I think you have a resistor of wrong resistance in there somewhere - I did this once when I built my Krell Clone - I accidentally put in a 220k instead of 220 ohm. Did the same thing - ran cool - not hot (class A). Took me a while to find it, but eventually found it...swapped it and all was good.
I was pretty certain i measured all the resistors one by one while soldering them in place, and i even measured again later when i found out it didn't produce heat but i feel kinda lost, so i'm gonna measure again to be sure.
Gonna focus on just one channel though, until we find the problem.
Thanks.
Ben.-
Sure.
I decided to use a split supply of 15 - 0 - 15VDC. I just like a split supply more and I had a center tapped transformer available.
It's already posted twice.
Imageshack this time.
http://img59.imageshack.us/img59/8619/jlh1969pcb.jpg
Here are the voltages i measured on my amplifier.
http://img26.imageshack.us/img26/3775/jlh1969schematicwithvol.jpg
Looks like the V-E voltage on TR2 is a bit low?
The voltages you measure is correct, (except you actually have a 22K in the bias line and not 39K as indicated on the circuit). No unless the power transistor you use has almost zero gain, the quescient current flowing is 1.4Amp.
Now place your ammeter in series with the supply, if you are not measuring 1.4 amps then replace the power transistors with something that has some gain like a 2N3773.
Now place your ammeter in series with the supply, if you are not measuring 1.4 amps then replace the power transistors with something that has some gain like a 2N3773.
The schematic says "(22K stereo) 39K" and since this is a dual-mono build it should be a 39K resistor then?
I have measured the current. The amplifier draws 0,59A per channel, which is about half of what it should be if i am right?
Are you saying i should replace the resistor?
Ben.-
I am not sure what stereo or mono has to do with the bias current. It could be considering the power supply volt drop if two amps are fed, by one power supply, but that is besides the point.
Seeing a quiescent current of only 0.59A can only mean that the gain of the power transistors are less than half of what they ought to be. At the moment you are lucky to get about 4 watt rms into 8 ohm resistance from the system.
Change the 560 ohm resistor for 330 ohms if you are using 2N3055. Just make sure that the driver can handle the extra dissipation.
Seeing a quiescent current of only 0.59A can only mean that the gain of the power transistors are less than half of what they ought to be. At the moment you are lucky to get about 4 watt rms into 8 ohm resistance from the system.
Change the 560 ohm resistor for 330 ohms if you are using 2N3055. Just make sure that the driver can handle the extra dissipation.
I am using MJ15003 but i have some 2n3055's laying around. Is there a big difference in gain between 2n3055 and MJ15003?
The PCB came with 3773's i think, but they looked fake
Uncomfortable? It's electronics. What could possibly go wrong... :-D
I'll try that. Thanks again for continued help.
EDIT: Tried with a 390 Ohm resistor instead of the 560 one. (The 22K resistor is still not replaced by a 39K one though. Should i?) It slowly became just hot enough to notice that the power is on. Like the pillow in the couch right after you get up. Nowhere near what i remember from the one we buildt in 92-93.
But do you think the outputs could be the problem, not providing enough gain?
Ben.-
Last edited:
I feel warm! :-D
(and so does my amp)
The amplifier have now been powered on for an hour after R2 was replaced with a 390 Ohms resistor. The heatsinks are maybe around 28-30 degrees now, and i guess it's like 5 degrees warmer than before. I assume it's supposed to run hotter still though. From what i remember, i could barely touch the old one we buildt in 92-93, but it had slightly smaller (and horisontal) heatsinks though.
Just so i know why we did this, we lowered the resistance of R2 to compensate for the low gain output transformers, right? So this is not the optimal solution, but a kind of rescue because i got low gain mj15003's?
I also notice the c2240 (Tr3) transistor is a bit warmer now. I dont want to push it too hard since it's a small transistor, but can i go even lower on R2?
I tried to replace Tr3 with a sc3421 earlier this week and although my multimeter showed significantly lower Hfe when i tested it (about 50 compared to 150 on the c2240). I couldn't hear or notice any practical difference while playing music though.
So now we produce more heat. What does this tell us? I need to replace the transistors with higher gain ones?
(and so does my amp)
The amplifier have now been powered on for an hour after R2 was replaced with a 390 Ohms resistor. The heatsinks are maybe around 28-30 degrees now, and i guess it's like 5 degrees warmer than before. I assume it's supposed to run hotter still though. From what i remember, i could barely touch the old one we buildt in 92-93, but it had slightly smaller (and horisontal) heatsinks though.
Just so i know why we did this, we lowered the resistance of R2 to compensate for the low gain output transformers, right? So this is not the optimal solution, but a kind of rescue because i got low gain mj15003's?
I also notice the c2240 (Tr3) transistor is a bit warmer now. I dont want to push it too hard since it's a small transistor, but can i go even lower on R2?
I tried to replace Tr3 with a sc3421 earlier this week and although my multimeter showed significantly lower Hfe when i tested it (about 50 compared to 150 on the c2240). I couldn't hear or notice any practical difference while playing music though.
So now we produce more heat. What does this tell us? I need to replace the transistors with higher gain ones?
What we have done is provide more drive current for the low gain MJ15003. I saw your heat sinks are pretty big from the photo. In reality the amp should run at around 45 watt and pretty hot, then again it is hard to tell if you have very efficient heat sinks. It would be best to measure the current drawn from the power supply than to judge the temperature by touching it.
The 2SA2240 can dissipate around one watt if I remember correctly, but the hotter it gets it should be de-rated.
I would not go much lower than 270 ohms which will run the 2SC2240 pretty hot.
