Hi, roger-k
Both of 18 and 22v transformers will be suitable. I prefer high voltages to feed my jlh, so if dissipated power is not a big deal you may start with 22v one.
There is no need to greatly increase the capacity of C8 and C7. I have not heard any noticeable changes in the sounding after increasing capacity from 1500u to 4700u per rail.
Both of 18 and 22v transformers will be suitable. I prefer high voltages to feed my jlh, so if dissipated power is not a big deal you may start with 22v one.
There is no need to greatly increase the capacity of C8 and C7. I have not heard any noticeable changes in the sounding after increasing capacity from 1500u to 4700u per rail.
Hi.
Thanks for the answers.
There is various opinions here...
I have send a mail to Geoff:
What transformer will be suitable for the 2x10W version, with capacitance multiplier, 22-0-22V or 18-0-18V?
Answer:
22V secondaries will be too high if you intend to use a capacitance multiplier so I suggest you go for 18V.
Thanks for the answers.
There is various opinions here...
I have send a mail to Geoff:
What transformer will be suitable for the 2x10W version, with capacitance multiplier, 22-0-22V or 18-0-18V?
Answer:
22V secondaries will be too high if you intend to use a capacitance multiplier so I suggest you go for 18V.
FaTTy said:
I think that, unless the transformer is undersized, you will find the rail voltage somewhat higher than this, particularly when the mains voltage is close to its maximum.
This means that, for a given output transistor dissipation, the quiescent current will need to be lower than ideal. Many speakers these days have an impedance dip to 4ohms or below, even those with a nominal 8ohm rating, so the ability to deliver current is, in my opinion, important. This is why I recommended 18V for the secondary voltage.
Yes, you right, Geoff. But when we go for class A, we should have huge heatsinks and power transformers to dissipate lots of heat and provide lots of current 🙂 Seems to me, jlh sounds better with +-25 rather than +-20 ,for example, supply rails.
As output transistor I will use MJ15003.
I think I also will use the same sort in the capacitance multiplier.
Will the transistors in the capacitance multiplier run to hot if they are mounted at the same heat sink as the output transistors?
Or will it be better to have a separate heat sink for the transistors in the capacitance multiplier?
I think I also will use the same sort in the capacitance multiplier.
Will the transistors in the capacitance multiplier run to hot if they are mounted at the same heat sink as the output transistors?
Or will it be better to have a separate heat sink for the transistors in the capacitance multiplier?
After reading this for ages now, and Geoff's excellent website, I now feel I am able to finally post on this thread.
However, its not (quite) so that I can report success, more to ask a sanity checking question.
I've been building the 1996 variant of the JLH class A for a while now.
I've deliberately tried to do it on the cheap. So far its cost me about £60, sourcing things mainly off of Ebay, and the more precise component requirements from Farnell.
This has meant sacrifices - like heatsinks that were designed for T-220 packages (which made drilling for, fitting and soldering T-03 packages very interesting!), and too high a voltage transformer, which I then began to unwind until my father found me one that would be more suitable - albeit a much lower voltage (my rails are +/- 16.7, not much headroom for capacitance multiplication or regulation, anyway it'll do for now, and it can deliver plenty of current).
Anyway, its been fun, and can't complain too much considering how cheap it was.
Anyway, I fired up one channel last night. Using a lightbulb at first was very useful as I soon noticed a few minor issues that were rectified.
Then I thought I'd risk it without the bulb. Not the most sensible idea I've ever made (with hindsight, the amp should essentially 'work' normally with a lightbulb in the liveline shouldn't it? if its a big enough wattage?). It sat there happilly for while, I was able to tweak the voltage offset to some extent, but didn't make much inroads on the bias current, and eventually R10 went pop - so I think the bias was too high and my attempts at tweaking the bias did not have the outcomes I had initially expected. As far as I can tell, I have my suspicions that the resistor was the only victim. I think I'll go back to the lightbulb for a bit!
I suspect I have a slight wiring issue with the trimmer for the bias setting, or I was very tired (both are likely 🙂 ).
So, to confirm, and sanity check, on the 1996 circuit the lowest bias setting is set when the voltage at the base of TR5 is closest to the positive voltage rail. The bias increases as the voltage at the base is lowered (or, voltage difference between TR5's base and the power rail is increased) with respect to the positive power rail. This being because, amongst other things, the voltage across R10 is Vb(Tr5)+Vbe.
