JLH 1969 10W Class A amplifier

[zog]

OK I've wandered off the "Class D" reservation and started building this old, simple Class A design from a Chinese kit.

 

I was going to post this in the "Cheap EBAY Class A" thread but my ebay order vanished, so I ended up getting the same kit from Alieexpress.

 

Enclosure is re-purposed from a crappy old busted Korean amp I found in the skip.

 

 

I just got one channel built and working, I have a big chunky Altronics Aussie made heat sink but I havent drilled it yet, so I'm using the heatsink from CrapAmp for test purposes - I had to pile some old PC heatsinks on top to stabilise the temp. The mutimeter is in temperature mode so it has stabilised @ 54 degrees.. and it's a HOT day today.

 

Link to the original reference docco, hosted on Elliot Sound Products site: http://sound.au.com/tcaas/index-1.htm

Another link: http://swansonc2012.blogspot.de/2012/05/jlh-class-amplifier-1969.html

 

This amp fairly easy to power, as it uses a single ended 24DC power supply. In fact a lot of people report very good results using a SMPS, so I got the best specced 120W SMPS I could find at reasonable price, an Altronics unit. This has made the build massively easier than it would have been.

 

I thought I'd try a Class A amp after borrowing a (very well made DIY) 5W SET amp and trying it on the HF section of my active speakers. Was pleasantly surprised as I didn’t expect much - but it's the only bit of gear I've swapped in that my wife has made a positive comment about the sound (and I matched EQ and levels exactly with miniDSP and measurement mic)... so seeing what I can do for a low cost solid state Class A amp to try.

 

 

I've left the kit "stock" except for the coupling capacitors, replaced the 1uF "Spirit" brand caps with ERO MKT8122 2.2uF caps.

I'll shorten the leads to the power transisters, ultimately I'll use one of these mounted across the front of the enclosure:
http://www.altronics.com.au/index.asp?area=item&id=H0545

Hoping the total build cost to come under $200 - plus a lot of fun fiddling with it.
I havent adjusted the voltage or resistors to bias it, although it appears to come with parts chosen for exactly 24VDC which is what I'm feeding it.

 

[davewantsmoore]

Exact same output at speakers... can pick the difference?   You could earn $10,000.   Heh.  Just kidding.

 

As you'll know... this is a big part of what happens when you're feeding a 100w amplifier such a small input V required for only < 0.5w   (noise and distortion suffer).

[zog]

I'll fully confess it wasn't double blind or scientific at all - but I think I have heard differences between various different class D modules I've tried in the 30-50W range and the $6 class AB chipamp I've also been playing with.. but these units have had measurable differences in the 2HD and 3HD profiles (measured with REW and a voltage divider tapped off the speaker).

[LogicprObe]

When I built one, years ago, I used the fan forced tunnel heatsink.

 

Like this one.

 

[thoglette]

When I built one, years ago, I used the fan forced tunnel heatsink.

 

Nice.   The JLH I have is living with a much, much smaller conventional heatsink (it was made in the 70s)  but is running about 20C above ambient.   

[LogicprObe]

I didn't have full voltage on the fan so it ran slow and quiet.

[Sub Sonic]

I'll be watching this build with great interest. :-)

I'm currently going through the throes of picking the next class A kit to try and this one was on the short list. However, I have a large enclosure with large heatsinks so the F5 is looking good.

Keep the instalments coming :-)

Regards,

SS

[3G_]

I'll be watching this build with great interest. :-)

I'm currently going through the throes of picking the next class A kit to try and this one was on the short list. However, I have a large enclosure with large heatsinks so the F5 is looking good.

Keep the instalments coming :-)

Regards,

SS

Hi Sub Sonic,

I am running Pass Labs Aleph J amps. My system should b hooked up and going again in a few weeks if you want to listen to some different Pass Labs designs.

[zog]

small update: tested the other channel, and all good.

 

one one channel using 40 - 45W at the powerpoint.

