Hello Apassgear, thanks again for your help. OK, since I do not have a reason to hide my lack of electronical knowledge I will ask this question: is regulated PSU needed for this particular preamplifier? I just do not have a ......feeling that it is necessary...
I see it to be used in headphone amplifier, which may deal with range of headphone impedances ( <32 and up to 600 Ohms). I would like to use the preamp with set of amplifiers like: Gainclone ( inverted version), Hiraga LeClasse 20W, and Pavel Macura Mosfet Power Follower ( http://www.pha.inecnet.cz/macura/follower_e.html ).
Could I escape regulated PSU idea without lowering quality of the signal? Would it be possible to maintain needed current supply with use of slightly higher capacitance? ( I still think about using mentioned above R_Core transformers).
I see it to be used in headphone amplifier, which may deal with range of headphone impedances ( <32 and up to 600 Ohms). I would like to use the preamp with set of amplifiers like: Gainclone ( inverted version), Hiraga LeClasse 20W, and Pavel Macura Mosfet Power Follower ( http://www.pha.inecnet.cz/macura/follower_e.html ).
Could I escape regulated PSU idea without lowering quality of the signal? Would it be possible to maintain needed current supply with use of slightly higher capacitance? ( I still think about using mentioned above R_Core transformers).
apassgear said:
Here are some answers from post 125.
The circuit is quite susceptible to power supply quality. The original uses a 3 leg regulator something on the line of an LM317 and to my thoughts that would be the minimum requirement for this preamp. I don’t think you will like it without regulation.
The LM317 is easy to setup but you would need a DCV at least of 28VDC before the regulation to reach a 24V rail.
Thanks, I want to make external PSU. I will start with what I have. If it will not work, I can always change trafos. Those are R-cores which I have ( can not beat the price):
http://www.herbach.com/Merchant2/me...AR&Product_Code=TM98XFR3088&Category_Code=XFR
http://www.herbach.com/Merchant2/me...AR&Product_Code=TM98XFR3088&Category_Code=XFR
OK, something just occured to me. Please, let me know if I am mistaken. Since those trafos have dual secondaries, I can wire them in series what would give me 26.5 V ( 19 + 7.5 ). After rectification, 37.36 Volts. But, in this setup current rating would be limited to the lower voltage, 7.5 V, which is 1.5 Amps ( 19 V secondaries are rated at 2 Amps).
Am I correct? Is 1.5 Amps output current enough? And if this is correct, What I have to do to bring it safely to 24 Volts?
There are three reasons why I want to use R-cores: I have them
, no problem with in-rush current, and they are really very quiet ( they are also small ).
Thanks
Am I correct? Is 1.5 Amps output current enough? And if this is correct, What I have to do to bring it safely to 24 Volts?
There are three reasons why I want to use R-cores: I have them
, no problem with in-rush current, and they are really very quiet ( they are also small ).Thanks
Hi TB,
Nice trafos
The circuit itself will only demand about 20mA each channel plus whatever the rail reg draw you decide, so 1.5A is enough for the stereo set.
As for the secondary voltage of the trafo is usually easy to change, but depends if you would like to do it and the way the coils were done. First check, by peaking beneath the outer shield paper and see if secondaries were wound on top of the primaries. Its an interesting thing to do and we can learn something on the way. Let me know.
Nice trafos
The circuit itself will only demand about 20mA each channel plus whatever the rail reg draw you decide, so 1.5A is enough for the stereo set.
As for the secondary voltage of the trafo is usually easy to change, but depends if you would like to do it and the way the coils were done. First check, by peaking beneath the outer shield paper and see if secondaries were wound on top of the primaries. Its an interesting thing to do and we can learn something on the way. Let me know.
The schematic shown on post 140 is what members could retrieve from this classic lineamp… yes, it also included Nelson’s memories of his work (please also read the foot note on the schematic).
One other refinement to the circuit that has come my way is to change C2 with a more up to date approach. Most of the caps used in those days were tantalum type which were regarded as best for this application but today these are not recommended in audio… even as Nelson has mentioned… so, instead of using a lytic on that place, as implied, I would suggest the use of a good film cap for C2 (MKP would be my choice) in the order of 0.22/0.47 uf.
