Andy, thank you for the link (which is the first time that I have seen this!)
I haven't yet had a chance to peruse the "Never-Connected" website, but to quickly outline our starting point, the idea was to apply bucket-brigade concepts to a power supply, with capacitors linked / separated by switches, and each capacitor transferring its charge to the next cap, and the timing of the switches staggered so that there was never an unbroken path back to the power transformer. I suppose that you could also consider this to be a type of battery and recharger system.
Since the goal was to always keep the path between the load circuit and the transformer partially broken, in our case I don't think that the source impedance of the incoming power supply should have had much effect.
Hmmm, managed to unearth some sort of schematic pretty quickly. There was definitely another circuit which was different from this one, and rather more ambitious. I can't recall where it may have hid itself, however. Actually, I think that the version that we built was the other one.
Oh, well.
Let me know what you think.
best, jonathan carr
I haven't yet had a chance to peruse the "Never-Connected" website, but to quickly outline our starting point, the idea was to apply bucket-brigade concepts to a power supply, with capacitors linked / separated by switches, and each capacitor transferring its charge to the next cap, and the timing of the switches staggered so that there was never an unbroken path back to the power transformer. I suppose that you could also consider this to be a type of battery and recharger system.
Since the goal was to always keep the path between the load circuit and the transformer partially broken, in our case I don't think that the source impedance of the incoming power supply should have had much effect.
Hmmm, managed to unearth some sort of schematic pretty quickly. There was definitely another circuit which was different from this one, and rather more ambitious. I can't recall where it may have hid itself, however. Actually, I think that the version that we built was the other one.
Oh, well.
Let me know what you think.best, jonathan carr
Come Fly With Me..... apologies to Frank S.
Well ........... since you asked.
Very interesting power supply circuit Monsieur Carr. It reminds me a bit of synchronous rectification and power factor both, in switching power supplies. I will address the first objection someone will raise, about it not maintaining the low impedance to the transformer, by pointing out that condition only exist during the forward conduction period of the diode when it is charging the filter capacitor. The rectification part of the circuit is a full wave center tap topology with usual single capacitor split into two caps.
http://www.powertronix.com/html/body_linear.html
This allows each switch between the first filters caps (C4,C5), and the next stage filter cap C8. to open on alternate phases of the AC during the charging phase for C4 and C5. The fast rising edge voltage of the ripple voltage of C4 and C5 are removed as well as AC line noise that would be transferred through the transformer to C8 and the input to the regulator.
The 4.7 ohm series resistor limits the peak current through the transistor switch when it is turned on the transfer charge from C4 or C5 to C8. It also slows the rise time of the charging currents to C8 to further reduce high frequency noise at the input to the regulator.
The schottky diodes between C4, C5, and C8 prevent from discharging back to C4, and C5 and shorten the interval in which the transfer of charge from C4 and C5 takes place. The transistors are being driving by a 60HZ signal so the interval during which they are conducting would be a fairly long period without these diodes. The bipolar switches and these schottkys actually look like another rectifier stage that switch at a different time that the first rectifier stage decoupling noise transferred from the AC line during the period that transformer is charging the filter caps.
It is a very sophisticated circuit with a lot of attention to detail in its design. The resistors across the transformers help to snub the voltage kick across the secondaries as the rectifiers reverse bias and the transformer inductance causes a voltage spike due to than rapid change in current. I have seen this done by one of my favorite designers Ben Duncan in England. I wonder if RC snubbers might be an even better approach. The capacitors C6 and C7 that serve a similar function may have a drawback by dumping their charge and resulting high frequency noise spike into C8 and the input to the regulator. Replacing them with a zener clamp or RC snubber would be very worth while. I noticed this problem because I used the same principle to speed up a telecom protection circuit.
I thank Mr. Carr for laying this one on us. I definitely plan to build this one. How about a circuit of the weekly contribution when time permits ,Jonathan?
