I have the latest version of the J/D super regulator from Audio Xpress but I don't know how to get the lower voltages out of it.I'm trying to get 5 vdc and 7vdc.I changed the reference diode from the LM329 to a TL 431.The voltage drops but doesn't go below 7.8vdc.What changes do I need to make to get 5 and 7 vdc.Thanks for any help in advance.
Mike
Mike
Hi,
The value of the zener diode should equal the value of the reference voltage at the + input of the regulator IC. The reference voltage should be lower than the output voltage (about half its value) from the regulator.
If your zener diode is 6.8V, it's not possible to get voltage lower than 7.4V at the regulator output. In many cases it may be possible to short the zener, which you may try as well to see what happens.
Also, in your case, the + power terminal of the regulator IC should be connected before the regulator transistor.
Could you post the schematic you're using to make things easier for me to visualize?
Regards,
Milan
The value of the zener diode should equal the value of the reference voltage at the + input of the regulator IC. The reference voltage should be lower than the output voltage (about half its value) from the regulator.
If your zener diode is 6.8V, it's not possible to get voltage lower than 7.4V at the regulator output. In many cases it may be possible to short the zener, which you may try as well to see what happens.
Also, in your case, the + power terminal of the regulator IC should be connected before the regulator transistor.
Could you post the schematic you're using to make things easier for me to visualize?
Regards,
Milan
janneman said:... why don't you feed the 825 from the prereg output, is there a voltage level issue?
Hi Jan,
You're right - there is a voltage level issue indeed. I'm not sure that the voltage at the prereg out is high enough for the regulator IC to work properly. The voltage at the prereg out exceeds the output voltage by about 2.5V, which is in this case only about 7.5V. However, Mike might also try out what you're suggesting and let us know if it works.
Regards,
Milan
Yes. With that zener in the opamp output, it doesn't have to swing more than some 2V BELOW Vout (=5V), so I see no need for a highish opamp supply voltage. The 825 is fully spec'd for 10V total supply, so that would be the min prereg output voltage to allow the 825 to be fed from the stable and clean prereg instead of the raw supply. And in that case the zener could go, and that ugly elcap as well, setting Vout of the opamp almost smack in the middle of that 10V (5.6V actually, +/- a dB. )
Jan Didden
Jan Didden
Hi,AndrewT said:can superreg be built up from alternative transistors?
Yes, it can. Almost any modern AF driver transistor will do the job.
Critical characteristics are VCEs and f(hfe/Ic), IMO.
Is D44 easily available in UK?
Yes. I've come across them in RS and Farnell catalogues.
Regards,
Milan
5V HQ regulator...
good day everyone...
I have been following the posts on the high quality voltage regulators. Thanks to the efforts of all involved, it has been extremely enjoyable to build my first prototype, along with the necessary 40dB Gain, 10Hz-3MHz "noise amplifier"... (trying for 60dB soon
Some comments:
- bootstrap circuit as suggested by W.Jung in EDN June/97, "Low Noise Power for Analog Circuits"
- 100uH / 120uF card filter implemented
- Am using an AD8651 for low noise & transient response requirements while satisfying GBW integrity with 5V single supply.
Had some problems with stability initially, corrected with a 1.5MHz RC filter on the Vin line. The "dead-bug" circuit uses long (~60mm) power supply wires which likely work well as antennae. AD8651 does not roll off very nicely with gain=1...
- Used a ST1803 pass-NPN, since that was what i had around at the time... had to increase bias currents in the driving circuit a bit to compensate for the low Hfe (10 or so). Ibias is now about 50mA, allowing 0.5A on the output.
Results:
- 4.2mVp-p output fluctuation with 68ohm load switched in and out at 100kHz (73mA pulses, fairly sharp edges, i suspect about 1µs rise/fall times): 54mohm output impedance at 1/1µs = 1MHz?
- reduces to within 1mVp-p after ~1µs
- 13.7µVrms noise from 10Hz-3MHz, 10mA static load. No shielding measures taken, actual noise should be less than this figure. When inputs to "noise amp" are shorted, noise decreases to below 1µV though, so probably not too much coupling in.
Conclusion/Ideas:
- very good noise performance
- would like to increase transient response
Questions:
- will a fast transistor like the BCP56(ft=130MHz) lower the output impedance at 100kHz and above? I certainly want to use a different pass transistor, to lower the bias currents - so I can can kill two flies in one swing...?
greeting from switzerland,
mike
ps: is my 1st post, feel free to correct me for poor form, etc...
good day everyone...