The benefit of this change could be quite audible so go ahead and test it. Steen could be the man to repot any findings since I’m still dragging my feet’s.
One other refinement to the circuit that has come my way is to change C2 with a more up to date approach. Most of the caps used in those days were tantalum type which were regarded as best for this application but today these are not recommended in audio… even as Nelson has mentioned… so, instead of using a lytic on that place, as implied, I would suggest the use of a good film cap for C2 (MKP would be my choice) in the order of 0.22/0.47 uf.
The benefit of this change could be quite audible so go ahead and test it. Steen could be the man to repot any findings since I’m still dragging my feet’s.
Question about apassgear's shunt reg (post 13)
Hallo folks,
May I ask a few questions about apassgear's shunt regulator schematic (post # 13)? Apparently, this circuit has not been tested yet and the schematic drawing seems to contain a few errors (or am I mistaken?)
1. Q1 is listed as a TIP30A. However, the TIP30A is a PNP transistor whereas the schematic shows a NPN transistor. I assume that Q1 should be a TIP31A, not a TIP30A. Am I correct?
2. Are the resistor values around the TL431A correct? According to the ON Semi datasheet, these resistor values would result in a cathode-anode voltage difference of 27.5 V for the TL431A which seems too high. Vref = 2.495 V, Iref = 1.8 mA, Vka = Vref*(1+R4/R5)+R4*Iref. For an output voltage of 24V, I would have expected a lower value for Vka, perhaps 20V or 22V. Can you answer this question?
3. Finally, I wondered if the polarities of the LEDs and of Q2 were correctly drawn. They may be correct, but I am not sure.
Thanks in advance for your answer!
PS I have already built one channel of the preamp, using apassgear's PCB design and this works great ! (I used low noise transistors from the German Democratic Republic, SC239 and SC309, and a LM301 op-amp).
Keep your ears happy and your soldering irons smoking.
arensattic (Aren van Waarde)
Hallo folks,
May I ask a few questions about apassgear's shunt regulator schematic (post # 13)? Apparently, this circuit has not been tested yet and the schematic drawing seems to contain a few errors (or am I mistaken?)
1. Q1 is listed as a TIP30A. However, the TIP30A is a PNP transistor whereas the schematic shows a NPN transistor. I assume that Q1 should be a TIP31A, not a TIP30A. Am I correct?
2. Are the resistor values around the TL431A correct? According to the ON Semi datasheet, these resistor values would result in a cathode-anode voltage difference of 27.5 V for the TL431A which seems too high. Vref = 2.495 V, Iref = 1.8 mA, Vka = Vref*(1+R4/R5)+R4*Iref. For an output voltage of 24V, I would have expected a lower value for Vka, perhaps 20V or 22V. Can you answer this question?
3. Finally, I wondered if the polarities of the LEDs and of Q2 were correctly drawn. They may be correct, but I am not sure.
Thanks in advance for your answer!
PS I have already built one channel of the preamp, using apassgear's PCB design and this works great ! (I used low noise transistors from the German Democratic Republic, SC239 and SC309, and a LM301 op-amp).
Keep your ears happy and your soldering irons smoking.
arensattic (Aren van Waarde)
Aren
Great to hear you have built the circuit and that it runs without problems.
As far as I can remember Steen (Steenoe) built this preamp using the shown PCB and the shunt reg.
Yes, the schematic for the shunt reg has some errors and you are correct in pointing that Q1 is a NPN (TIP31A). I still have not tested the circuit so adjustment has to be made to the resistors around the TL431 to set the correct voltage at the output (24V).
Steen did found some problems setting the voltage reference (TL431) and I don’t know which the actual difficulty was but he replaced it with a simple zener and worked well as far as I know. But I know the TL can be made to work as intended. Maybe Steen will chip in and tell us what happened.
Regarding the LED’s I think they are correctly polarized as shown in the schematic.
I’m encouraging you to build this reg, since it should work very well with this line amp. Here are some tips on how I would fine tune it to the preamp:
- Build the circuit with the resistors you have calculate for setting the 431 and test it with a dummy load instead of connecting it to the line amp. Since the estimated current draw of one preamp circuit is around 25 mA connect a 1K resistor as load, assuming you will use one regulator per channel.
