With 24V Zener you get about 24.7-25V. 28V you quote is fishy. That much goes to Q1's gate with about 33V DC In and accounting the Leds drop. Maybe you got Q1 wrongly wired. The Mosfets are G,D,S from left to right as you face them and you can read the part numbers. S is the arrow pin on the schematic. Please check everything again.
staggerlee said:
How about a 22v zener?
![]()
That will give about 23V, that Zener number shown was a leftover when drawing. 1N4749A is the correct part.
mikelm,
I am not condemming CFP's but if you notice not many of the worlds best designers use CFP's. Most of them have avoided them because of potential problems.
Sometimes getting a CFP to work requires a heroic amount of compensation(which usually damages it's sonic qualities) and I myself try to aviod any circuitry that is on the verge of instability even if my choice has slightly higher distortion.
Countless papers have been written on the subject but no one has come up wiyh a total solution. CFP's enclosed in feedback loops can be particularly problematic. CFP's have their place but in most cases they can be subsituted for another design element that sounded better (note I said sounded not measured)
Regards,
Jam
I am not condemming CFP's but if you notice not many of the worlds best designers use CFP's. Most of them have avoided them because of potential problems.
Sometimes getting a CFP to work requires a heroic amount of compensation(which usually damages it's sonic qualities) and I myself try to aviod any circuitry that is on the verge of instability even if my choice has slightly higher distortion.
Countless papers have been written on the subject but no one has come up wiyh a total solution. CFP's enclosed in feedback loops can be particularly problematic. CFP's have their place but in most cases they can be subsituted for another design element that sounded better (note I said sounded not measured)
Regards,
Jam
mikelm said:I use irf9530 successfully for current source - so far so good
I used two matched zvp3310a's in parallel for shunt.
had some oscillation - ran out of time - needs more work
I'll report if I get it working 🙂
Try a 100pF ceramic cap on the B-C pins of the NPN driver. Also, try different NPNs if you have them around. I successfully used 2SC828s, 2N3904, etc.
Complementary Feedback Pair, generally hates parallel current elements. Much more prone to oscillate. If suggestions like double stoppers, big output cap, and what Iko said fail, try a single heavier and slower Mosfet instead of the 2X TO-92 ZVPs.
jam said:CFP's have their place but in most cases they can be subsituted for another design element that sounded better (note I said sounded not measured)
Regards,
Jam
I agree for AF amps, where the distortion spectrum shape counts more than the CFP's lower %. For DC regulation, CFP mates well.
Hi Salas,
Checked and found out that I did not connect the emitter (BC550C) to the drain of IRFP9240. Other areas seemed to be correctly connected. However, the output now measures just over 12VDC 😕
I used the schematic on post #255 which is for 37VDC output. Do I need to make changes to component values on that schematic to get 24VDC?
Regards
Tham
Checked and found out that I did not connect the emitter (BC550C) to the drain of IRFP9240. Other areas seemed to be correctly connected. However, the output now measures just over 12VDC 😕
I used the schematic on post #255 which is for 37VDC output. Do I need to make changes to component values on that schematic to get 24VDC?
Regards
Tham
Are you sure that you don't have just a 12V Zener by mistake?
No, you don't need more changes than a DC In about 30-35V and a Vref of 24V. I.e. a 24V Zener or 2X12V in series.
No, you don't need more changes than a DC In about 30-35V and a Vref of 24V. I.e. a 24V Zener or 2X12V in series.
Hi Salas,
It is a 1N4749A. I use BC550 and I will just turn it the other way to see if it is the culprit.
It is a 1N4749A. I use BC550 and I will just turn it the other way to see if it is the culprit.
Built a Le Pacific MM RIAA clone and was thinking of powering it with your V1 shunt regulator design.
Will post photo once it is working.
Will post photo once it is working.
My scope just showed straight lines indistinguishable in thickness from its GND reference line.
P.S. James has made various V1s for current and Voltage, feeding his headphone amp (about half an Amp CCS), his lightspeed etc. He is chuffed as he wrote me and plans to give to Than for detailed examination that has access to some lab.
Salas,
I built a version of V1 and it was very quiet, I thought that the zener would compromise the noise performance buy apparently not. Same results with my scope.
Thanks,
Jam
I built a version of V1 and it was very quiet, I thought that the zener would compromise the noise performance buy apparently not. Same results with my scope.
Thanks,
Jam
If you up the Zener bypass cap to 1000uF, it will guard even better against its residual noise in case it is not the best Zener you could find. I had once some plastic cased 36V ones that upped the noise floor of the phono on FFT. 1000uF did better on them, but not better than the 100uF on the 1NXXXXA glass encapsulated ones. So I changed them.
Did you listen to it too? What you power?
Did you listen to it too? What you power?
Salas,
I have built several shunts before and I can honestly say that your design is one of the best if not the best that I have heard.
I only ran about 150ma through it and the current source was a bipolar with a led reference that was fed by a jfet current source.I like this better than the two transistor current source by a small amount. For higher currents the fet current is probably better.
I hear that temperature compensated zeners that are used in aerospace applications are quiter but they cost between $35 and $50 and have to be specially ordered. Talking to one of the engineers at the plant he let me know that a temp. compensated zener was made up of a zener and some reversed biased diodes in series.
Sonically your v1 sounds the most natural in the midrange and seems to have a nice balance to the sound and before I forget ,it is very quiet. I prefer it to any series regulator I have tried. Increasing the loop gain seems to improve the bass and extreme high end ( by a small amount) but at the expense of the midrange (not so good).
What brand of Zeners do you use?
Regards,
Jam
P.S. Sorry no pictures.
I have built several shunts before and I can honestly say that your design is one of the best if not the best that I have heard.
I only ran about 150ma through it and the current source was a bipolar with a led reference that was fed by a jfet current source.I like this better than the two transistor current source by a small amount. For higher currents the fet current is probably better.
I hear that temperature compensated zeners that are used in aerospace applications are quiter but they cost between $35 and $50 and have to be specially ordered. Talking to one of the engineers at the plant he let me know that a temp. compensated zener was made up of a zener and some reversed biased diodes in series.
Sonically your v1 sounds the most natural in the midrange and seems to have a nice balance to the sound and before I forget ,it is very quiet. I prefer it to any series regulator I have tried. Increasing the loop gain seems to improve the bass and extreme high end ( by a small amount) but at the expense of the midrange (not so good).
What brand of Zeners do you use?
Regards,
Jam
P.S. Sorry no pictures.
- Status
- Not open for further replies.
- Home
- Amplifiers
- Power Supplies
- The simplistic Salas low voltage shunt regulator