What rail voltage, bias and devices did you use?
PMA seems quite taken with the original and has some nice testing to back it up. Hugh likes it too!
I already pulled it apart a year ago, it was a JLH 69 kit off ebay. I think my rails were arround 25-28V. The amp wasn't terrible but compared to my pass zen v1 it lost out. I might even have the boards arround somewhere if you want me to look for them.
Assesears AFAIK the dual rail version is the 96 revision.
Last edited:
I have to say that when I first finished my JLH 1969 I was a little disappointed. Mine is powered by a 25V 42A Server power supply. I threw it on the shelf for a while. Then after I built an A/B setup I started going through all my amps and when I got back to the JLH I was quite surprised to find that I really like it. I has a nice full bodied sound with sweet highs and tight bass. I was also surprised how loud 10W can get. Mine is nothing fancy. Built on the Jim's Audio boards. I did substitute MJ15003 for the 3055.
Attachments
Hi, all.
I would like to build this amplifier and, not being a electrician, I have some questions.
I've assumed that in the schematic all transistors have the emitter down and the collector up. Looking at some pcb's I've noticed that on Q4 and Q8 the emitter in up and the collector is down. Is that right, or am I not understand something.
Where can I integrate a volume control in the schematic? I'm thinking bout this volume control. And what values should it have?
Thanx.
I would like to build this amplifier and, not being a electrician, I have some questions.
I've assumed that in the schematic all transistors have the emitter down and the collector up. Looking at some pcb's I've noticed that on Q4 and Q8 the emitter in up and the collector is down. Is that right, or am I not understand something.
Where can I integrate a volume control in the schematic? I'm thinking bout this volume control. And what values should it have?
Thanx.
Last edited by a moderator:
You need to understand and appreciate the difference between NPN and PNP transistors.
Look at the arrow on the emitter. It points outward for NPN devices and inward for PNP. So the PNP devices with the emitter at the top are correct because for a PNP the emitter is biased to be more positive than the collector. The NPN is the opposite, the collector is the most positive.
Your volume control is fine. That picture is from ESP and is subject to copyright. All the info is here
(and a 10k or 22k pot is ideal)Potentiometers (Beginners' Guide to Pots)
Thank you for your quick answer. I've read the link, but like always, reading calls for more reading. So I'll try to give to you all the data so you could give me the right solution. The resistance measured on my source is 19.5kohms and I could change that to what is needed if you teach me how. Then I'll have the above amplifier, and somewhere in between I need to insert a volume control. Please give me the right solution to my issue. Thanx.
So I'll try to give to you all the data so you could give me the right solution. The resistance measured on my source is 19.5kohms and I could change that to what is needed if you teach me how. Then I'll have the above amplifier, and somewhere in between I need to insert a volume control. Please give me the right solution to my issue. Thanx.
The input impedance of the amp in your link is around 50k. A good choice of volume control would be a 10k pot providing your preamp or source component can drive that. Most solid state would have no problem with that.
I'm not sure what your comment of 19.5k refers too
The volume control can be either a normal log pot (log type is used for audio because it matches the ears non linear amplitude response and so makes the progression of sound seem constant as you turn the control).
Linear pots can be better matched between channels though, and that is where that extra resistor comes in. You can experiment with that resistor value, anything from around 1k8 upwards and see what "feels" the best as the pot is rotated. The resistor has no effect on the sound, only the sound vs rotation angle of the control.
I'm not sure what your comment of 19.5k refers too
The volume control can be either a normal log pot (log type is used for audio because it matches the ears non linear amplitude response and so makes the progression of sound seem constant as you turn the control).
Linear pots can be better matched between channels though, and that is where that extra resistor comes in. You can experiment with that resistor value, anything from around 1k8 upwards and see what "feels" the best as the pot is rotated. The resistor has no effect on the sound, only the sound vs rotation angle of the control.