- Connect the raw DC to the input, 29/30V and check the voltage at the output between the leads of the dummy load resistor and adjust the biasing resistors –R4/5- to the 431 if needed to get the 24 or so volts.
- Next we need to know how much current the whole thing is drawing, for this we would check the voltage drop on R2. I would shoot to something like 50 to 80 mA.
Would love to see some pics of the populated PCB if possible
P.S. R2 should have 1 watt rating to be on the safe side.
Great to hear you have built the circuit and that it runs without problems.
As far as I can remember Steen (Steenoe) built this preamp using the shown PCB and the shunt reg.
Yes, the schematic for the shunt reg has some errors and you are correct in pointing that Q1 is a NPN (TIP31A). I still have not tested the circuit so adjustment has to be made to the resistors around the TL431 to set the correct voltage at the output (24V).
Steen did found some problems setting the voltage reference (TL431) and I don’t know which the actual difficulty was but he replaced it with a simple zener and worked well as far as I know. But I know the TL can be made to work as intended. Maybe Steen will chip in and tell us what happened.
Regarding the LED’s I think they are correctly polarized as shown in the schematic.
I’m encouraging you to build this reg, since it should work very well with this line amp. Here are some tips on how I would fine tune it to the preamp:
- Build the circuit with the resistors you have calculate for setting the 431 and test it with a dummy load instead of connecting it to the line amp. Since the estimated current draw of one preamp circuit is around 25 mA connect a 1K resistor as load, assuming you will use one regulator per channel.
- Connect the raw DC to the input, 29/30V and check the voltage at the output between the leads of the dummy load resistor and adjust the biasing resistors –R4/5- to the 431 if needed to get the 24 or so volts.
- Next we need to know how much current the whole thing is drawing, for this we would check the voltage drop on R2. I would shoot to something like 50 to 80 mA.
Would love to see some pics of the populated PCB if possible
P.S. R2 should have 1 watt rating to be on the safe side.
Pictures will come eventually
Hallo Apassgear,
Thanks for your kind advice. I have just etched and drilled the PCB for the shunt reg. Hopefully I can do the experiments which you suggest next Tuesday (when I have a day off from work ).
If both the preamp and PSU boards are working I will post some pictures of them.
There is one thing about the preamp board which I do not quite understand: the current draw is a lot lower than you suggest, about 14 mA at 24V. This agrees with my PSPICE simulation, though. Measurement-wise the preamp board is flawless, even 100kHz squarewaves look very good. I have not done any listening tests yet, that will come later after I have made a stereo version.
Will keep you informed.
Aren van Waarde aka Arensattic
Hallo Apassgear,
Thanks for your kind advice. I have just etched and drilled the PCB for the shunt reg. Hopefully I can do the experiments which you suggest next Tuesday (when I have a day off from work
).If both the preamp and PSU boards are working I will post some pictures of them.
There is one thing about the preamp board which I do not quite understand: the current draw is a lot lower than you suggest, about 14 mA at 24V. This agrees with my PSPICE simulation, though. Measurement-wise the preamp board is flawless, even 100kHz squarewaves look very good. I have not done any listening tests yet, that will come later after I have made a stereo version.
Will keep you informed.
Aren van Waarde aka Arensattic
Aren,
As you know I have not built the circuit but if you look to this and other NS10 threads you will see that we had a gross estimate of around 20mA of current draw per channel, using different transistors could be somewhat else.... anyway, I think that 14mA is in the ball park of the guess-estimated figure.
As you know I have not built the circuit but if you look to this and other NS10 threads you will see that we had a gross estimate of around 20mA of current draw per channel, using different transistors could be somewhat else.... anyway, I think that 14mA is in the ball park of the guess-estimated figure.
apassgear said:Aren and All,
There are issues related to the Shunt Reg, will make some tests hopefully during next week and report back.
I apologize for this inconvenience specially to you Aren and Steen that I know have etched boards.
Not to worry, Tony.
The shuntreg had some problems, mainly Q1 as mentioned, and the precision reg TL431. I changed the TL and the related resistors to a zener, and that worked fine. I didnt try to use the TL again later, but if you guys make it work, I will build it again.
