Like I stated in the other thread...
I used it with great success on removing noise generated by alternators in automotive application.
The only thing that may be of concern is emi radiated from the toroid and the choke acting as an "antenna" and instead of helping making things worse.
And....
Here's a Regulated version of the earlier post :
An externally hosted image should be here but it was not working when we last tested it.
its not center tapped...
Has dual secondaries, but Xpress Schematic has only single configuration or if you stack them then you get a center tapped one.
I will be using a Plitron 1KVA unit with dual 55V secondaries rated at 9A per side
Outputs will be set at 50VDC through the regulators.
That's in excess of 500W of dissipation in the two regulators!
Yes i know, but that output voltage is not set in stone either.
Also the regs are buffered by multiple transistors ( 4 on each side to be more precise, but were not inlcuded on the diagram as it would extend out of the visible area) and most likely will lower the capacitance pre-reg to 8600uF. Will most likely add a reg bypass relay to allow fast charge up of caps upon start up that will be latched until caps reach 70% of charge where it would latch in and let the regs take over the stabilization.
Now my main issue that I'm stuck at is what devices to use to use as the ones used in the data sheet Q1,Q2 - 2N2905 x 2 and Q3,Q4 -LM195 x 6 ( 3 on each rail )and both are low power. Looked for a suitable higher power units with no luck, but I'm sure I'm looking in the wrong area.
Any idea's ???
Last edited:
That's the opposite to what most builders would WANT to do.
We generally want to reduce the peak charging current into the capacitors.
Big surges reduce the life/reliability of caps and blow fuses at start up.
Haha, yes, it will also weld the relay contacts together so the regulator has really an easy life
jd
Haha, yes, it will also weld the relay contacts together so the regulator has really an easy life
jd
Dammit, never thought of that.
Back to the drawing board
...It is generally a bad idea to regulate the main rails of a power amplifier to a fixed voltage. This will cause very high power dissipation and will force the design to regulate for worst-case low mains and heavy load. It also deprives the amplifier of dynamic headroom.
I prefer to use what I like to call soft rail regulation. It is a regulated output voltage that floats with the available voltage at the input to the regulator. This is really nothing more than a fancy high-current capacitance multiplier. This allows the amplifier to adapt to the available voltage and charge without falling out of regulation. It greatly reduces the penetration of rectifier ripple and other sources of noise into the output stage and the rest of the amplifier. There does not need to be much voltage drop across such a regulator to make it very effective. In fact (especially if implemented with MOSFETs), it can be designed so that there is very little penalty or problem if it does fall briefly out of regulation in a particularly demanding passage.
Cheers,
Bob
I'm understanding what your saying, but I'm having a hard time making up a schematic in my head. Willing to share an rough idea ???
I'm understanding what your saying, but I'm having a hard time making up a schematic in my head. Willing to share an rough idea ???
Google capacitance multiplier
jd
Thanks Bob for the idea
looked it up but i still don't see how more efficient it would be.
SMPS is crappy for reactive loads and that leaves a design with stepped rails to follow the input signal . Wouldn't this be a better suited alternative for Class AB, even though there would still be some noise at the switching point, but it still would be a viable option as far as efficiency goes.
I see all this talk about standard power supplies and noise rejection, but i don't see a single person mention about stepped rail supply's used in Class H amplifiers.
wg_sky had posted a while back in my thread a design for a multiple rail power supply. i will have to go and dig it up and post it up here.
OK...
Found it !!
Low side and High side.
I would say it would be a bit more complex, but its worth a try for someone that has multiple output devices and is well with in their SOA.
looked it up but i still don't see how more efficient it would be.
SMPS is crappy for reactive loads and that leaves a design with stepped rails to follow the input signal . Wouldn't this be a better suited alternative for Class AB, even though there would still be some noise at the switching point, but it still would be a viable option as far as efficiency goes.
I see all this talk about standard power supplies and noise rejection, but i don't see a single person mention about stepped rail supply's used in Class H amplifiers.
wg_sky had posted a while back in my thread a design for a multiple rail power supply. i will have to go and dig it up and post it up here.
OK...
Found it !!
Low side and High side.
I would say it would be a bit more complex, but its worth a try for someone that has multiple output devices and is well with in their SOA.
Last edited:
Finally had some time and re-drew a few things and...
here she is
INPUT ????
An externally hosted image should be here but it was not working when we last tested it.
here she is
INPUT ????
LEDs are not forward biased
Dammit...
LOL
I thought i had them right the first time around!
Anything else ???
Last edited:
Regulated supplies
I changed my mind about the amplifier power (and complexity) i needed and reduced the rail voltage by adding a high current regulator after the reservoir capacitor. I had heard before that this can leave an amplifier with a sonic signature of the regulator and to avoid that there should be as much capacitance after the regulator as before it, so i added 100,000uf to the output of the regulator and a protection diode across the regulator "just-in-case". All works ok so far but now i am a little puzzled.
Bob Cordell said that regulation limits the dynamic headroom - sorry to be ignorant, but what does this mean?
How does a capacitance multiplier get over the problem?
I changed my mind about the amplifier power (and complexity) i needed and reduced the rail voltage by adding a high current regulator after the reservoir capacitor. I had heard before that this can leave an amplifier with a sonic signature of the regulator and to avoid that there should be as much capacitance after the regulator as before it, so i added 100,000uf to the output of the regulator and a protection diode across the regulator "just-in-case". All works ok so far but now i am a little puzzled.
Bob Cordell said that regulation limits the dynamic headroom - sorry to be ignorant, but what does this mean?
How does a capacitance multiplier get over the problem?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.