Hi,
What headphones are you using as a load? Got any technical specs? I have some Sennheiser HD600's and AKG701's. Around £200 here when they first come out. Great thing with the JLH class A amp is you can customize them for your load requirements. Have also thought myself of using a JLH headphone amp for tuning purposes with headphones.
For some transistors check out Linear Systems website. Some of these are IC type/based transistors. The snag with the JLH amp is the compensation problem not so much using ruler flat linearity transistors!
http://www.linearsystems.com/datasheets/LS301-3.pdf
Some of these are in discrete packages also.
A real problem to me with the JLH amp is the power supply. I would go with some Walt Jung type Super Regs as the PSRR of the amp is not so hot.
Hope this helps
Kevin
What headphones are you using as a load? Got any technical specs? I have some Sennheiser HD600's and AKG701's. Around £200 here when they first come out. Great thing with the JLH class A amp is you can customize them for your load requirements. Have also thought myself of using a JLH headphone amp for tuning purposes with headphones.
For some transistors check out Linear Systems website. Some of these are IC type/based transistors. The snag with the JLH amp is the compensation problem not so much using ruler flat linearity transistors!
http://www.linearsystems.com/datasheets/LS301-3.pdf
Some of these are in discrete packages also.
A real problem to me with the JLH amp is the power supply. I would go with some Walt Jung type Super Regs as the PSRR of the amp is not so hot.
Hope this helps
Kevin
nelsonvandal said:
Re-building left channel after big blown out
Hello Folks,
I am refering to the Higher Power Circuit of the JLH. Besides the unmatched 2N3055 output transistors, what other possible causes that could produced this? No input or output connected though.
Measured across the emitter resistors (100m) after 20 minutes warm up.
For Q1 R14 ---> 49 mV
For Q1a R16 ---> 86 mV
For Q2 R13 ---> 37 mV
For Q2a R15 ---> 97 mV
DC offset stable around 0mV
Something is not quite right there I think. The output transistors are not sharing the current.
Hello Folks,
I am refering to the Higher Power Circuit of the JLH. Besides the unmatched 2N3055 output transistors, what other possible causes that could produced this? No input or output connected though.
Measured across the emitter resistors (100m) after 20 minutes warm up.
For Q1 R14 ---> 49 mV
For Q1a R16 ---> 86 mV
For Q2 R13 ---> 37 mV
For Q2a R15 ---> 97 mV
DC offset stable around 0mV
Something is not quite right there I think. The output transistors are not sharing the current.
hello all, sorry to restart a tired old threadl have read through the whole thread and gleaned that the penultimate circuit is the one using MJ 15003 devices but l also came across an obscure reference to the MJ 802 as well someone having tried them in liue of the 2n3055, what l would like to know what difference in sound quality betwen these two? could anyone please enlighten me please?: the reason l ask is l have 7 x 2n3055's already and 2 x 802's l could use so before spending anymore than l have to ona bunch of MJ 15003's, l would rather get some more 802's if they are a little better than the cooking/common/ garden variety 2n3055's
thanx in anticipation to some considered opinions
T.C.
While it is open, I completed and populated a board (1969 version) but did not add power transistors yet... I have the 2n3055s but no adequate angle stock to mount it with, but I do have some 2sc5200... which I believe there is some problem useing with the amp due to its higer speed...
Is there a way to make it work?
Is there a way to make it work?
...tiny contribution if it can help...
I have built a JLH version 2003 with MJ15003 output transistors.
It was slightly unstable when I first built it (transistors really hot, occasional long-wave-radio sweeps audible).
Shortening the leads to the transistors (they're now 8 or 10 cm) and also untwisting them (contrary to what suggested by some about the JLH) solved these problems.
I'm also planning to try 2SC5200, leftovers from building a GEM amplifier. I really hate to say it, but it didn't stand the competition with the JLH. Certainly more powerful, excellent highs and lows, probably better for playing rock; but it doesn't render the pianos and voices like a class-A... so for my tastes it's the JLH that will stay.