Well ........... since you asked.
Very interesting power supply circuit Monsieur Carr. It reminds me a bit of synchronous rectification and power factor both, in switching power supplies. I will address the first objection someone will raise, about it not maintaining the low impedance to the transformer, by pointing out that condition only exist during the forward conduction period of the diode when it is charging the filter capacitor. The rectification part of the circuit is a full wave center tap topology with usual single capacitor split into two caps.
http://www.powertronix.com/html/body_linear.html
This allows each switch between the first filters caps (C4,C5), and the next stage filter cap C8. to open on alternate phases of the AC during the charging phase for C4 and C5. The fast rising edge voltage of the ripple voltage of C4 and C5 are removed as well as AC line noise that would be transferred through the transformer to C8 and the input to the regulator.
The 4.7 ohm series resistor limits the peak current through the transistor switch when it is turned on the transfer charge from C4 or C5 to C8. It also slows the rise time of the charging currents to C8 to further reduce high frequency noise at the input to the regulator.
The schottky diodes between C4, C5, and C8 prevent from discharging back to C4, and C5 and shorten the interval in which the transfer of charge from C4 and C5 takes place. The transistors are being driving by a 60HZ signal so the interval during which they are conducting would be a fairly long period without these diodes. The bipolar switches and these schottkys actually look like another rectifier stage that switch at a different time that the first rectifier stage decoupling noise transferred from the AC line during the period that transformer is charging the filter caps.
It is a very sophisticated circuit with a lot of attention to detail in its design. The resistors across the transformers help to snub the voltage kick across the secondaries as the rectifiers reverse bias and the transformer inductance causes a voltage spike due to than rapid change in current. I have seen this done by one of my favorite designers Ben Duncan in England. I wonder if RC snubbers might be an even better approach. The capacitors C6 and C7 that serve a similar function may have a drawback by dumping their charge and resulting high frequency noise spike into C8 and the input to the regulator. Replacing them with a zener clamp or RC snubber would be very worth while. I noticed this problem because I used the same principle to speed up a telecom protection circuit.
I thank Mr. Carr for laying this one on us. I definitely plan to build this one. How about a circuit of the weekly contribution when time permits ,Jonathan?
Re: Come Fly With Me..... apologies to Frank S.
At the risk of starting a rash of bad puns, I resonate with what Fred is saying.
I needed something quick so, I am just finishing up on a version of this circuit to play with. Just because I don't have any 79xx and 78xx, I will use some LT3xx parts instead. I also plan to experiment by switching rects: between ON Semi MUR860 and MSR860, and maybe the Fairchild Stealths if they show up over the weekend. I want to see if I can hear or see (via scope) a difference.
Now that I see it, I want to build Msr. Carr's circuit also, but my allotment of time for hobby stuff is already spoken for the week
(the wife has spoken)
Hey, I haven't seen Ben Duncan's name in these parts for a while. Maybe somebody can coax him into wandering over ...
We only bite on request
mlloyd1
(who is still stuck at work at the moment but can smell the solder fumes coming from his basement in a couple of hours...)
At the risk of starting a rash of bad puns, I resonate with what Fred is saying.
I needed something quick so, I am just finishing up on a version of this circuit to play with. Just because I don't have any 79xx and 78xx, I will use some LT3xx parts instead. I also plan to experiment by switching rects: between ON Semi MUR860 and MSR860, and maybe the Fairchild Stealths if they show up over the weekend. I want to see if I can hear or see (via scope) a difference.
Now that I see it, I want to build Msr. Carr's circuit also, but my allotment of time for hobby stuff is already spoken for the week
(the wife has spoken)Hey, I haven't seen Ben Duncan's name in these parts for a while. Maybe somebody can coax him into wandering over ...
We only bite on request
mlloyd1
(who is still stuck at work at the moment but can smell the solder fumes coming from his basement in a couple of hours...)
Fred Dieckmann said:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.