide to go with an LM-317 so I can slow the fan down some if it is noisy.

I found this fan which may work for what I need. http://www.newegg.com/Product/Product.a ... 6835191005