JLH 10 Watt class A amplifier

I won't comment on your second paragraph, but I know of no evidence that a properly designed PCB can damage the 'sound' of an amplifier, unless we believe that electricity somehow behaves differently if the signal being processed is audio - and if so, how does the electricity know that there's a microphone, a loudspeaker and an ear outside of the board?
Describe not current,
describe (the) flowing, (the) streaming;-)
 
To reduce audible (!) distortions use base Q3 as input
To reduce audible (!) distortions use emitter Q4 (R5) as input...
 

Attachments

  • c0f593a7dbcb.gif
    c0f593a7dbcb.gif
    13.2 KB · Views: 294
Has anyone ever tried 2N3773 instead of 2N3055?
2N3773 - (Pc): 150 Вт, (Ic): 16 А, (Tj): 200 °C, (ft): 0,2 МГц, (hfe): 15
2N3055 - (Pc): 117 Вт, (Ic): 15 А, (Tj): 200 °C, (ft): 0,2 МГц, (hfe): 20
When using these vintage transistors, it is necessary to limit the frequency spectrum in the amplifier, i.e. the unity gain frequency of the circuit is higher than the cutoff frequency of these output transistors, and no a even harmonics will save the purity of the sound.
 
Modern replacements of original MJ480, such as 2N3055/3773 and TO247 versions like TIP3055, will actually have significantly higher Ft than the numbers suggested by original, 50 years ago, Motorola spec. datasheets. At the low bias current of only 1A, the small-signal Ft will also be significantly higher than 4MHz. 15MHz would be more realistic at 10W audio output and many DIY builders can verify that at 10W peak power, the JLH amplifier still sounds great and shows little if any degradation of HF. As always, listening is the real validation of adequate performance.
 
Has anyone ever tried 2N3773 instead of 2N3055?
The first amplifier I built when I was very young was with Motorola 3055, which I upgraded to... Motorola 3773. That was an impressive upgrade with fantastic results. Weird, the spec sheets do not suggest that much of a difference, but sound-wise, 3773 was fast, natural, pleasing, extended... beautiful. Later in my life, I worked for a large HiFi and PA amplifier manufacturing company. My favourites from that era were Toshiba 2SC5200/2SA1943 and, in particular - the Exicon 10P20/10N20.
 
2N3773 - (Pc): 150 Вт, (Ic): 16 А, (Tj): 200 °C, (ft): 0,2 МГц, (hfe): 15
2N3055 - (Pc): 117 Вт, (Ic): 15 А, (Tj): 200 °C, (ft): 0,2 МГц, (hfe): 20
When using these vintage transistors, it is necessary to limit the frequency spectrum in the amplifier, i.e. the unity gain frequency of the circuit is higher than the cutoff frequency of these output transistors, and no a even harmonics will save the purity of the sound.
The original RCA ft spec. was 800kHz. I assume that this is not from a genuine supplier's spec if the 2N3055 is rated at 200kHz.
Some hometaxial devices were 200kHz, like the 2N3773, which I would recommend would never make a good audio device.
However, today, older transistors like the 2N3055 are manufactured (seems ON semi are the main manufacturer still selling this) on an epi process, which has been pointed out many times on these threads. So anyone "cloning" a 2N3055 with only a 200kHz spec is not even trying to match the registered spec.
The epi base transistors are better, acoustically, than the old RCA transistors as they have a better frequency response (2.5MHz).
For some reason, (actually, obviously to sell more poorer quality parts) the 2N3773 spec was never changed, though that too will be made on an epi process probably around 2MHz now.
The RCA 2N3055 did not make a "bad" quality - 800kHz is fair, but not as good as the 4MHz parts around at the time. (Note to Ian F- the MJ480 was 4MHz, and the 2N3055 was not really a replacement, it was more a case that the MJ480 was obsoleted sooner). The 3055 was robust. The recent 2N3055 is respectable (it's the fhfe that is really more important in the JLH design) and that is around 60kHz on devices I have measured. An epi 3773 measured at only around 20kH, but I have not measured devices from recent production as I have not considered them suitable.
 
Last edited:
I won't comment on your second paragraph, but I know of no evidence that a properly designed PCB can damage the 'sound' of an amplifier, unless we believe that electricity somehow behaves differently if the signal being processed is audio - and if so, how does the electricity know that there's a microphone, a loudspeaker and an ear outside of the board?
You've not done much work with wide signal range analog electronics then?

Your key words above (properly designed) are key: every trace is coupled to every other trace, particularly its neighbors. Component and track layout are critical to analog design, be it instrument, RF or audio. For complex circuits, it's close to a black art. That's where myth, legend (Bob Pease always said...) and cargo-cult design sets in
 
I have seen some fancy PCB versions of JLH's amp. but It's a very simple design and hard to screw up unless you are silly about what parts are located in close proximity to others or try to make it as small as possible for some reason. Most clone PCBs also feature quite generous tracks and spacings which circumvent a lot problems. Take a look at the original prototype for comparison. It was literally wired up on an early type of "Perfboard" or matrix board etc.. https://www.keith-snook.info/wireless-world-magazine/Wireless-World-1969/Simple Class A Amplifier - .pdf
 
Last edited:
  • Like
Reactions: sumeshak
I have seen some fancy PCB versions of JLH's amp. but It's a very simple design and hard to screw up unless you are silly about what parts are located in close proximity to others or try to make it as small as possible for some reason. Most clone PCBs also feature quite generous tracks and spacings which circumvent a lot problems. Take a look at the original prototype for comparison. It was literally wired up on an early type of "Perfboard" or matrix board etc.. https://www.keith-snook.info/wireless-world-magazine/Wireless-World-1969/Simple Class A Amplifier - .pdf
I have read and re-read it dozens of times but I never tire of it.
 
You've not done much work with wide signal range analog electronics then?

Your key words above (properly designed) are key: every trace is coupled to every other trace, particularly its neighbors. Component and track layout are critical to analog design, be it instrument, RF or audio. For complex circuits, it's close to a black art. That's where myth, legend (Bob Pease always said...) and cargo-cult design sets in
I'm not sure whether you're agreeing or not. Of course I'm not suggesting that a badly designed board is going to work.

For the record, I worked with boards that simultaneously carried GHz RF, microprocessors and DSPs, analogue audio, and switch-mode power regulators. Then I had to get them through EMC testing. So I'm well aware of parasitic coupling 🙂