 

was using a low sensitivity sacrificial DSE speaker so no idea of the noise level yet, which will be important for the 110dB (103dB with LPAD) compression drivers I intend to drive with it.

 

next step,  drill holes in the big heatsink and mount it to the "Pioneerish" aluminium faceplate. there are big holes in the aluminium where I can mount the transistors directly on the heatsink, but hopefully the extra black aluminium will help anyway:

http://pix.minirig.org.au/main.php?g2_view=core.DownloadItem&g2_itemId=6025&g2_serialNumber=2

 

 

 

 

[tassie tiger]

Hi zog,

 

Contemplating modest-power Class A to pair with full-range Fostex drivers in back-loaded horn. Currently using Quad2 monoblocks built in late 1950's/early 1960's, so a 1969 JLH seems appropriately old.

 

Do you find any problems with the SMPS? e.g. hum?

 

Thanks,

J

[zog]

Do you find any problems with the SMPS? e.g. hum?

 

Right now I do BUT, I'm in the "quickly wire it together and get it working" phase.. I havent optimised the wiring at all, and used whatever junk cable for the line inputs, soon to be replaced with shielded cable.

 

I also wired up overly long leads to the power transistors to help while I tested the heatsink, and enclosure assembly stuff, which wouldnt help.

 

It also seems hyper sensitive to what kind of device I hook up to it, but again, wiring and layout is not the final form.

 

With 24 VDC it is consuming about 85W on the wall. I've dropped the SMPS down to 22VDC (the minimum, it has a trimpot to adjust) and and it's using 70W. I guess that means I'm closer to 8W of power, but that's fine.

 

Temperature with all 4 transistors on the large heatsink acceptable.

 

My first try at using an M3 "tap" to thread drilled holes in the heatsink worked out OK as well.

 

Between the two power transistors I measure 10.4VDC which is 47% of 22VDC, not exactly 50%, need to determine if "close enough is good enough", or get that to exactly half.

[LogicprObe]

Only going by memory here..........but don't you have to take into account the voltage drop across the junction?

[thoglette]

Between the two power transistors I measure 10.4VDC which is 47% of 22VDC, not exactly 50%, need to determine if "close enough is good enough", or get that to exactly half.

 

It'll do.   That's what the output capacitor is for.

 

The main areas for tweaking are the various capacitors and power supply options

 

On power supplies, the JLH at home has a vintage regulated power supply (separate box).  

 

I've been itching to try a choke input supply but I've got to build my Hay bridge inductance meter first.

 

If you've not already noticed there's a pile of JLH stuff out there on the web, including an Australian one http://sound.au.com/tcaas/

 

(Some interesting distortion spectra images here )

Edited February 6, 2014 by thoglette

[zog]

If you've not already noticed there's a pile of JLH stuff out there on the web, including an Australian one http://sound.au.com/tcaas/

 

(Some interesting distortion spectra images here )

 

Yes, I linked that in the first post . As I was waiting 2 months for the $30 kit to arrive I think I googled every JLH web page and picture I could find.

Finally got the chassis assembled, it's a bit of a puzzle trying to get the power transistors mounted inside the front panel with the heatsink snug in such a way that I can disassemble it in future but I think I've figured it out. And the black aluminium front panel now acts as part of the heatsink.

 

I initially had a LOT of buzz and noise, but after carefully going through:

http://www.diyaudio.com/forums/diyaudio-com-articles/163575-audio-component-grounding-interconnection.html

and

http://sound.westhost.com/earthing.htm

I redid where the ground loop connector was, and more importantly studied the PCB details here:

http://www.aliexpress.com/item-img/SC-HOOD-JLH-1969-10W-10W-Class-A-amplifier-kit/754337698.html

and realised I had the audio input inverted (as the orientation directions on the PCB are covered once you solder on the connectors)

I now have it running silently with no signal, and sounding nice and clean.