Is that right? Before now, I haven't really dealt with currents above 500mA tops, so this is a completely new area for me. The theory's the same, I'm just suitably paranoid.
Cheers,
Phil
However, its not (quite) so that I can report success, more to ask a sanity checking question.
I've been building the 1996 variant of the JLH class A for a while now.
I've deliberately tried to do it on the cheap. So far its cost me about £60, sourcing things mainly off of Ebay, and the more precise component requirements from Farnell.
This has meant sacrifices - like heatsinks that were designed for T-220 packages (which made drilling for, fitting and soldering T-03 packages very interesting!), and too high a voltage transformer, which I then began to unwind until my father found me one that would be more suitable - albeit a much lower voltage (my rails are +/- 16.7, not much headroom for capacitance multiplication or regulation, anyway it'll do for now, and it can deliver plenty of current).
Anyway, its been fun, and can't complain too much considering how cheap it was.
Anyway, I fired up one channel last night. Using a lightbulb at first was very useful as I soon noticed a few minor issues that were rectified.
Then I thought I'd risk it without the bulb. Not the most sensible idea I've ever made (with hindsight, the amp should essentially 'work' normally with a lightbulb in the liveline shouldn't it? if its a big enough wattage?). It sat there happilly for while, I was able to tweak the voltage offset to some extent, but didn't make much inroads on the bias current, and eventually R10 went pop - so I think the bias was too high and my attempts at tweaking the bias did not have the outcomes I had initially expected. As far as I can tell, I have my suspicions that the resistor was the only victim. I think I'll go back to the lightbulb for a bit!
I suspect I have a slight wiring issue with the trimmer for the bias setting, or I was very tired (both are likely 🙂 ).
So, to confirm, and sanity check, on the 1996 circuit the lowest bias setting is set when the voltage at the base of TR5 is closest to the positive voltage rail. The bias increases as the voltage at the base is lowered (or, voltage difference between TR5's base and the power rail is increased) with respect to the positive power rail. This being because, amongst other things, the voltage across R10 is Vb(Tr5)+Vbe.
Is that right? Before now, I haven't really dealt with currents above 500mA tops, so this is a completely new area for me. The theory's the same, I'm just suitably paranoid.
Cheers,
Phil
Oh, and on first impressions...
The sound was awesome, despite being short-lived, surprisingly loud for such a low powered output, the attack was phenomenal. Literally made me jump!
Alas, there was no speaker attached.
And the smell was quite poignant (do you describe smell when reviewing an amplifier? I'm a bit new to amplifiers).
😀
The sound was awesome, despite being short-lived, surprisingly loud for such a low powered output, the attack was phenomenal. Literally made me jump!
Alas, there was no speaker attached.
And the smell was quite poignant (do you describe smell when reviewing an amplifier? I'm a bit new to amplifiers).
😀JPeitzman said:FaTTy: IC, the Mk.132 is a 3/4watt precision resistor from Caddock that gets a lot of talk for it's audio qualities. I have yet to try them but they are said to sound better than the Vishay/Dale RN's and close to or better than tantalums(?). Very clean and lots of detail, but they cost anywhere from $3-$7 USD. Borbely Audio and Welborn Labs use them in some of their premium kits.
http://www.caddock.com/Online_catalog/power/Power.html
Bottom of the page on the left, the MK Series. I have always wanted to try Caddocks and was wondering if you have tried wire wounds or anything else and liked the sound of the Caddocks, or if it was just what was available. Either way, again a very nice looking build.
"I would expect the same or even more difference with the JLH" - Paulb
I saw...edit heard, a pretty large difference moving from the 3055's to the 15003's. Keep in mind though that my 3055's where RCA's, old transistors, older than I am, of course I am just a youngin'. I was kind of surprised though that I sampled them, got them, dropped them in, and it worked just fine! I switched back and forth once, and then kept 3055's in one channel and 15003's in the other to compare them. The 3055's seemed almost to mask some of the sound, it sounded almost muffled and just as not quite right. The sound was still better than my Sony, but it was still kind of flat and muffled. The top end wasn't bad but the low end wasn't quite as tight and focused. When I went to the 15003's, honestly, this all went away. The sound was open and natural, yet clean, focused and a lot tighter, especially in the low end. The sound stage was much more 3D, the curtain was lifted, literally it sounded like something was draped in front of the speakers before with the RCA's. I noticed it a lot more too when I watch movies on the amplifier. Everything was tight and controlled, it seemed like I could connect more with the actors. The movie was also more "around me" without having more than two speakers. Just the other day I watched Smokin' Aces and was impressed, even during the chaotic gunfight scenes where there was lots and lots of sound nothing was muffled or left out, it was all there. I like it when an amp I built almost two years ago that needs lots of improvements can still impress me.