If the board is re-designed, I also found that there was a tad too little room for the big smoothing cap.
Steen
Tests on Shunt Reg: Found other errors
Hallo Apassgear and Steen,
Today I have tinkered with the proposed Shunt Reg circuit and run some tests. When populating the Shunt Reg board, I spotted a few more errors:
a. The position of the TL431 on the board is not shown correctly. The pinout of the 3-pin version of this i.c. when you look at it from the top and the flat part of the i.c. points upwards is, from left to right, cathode (pin 1), anode (pin 2), and reference (pin 3). The cathode should go to node R3, R4, R6; the anode to ground and the reference to node R4, R5. See datasheet. When the i.c. is mounted correctly, it works well. It may be necessary to trim the values of R4 and R5 somewhat (e.g., to 4k7 and 560R) to get the correct output voltage.
b. The value of R2 seems too low. The current through the shunt transistor might increase to 5 (voltage drop) divided by 10 (ohmic resistance) = 0.5 Ampere with the proposed value. In reaility somewhat less since there will also be some voltage drop across Q1. I have done a few experiments both with a mosfet and with a bipolar (PNP) transistor as the shunt regulating element, and with a 10R value for R2, the components Q1, R2 and Q2 get blazingly hot. Therefore I propose to increase the value of R2 from 10R to 100R or 120R.
c. A feature of the schematic which I don't quite understand is the high value for R6 (i.e., 100K), which results in a very high time constant for the R6/C3 combination. With the values shown, it takes several MINUTES before the proper output voltage is reached. This is a problem, since as long as the output voltage is low, the current through Q1, R2 and Q2 is far above the operating value. I assume the value of R6 may be decreased, e.g. to 10K or even less, although this may affect the regulatory properties of the circuit (?). With a value of 10K for R6, the operating voltage is reached within 1 min.
d. Finally, I assume it is necessary to use at least a 5W resistor for R2 and a small heatsink for Q2 (perhaps also for Q1).
Will do additional tests in the future and keep you informed. I am not an electronic engineer, so hopefully I have not done anything stupid. Also I hope to have expressed myself clearly (English is not my native tongue).
Best regards,
Aren
Hallo Apassgear and Steen,
Today I have tinkered with the proposed Shunt Reg circuit and run some tests. When populating the Shunt Reg board, I spotted a few more errors:
a. The position of the TL431 on the board is not shown correctly. The pinout of the 3-pin version of this i.c. when you look at it from the top and the flat part of the i.c. points upwards is, from left to right, cathode (pin 1), anode (pin 2), and reference (pin 3). The cathode should go to node R3, R4, R6; the anode to ground and the reference to node R4, R5. See datasheet. When the i.c. is mounted correctly, it works well. It may be necessary to trim the values of R4 and R5 somewhat (e.g., to 4k7 and 560R) to get the correct output voltage.
b. The value of R2 seems too low. The current through the shunt transistor might increase to 5 (voltage drop) divided by 10 (ohmic resistance) = 0.5 Ampere with the proposed value. In reaility somewhat less since there will also be some voltage drop across Q1. I have done a few experiments both with a mosfet and with a bipolar (PNP) transistor as the shunt regulating element, and with a 10R value for R2, the components Q1, R2 and Q2 get blazingly hot. Therefore I propose to increase the value of R2 from 10R to 100R or 120R.
c. A feature of the schematic which I don't quite understand is the high value for R6 (i.e., 100K), which results in a very high time constant for the R6/C3 combination. With the values shown, it takes several MINUTES before the proper output voltage is reached. This is a problem, since as long as the output voltage is low, the current through Q1, R2 and Q2 is far above the operating value. I assume the value of R6 may be decreased, e.g. to 10K or even less, although this may affect the regulatory properties of the circuit (?). With a value of 10K for R6, the operating voltage is reached within 1 min.
d. Finally, I assume it is necessary to use at least a 5W resistor for R2 and a small heatsink for Q2 (perhaps also for Q1).
Will do additional tests in the future and keep you informed. I am not an electronic engineer, so hopefully I have not done anything stupid. Also I hope to have expressed myself clearly (English is not my native tongue).