I have been following the posts on the high quality voltage regulators. Thanks to the efforts of all involved, it has been extremely enjoyable to build my first prototype, along with the necessary 40dB Gain, 10Hz-3MHz "noise amplifier"... (trying for 60dB soon
Some comments:
- bootstrap circuit as suggested by W.Jung in EDN June/97, "Low Noise Power for Analog Circuits"
- 100uH / 120uF card filter implemented
- Am using an AD8651 for low noise & transient response requirements while satisfying GBW integrity with 5V single supply.
Had some problems with stability initially, corrected with a 1.5MHz RC filter on the Vin line. The "dead-bug" circuit uses long (~60mm) power supply wires which likely work well as antennae. AD8651 does not roll off very nicely with gain=1...
- Used a ST1803 pass-NPN, since that was what i had around at the time... had to increase bias currents in the driving circuit a bit to compensate for the low Hfe (10 or so). Ibias is now about 50mA, allowing 0.5A on the output.
Results:
- 4.2mVp-p output fluctuation with 68ohm load switched in and out at 100kHz (73mA pulses, fairly sharp edges, i suspect about 1µs rise/fall times): 54mohm output impedance at 1/1µs = 1MHz?
- reduces to within 1mVp-p after ~1µs
- 13.7µVrms noise from 10Hz-3MHz, 10mA static load. No shielding measures taken, actual noise should be less than this figure. When inputs to "noise amp" are shorted, noise decreases to below 1µV though, so probably not too much coupling in.
Conclusion/Ideas:
- very good noise performance
- would like to increase transient response
Questions:
- will a fast transistor like the BCP56(ft=130MHz) lower the output impedance at 100kHz and above? I certainly want to use a different pass transistor, to lower the bias currents - so I can can kill two flies in one swing...?
greeting from switzerland,
mike
ps: is my 1st post, feel free to correct me for poor form, etc...
ST1803 is not particulary well suited, not very fast and has a poor current gain.
I have used BCP56 with good results. If you are looking for holemounted types I think BD139/140 and D44H11/D45H11 are very good.
Here is a regulator with BCP56.
http://home5.swipnet.se/~w-50674/hifi_pics/hifi_100pr/jsr02r0_overview.jpg
I have used BCP56 with good results. If you are looking for holemounted types I think BD139/140 and D44H11/D45H11 are very good.
Here is a regulator with BCP56.
http://home5.swipnet.se/~w-50674/hifi_pics/hifi_100pr/jsr02r0_overview.jpg
Re: 5V HQ regulator...
http://waltjung.org/PDFs/Improved_PN_Regs.pdf
http://waltjung.org/Regs.html
Have you seen the later articles?madstamm said:- bootstrap circuit as suggested by W.Jung in EDN June/97, "Low Noise Power for Analog Circuits"
http://waltjung.org/PDFs/Improved_PN_Regs.pdf
http://waltjung.org/Regs.html
BC5xx types and BD139 types will be just fine.AndrewT said:Hi,
can superreg be built up from alternative transistors?
Is D44 easily available in UK?
regards Andrew T.
I have used BC550C, BC560C, BC850C, BC860C, BCP56, BCP53, BD139, BD140, D44H11 and D45H11.
peranders,
thanks for the tips!
I am aware of the newer articles and have read them. I also realize the ST NPN as not being optimal, I did mention that.
I will try to source a BCP56 or similar speed/gain unit as soon as i can. I am interested in SMD (always, through hole is a bit dated, IMHO , but would need a package capable of more power, ie TO263 or similar...
end load would be in the range of 1A, driving several sigmadelta ADCs and their respective preamps. (16bits at 2MHz)
gruess,
M
thanks for the tips!
I am aware of the newer articles and have read them. I also realize the ST NPN as not being optimal, I did mention that.
I will try to source a BCP56 or similar speed/gain unit as soon as i can. I am interested in SMD (always, through hole is a bit dated, IMHO , but would need a package capable of more power, ie TO263 or similar...
end load would be in the range of 1A, driving several sigmadelta ADCs and their respective preamps. (16bits at 2MHz)
gruess,
M
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
- Home
- Design & Build
- Parts
- 5 volt version of the Updated super regulator.