Thank you. I've been under the illusion that 19.5k (the resistance measured between ground and signal of the source) is the source impedance. If not, please explain to me. Bout the 50k input impedance of the amplifier in the schematic I have no respect if the amplifier has no respect also. If the amplifier don't care bout it, it can be 200k or whatever. One thing that I don't understand is how input impedance can be 50k if i bring in parallel with it a 1k8 resistor and a 10k variable resistor. I understand how the variable resistor will work depending of the source output impedance. I do not understand who is judging and restrict the input impedance of the amplifier. Is that the Q4? What if I put a 100k variable resistor fallowed of a 20k that could replace the 22k resistor for the logarithmic simulation of the pot and, to say, 200k input impedance of the amplifier? Sorry for this stupid questions, but trying to skip past years of learning I was unable to find answers for this kind of questions. Thanx.
Source impedance. You can't measure this on a meter (not by measuring resistance anyway).
If you don't have an oscilloscope then you can get an idea of source impedance as follows. You need a test tone to play through the device being tested. Lets say you played a 0db 1kHz tone on a CD player and measured 2.2 volts rms at the player output (so that's measuring AC voltage on your meter). If you now load the output with a resistor, and keep dropping that resistor until the output falls to halve its value (so you now see 1.1 volts AC), then the output impedance of the player would equal the value of the resistor you have shunted the output with. Any CD player is going to have a low output impedance and so to any active preamplifier.
When you add a volume control then ideally you want the control to present only a light load to the source. So if the source impedance were high at say 50kohm (which a valve preamp might be) then a 10k pot would be to low. You would need something like a 200k or higher in that case.
For your amp (and most solid state amps) then using a high value pot isn't a good idea because the input capacitance of the amp will form a filter as it combines with that high value of resistance. That will have the effect of reducing the treble.
I don't think you have anything to worry over, use a 10k pot and it will be fine
If you don't have an oscilloscope then you can get an idea of source impedance as follows. You need a test tone to play through the device being tested. Lets say you played a 0db 1kHz tone on a CD player and measured 2.2 volts rms at the player output (so that's measuring AC voltage on your meter). If you now load the output with a resistor, and keep dropping that resistor until the output falls to halve its value (so you now see 1.1 volts AC), then the output impedance of the player would equal the value of the resistor you have shunted the output with. Any CD player is going to have a low output impedance and so to any active preamplifier.
When you add a volume control then ideally you want the control to present only a light load to the source. So if the source impedance were high at say 50kohm (which a valve preamp might be) then a 10k pot would be to low. You would need something like a 200k or higher in that case.
For your amp (and most solid state amps) then using a high value pot isn't a good idea because the input capacitance of the amp will form a filter as it combines with that high value of resistance. That will have the effect of reducing the treble.
I don't think you have anything to worry over, use a 10k pot and it will be fine
I have not used the half voltage method, but have seen it described by a few Members.
I don't like it !
I suspect that many Sources will reach current limiting by the time the load resistance is down to the value of the Source's output impedance.
eg.
Source has a 200r output impedance.
Test signal into open circuit is 1Vac.
The test resistor would need to be 200r for the half voltage method.
Add that 200r to the output and the Sources sees a total impedance of 400ohms (it's own 200ohms plus the 200r test load)
The peak current is 2Vac /400ohms * sqrt(2) for sinewave test signal.
Ipk from the output is ~ 7mApk.
I wonder how many Sources can give an undistorted 7mApk.
A single ended EF buffer biased to 10mA would be struggling with a 200r test load.
I don't like it !
I suspect that many Sources will reach current limiting by the time the load resistance is down to the value of the Source's output impedance.
eg.
Source has a 200r output impedance.
Test signal into open circuit is 1Vac.
The test resistor would need to be 200r for the half voltage method.
Add that 200r to the output and the Sources sees a total impedance of 400ohms (it's own 200ohms plus the 200r test load)
The peak current is 2Vac /400ohms * sqrt(2) for sinewave test signal.
Ipk from the output is ~ 7mApk.
I wonder how many Sources can give an undistorted 7mApk.
A single ended EF buffer biased to 10mA would be struggling with a 200r test load.