 

Still need to drill the chassis for the PCBs and tidy up the wiring. But first to integrate it properly with the active system with some miniDSP adjustment so I can have a listen.

 

[henry218]

cool   

[zog]

This is the PSU:

 

http://www.altronics.com.au/index.asp?area=item&id=M8410

[thoglette]

Yes, I linked that in the first post . As I was waiting 2 months for the $30 kit to arrive I think I googled every JLH web page and picture I could find.

Ah, yes :o .

 

There's a lot out there, that's for sure.  Between the JLH and the various Hiraga there's months of reading....

  

 

Finally got the chassis assembled, it's a bit of a puzzle trying to get the power transistors mounted inside the front panel with the heatsink snug in such a way that I can disassemble it in future but I think I've figured it out. And the black aluminium front panel now acts as part of the heatsink.

Nice renaissance of a dead chassis.  

Q: What material have you mounted the input RCA connectors on?

[zog]

cool   

 

It's 28 degrees inside (on a 39 degree day) - I measure 57 degrees at the top of the heatsink, so cool enough I hope!

 

 

 

Nice renaissance of a dead chassis.  

Q: What material have you mounted the input RCA connectors on?

 

Yes the "Glenwood" of Korea seemed quite junky, but its chassis has worked out well. The blue plastic the RCA connectors is mounted on is a bit cut from a plastic paint scraper cut to size. This was chosen as it happened to be suitable, and within reach in the garage at the time.

[Sub Sonic]

There's a lot out there, that's for sure.  Between the JLH and the various Hiraga there's months of reading....

 

 

  How right you are

 

I've been reading the JLH, Hiraga and now Pass threads, one of the Pass threads is over 1K pages long!!!! Still working on that one. Lots of good info though!

 

SS

[zog]

And it's finished. It sounds really good. And it runs more than quiet enough to be drive the compression drivers.

Looks suitably black and sinister with the lid on.

 

 

[Sub Sonic]

So... how does it sound?

 

Regards,

 

SS

[zog]

So... how does it sound?

 

Early days yet, no A/B level matched comparisons or measurements .. and I've swapped over from a $6 Class AB 15W chipamp that I've been tinkering with since Xmas, and I'm fully conscious "new toy" syndrome may be colouring my judgement here..

 

But so far GREAT. Luminous. Big. Pleasant.

Electric guitars sound wonderful.

 

(using this as the HF amp crossing over @ 600Hz, with a TA3020 Tripath amp, approx 120W, as the LF amp)

[LogicprObe]

A lot of great guitar amps used horn speakers.

They added to the expense and some nong headed roadies didn't know how to mike them up.

(even though most of these high end amps had a line out.)

[tassie tiger]

It'll do.   That's what the output capacitor is for.

 

The main areas for tweaking are the various capacitors and power supply options

 

On power supplies, the JLH at home has a vintage regulated power supply (separate box).  

 

So are there any sonic advantages with 'traditional' toroidal transformer + regulated PSU over the SMPS option that Zog's gone for?

Has 'traditional' design been overtaken by technological advancement in SMPS design, or are they both good options but different?

I'm new to this game, and keen to learn.

 

Thanks for any advice

Edited February 18, 2014 by tassie tiger

[thoglette]

So are there any sonic advantages with 'traditional' toroidal transformer + regulated PSU over the SMPS option that Zog's gone for?

Has 'traditional' design been overtaken by technological advancement in SMPS design, or are they both good options but different?

I'm new to this game, and keen to learn.

 

Thanks for any advice

Traditional would be EI cored transformers with split bobbins. 

 

Torroids have some fundamental noise problems (best summed up by Rod Coleman about halfway down the first page of this DIYAudio thread ) 

 

And SMPS still (and by definition) radiate bucket loads of noise.

 

As I've said elsewhere, I'm keen to try the JLH with both batteries power and with a choke input supply.  

 

The later is having a bit of a renaissance in the tube world as it both generates less noise and self regulates much better than capacitance input supplies.