You could use 2N3773 for output transistors. They are a more powerful version of 2N3055 but probably no better and possibly worse at <10W than later versions (like ST Micro) of 2N3055 or TIP3055.
I would definitely not try a small signal transistor like BC109 to drive lazy slugs like 2N3055s. You need a specific driver transistor that is fast but can safely deliver relatively high peak currents at full supply voltage. If you want a metal can transistor for the appearance, ST Micro 2N1711 and possibly Chinese copies or most fast TO39 types will probably work OK but I would use a specific, modern driver type as already discussed here over the last few pages.
I would definitely not try a small signal transistor like BC109 to drive lazy slugs like 2N3055s. You need a specific driver transistor that is fast but can safely deliver relatively high peak currents at full supply voltage. If you want a metal can transistor for the appearance, ST Micro 2N1711 and possibly Chinese copies or most fast TO39 types will probably work OK but I would use a specific, modern driver type as already discussed here over the last few pages.
Last edited:
Hi Ian,
Have been wading through this thread (now 21 years old), I have some boards to build the original JLH69 and have selected the following transistors as subs. Any comments as to what you think on these, thanks. The pcb's were done by Prasi with the onboard TO3 2N3055 adj. reg transistor, with a trimpot to set the output volts to 1/2 Vcc and a trimpot to adjust quiescent current - will stick with orig value of 1.2ADC with a supply at 28VDC.
2N3906 = PNP input
2N1711 = NPN driver ( have original NOS)
MJ15003 NPN outputs. (Although I do have a range of 2N3055's, I feel one could do better than these)
Cheers,
Gary.
Have been wading through this thread (now 21 years old), I have some boards to build the original JLH69 and have selected the following transistors as subs. Any comments as to what you think on these, thanks. The pcb's were done by Prasi with the onboard TO3 2N3055 adj. reg transistor, with a trimpot to set the output volts to 1/2 Vcc and a trimpot to adjust quiescent current - will stick with orig value of 1.2ADC with a supply at 28VDC.
2N3906 = PNP input
2N1711 = NPN driver ( have original NOS)
MJ15003 NPN outputs. (Although I do have a range of 2N3055's, I feel one could do better than these)
Cheers,
Gary.
While thinking about it, I found this high quality write-up of a JLH build that uses MJ15003 transistors here: https://theslowdiyer.wordpress.com/2015/06/08/more-jlhs/ . It's good reading and there is more to the site and JLH project than what you see at that page, but the index is a bit confusing. Anyway, with patience, I expect it will give others here, the enthusiasm to start yet another amp. and dispel any doubts you may still have about yours.
Last edited:
Hi. I measured 10uA (Ib) with my chinese hfe meter and Hfe's are between 100 and 300 and mostly around 175. Maybe i can use around 250 Hfe ones which i found about 10pcs from about 100pcs.
My supplier told that he has oroginal st bd237 but it seems Ft is not enough for Jlh
My supplier told that he has oroginal st bd237 but it seems Ft is not enough for Jlh
Attachments
I have read, many times, something along the lines that "I replaced the 2N3055 with MJ15003 and it sounded a lot better".
Hmmmm. When people say this, I'd like to know which 2N3055 they replaced. If these were the old RCA hometaxial base types, or equivalent (BDY20, BD130, BD184) then I would agree. It is harder to conclude that there is a better result using recent 2N3055's which are epitaxial, with a spec. min. ft of 2.5MHz (MJ15003 2MHz). Even more so that the OnSemi 2N3055 appears to be built on the same manufacturing process as the MJ15003.
So, which version of the 2N3055 did you replace, those of you who did?
Another possibility for the (apparently) better result is that the MJ15003 has a higher current gain (slightly - 25 @5A rather than 20@4A) and might just have a better gain hold-up than 3055 over the 1...2A range.
Next time I get one of my JLH's out I may just test between 2N3055 and MJ15003.
Hmmmm. When people say this, I'd like to know which 2N3055 they replaced. If these were the old RCA hometaxial base types, or equivalent (BDY20, BD130, BD184) then I would agree. It is harder to conclude that there is a better result using recent 2N3055's which are epitaxial, with a spec. min. ft of 2.5MHz (MJ15003 2MHz). Even more so that the OnSemi 2N3055 appears to be built on the same manufacturing process as the MJ15003.
So, which version of the 2N3055 did you replace, those of you who did?
Another possibility for the (apparently) better result is that the MJ15003 has a higher current gain (slightly - 25 @5A rather than 20@4A) and might just have a better gain hold-up than 3055 over the 1...2A range.
Next time I get one of my JLH's out I may just test between 2N3055 and MJ15003.
I have to agree with John Ellis's comments, particularly as my own hearing isn't so good above 12 kHz any more. However, the comments probably also refer to class AB designs, where the crossover region brings its own problems. I didn't bother with comparing alternative TO3 types but I'd say many 2N3055 transistors would have first arrived in our mailboxes as components of a too cheap Chinese kit and hence they're suspect to begin with. Some kits did come with genuine ST Micro 2N3055 but I saw plenty used by my group of builders, that looked identical to ST product but the marking could be rubbed off without much effort. Draw your own conclusion.
Buying any replacement TO3 semis from a reputable distributor may well give you an apparent improvement but in my experience, there are usually expectancy factors with auditioning substituted parts. That seems to be unavoidable among newbies and it takes a committed sceptic to get down to blind listening tests as JLH and friends did. So, if trust his design, get your family or friends to lend a critical ear (no prompting or hints!) when making comparisons and take a leaf from JLH's articles too.
Buying any replacement TO3 semis from a reputable distributor may well give you an apparent improvement but in my experience, there are usually expectancy factors with auditioning substituted parts. That seems to be unavoidable among newbies and it takes a committed sceptic to get down to blind listening tests as JLH and friends did. So, if trust his design, get your family or friends to lend a critical ear (no prompting or hints!) when making comparisons and take a leaf from JLH's articles too.
Hello Ian,
My own switched tests would sometimes take a week to discern differences with various chip amps
LM3886 vs JLH69 took just 4 days, I noticed straightaway that slight rises in volume (piano keys) were faster and and louder with the JLH.
Was this just wishful thinking on my part? After the 4 days I realised that it wasn't.
This may sound a bit odd but the music comes out of total blackness, it's the only way I can think of to describe it.
I no longer listen to the amplifier it's as though it's not there.
Cheers - J
My own switched tests would sometimes take a week to discern differences with various chip amps
LM3886 vs JLH69 took just 4 days, I noticed straightaway that slight rises in volume (piano keys) were faster and and louder with the JLH.
Was this just wishful thinking on my part? After the 4 days I realised that it wasn't.
This may sound a bit odd but the music comes out of total blackness, it's the only way I can think of to describe it.
I no longer listen to the amplifier it's as though it's not there.
Cheers - J
'Good that you can arrive at a conclusion in your listening tests. You are indeed blessed if you get reliable and repeatable results from listening with just your own ears. I often find myself fuddled by differences from one day to the next whilst repeated tests on the same day are usually quite consistent.....hmm. I'm fortunate though, that I have a few musician friends to lend an ear occasionally. Their long experience with playing in concert, seems to give them quite an edge when it comes to analysing what's going on - musically at least.