Interestingly though, when running the 2SC2240 at this current then the third harmonic becomes dominant and may change the overall sound character of the amp. If you have one then replace it with a BD139 which can handle the higher power dissipation.
However, in keeping with the original design, the 2N3773 may be a better option - what makes you think that you have a counterfeit part? You may as well have a counterfeit MJ15003.
The 2SA2240 can dissipate around one watt if I remember correctly, but the hotter it gets it should be de-rated.
I would not go much lower than 270 ohms which will run the 2SC2240 pretty hot.
Interestingly though, when running the 2SC2240 at this current then the third harmonic becomes dominant and may change the overall sound character of the amp. If you have one then replace it with a BD139 which can handle the higher power dissipation.
However, in keeping with the original design, the 2N3773 may be a better option - what makes you think that you have a counterfeit part? You may as well have a counterfeit MJ15003.
I will measure the current later tonight and yes, i do have bd139's. The PCB came with all of the components of which i have replaced some with higher quality components. I bought the kit from China and the 3773's were almost round as a rocking chair on the lead side that is supposed to be flat for best possible thermal contact with the heatsink. Also it had a weird circle between the legs.
The mj15003's i buy from Elfa.se which is one of Europes largest distributors of electronic components and tools. Helps the credibility a bit more.
The mj15003 is a better transistor but have lower gain, so even though the 3773 is older, i will benefit from the extra gain? Or am i underestimating the 3773?
Thanks a lot for getting me this far.
Ben.-
Hi Ben,
There is absolutely nothing wrong with the 2N3773, neither with the MJ15003, but when changing components in a design you always stand the chance to make things worse rather than better. Component values are calculated based on the semiconductor characteristics.
One may think that one transistor is "better" than another, but in the circuit context may perform worse as the design does not tolerate the changes.
JLH favored certain components (most of us do I guess) and the designs are tweaked around the specific component characteristics.
I built that circuit this afternoon on my makeshift PCB using BD139-16 and MJL3281A. All measured fine except my heatsink was way too small for dissipating the heat so I could not test or listen to it for more than a few minutes at a time.
What I can say is that there is nothing wrong with JHL design, but don't haphazardly change semiconductors, some designs are much less tolerant to changes than high gain amps using large amounts of feed-back.
Anyway, I am sure that you will have the amp running as it should be tomorrow you have a good idea where the potential problems are.
There is absolutely nothing wrong with the 2N3773, neither with the MJ15003, but when changing components in a design you always stand the chance to make things worse rather than better. Component values are calculated based on the semiconductor characteristics.
One may think that one transistor is "better" than another, but in the circuit context may perform worse as the design does not tolerate the changes.
JLH favored certain components (most of us do I guess) and the designs are tweaked around the specific component characteristics.
I built that circuit this afternoon on my makeshift PCB using BD139-16 and MJL3281A. All measured fine except my heatsink was way too small for dissipating the heat so I could not test or listen to it for more than a few minutes at a time.
What I can say is that there is nothing wrong with JHL design, but don't haphazardly change semiconductors, some designs are much less tolerant to changes than high gain amps using large amounts of feed-back.
Anyway, I am sure that you will have the amp running as it should be tomorrow you have a good idea where the potential problems are.
Sorry I did not answer the last question, you will benefit from the higher gain of the 2N3773. As is now evident, the MJ15003 is not a "better" transistor in this particular design, but it could be made to work by changing the design. But changing the design may change some sonic character that made the design popular.
Last edited:
I have replaced R2 with a 300Ohm resistor and Tr3 (c2240) with a BD139. The c2240 had a hfe @ 250 and the bd139 got a hfe @ 93. So even though i lowered the resistance in R2 even further, i guess the gain in the bd139 just isn't enough.
I guess the next logical step is to replace the new mj15003's with the old 2n3773's and replace the bd139 with the old c2240 again and see where that takes us. If that doesn't work right out of the box, i'll experiment with R2 again. My guess is that i need less manipulation of the resistor with the higher gain the c2240 delivers.
The amp became hotter with R2 @ 370 Ohm and with the c2240 than R2 being 300 Ohm and with the bd139.
I also measured the current each channel draws from the power supply and it rests steadily around 1.05A. Not far from the 1.2A where it's supposed to be.
Sounds about right?
Ben.-
I guess the next logical step is to replace the new mj15003's with the old 2n3773's and replace the bd139 with the old c2240 again and see where that takes us. If that doesn't work right out of the box, i'll experiment with R2 again. My guess is that i need less manipulation of the resistor with the higher gain the c2240 delivers.
The amp became hotter with R2 @ 370 Ohm and with the c2240 than R2 being 300 Ohm and with the bd139.
I also measured the current each channel draws from the power supply and it rests steadily around 1.05A. Not far from the 1.2A where it's supposed to be.
Sounds about right?
Ben.-
Yes, some progress. I've not played more with the resistors yet, but i bougth a kit that used another transistor for tr3 and tr4 with more gain. this has lifted the current from 1A to 1.5A. Also running it a bit high on voltage at the moment so it's getting comfortably hot in here. The heatsinks stays about 50-60Degrees Celcius. I can touch and hold for 2-3 seconds. :-D
New kit i tested. (replaced crap components with better quality ones)
2pcs JLH 1969 A CLASS A amplifier kit sc | eBay
I've just started in a new job (dream job for me) on a service workshop for everthing between fridges to amplifiers so i will hopefully refresh some forgotten knowledge and gain some new.
New kit i tested. (replaced crap components with better quality ones)
2pcs JLH 1969 A CLASS A amplifier kit sc | eBay
I've just started in a new job (dream job for me) on a service workshop for everthing between fridges to amplifiers so i will hopefully refresh some forgotten knowledge and gain some new.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.