power supply questions

[dentaku]

I salvaged some parts form a Sony amp/receiver and everything is there to build a decent power supply all broken into separate sub-boards. All it needs is some positive and negative regulators.

I have three questions about these parts.

1 - What's the purpose of the two 22nF caps?

2 - The transformer has multiple taps, the two red wires put out around 43V and blue is 34.5, it's all connected together and has a center tap. There was a switch on the back of the device so you can switch between them. I don't have the original enclosure anymore so I'm not sure what the switch is for. Could it have been a switch for 110 vs 120?

3 - I'm wondering why the center tap of the transformer isn't connected to the point between the two big caps and the - side of the two small caps on what is labeled DC1 board instead of the way it is now where they're on two separate wires using an extra pin on the connector and finally connecting together on the main board at the point I labeled with the yellow circle. As you can see, it uses a big copper thing I'm assuming has to do with start grounding.

[Paul Price]

If you are going to build a power supply, re-use the transformer, the bridge rectifier and the filter caps, forget about the schematic..you are not going to build another receiver.

The small caps are for suppressing RFI that the rectifier diodes can create.

The common ground connection you've found is Sony's way to prevent power line hum from ground loops. You needn't use the same technique if you want to build your own power supply to achieve a good common ground connection.

The taps are likely to switch between power line voltages, but again, you will be making your own circuit and you can choose to use either winding.

Your greatest concern with this transformer is the peak no-load output voltage which is higher than most low-voltage regulator chips can handle. There are switching regulator chips that can easily handle this higher voltage and deliver 3 to 5 A  to create a step-down regulator and you can use a large heatsink from the old chassis to other make a zener-diode pass-transistor pre-regulator to develop well-filtered power at a voltage level that doesn't exceed the Max. V of the regulator and other chips of your power supply design.
If you still have the rest of the receiver, you can find you can obtain the zener diode pre-regulator pass transistor you need for free if the receiver used discrete high power transistors for the audio output stages.


Don't be afraid of creating a nice bench power supply of your own design that can double as a space-heater in the cold days of winter. If you chose in your design to go the way of a quiet pass-transistor pre-regulator, just recycle a small fan and other heatsinks and large toroids useful for your new original custom designed power supply. You might not even need a fan for cooling, but if you discover during development that things start to run a bit hot,  a fan can be salvaged for free from an orphan or defective desktop PC power supply.

Don't worry, this global-room-warming effect will only be noticed when your custom-built power supply is outputting low-voltages at high currents with this transformer.
 

[Kleinstein]

The switch could have been to get the best power for 4 or 8 Ohms load. For lower impedance the amplifier would work better with a slightly lower supply.

For a power supply you could also use the heat sinks of the power amplifier, especially in combination with the enclosure.

43 and 34 V are rather high voltages. So after rectifier and filter the voltage could be up to about 60 V / 50 V. So the supply is more about high voltage, not so much current. Passive PFC filtering with a rather bulky inductor could help to get a little more current at lower voltage.

[Paul Price]

While a PFC type of pre-regulator might be a good idea for NASA, it would be beyond the capability of most undergraduate electrical engineering students to design or else find a reference circuit without custom parts costing so much you could instead buy a new used car and a Chinese made power supply.

Keep things simple.

Think pass-transistor pre-regulator, or if you must keep things cool and spend more money, you can buy high-voltage versions of Simple Switcher IC's from National (now T.I.). Much lower in complexity, cost, and their specsheets offer well-documented design, all steps included, cheap enough to buy, easier to use.

My advice would be (for those without a Masters Degree in EE) to always stay with a good simple-fewer-parts design.

[dentaku]

If you are going to build a power supply, re-use the transformer, the bridge rectifier and the filter caps, forget about the schematic..you are not going to build another receiver.

The small caps are for suppressing RFI that the rectifier diodes can create.

The common ground connection you've found is Sony's way to prevent power line hum from ground loops. You needn't use the same technique if you want to build your own power supply to achieve a good common ground connection.

The taps are likely to switch between power line voltages, but again, you will be making your own circuit and you can choose to use either winding.