I would say if you are thinking about the 15003's, just do it and don't look back. They are a bit on the pricey side...unless you sample them for $6 per 4, $1.50USD apiece isn't to bad when Mouser has them for $3.29 each plus shipping 🙂 . I would just do it, pop them in, listen to the new wonderful sound, and not look back. Though now I have an itch to try John's version with high speed trannies, MJL3281's sounds good but maybe some 3SC5200's or some 2SC3264's with some fast and quiet Toshiba front end transistors, the JLH from hell.
Cheers
James
BTW: Paulb - You can use transistors like the MJL4281 in the ESP P3A without any problem. They are super fast, linear, and according to Mr. Elliott the P3A sounds it's best with these kinds of transistors. You could also try the well regarded 2SC5200 (watch out for fakes) or the super fast (faster then the MJL4281) 2SC3264, it's in a funny package though, Sanken MT-200 package instead of TO-247 / TO-264. You would have to use the compliments in the P3A of course as well.
Today i powered up (one chanel) JLH with sanken output transistor 2SC3264.(higher power version)
Amplifier work very stable and totaly quiet.
DC offset is very small (no fedback capacitor), i am totaly surprising 20mV
And sound is faster,more dinamik,fantastic open and clear treble.
Sanken work much beter than toshiba 2sc5200
2SC3264 is maybe the best choice for JLH
I am a happy man javascript:smilie('😀')
Hi Dary my first version was alsoi without the cap... however my DC would drift over time .... too small heatsink... Behaves much better with the cap, but I agree it sounded better...
Using this schematic http://www.tcaas.btinternet.co.uk/jlhupdate.htm
at Pic 1.
is this correct VR2 is for setting Quiescent Current and R10 to measure Quiescent Current (I= V/R) ??
at Pic 1.
is this correct VR2 is for setting Quiescent Current and R10 to measure Quiescent Current (I= V/R) ??
Correct, and according to some the amp sounds better with than without R10, but I don't recall where I've read that.
Remember that the Quiescent Current is dependent on temperature as well (read the start up procedure earlier in this thread).
VR1 sets DC offset (and this one is dependent on temperature as well).
Marco
ok thanks you.. very much...deduikertjes said:
My Amps now is back from dead...
I use WD 9.1 Speaker with 6 Ohm impedance.
Can this amps work in 10W ? Using lower voltage (15v) and lower Current because my heatsink is too hot .
I think here:deduikertjes said:
http://www.diyaudio.com/forums/showthread.php?postid=1119174#post1119174
dary...
Wow, I honestly haven't checked this thread for a very long time. I am glad to hear your JLH is playing nicely with the Sanken's! I haven't messed with mine for a while now, it is on my list but school, work, and other projects have kept me busy. I just got done tweaking on my JBOZ line amp, sweetly singing now. Right now I am in the process of starting a build of a 6V6 Musical Machine (AudioTropic). I am going to be getting to the JLH soon though I hope. I plan on building the updated version (2003) to replace my 1969 boards and get that ugly cap out of the output.
Still using the 15003's for the output, but I may have to take a taste of my own medicine now and whip out some of those others. I have given some serious thought to toying with a P-Channel JFET in the front end (with some modification of course) too instead of the BC560.
Happy listening and happy holidays everyone!
Cheers!
James
Wow, I honestly haven't checked this thread for a very long time. I am glad to hear your JLH is playing nicely with the Sanken's! I haven't messed with mine for a while now, it is on my list but school, work, and other projects have kept me busy. I just got done tweaking on my JBOZ line amp, sweetly singing now. Right now I am in the process of starting a build of a 6V6 Musical Machine (AudioTropic). I am going to be getting to the JLH soon though I hope. I plan on building the updated version (2003) to replace my 1969 boards and get that ugly cap out of the output.
Still using the 15003's for the output, but I may have to take a taste of my own medicine now and whip out some of those others. I have given some serious thought to toying with a P-Channel JFET in the front end (with some modification of course) too instead of the BC560.
Happy listening and happy holidays everyone!
Cheers!
James