It gives a good idea I think although the results are open to being interpreted correctly. I suggested a high level simply because the resolution of many DVM's is poor at low levels.
If you load an output with 10k and it dramatically reduces then you need to think twice about what you are doing, if there is no change with say 1k loading then there is no problem.
If you load an output with 10k and it dramatically reduces then you need to think twice about what you are doing, if there is no change with say 1k loading then there is no problem.
Only slightly off topic - has anyone tried the Toshiba 2SC5200 in the JLH?
It seems to be rather linear, at least w.r.t. hFE vs Ic (nominally 100, Ft 30MHz)
It seems to be rather linear, at least w.r.t. hFE vs Ic (nominally 100, Ft 30MHz)
A 1k0 test load brings the total load to 1200ohms for the example I gave in post2932
That reduces the current to ~2.4mApk.
If a Source cannot drive 2.4mApk into a test load when biased into single ended ClassA then it does not deserve to be used as a Source.
But 200r and 1000r are very far apart in the stress they apply to a Source.
It is not reasonable to assume that a Source that can be loaded with 10k and cable capacitance will properly drive a 200r test load.
That reduces the current to ~2.4mApk.
If a Source cannot drive 2.4mApk into a test load when biased into single ended ClassA then it does not deserve to be used as a Source.
But 200r and 1000r are very far apart in the stress they apply to a Source.
It is not reasonable to assume that a Source that can be loaded with 10k and cable capacitance will properly drive a 200r test load.
Last edited:
Worth trying.
Be prepared to working out, if and what stability compensation may be required.
G'Day,
I tried onsemi's MJL4281 same sort of spec in the 96 version all I got was oscillatation even after compensation, base resistors etc & gave up with it. Stuck them in 69 dual rail version and all is good. Not tried 2SC5200 though, i believe alot of fakes out there
Regards
Bruce
I vaguely recalled that someone has built the high power version using 2SC5200 (it might be John65B) without problem
Regards,
yup. that was me. I have not listened to it in a few months, but will be rehousing it shortly....no problems if I remember correctly...it is the high power version - two pairs of 2SC5200 (not fakes, I think) per channel...when rehoused, it will have +/-34VDC dual rails with single 570VA Piltron toroid this time, while previously it was +/-24VDC 400VA toroid...
Last edited:
Thanks all. Sounds like a path worth pursuing further.
Agree that'll need checking. Now for Tr3
Last edited:
Sorry to disturb you with few more questions, but I'm in need of answers to finish the BOM and PCB.
1. Where is the place for the pot into the schematic? Can I combine R2 with the resistor after the pot into one resistor?
2. I understood that Q3 should be BD139-16 for more gain. What about the Q8?
3. I understood that Q8 is in need of a heat sink of 6C/W. Should it be more or less? I've seen on others PCB's that Q3 also has a heat sink. What thermal resistance should it have?
4. I've found
Manufacturer Part Number PF2205-0R33F1
Description RES 0.33 OHM 50W 1% TO-220 for R10. Do you think 'tis in need of a heat sink, what thermal resistance? - RESOLVED
5. Using the capacitance multiplier power supply with BD139, BD140, TIP2955, TIP3055. Transistors are in need of heat sinks? What thermal resistance?
Thank you for all your help.
1. Where is the place for the pot into the schematic? Can I combine R2 with the resistor after the pot into one resistor?
2. I understood that Q3 should be BD139-16 for more gain. What about the Q8?
3. I understood that Q8 is in need of a heat sink of 6C/W. Should it be more or less? I've seen on others PCB's that Q3 also has a heat sink. What thermal resistance should it have?
4. I've found
Manufacturer Part Number PF2205-0R33F1
Description RES 0.33 OHM 50W 1% TO-220 for R10. Do you think 'tis in need of a heat sink, what thermal resistance? - RESOLVED
5. Using the capacitance multiplier power supply with BD139, BD140, TIP2955, TIP3055. Transistors are in need of heat sinks? What thermal resistance?
Thank you for all your help.
Last edited: