I'm using 2SC3421's for some months. I built two separate power supplies for each channel with 2x 90W EI transformers. I replace all electrolyrics with nichicon muse es, all supply decoupling caps with nichicon fine gold, input capacitor with 1uF paper in oil, highpass filter capacitor with polystyrene ones. I used 1% metal film resistors everywhere too.
I bulit rod elliot's doz preamp as a preamplifier by using separate transformer and regulator for it.
I built rod elliot's DC protection and muting circuit to protect my speakers in case of failure.
amp is set to ~8W@6ohm (it was 10W@6ohm but i'm not listening loud so it was unnecessary to run outputs too hot) output transistors are between 50-55C when ambient temperature is around 22-23C.
I'm content with the results. This is my first hi-fi grade amplifier and love it.
I bulit rod elliot's doz preamp as a preamplifier by using separate transformer and regulator for it.
I built rod elliot's DC protection and muting circuit to protect my speakers in case of failure.
amp is set to ~8W@6ohm (it was 10W@6ohm but i'm not listening loud so it was unnecessary to run outputs too hot) output transistors are between 50-55C when ambient temperature is around 22-23C.
I'm content with the results. This is my first hi-fi grade amplifier and love it.
I have the need for a unity gain output buffer to drive 8 ohms. I'd like to use a JLH but don't want to waste the gain. Has one been made into a buffer, or is there maybe a way I can strip the first stage or maybe even eliminate the driver and drive an output stage directly?
Why not just use a Taylor source follower if you only need a buffer ?
(I assume you want to stay with single device types.)
http://www.tubecad.com/2004/blog0023.htm
Patrick
(I assume you want to stay with single device types.)
http://www.tubecad.com/2004/blog0023.htm
Patrick
Last edited:
Most up to date JLH Class A - High Power Circuit?
After lurking here for almost a week, i've read trough this whole thread, gatheried info, builds, suggestions, etc... I guess it's time i say Hi! :-D
It's been a while since i "played" with electronics. I studyed electronics in 94, and we build the JLH with 2n3055's as a project back then. I gave it away, but i always regretted it so i think it's time i make another. 🙂
Most, if not all my knowledge about electronics are forgotten, also the most advanced equipment i got, is a soldering station and a multimeter. I consider this a fun way to pick up electronics as a hobby again. 🙂
From what i've found, this is the most recent update to the JLH Class A - High Power Circuit? (Figur 3)
The Class-A Amplifier Site - JLH Class-A Update
And figur 2 is the most recent "normal" one?
I can't wait. I'm getting the PCB's etched today, using the Gerber files by Geoff from 2005.
I've read a lot of helpful posts by so many of you, you have already helped me tons 🙂
Ben.-
After lurking here for almost a week, i've read trough this whole thread, gatheried info, builds, suggestions, etc... I guess it's time i say Hi! :-D
It's been a while since i "played" with electronics. I studyed electronics in 94, and we build the JLH with 2n3055's as a project back then. I gave it away, but i always regretted it so i think it's time i make another. 🙂
Most, if not all my knowledge about electronics are forgotten, also the most advanced equipment i got, is a soldering station and a multimeter. I consider this a fun way to pick up electronics as a hobby again. 🙂
From what i've found, this is the most recent update to the JLH Class A - High Power Circuit? (Figur 3)
The Class-A Amplifier Site - JLH Class-A Update
And figur 2 is the most recent "normal" one?
I can't wait. I'm getting the PCB's etched today, using the Gerber files by Geoff from 2005.
I've read a lot of helpful posts by so many of you, you have already helped me tons 🙂
Ben.-
Member
Joined 2009
Paid Member
I think figure 2 is the most advanced, the constant current sources are a big technical improvement to the design together with modern device choices.
I nearly built one but found that my heatsinks were too small, aaargh!
One thing you'll find very handy if you get bitten by the audio bug is some kind of oscilloscope. I have an old one, but it's invaluable in setting up and testing a new amplifier. You maybe able to pick up an old one used.
I nearly built one but found that my heatsinks were too small, aaargh!
One thing you'll find very handy if you get bitten by the audio bug is some kind of oscilloscope. I have an old one, but it's invaluable in setting up and testing a new amplifier. You maybe able to pick up an old one used.
Hi to all, i am starting to build the JLH 1969. Can anybody tell me if i need a small heatsink for the 2N1613? In the article J.L. Hood says a finned heatsink of 120 x 100 mm will do for the power transistors (i will use RCA 16622 -matched 2N3055 for low VCEsat-)
My heatsinks are of exactly that size. Will they be okay? Thx for your answers.
Best regards Thorsten
My heatsinks are of exactly that size. Will they be okay? Thx for your answers.
Best regards Thorsten
That seems rather on the small size to me... the bigger the better 😀
I use fan cooling on my 25WPC version and it stays at about 50C.
A small heatsink on the "driver" would be a good idea yes.
I use fan cooling on my 25WPC version and it stays at about 50C.
A small heatsink on the "driver" would be a good idea yes.
Simply eliminate the input transistor and trade the npn vas transistor for a pnp.
If you remove active devices you will have to look at all dc conditions again.
You could remove the negative feedback components from the input Tr to ground R6 and C4 . That will make it a unity gain buffer . However you might have to check it's stability . I tried my sim on the design with other output transistors and it appears to be stable with unity gain. But you need to check it out in practice as the spice models may not be accurate and give wrong results.