Best regards,
Aren
Re: Tests on Shunt Reg: Found other errors
Yes, you have addressed most of the problems that are there which I didn't want to anticipate before testing and giving the new values.
Time constant for R6/C3 is very high, as you noted, R6 should be between 220R and 1K.
Q1 as a current limiter dose not work and apparently here is no way that will work with that topology. I'm not very inclined to to make R2 that big. If we want a current limiter at that position the best way out is to use LM317 IC reg biased for that purpose.
There seem to be variations on pinout for the 431 depending on manufacturer, as you noted, so the best way to make sure is downloading the corresponding data sheet. Also be aware that some 431 are quite noisy so care should be taken as for were you buy them and brands, I'm sure there are fake ones on the market too. A noisy 431 will ruin the party as a voltage ref even though we have an RC filter. Ditto for zeners on that voltage range, all quite noisy.
Yes Aren, please report your findings
P.S. I'm working on a different reg which should provide a better solution for this line amp.
arensattic said:Hallo Apassgear and Steen,
Today I have tinkered with the proposed Shunt Reg circuit and run some tests. When populating the Shunt Reg board, I spotted a few more errors:
a. The position of the TL431 on the board is not shown correctly. The pinout of the 3-pin version of this i.c. when you look at it from the top and the flat part of the i.c. points upwards is, from left to right, cathode (pin 1), anode (pin 2), and reference (pin 3). The cathode should go to node R3, R4, R6; the anode to ground and the reference to node R4, R5. See datasheet. When the i.c. is mounted correctly, it works well. It may be necessary to trim the values of R4 and R5 somewhat (e.g., to 4k7 and 560R) to get the correct output voltage.
b. The value of R2 seems too low. The current through the shunt transistor might increase to 5 (voltage drop) divided by 10 (ohmic resistance) = 0.5 Ampere with the proposed value. In reaility somewhat less since there will also be some voltage drop across Q1. I have done a few experiments both with a mosfet and with a bipolar (PNP) transistor as the shunt regulating element, and with a 10R value for R2, the components Q1, R2 and Q2 get blazingly hot. Therefore I propose to increase the value of R2 from 10R to 100R or 120R.
c. A feature of the schematic which I don't quite understand is the high value for R6 (i.e., 100K), which results in a very high time constant for the R6/C3 combination. With the values shown, it takes several MINUTES before the proper output voltage is reached. This is a problem, since as long as the output voltage is low, the current through Q1, R2 and Q2 is far above the operating value. I assume the value of R6 may be decreased, e.g. to 10K or even less, although this may affect the regulatory properties of the circuit (?). With a value of 10K for R6, the operating voltage is reached within 1 min.
d. Finally, I assume it is necessary to use at least a 5W resistor for R2 and a small heatsink for Q2 (perhaps also for Q1).
Will do additional tests in the future and keep you informed. I am not an electronic engineer, so hopefully I have not done anything stupid. Also I hope to have expressed myself clearly (English is not my native tongue).
Best regards,
Aren
Yes, you have addressed most of the problems that are there which I didn't want to anticipate before testing and giving the new values.
Time constant for R6/C3 is very high, as you noted, R6 should be between 220R and 1K.
Q1 as a current limiter dose not work and apparently here is no way that will work with that topology. I'm not very inclined to to make R2 that big. If we want a current limiter at that position the best way out is to use LM317 IC reg biased for that purpose.
There seem to be variations on pinout for the 431 depending on manufacturer, as you noted, so the best way to make sure is downloading the corresponding data sheet. Also be aware that some 431 are quite noisy so care should be taken as for were you buy them and brands, I'm sure there are fake ones on the market too. A noisy 431 will ruin the party as a voltage ref even though we have an RC filter. Ditto for zeners on that voltage range, all quite noisy.
Yes Aren, please report your findings
P.S. I'm working on a different reg which should provide a better solution for this line amp.
Zen Mod said:
you can always use positive half ofreg , with somewhat tweaked values
Oooooops Choky.... looks like a full amplifier... maybe good for pumpkins. Have you tested it?
I'm looking for something simple and better than the original which will preserve the spirit of the NS10. Holy grail? ... don't think so.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.