I really hate to say this, because I liked the GEM project a lot, and the guys who contributed to it here on diyaudio. And that's where I learned, from Graham, to NOT twist the leads to the output transistors.
Anyhow, I'll let you know (one day) what happens with those fast transistors on the 2003 version.
_
I have built a JLH version 2003 with MJ15003 output transistors.
It was slightly unstable when I first built it (transistors really hot, occasional long-wave-radio sweeps audible).
Shortening the leads to the transistors (they're now 8 or 10 cm) and also untwisting them (contrary to what suggested by some about the JLH) solved these problems.
I'm also planning to try 2SC5200, leftovers from building a GEM amplifier. I really hate to say it, but it didn't stand the competition with the JLH. Certainly more powerful, excellent highs and lows, probably better for playing rock; but it doesn't render the pianos and voices like a class-A... so for my tastes it's the JLH that will stay.
I really hate to say this, because I liked the GEM project a lot, and the guys who contributed to it here on diyaudio. And that's where I learned, from Graham, to NOT twist the leads to the output transistors.
Anyhow, I'll let you know (one day) what happens with those fast transistors on the 2003 version.
_
qwad said:
Well, why don't you build one channel with the two MJ802 (not the high-power version that takes four) and one with the 3055.
If the difference is big enough, buy two more MJ802.
Since these are all TO-3 packages, you only mount and dismount from the heatsinks, but the rest of the amplifier will remain.
I think the MJ802 costs about as much as an MJ15003, below 4 Euros around here.
Edit: looking at the datasheets, the hFE variation with output current semms favourable to the 15003 over the MJ802, apparently flatter in the 1 to 3 A region.
The 2SC5200 looks even better, the 2N3055 not too bad
(apparently only, didn't compare very precisely.)
And this is only one parameter.
Of course let us know.
_
...while suggesting to qwad that the output transistors can be kind of independent from the amplifier board (usually the case with TO-3), I thought that it was true for me too and luckily my heatsink has enough holes that it was no big deal to try the 2SC5200s on it (the advantage of those flat packs...)
And it works fine.
No capacitors added, just connected with leads as short as possible (say 5cm). The transistors are Toshibas.
I biased at 1 ampere, for now.
At a very first impression, the sound gained in the bass.
No signs of instability or oscillation.
No "low level hum" as reported with fast transistors on the JLH site.
So thanks for pushing me to do this experiment that I've had in mind for a while.
Edit: it might be useful to say that the connected loudspeaker is 8 ohm two way floorstander from the brand "Opera".
_
and luckily my heatsink has enough holes that it was no big deal to try the 2SC5200s on it (the advantage of those flat packs...)And it works fine.
No capacitors added, just connected with leads as short as possible (say 5cm). The transistors are Toshibas.
I biased at 1 ampere, for now.
At a very first impression, the sound gained in the bass.
No signs of instability or oscillation.
No "low level hum" as reported with fast transistors on the JLH site.
So thanks for pushing me to do this experiment that I've had in mind for a while.
Edit: it might be useful to say that the connected loudspeaker is 8 ohm two way floorstander from the brand "Opera".
_
Greetings to one and all who bothered to reply to my query, and much thanks for this, but this still does not answer my question.................. perhaps l should have addressed it to TIM A in the first instance: who l understand built his first version with the 802 and later changed to 2n 3055's and on to the mj 15003's, so l was wondering if he could recall the difference between the first two devices please
cheers to all and thank you again
cheers to all and thank you again
Hehe just powered up the JLH now... got it biased at 1.3A (well it crawled from 1.3 to 1.39A in about 10 minutes) with 35V supply
Gets realy hot, but it is just a test sink, half the size of the full ones I will use... can put my finger on the outputs for about 10 seconds...
Wife is sleeping allready, so it will have to wait untill tommorrow for testing... I suspect it may hum a little as I went totaly minimalist on it so that I can easily zoom in on areas that makes for best tweaking.
Gets realy hot, but it is just a test sink, half the size of the full ones I will use... can put my finger on the outputs for about 10 seconds...
