some further suggestions
Nice work!
I took it to one of my firends who is a pro PCB designer, he now design PCBs for the PC computer.
His said it is a very good effort, but the board is a bit too big, one should able to fit everything nice and tidy on the PCB and only using 50% of the board size, this will reduce the all the circuit paths and improve the S/N ratio.
Further, I read that Thorsten's super regulator is much better in terms of performance then Jung's, especially for digital circuit, so you may want to try it out also.
Have fun.
Nice work!
I took it to one of my firends who is a pro PCB designer, he now design PCBs for the PC computer.
His said it is a very good effort, but the board is a bit too big, one should able to fit everything nice and tidy on the PCB and only using 50% of the board size, this will reduce the all the circuit paths and improve the S/N ratio.
Further, I read that Thorsten's super regulator is much better in terms of performance then Jung's, especially for digital circuit, so you may want to try it out also.
Have fun.
PCB too big
Thanks for your comments. One reason the PCB seems to big is the traces width issue. I used the biggest trace that I could to reduce trace resistance and improve current drive from the regulators. I'm using a thermal transfert method for the PCB and the resolution is not up to the photographic method. My experience with homemade PCB told me to try to have large traces and via.
This regulator will be used for my audio preamp.
I'm not considering a batch built for now. I want to build the prototype and listen to them, then decide if it is worth to build more. For commercially made PCB you should contact ALW. His PCB and kit are great and he did all the desing.
Thanks ALW for sharing your desing with us...
Thanks for your comments. One reason the PCB seems to big is the traces width issue. I used the biggest trace that I could to reduce trace resistance and improve current drive from the regulators. I'm using a thermal transfert method for the PCB and the resolution is not up to the photographic method. My experience with homemade PCB told me to try to have large traces and via.
This regulator will be used for my audio preamp.
I'm not considering a batch built for now. I want to build the prototype and listen to them, then decide if it is worth to build more. For commercially made PCB you should contact ALW. His PCB and kit are great and he did all the desing.
Thanks ALW for sharing your desing with us...
Some measurements
Both my prototypes are working fine now. I had some little problems with the pcb but this fixed.
No load, Power consumption, approx 20ma (mostly from the led)
Noise: Lower than the noise floor of my HP334A , <0.01mv
Rejection: 90 dB (DC Vin + 0.4V-1Khz, Vout ac: 0.5mv)
Using an external relay and a pulse generator, I switched at high speed a 50 ohms resistive load connected at the output and watch the output using my scope. Just DC, very good.
Listening impressions next...
Both my prototypes are working fine now. I had some little problems with the pcb but this fixed.
No load, Power consumption, approx 20ma (mostly from the led)
Noise: Lower than the noise floor of my HP334A , <0.01mv
Rejection: 90 dB (DC Vin + 0.4V-1Khz, Vout ac: 0.5mv)
Using an external relay and a pulse generator, I switched at high speed a 50 ohms resistive load connected at the output and watch the output using my scope. Just DC, very good.
Listening impressions next...
LT1963 and currently availble regulators
I've been wonder now for a few day and have to ask. Isn't the SuperReg circuitry now built in to some of the currently available regulators i.e. LT1963 (http://www.linear.com/prod/datasheet.html?datasheet=886)?
It has <= 40uV noise. If you look at the circuit description of the TI/Burr-Brown equivalent the diagram looks much like the Jung SuperRegulator.
I too was going to build my own version of the SuperReg, but find it hard to be the Linear Tech specs.
JF
I've been wonder now for a few day and have to ask. Isn't the SuperReg circuitry now built in to some of the currently available regulators i.e. LT1963 (http://www.linear.com/prod/datasheet.html?datasheet=886)?
It has <= 40uV noise. If you look at the circuit description of the TI/Burr-Brown equivalent the diagram looks much like the Jung SuperRegulator.
I too was going to build my own version of the SuperReg, but find it hard to be the Linear Tech specs.
JF
LT1963 and Burr-Brown
Yes, the 1.5A Linear Tech LT1963 is hard to beat as far as low noise at 40 uV. The 0.5A and lower LT's are even better at 20uV. A good voltage reference for a SuperReg such as the REF02 is 10 uV itself. Add the resistors and op-amp and it seems that a person would find it hard to beat the $4.50 LT1963. Plus low drop out, good CMRR. (I think the wonder SuperReg idea has been assimulated by the semiconductor industry for a few years now.)
Here is the TI datasheet: http://focus.ti.com/lit/ds/symlink/tps78601.pdf
See page four for a diagram that looks to me like the Jung SuperRegulator circuit. The TI part is >=48uV (not quite as good as the LT). I'm interested in 1.5A regs (the lower current versions have better specs.)
If the SuperReg circuit is better, I'd like to know and I'll go back to that flavor.
Here is a good LT app note: http://www.linear.com/pdf/an83f.pdf
Page two shows the LT family.
JF
Yes, the 1.5A Linear Tech LT1963 is hard to beat as far as low noise at 40 uV. The 0.5A and lower LT's are even better at 20uV. A good voltage reference for a SuperReg such as the REF02 is 10 uV itself. Add the resistors and op-amp and it seems that a person would find it hard to beat the $4.50 LT1963. Plus low drop out, good CMRR. (I think the wonder SuperReg idea has been assimulated by the semiconductor industry for a few years now.)
Here is the TI datasheet: http://focus.ti.com/lit/ds/symlink/tps78601.pdf
See page four for a diagram that looks to me like the Jung SuperRegulator circuit. The TI part is >=48uV (not quite as good as the LT). I'm interested in 1.5A regs (the lower current versions have better specs.)
If the SuperReg circuit is better, I'd like to know and I'll go back to that flavor.
Here is a good LT app note: http://www.linear.com/pdf/an83f.pdf
Page two shows the LT family.
JF
This is a quite good regulator as IC regulators go, IMHO. But I don't think it is quite up to the Jung or ALW standard.
Since the noise is determined largely by the internal ref, which is the amplified by the error amp, you get similar performance as the discrete ones.
If you look at the input supply rejection ratio, it is nicely constant across the audio band, but only 50 or 60dB, and the dicretes can be significantly better, IF your PCB layout is good enough, that becomes the limiting factor (star ground!).
If you look at the curves for line and load pulse regulation, you see relatively large switching noise at the output, pointing to a limited regulation capability at higher frequencies. This would be no problem for analog audio work, but would limit performance as digital supply regulators.
For analog work I would not go to the trouble of a full discrete regulator, but use one like this if it fits the bill. For digital applications I think discretes like ALW's implementation outperforms them. Whether you can hear that is of course another story.
Jan Didden
Since the noise is determined largely by the internal ref, which is the amplified by the error amp, you get similar performance as the discrete ones.
If you look at the input supply rejection ratio, it is nicely constant across the audio band, but only 50 or 60dB, and the dicretes can be significantly better, IF your PCB layout is good enough, that becomes the limiting factor (star ground!).
If you look at the curves for line and load pulse regulation, you see relatively large switching noise at the output, pointing to a limited regulation capability at higher frequencies. This would be no problem for analog audio work, but would limit performance as digital supply regulators.
For analog work I would not go to the trouble of a full discrete regulator, but use one like this if it fits the bill. For digital applications I think discretes like ALW's implementation outperforms them. Whether you can hear that is of course another story.
Jan Didden
Low Noise Chip Regulators
Carlos: To me, the diagram on page 4 of the BB/TI datasheet resembled the Jung SuperRegulator; however, as Jan pointed out there are a few differences. For example, the voltage reference and the error correcting amp are not bootstrapped with the output.
Jan: Thanks, for pointing out the weakness of the chip regulator--I figured that someone knew more about this than I did. The Linear Tech regulators, in particular, are hard to beat. However, it is possible with discrete components to improve a bit on noise and probably as Carlos mentions lower output impedance. For a low noise voltage reference, Linear Tech makes the quietest one I could find 2 uV max (in-stock at Digikey for a mere $64.95).
Though it is tempting to try the Jung Regulator to squeeze out a bit more performance, I think I'm going to try to optimize use of the LT1963.
Thanks for the replies!
JF
Carlos: To me, the diagram on page 4 of the BB/TI datasheet resembled the Jung SuperRegulator; however, as Jan pointed out there are a few differences. For example, the voltage reference and the error correcting amp are not bootstrapped with the output.
Jan: Thanks, for pointing out the weakness of the chip regulator--I figured that someone knew more about this than I did. The Linear Tech regulators, in particular, are hard to beat. However, it is possible with discrete components to improve a bit on noise and probably as Carlos mentions lower output impedance. For a low noise voltage reference, Linear Tech makes the quietest one I could find 2 uV max (in-stock at Digikey for a mere $64.95).
Though it is tempting to try the Jung Regulator to squeeze out a bit more performance, I think I'm going to try to optimize use of the LT1963.
Thanks for the replies!
JF
Jung Pos reg - Corrected Eagle PCB
Hi. Following an other user comments, thanks Raja, here the corrected Eagle PCB files. The positive PCB had an error on it for C3. Sorry for the mistake. I built 2 PCB with the defective PCB desing and I had to correct them by hand.
I plan to use these Jung +/-Reg in one of my ML38 preamp.
Bye...
SB
Hi. Following an other user comments, thanks Raja, here the corrected Eagle PCB files. The positive PCB had an error on it for C3. Sorry for the mistake. I built 2 PCB with the defective PCB desing and I had to correct them by hand.
I plan to use these Jung +/-Reg in one of my ML38 preamp.
Bye...
SB
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.