Hi gang, I got the replacement parts for my super reg board yesterday. I replaced the chips, caps and LEDs, and now it seems to be working!
However, since I've already blown it up once, I have a couple of questions before I hook it up to anything expensive:
- I have the reg set for +/- ~12v with the divider resistors set to 1K and 750R. I measured the output with sense lines tied together but no load attached and got something like +/- ~13.5v. Does that sound reasonable?
- My supply voltage for the reg is a little high at +/- 25v. I'm using 35v capacitors, and have the recommended heat sinks. Should the board otherwise be OK?
Thanks in advance. If this checks out, I'm hoping to get time to hook it up to my CD63's output stage this evening.
Cheers,
Ben
However, since I've already blown it up once, I have a couple of questions before I hook it up to anything expensive:
- I have the reg set for +/- ~12v with the divider resistors set to 1K and 750R. I measured the output with sense lines tied together but no load attached and got something like +/- ~13.5v. Does that sound reasonable?
- My supply voltage for the reg is a little high at +/- 25v. I'm using 35v capacitors, and have the recommended heat sinks. Should the board otherwise be OK?
Thanks in advance. If this checks out, I'm hoping to get time to hook it up to my CD63's output stage this evening.
Cheers,
Ben
Hi Jan, thanks for getting back to me. No, I only measured the input and output voltages so far. I wondered if a slightly higher reading was expected with no load attached. The positive and negative boards are both out by the same magnitude, so if I have done something wrong then it appears I've done it wrong twice. I'll also check the battery in my DMM, because that's got me before.
The regulator will be supplying 40mA max.
The regulator will be supplying 40mA max.
What IC are you using? I didn't notice much difference loaded vs. unloaded in mine.
Update: I replaced that battery in my DMM and checked its accuracy. It seems fine.
The voltage divider resistors measure within 1% of their expected values.
The LM329 in the positive reg reads as putting out 7.1v, which is within the quoted 5% tolerance.
I checked the output voltage on the positive reg with two 9v batteries. With ~18v in, the output is measuring 13.1v.
I'll try attaching some kind of load to it and see if it varies at all.
The voltage divider resistors measure within 1% of their expected values.
The LM329 in the positive reg reads as putting out 7.1v, which is within the quoted 5% tolerance.
I checked the output voltage on the positive reg with two 9v batteries. With ~18v in, the output is measuring 13.1v.
I'll try attaching some kind of load to it and see if it varies at all.
Hi,
I was checking the thread and if you are looking for a high voltage op-amp check the LTC6090. The output voltage can go up to +/- 70 volts 50 ma or +/140 volts. I used it to control the rails voltage output of my amplifier using a micro. If the current goes too high then the program will shutdown the voltage output to protect the speakers from high current.
I was checking the thread and if you are looking for a high voltage op-amp check the LTC6090. The output voltage can go up to +/- 70 volts 50 ma or +/140 volts. I used it to control the rails voltage output of my amplifier using a micro. If the current goes too high then the program will shutdown the voltage output to protect the speakers from high current.
Another update: I hooked it up to my +/-25v supply. Still reads +/-13.1v out, so it seems it is setting the output to a constant value. I then hooked it up to my CD player and it seems quiet. Too late to listen, but it would seem it's working more or less as designed. I just wish I could understand where this extra volt is coming from. Could the LEDs cause it if the forward voltage or drop is non-standard? They look regular green to me.
Would it be possible to re-do the boards be a lot smaller and have a lot, or all, of the components in surface mount?
Everything is possible, in principle
Vout should not vary with/without load. With 7.1V ref the design Vout with your resistor values is about 12.45V. Still a difference, don't know why.
With a Vce of 25-13=12V across the pass device, at 40mA, is 480mW. No sweat.
Edit: what's the Vo of the opamp?
jan
Last edited:
if contemplating a smt version -- you might want to allow for a power resistor in series to lower the dissipation of the pass transistor heat sink, the idea of a power diode to prevent tears is good idea as well. i don't really think that the pre-regulator is a necessity, it raises the impedance a bit -- but it would also take off some of the stress on the pass transistor.
could also be reconfigured for Walts idea as a shunt regulator, although i haven't listened to this yet
could also be reconfigured for Walts idea as a shunt regulator, although i haven't listened to this yet
You could even combine these suggestions and recommend the protective power diode goes in series with the input.
(Diode #1) drops 0.35V at 1 ampere (room temp)
(Diode #2) drops 0.95V at 1 ampere (room temp)
(Diode #3) drops 1.5V at 1 ampere (room temp)
Noting that the ref is also high (although within spec), I now believe your DMM is off.
Remember, DMMs (except the really expensive ones) spec as anywhere from 2 to 5% PLUS say 2 digits.
So if you measure 13.1 it could be 13.3 or 12.9 aside from the basic tolerance.
