The picture above show traces of the negative power supply rail. Therefore, a dip in the negative rail voltage goes up, or in the positive direction. Sorry if I did not make that clear. So, lets say a drum, or any moderately loud bass will cause a dip in the voltage, and then a recovery as the level of charge in the reservoir caps is restored.
The test shows that the typical power supply dips caused by bass passages are suppressed, along with the ripple (assuming of course that they are within the cap multiplier dropout voltage, which is adjustable by a trim pot).
The test shows that the typical power supply dips caused by bass passages are suppressed, along with the ripple (assuming of course that they are within the cap multiplier dropout voltage, which is adjustable by a trim pot).
For the VSSA circuit we can make voltage drop higher, so along with the excellent noise suppression we can use cap multiplier to trim front end bias in case that we use PeeCeeBee version of the circuit without CCS. I was foolish to dismantle the pcb from heatsink in order to change current feeder resisitor, while all that was needed is cap multiplier set for higher voltage drop. Since VSSA standing current is around 150mA per channel, even if we set dropout voltage 3V we are still well withing ratings of all used components and there will be no too much dissipation.
It seems to me that Mr Evil/PMI cap multiplier is genuine budget priced alternative for VSSA PSU, and much safer for the average user. I am not saying that I shall never use SMPS, but we can build several cap multipliers for the price of single SMPS recommended by LC.
Yes, within a reasonable range, with typical 8-ohm speakers, but you always have a greater range of adjustment with the resistor values. So, I don't think you were wrong to do that.
For anyone else: For a power amp which has a very high standing current (or class A), the dropout voltage should not be set any higher than needed for ripple and noise suppression.
If you double the dropout voltage, the power dissipation in the pass transistor heatsink will also double. Even with class A-B, the power dissipation will increase in use, especially with low impedance speakers, high volume, etc.
...plus the cost of the transformer, of course,
Hi Pete,
Thanks again for all you do (and thanks to Mr Evil, of course!). Can you discuss suitability of low power applications? I basically need +/-15v to power a preamp and crossover at about 200 - 500mA. What size transformer and what other tweaks are needed? Many thanks. Regards. -Darrell
Thanks again for all you do (and thanks to Mr Evil, of course!). Can you discuss suitability of low power applications? I basically need +/-15v to power a preamp and crossover at about 200 - 500mA. What size transformer and what other tweaks are needed? Many thanks. Regards. -Darrell
+/-50V, 6A max, pk.
Cost will be $10, + shipping + Paypal fee.
The maximum voltage is limited by the voltage rating of the pass transistor, because the full supply voltage will appear across it briefly at startup.
Maximum average current depends on dropout voltage setting and size of pass transistor heatsink.
As discussed and shown above, I designed for class A-B, up to 100W per channel, 8-ohm load.
The board is also designed to allow mounting the rectifier diodes and pass transistors from below the board to allow higher current/dropout voltage combinations, but in most cases this would not be needed.
I'll do some calculations over the weekend. In general, I would probably use smaller value and lower profile electrolytic caps and heatsink, both to save some money, and lower the height of the power supply a bit for a smaller chassis. Up to 250 mA, I would probably use a small encapsulated EI transformer, for 500mA, a small toroid.
I thought that I was making layouts with plenty of room! ...It was so difficult to remove resistor from the compact pcb that many "F" words were used.
For the CLCRC filter, I used a 10uH 3.5A inductor (under $0.50) and a 0.33R 5W resistor ($1.50). The 5W is overkill, but I had those on hand.
It may be useful to think of this as a classic PI filter, followed by an RC filter. The purpose of the pi filter with the inductor is to filter out rectifier switching noise in the audio frequency range. The RC hits the ripple, which is what is left after the PI filter.
Mr. Evil,
I would like to use this cap multiplier circuit in following situation, do you see any reason's why it would not work:
Main xfmr r-core 500va 2x36v for OPS
A second xfmr r-core 30va 2x9v for front end drivers
For front end and drivers psu, the secondaries of both xfmr's will be in series.
I would like to use your cap multiplier (with higher voltage rating components) for the driver stage with an adjustable voltage drop to about 8v.
A second cap multiplier would be used for the front end with a voltage drop set at about 4v.
Greetz
Ben
I would like to use this cap multiplier circuit in following situation, do you see any reason's why it would not work:
Main xfmr r-core 500va 2x36v for OPS
A second xfmr r-core 30va 2x9v for front end drivers
For front end and drivers psu, the secondaries of both xfmr's will be in series.
I would like to use your cap multiplier (with higher voltage rating components) for the driver stage with an adjustable voltage drop to about 8v.
A second cap multiplier would be used for the front end with a voltage drop set at about 4v.
Greetz
Ben
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.