Note Q1 Vce must be able to handle the full supply voltage . So for a +/- 30 volt supply it should handle more than 60 volts across collector and emitter to avoid breakdown!
You could remove the negative feedback components from the input Tr to ground R6 and C4 . That will make it a unity gain buffer . However you might have to check it's stability . I tried my sim on the design with other output transistors and it appears to be stable with unity gain. But you need to check it out in practice as the spice models may not be accurate and give wrong results.
Note Q1 Vce must be able to handle the full supply voltage . So for a +/- 30 volt supply it should handle more than 60 volts across collector and emitter to avoid breakdown!
Attachments
Last edited:
Hi, everyone, this is my first post to this thread. Last year I constructed 3 class A amplifiers to run listening comparison tests. These were Hiraga 20W Krell KSA50 ( which I limited to 25W ) and JLH 2005. I used the same power supply for all of them, which was 100,000micro farad resovoir followed by a capacitance multiplyer for both + and - rails. the voltage was +-22V.
I decided from listening that the amplifier I would be most happy with was the JLH.
Unfortunately for personal reasons I had to abandon my hobby, but I am now in the position to start again with the final product. I need some advice regarding the power supply.
Should I stay with the capaciance multiplyer.
Should I use the series regulator from 2005
Should I use a simple C L C.
Any advice or views would be much appreciated.
Thanks
Alan.
I decided from listening that the amplifier I would be most happy with was the JLH.
Unfortunately for personal reasons I had to abandon my hobby, but I am now in the position to start again with the final product. I need some advice regarding the power supply.
Should I stay with the capaciance multiplyer.
Should I use the series regulator from 2005
Should I use a simple C L C.
Any advice or views would be much appreciated.
Thanks
Alan.
Hi Alan,
KSA50 running +/-45~50V on rails will be much much better than +/-22V. I'm running +/-48V, 650mA per transistor. It is hot, but it is worth to.
On my own idea, for an amplifier, a bigger transformer is rather better than more capacitance. For instance, my machine has 28000uf per rail, but transformer has about 500W.
CLC is good. But inductance is expensive normally. So I use a CRC, i.e. 14000uf+0.47Ohm 10W+14000uf. That will be also a good solution.
Hope above helps. Have fun. 😉
Thanks la Mer for the reply. I purposely built the KSA50 to run on +-22v to keep heat dissapation down to an acceptible level ( the people on the KSA forum were very helpful in guiding me through component alterations to suit the lower voltage). I found the Krell to be very easy to set up with offset voltage at near zero and not altering during warm up. The Hiraga seemed to be the most sensitive to offset problems, moreso than the JLH.
I tried to be subjective when comparing, and I must say I liked them all, but there was something about the JLH that made me think i could listen to it for a long time. ( just me I expect )
I will start building the final version after Christmas and I think I will go with the C R C or C L C unless someone can persaude me differently.
Thanks for your help.
Alan
I tried to be subjective when comparing, and I must say I liked them all, but there was something about the JLH that made me think i could listen to it for a long time. ( just me I expect )
I will start building the final version after Christmas and I think I will go with the C R C or C L C unless someone can persaude me differently.
Thanks for your help.
Alan
Hi Alan, It's ok because I'm listening to KSA50 and JLH too for several years. But finnally my JLH was made into 3W not 10W at this subject #1802 and #1807, to fit desktop purpose. My 10W JLH was stated at #1813.
I didn't make an OCL 1969 finnally, becase for myself it sounds not same sweet as OTL.
But anyway, maybe you will find a way to make OCL1969 same sweet as OTL one.
la Mer 😀
Anyone build the JLH High Power with Toshiba 2SA1943/2SC5200 output devices??
I have the KSA50 Clone and it is my reference. It as one great amp.
I have the KSA50 Clone and it is my reference. It as one great amp.
Hello everyone and a happy new year.
Does anyone know if there is any advantage in using non inductive resistors for the emmiters of the output transistors instead of standard wirewound resistors. Any views or comments would be much appreciated.
Thanks
Alan
Does anyone know if there is any advantage in using non inductive resistors for the emmiters of the output transistors instead of standard wirewound resistors. Any views or comments would be much appreciated.
Thanks
Alan
Non inductive resistors are recommended for emitter resistors but I have never had a problem uisng generic wirewound resistors in that location. Some circuits with very wideband performance might need 'genuine' non inductive resistors.
I'd say don't make the circuits so 'wideband' if the use is for audio signals and standard wirewounds would probably be good enough in that case.
Cheers.
metal film type resistors ( expensive and harder to get in small quantities here ) probably are better than "non-inductive wirewounds ".
I'd say don't make the circuits so 'wideband' if the use is for audio signals and standard wirewounds would probably be good enough in that case.
Cheers.
metal film type resistors ( expensive and harder to get in small quantities here ) probably are better than "non-inductive wirewounds ".
Last edited:
I think this amp sounds good because it is wideband.
I think regular metal film resistors sound dreadful especially in this kind of high current position - I have tried and I really not like.
I would strongly recommend a good non-inductive wire wound like mills etc
Myself I used a short length of suitable value resistance wire ( constantan wire ) - a bit fiddly but very cheap & sounds great
cheers
mike