Your greatest concern with this transformer is the peak no-load output voltage which is higher than most low-voltage regulator chips can handle. There are switching regulator chips that can easily handle this higher voltage and deliver 3 to 5 A  to create a step-down regulator and you can use a large heatsink from the old chassis to other make a zener-diode pass-transistor pre-regulator to develop well-filtered power at a voltage level that doesn't exceed the Max. V of the regulator and other chips of your power supply design.
If you still have the rest of the receiver, you can find you can obtain the zener diode pre-regulator pass transistor you need for free if the receiver used discrete high power transistors for the audio output stages.


Don't be afraid of creating a nice bench power supply of your own design that can double as a space-heater in the cold days of winter. If you chose in your design to go the way of a quiet pass-transistor pre-regulator, just recycle a small fan and other heatsinks and large toroids useful for your new original custom designed power supply. You might not even need a fan for cooling, but if you discover during development that things start to run a bit hot,  a fan can be salvaged for free from an orphan or defective desktop PC power supply.

Don't worry, this global-room-warming effect will only be noticed when your custom-built power supply is outputting low-voltages at high currents with this transformer.
 

Good stuff, thanks 

I was asking more for educational purposes. I already have a power supply I built out of a few parts from an old amp (transformer, power button, fuse) and some new parts (the caps and pots of course).
I just like the way this is broken into separate modules.

I was also wondering about the quite high voltage coming from that transformer. It has another secondary winding that puts out 21V AC so that would be more reasonable.
I have an LM338T I've never used so at least it's more capable than a plain old 317. I don't have a negative regulator better than a 337 though.
I know this is old fashioned stuff but for some reason I find it interesting to learn about.

So... Do you think it's reasonable that the switch would be for 110V and 120V input? at 120V there's barely a 10V difference between both taps.
and yes, I wasn't planning to use the switch anyway

[dentaku]

The switch could have been to get the best power for 4 or 8 Ohms load. For lower impedance the amplifier would work better with a slightly lower supply.

interesting

[Kleinstein]

The 21 V tap is likely there to power the preamplifier or similar low power parts. The rectifier and filter cap and fuse used there will show how much current you can expect.

So for the supply the best choice is more likely a floating regulator like the cheap Chinese supplies (and many HP brand lab supplies) use. The 21 V winding for the control circuit and the 34 V / 43 V or even only the 9 V winding(s) for the power. So the main regulation uses OPs and power BJTs. There might be an LM317 for the supply of the OPs.

The LM317 / LM338 are limited to rather low voltages (e.g. 9 V between the 43 and 34 v taps).

[Paul Price]

Although the 21V winding seems inviting, it might not be a high-current winding like the higher voltage windings, but could only supply a few tenths of an Amp for  something like a VFD front-panel display.

[dentaku]

Although the 21V winding seems inviting, it might not be a high-current winding like the higher voltage windings, but could only supply a few tenths of an Amp for  something like a VFD front-panel display.

What's the best way to tell. It's the same size wire and connector as the other winding. The orange brown orange are the the 21V  0V  21V .
It's connected to a board that just uses individual small-ish diodes instead of a big bridge rectifier and goes through 125V 4A fuses. The 34V, 43V windings go through 8A rated fuses and a 6A rated rectifier so that make sense.
The board with the VFD is well labeled and only has 32V and +5V on it coming through a flex ribbon.

I have another more reasonably sized transformer I can use but I think it's still rather high voltage, like 27V. The transformer this amp came with is unreasonably large for me anyway

[Audioguru]

34.5V is 80% of 43V. The 100VAC electricity in Japan is 80% of the 120VAC over here. So the switch selects being powered in Japan or in North America.

[dentaku]

34.5V is 80% of 43V. The 100VAC electricity in Japan is 80% of the 120VAC over here. So the switch selects being powered in Japan or in North America.

Yeah, 100V not 110V like I was saying.
That's totally logical.

[David Hess]

1 - What's the purpose of the two 22nF caps?

The small capacitors suppress noise produced by reverse recovery of the rectifiers.