JLH 10 Watt class A amplifier

I powered the amps for the first time today on a +/- 16V supply, both worked perfectly. Unfortunately I only had a small low sensitivity on hand speaker to test with so did not get a good idea of the sound.

The thermal pads were not up to the task so I swapped them for old fashioned washers and thermal grease. The heat sink handled the load fine the chips were cool enough to touch continuously. I'm using MJE253G/MJE243G instead of bd139/bd140 and NJW0281G's for outputs.

I will be ordering a case and some more caps soon and will pick up a pair of 18V-0-18V 160VA toroidal transformers next week.
 
Cheers,

That board of yours looks very tidy!

Thank you.

Shaan,

Some great wiring of those boards, really neat.

p.s. you could use a little solvent to clean off the solder flux.

Thanks Bigun. I had a headache after finishing it. It was too complex for me. And in fact I expected it to malfunction, but it's running fine from the first run. LOL

Shann what is your power transistor? How much Iq? I'm afraid the heatsink is too small for a JLH. My heatsink is 5 times bigger than yours for each transistor but still very hot.

These are only the drivers. The power trannies are 2SC5200(genuine 🙂 ) mounted on pentium 4 heatsinks with minimal airflow. Runs okay all day all night.
 

Attachments

  • DSC09288.JPG
    DSC09288.JPG
    395 KB · Views: 675
  • DSC09289.JPG
    DSC09289.JPG
    278.2 KB · Views: 647
Next project:- stereo '69 JLH on veroboard with 1) dual supply with CRC pi filter 2) opamp offset servo powered from ps rails 3) 2SC5200 output bjts 4) 0.047uF+10ohm zobel 5) 5pF silver-mica parallel with feedback resistor.

Coz, '69 JLH with these damn fast BJTs sounds better and more "right" with my fullranges, than DOZ.

I found DOZ to be happiest with TIP3055s, and I'm building a stereo system with this config, halfway done, for sale.
 
2N3055's are old designed, cheap transistors, totally unsuitable,
unusable for high fidelity amplifiers...

Regards, PL.

What about in Rod Elliots P3?

I built that project and the TIP3055, and 2955s sounded rather good, despite the TIPs being in plastic cases rather than TO3 cases, heat wasnt an issue either. I also followed the suggestion of paralelling output devices, which may have helped somewhat. Not lacking in bass, very balanced indeed. The only faults in my implementation of the design, was omitting the PSU snubber, which raised the noise floor somewhat, and the fact I used a +/-45V PSU, so they were operating at very close to the max voltage. They may be virtual antiques, but the ONLY thing that once concerned me about that particular transistor, was the bandwidth. I searched around and found the best quality, highest BW versions I could; I believe they were Bourns TIPs, which dont seem to be available anymore. These seemed to have better bandwidth, and TO3P case, rather than the T220 types(urgh!). They cost 3 times that of the SMicro ones, but they were closely matched, and better to heatsink.

The only problem with using a transistor of this age, AFAIK is HOW you implement them, they still do a very respectable job IMHO.
 
Hehehe... I don't know about "right" or "wrong" but by ears or by eyes it should be clear to everyone why people are building JLH not DOZ 🙂

I know that the DOZ simply likes slower transistors in output. 30MHz bjts make DOZ too bright in the highs. Some may like this, I don't.

I noticed the DOZ sounds exactly as warm and relaxed as the JLH when slow transistors are used in output. But this also puts more stress on the driver and CCS transistors(deliver more Ib for low gain bjts). If this is okay with the builder the DOZ works like a charm and produces great music even with those 800KHz 3055s, just like JLH. I am very satisfied with DOZ. Well it is nothing but JLH in disguise.

And by "right" I meant what my ears feel right. You know well what your "right" is, don't you?
 
"What about in Rod Elliots P3?"

The P3A is pretty good but I prefer the JLH for its more natural way with music. I use the P3A as a sub amp.

Tim.

ok, well P3A uses the MJE**** (or is that the drivers?) as far as i recall, P3 uses the 2n3055/TIP3055.