Wife is sleeping allready, so it will have to wait untill tommorrow for testing... I suspect it may hum a little as I went totaly minimalist on it so that I can easily zoom in on areas that makes for best tweaking.
Thanks for your reply TimA, mustév misread geoffs writings, çause l recall reading on there that you built your first version with 802's as you already had them, then you went to the 2033's and then on to the 15003's, so l thought you may have been able to shed some light on the differences, if any at all, obviously you cant, which is sad, as it would save me a lot of stuffing around: out of interest to you and anyone else; l intend to build my 1st version with 2033's as l already have 7 of them on hand and intend to get some more so l can select my devices bit more carefully and theyre only about 2 dollars a peice compared to 5 times as much for the 15003's, and after l have my amp up and running successfully l will try the 802's which are only 3 times the price of 2033s, nyhow we'll see how it transpires, btw how is Geoff these days?, trust he's better than l am as l recently had another stroke luckily it hasnt disabled me anymore than l am already........... cheers and enjoy your nice warm summer...... T.C.
Well who would have guessed. Works perfect right from the start...
I used 2n5401 on the input , bd139 in the "driver" and 2sc5200 on the outputs...
I suspect clever layout played a part, but I have no hum.
PSU is simply half of one of my universal PSU boards (not even with inductor installed).. with 2 10000uf in parallel, and another 1000uf axial cap from V+ to GND...
Lol getting goosepimples from the sound now... and that's just one channel... PCB drawn by hand with a black marker on the copper.
How do these amps prevent themselves from destruction, by bias drifting with temperature????????
Sound... I am realy impressed by the honesty of the reproductions.
Finaly something that does the T ans S sounds more realisticaly than the HRII...
But I would never believe how hot that sink is getting.... can probably hold on to it for 5 seconds. Which I guess means time to swithch off and get the bigger sinks or get a lower supply.
I used 2n5401 on the input , bd139 in the "driver" and 2sc5200 on the outputs...
I suspect clever layout played a part, but I have no hum.
PSU is simply half of one of my universal PSU boards (not even with inductor installed).. with 2 10000uf in parallel, and another 1000uf axial cap from V+ to GND...
Lol getting goosepimples from the sound now... and that's just one channel... PCB drawn by hand with a black marker on the copper.
How do these amps prevent themselves from destruction, by bias drifting with temperature????????
Sound... I am realy impressed by the honesty of the reproductions.
Finaly something that does the T ans S sounds more realisticaly than the HRII...
But I would never believe how hot that sink is getting.... can probably hold on to it for 5 seconds. Which I guess means time to swithch off and get the bigger sinks or get a lower supply.
Attachments
to be multiplied by the dissipation factor)
I think my last calculation showed about 50W for the two outputs together, which seems to match the rough 1C/W I calculated the sinks as being.
The voltage was a bit low last night as I was only useing one secondary of the transformer... which is tapped for 0-24-0-24-0-6 VAC. As you can see I was pulling 1.45A at the time of the picture (that is from PSU, you can take off 10mA for the front of the amp to calculate bias setting.
The full size heatsinl is identical, but twice the height (20cm) so I guess in the region of .6C/W.
I plan to first make some prototypes of the later versions too, before I decide which one to build ultimately.
I just wanted to see for myself if the whole PSU and PSSR thing was realy such a big problem or just a case of bad layouts... I see no insurmountable problem or anything expensive needed.
The voltage was a bit low last night as I was only useing one secondary of the transformer... which is tapped for 0-24-0-24-0-6 VAC. As you can see I was pulling 1.45A at the time of the picture (that is from PSU, you can take off 10mA for the front of the amp to calculate bias setting.
The full size heatsinl is identical, but twice the height (20cm) so I guess in the region of .6C/W.
I plan to first make some prototypes of the later versions too, before I decide which one to build ultimately.
I just wanted to see for myself if the whole PSU and PSSR thing was realy such a big problem or just a case of bad layouts... I see no insurmountable problem or anything expensive needed.