Subtract a few % and you're back at your design value.
Since both channels measure the same, this is a probable situation.
I wouldn't worry.
Jan
Thank you, Jan, I really appreciate it.
I have just finished installing the regulator boards in my player. Once separated, the two halves fit perfectly mounted to the side wall of the chassis.
First impressions are indeed excellent. I have no doubt that it's functioning perfectly. Time to pour a drink and listen some more, cheers!
I have just finished installing the regulator boards in my player. Once separated, the two halves fit perfectly mounted to the side wall of the chassis.
First impressions are indeed excellent. I have no doubt that it's functioning perfectly. Time to pour a drink and listen some more, cheers!
There was sugestion
http://www.diyaudio.com/forums/power-supplies/20937-super-regulator-collecting-facts-28.html
to change the 10k and 5K ( biasing the led and voltage reference ) with some curent source diodes ( J508 - J511)
(....The original point was that the substitution of the Jfet current source for the 10 K resistor brought about more than an order of magnitude improvement in rejection for the current source with a very simple change requiring no PCB modifications. I think all of this does show how underfiltered most voltage references in current are and should make one spend a bit more effort designing the current source for biasing the diff pair in your next amplifier or preamp circuit. ...)
(...Replacing a 5kΩ R with a >MΩ current source should improve the noise at the + input of the error amp, but I'm not sure how much improvement can be expected here as the Vout line is quiet. On the other hand we want absolute silence on our reference voltage. I biased the LM329 that way in a Sulzer regulator with some success. ...)
My question is did someone made such modifications without facing with some stability issues ??
http://www.diyaudio.com/forums/power-supplies/20937-super-regulator-collecting-facts-28.html
to change the 10k and 5K ( biasing the led and voltage reference ) with some curent source diodes ( J508 - J511)
(....The original point was that the substitution of the Jfet current source for the 10 K resistor brought about more than an order of magnitude improvement in rejection for the current source with a very simple change requiring no PCB modifications. I think all of this does show how underfiltered most voltage references in current are and should make one spend a bit more effort designing the current source for biasing the diff pair in your next amplifier or preamp circuit. ...)
(...Replacing a 5kΩ R with a >MΩ current source should improve the noise at the + input of the error amp, but I'm not sure how much improvement can be expected here as the Vout line is quiet. On the other hand we want absolute silence on our reference voltage. I biased the LM329 that way in a Sulzer regulator with some success. ...)
My question is did someone made such modifications without facing with some stability issues ??
The reference is not within the feedback loop so has absolutely no impact on stability.
Jan
For some comments on this, see post 112 of:
http://www.diyaudio.com/forums/power-supplies/90502-about-jung-super-regulator-12.html
Could someone please explain to me the purpose of the current source (R1, Q2 in the positive regulator)?
I note that these components were in place in the original version and so their original purpose was not related to the bootstrapping (which presumably could have been effected in a few different ways). Also, I read Walt's article about the required changes to the circuit for bootstrapping and note, in particular, the required change to diode D2 to a Zener. Do I understand it correctly that the Zener is there only to ensure the base voltage of the pass transistor is at least 6.8V? What does C1 do?
I apologise if these are very basic questions but I am new to all this and particularly series pass regulators.
I note that these components were in place in the original version and so their original purpose was not related to the bootstrapping (which presumably could have been effected in a few different ways). Also, I read Walt's article about the required changes to the circuit for bootstrapping and note, in particular, the required change to diode D2 to a Zener. Do I understand it correctly that the Zener is there only to ensure the base voltage of the pass transistor is at least 6.8V? What does C1 do?
I apologise if these are very basic questions but I am new to all this and particularly series pass regulators.
Not having the circuit in front of me, I guess you mean the current source that supplies the base current to the pass transistor. That is it's purpose.
The opamp regulates the output by siphoning off any excess current that would cause the output to rise above the set value.
On the zener: if you mean the zener in series with the opamp output, that is there to keep the opamp output somewhere in the middle of the supply voltage so it works correctly.
Without it the opamp output would need to be one Vbe below the supply (the base voltage is NOT 6.8V of course - it's one Vbe, some 0.62V - below the emitter which is Vout).
Not many opamps would be comfortable with that; it would restrict the opamp choices.
Hope this helps,
Jan
The opamp regulates the output by siphoning off any excess current that would cause the output to rise above the set value.
On the zener: if you mean the zener in series with the opamp output, that is there to keep the opamp output somewhere in the middle of the supply voltage so it works correctly.
Without it the opamp output would need to be one Vbe below the supply (the base voltage is NOT 6.8V of course - it's one Vbe, some 0.62V - below the emitter which is Vout).
Not many opamps would be comfortable with that; it would restrict the opamp choices.
Hope this helps,
Jan
Last edited: