I bumped C2 up to 1200uF. Voltages remain flat.
I even changed C2 to 3000uF, just to see what happens. Nothing noteworthy.
I suspect that it is smooth because the caps are being charged and discharged on opposing cycles, but never are they actually drained much.
With a standard bridge rectifier and smoothing cap setup, the smoothing cap (or group) is charged and discharged at 120Hz, resulting in 120Hz ripple.
With the above arrangement, the AC current is pumping and draining the caps at 60Hz each, but at precisely opposite timing which provides stability to the entire circuit.
But that's just a layman's theory of what is going on. There's probably a much better explanation.
I even changed C2 to 3000uF, just to see what happens. Nothing noteworthy.
I suspect that it is smooth because the caps are being charged and discharged on opposing cycles, but never are they actually drained much.
With a standard bridge rectifier and smoothing cap setup, the smoothing cap (or group) is charged and discharged at 120Hz, resulting in 120Hz ripple.
With the above arrangement, the AC current is pumping and draining the caps at 60Hz each, but at precisely opposite timing which provides stability to the entire circuit.
But that's just a layman's theory of what is going on. There's probably a much better explanation.
There is no load to discharge the caps...try adding a resistor and you will start to see ripple.
Crap. You are right DUG. I thought R2 was the "load", but when I add a real load the ripple starts up again.
I got the bright idea to try to use a voltage divider to bring my 37.x volts back down to around 24v. Result; very odd behaviour in the circuit. I end up with millivolts throughout the entire circuit. This stuff is perplexing.
No free lunch, as they say.
Question: is there any way to isolate one part of a circuit from another? I find it frustrating that adding a voltage divider has such a dramatic affect on the entire circuit.
I got the bright idea to try to use a voltage divider to bring my 37.x volts back down to around 24v. Result; very odd behaviour in the circuit. I end up with millivolts throughout the entire circuit. This stuff is perplexing.
No free lunch, as they say.
Question: is there any way to isolate one part of a circuit from another? I find it frustrating that adding a voltage divider has such a dramatic affect on the entire circuit.
Last edited:
In case you are curious about my power supply adventures, I am planning to simply use a reasonably smart implementation of LM317 to regulate. Hopefully that'll be "good enough".
Curious, I will review your PS design however but I think a LM317 or any linear regulator is unnecessary!! Reasons stated earlier.
If anything, I suggest a capacitance multiplier, which will reduce ripple and have the least amount of power dissipation for a linear design.
Another option is a SMPS, followed by a low V drop linear regulator or cap multiplier.
Problem with a straight (regular) linear reg is having enough Vin-vout margin, so that the reg still operates at the trough (low side) of the ripple having a low line V.
Lots to play with!!
If anything, I suggest a capacitance multiplier, which will reduce ripple and have the least amount of power dissipation for a linear design.
Another option is a SMPS, followed by a low V drop linear regulator or cap multiplier.
Problem with a straight (regular) linear reg is having enough Vin-vout margin, so that the reg still operates at the trough (low side) of the ripple having a low line V.
Lots to play with!!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.