I would personally love to try the TIP3055s/2955s in Class A, but id probably need 16 of them in // and a multi kVA PSU
 
Last edited:
Ahhh ok. I thought Gain BW, my error. Also I have TIP3055, only 90W. TO3 types are extinct here as far as i know.😀

Yes I was also talking about gain BW. Is there a difference between fT and gain BW?

http://en.wikipedia.org/wiki/Gain–bandwidth_product#Transistors

See the third page of the pdf which franzm posted in the previous page of this thread. 800KHz, TO3. Available here along with 3000KHz TIP3055s. Made(imported) by CDIL/BEL etc, as cheap as $0.5. Also Motorola and Onsemis are available. Expensive, $0.75. 🙂
 
Last edited:
Yes I was also talking about gain BW. Is there a difference between fT and gain BW?

Gain?bandwidth product - Wikipedia, the free encyclopedia

See the third page of the pdf which franzm posted in the previous page of this thread. 800KHz, TO3. Available here along with 3000KHz TIP3055s. Made(imported) by CDIL/BEL etc, as cheap as $0.5. Also Motorola and Onsemis are available. Expensive, $0.75. 🙂

haha ok smartypants, i dont often look at specs for silicon, IM AN AMATEUR.

jeez.......

I considered TIP3055-2955 as 'modern equivalents', and didnt expect Gain BW to be so much different between the 2 types.....if I had more knowledge id probably just go for lateral FETS anyway.....there is such a thing as TOO cheap 😀
 
Last edited:
lol By amateur, I mean that I could probably design a BASIC single stage, or use an opamp for a pre amp. Putting the stages together, servos, etc is all abit out of my scope of experience at the moment lol. I tend to find a design I like, and maybe mod it slightly. Either that, or chip amps...which I can handle, since they generally only need very few external components.

Put it this way, any reference I have presently, goes back to the 60s and a dog-eared book that ive owned since I was about 7 years old....

CLASS_A.jpg


Don't look at specs for silicon. Wow, mondogenerator is a brave soldier!

Perhaps I shouldve said: "I dont OFTEN, study specs for silicon", since I dont often build/design amplifiers lol 😀
 
Last edited:
One thing I observed during listening the last direct coupled jlh is that it is completely silent at turn on and off. Previously I had made another couple of DOZ with the servo powered from an external supply. It worked, but it had a "thumb" and a "tick" at on and off respectively.

This time I supplied power to the servo from the amp rails and that problem went away. I used a diode bridge, two zeners, two resistors and two bjts to regulate the servo supply to 12 volts. The diode bridge's inputs are connected to both channel's supplies, so powering any channel separately supplies power to the opamp as well as powering both channels.

This is my first time doing this and it is working like a charm! I am very happy.
 
I'm a fan of the JLH school of thought but I'm afraid it isn't that obvious to be why Simple Class A is better than DOZ

I don't think I fully understand your words (re: Simple Class A is better than DOZ), but may be I understand the question. But this difference is only subtle for most and can easily be considered as a voodoo thing. In essence, I don't want to discuss the voodoo thing 😀

The DOZ should have been named the DOM (Death Of Mosfet) to be relevant (not DOJ of course). Trying to compare a one stage mosfet (Zen) with a more complex bipolar is just improper imho. Then, why not make it as simple as the JLH? Or may be the JLH is too simple as the Zen? Do you think the DOZ can also be called a DOJ (Death Of JLH)? 😀

For small power amp, why not go for the best possible. Everything (audible or "not") counts. It may be inaudible for one but audible for everyone else.
 
And by "right" I meant what my ears feel right. You know well what your "right" is, don't you?

Hmm not really... I mean, I don't know what "right" is in (good) amplifiers. I have never seen it that way (If you switch one of the speaker cable, the sound would be not "right", of course 😀).

Or may be I use different wording? Oscillate may be? Or the sound correlates with being in the border of high frequency instability? I'm familiar with that kind of sound. Might happens with 30MHz transistor.