Member
Joined 2009
Paid Member
One of the original JLH articles mentions that it's important to get close to the rail midpoint for lowest distortion - I remember reading it but i don't remember any explanation for it.
The obvious would seem to be largest symmetrical swing, but this is mainly an issue if the amp power is limited for the application.
I would like to see the virtual ground at a stable Voltage.
There was mention of the relative hFE of the upper and lower output transistors as to which imposes the most 'character'.
I would like to see the virtual ground at a stable Voltage.
There was mention of the relative hFE of the upper and lower output transistors as to which imposes the most 'character'.
Re: EI transformer vs sound: Your power must be cleaner than mine!😛😛
Unless you spend $$$ toroids generally come with lots of interleaf capacitance which means mains hash gets a free ride from primary to secondary. Both measurable and audible in my neck of the woods.
Any EI transformer is likely to be better in this regard and twin bobbin mains transformers (still pretty common) are miles ahead. C core is even better, if you can find it.
As a bunch of other posts between ours point out, the power supply is half the amplifier in Class A designs. And despite the simplicity of the JLH, layout still needs consideration.
Do you want to explain further?
At the moment your statement runs contrary to the experience of the majority of JLH users.
Yes, there is a 6db roll off below a certain point due to the output capacitor but that's a design choice which can be placed below your program material.
Like every amplifier made, there are limits to the voltage & currents it produces. I can make any amplifier sound bad (especially the bass response) by selecting speakers which are a mismatch.
In 1980 I built a 225WRMS Maplin disco amplifier with 2N3055 and MJ2955's.
It sounded great and was very loud.
It worked well for many years.
For some reason they used +/-55VDC volts rails and the transistors didn't blow up. I think they are rated about 60 volts rail to rail.
You can still get 2N3055/MJ2955 and they are very cheap.
I hated the 3055 for years and thought it was only good to have stabilized power.
and then one day, an elder put my nose in my poop by making me listen to a jlh he had built in the 70s.
it's been 10 years now since I've been chasing the real 3055s and the mj15003s and eight years now that I'm assembling jlhs by trial / error / learn and still learn today.
on the other hand, I have never heard a good amp in AB with 3055, but some amp in B yes
and then one day, an elder put my nose in my poop by making me listen to a jlh he had built in the 70s.
it's been 10 years now since I've been chasing the real 3055s and the mj15003s and eight years now that I'm assembling jlhs by trial / error / learn and still learn today.
on the other hand, I have never heard a good amp in AB with 3055, but some amp in B yes
Some of TO3 3055's ("H" suffix or what was it?) were rated specially to 100V Vce, the regulars were 60V ... and some of those available in local shops ("industrial rejects"?) were only good up to 45-50V.
But even with small plastic MJE or TIP 3055 variants you can make a nice little amp (if "nice" is your thing) ...
there are actually a lot of 3055 different,
and according to the manufacturers, the operating limits are also different
and according to the manufacturers, the operating limits are also different
Actually, it was Vcbo that was 100V for the original RCA 2N3055. Vceo was always 60V. If you connected a resistor between base and emitter you could also work with Vcer=70V.
I don't know who makes 2N3055's now other than ON Semi. Any others?
ON Semi's devices still have the 100V spec, so if you could arrange for the base to be reverse biased or biased with a very low impedance you might have had no problems at rail voltages up to +/-40V, but I would not push them further.
And the SOA for ON Semi's 2N3055 has improved since the earlier epi 2N3055, but does not quite match the original RCA spec.
the output capacitor is too small for good bass. Many 70's amps used 2000uf but simulations show how poor that is well above the -3dB point. At least 10mF is needed and I agree with one suggestion that 22mF is preferable still.
You may then also need to change the feedback decoupling capacitor and the input capacitor too. (and the bootstrap ...)
yes but with a increased output capacitor,the mid and high are impacted .
for me,6800uf to 10000uf are the best compromise for the jlh
for me,6800uf to 10000uf are the best compromise for the jlh
There were some model of 3055s specifically spec. for higher Vceo.
Vcbo was rated at 100V, but Vce only 60V in regulars.
Vcbo was rated at 100V, but Vce only 60V in regulars.
Hi Nigel
Yes, and BDY20 and BD130 and 2N3055A, but these all represent the 2N3055H era with ~800kHz fT.
epi 2N3055 ft is 2.MHz (which is why I suggest it is now not as bad as some say for audio amps, though of course the newer devices are better) and TIP3055 is 3MHz.
TIP 3055 has 100V Vcbo too!
I'd suggest MJ15003/4 rather than 15015/6 for audio, but I've not used any15015 in an audio amp to try.
Yes, and BDY20 and BD130 and 2N3055A, but these all represent the 2N3055H era with ~800kHz fT.
epi 2N3055 ft is 2.MHz (which is why I suggest it is now not as bad as some say for audio amps, though of course the newer devices are better) and TIP3055 is 3MHz.
TIP 3055 has 100V Vcbo too!
I'd suggest MJ15003/4 rather than 15015/6 for audio, but I've not used any15015 in an audio amp to try.
huggygood - but you can always shunt the OP capacitor with an HF high speed one too, e.g. 2.2uF double metal unit intended for pulse applications (quite high current, that is).
Are you saying the mid-high is impacted by ESR, ESL or distortion?
Are you saying the mid-high is impacted by ESR, ESL or distortion?
Plastic MJE/TIP 3055's are quite sweet toys for what they are.
I know that neve preamp cloners often use a TO220 MJE 3055 instead of the TO3 part.
It is funny how audiophiles massage their rigs to listen to recordings that for most part passed thru a bunch of 5532/34's, 2N3055s, and God forbid even TL072's ...
I know that neve preamp cloners often use a TO220 MJE 3055 instead of the TO3 part.
It is funny how audiophiles massage their rigs to listen to recordings that for most part passed thru a bunch of 5532/34's, 2N3055s, and God forbid even TL072's ...
i never try this but i will try ,thanks 😉
and ,i don't know but I think it's easier to choose an output transformer than a capacitor (from a technical point of view)
I like the tl072 in some application and especially with the philips cd player + tda 1541 NOS mode
Although I wasn't really comparing E-I virtues with toroidal in that comment, point taken. My personal JLH'69 though, has the cap. multiplier supply which makes quite a difference to noise levels, not at they're all that noticeable in my location 😉