advertisement


John Linsley Hood Class A or similar

stevec67

pfm Member
I fancy building an amp, a Class A design would be interesting. The Nelson Pass designs look good but comments from Arkless of this parish suggest that all Class As operate on similar principles. This has led me to the JLH 10W Class A, which appears very simple. I have even seen prebuilt populated PCBs for under £20 a side, and kits for half that. This would make for an easy build, I can design and build a linear PSU or use a computer/printer type SMPS, play with regulators, etc, and build my own case easily enough. Is this a sensible choice for a first build, bearing in mind I haven't played with DIY electronics since I was in 6th form? I can read a schematic but I don't have a scope or a sig gen.
 
I built my first JLH on veroboard soon after the original article appeared in 1968 with the wrong output transistors and several other slight variations due to parts availability - worked a treat. At the time I had no testgear beyond an uncalibrated oscillator and a slightly iffy Cossor scope, later I was able to test it on real testgear and the distortion was better than my colleagues Bailey, the hot diy amp of the time.
 
That's encouraging. I was toying with building it on Vero myself, but if there is a PCB that will be easier. Vero layouts aren't that easy.
 
Yes, there are really cheap kits available, via diyaudio or china (try the first, first) and - it's still a really good little amp.
You can still build a damn good one 'dead-bug' style, it's only (at simplest) about four transistors a side...

Everything you need to read is here:
http://sound-au.com/tcaas/

Nelson Pass also did a nice article a decade or so on the 'PLH': his homage to it ( - that tells you something!) - but using mosfets instead: do read this, it's a great round-up and commentary too:
https://www.passdiy.com/gallery/amplifiers/the-plh-amplifier

ETA: I do think it's interesting, NP's mosfet approach doesn't actually measure as well as the simple JLH did out in the 1-10w range, by no means.
 
I'd hard wire the output stages and power supply and put the small signal stuff on a pcb. Happy to let you use the workshop if you need to measure or fault find. There's a cnc router in there if you fancy knocking up some pcb's. Be interesting to hear a jllh design hanging off a modern smps. You're going to need some big heatsinks.
 
Be interesting to hear a jlh design hanging off a modern smps.
A local mate built a couple - works really well (perhaps because the largish idle draw dwarves the output signal, something even cheapo smps thingummies can cope with)

Also fair to say, while a JLH isn't actually the quietest-possible amp: it's still true that the noise/the failings it contributes are ... really very consonant /unobtrusive in musical senses.

tl;dr: still a great 'simple' amp build and as such, still an object lesson in what really matters (pleasure in music). I'd suggest it takes a lot more complexity & finesse in layout & parts to really 'better' it if honest power needs are modest- and that's all well into the long grass of 'taste', anyway.
 
Great stuff, thanks for the encouragement gents. Thanks also David for the offer of workshop help, that's really kind. My own workshop isn't suitably equipped, CNC routers being in shorter supply than old fridges and half-rebuilt bike brake calipers. I shall have a think about the best way of tackling this. The commercial kits use 2N3055 devices throughout, I have seen versions with PNP trannies besides. Are tere any advantages to the latter? Re heat sinks, why will they need to be huge? Because it's Class A, of course, but did I read that typically a Class A amp dissipates 2x its output power as heat? That should indicate 2x20W which I should be able to heat sink fairly easily, I have a heat sink that I can cut up and hang off the back of an old DVD player case.
 
The original suggests idling at 32W per side. I guess with modern components you might be tempted to raise the voltage and current, so budget for a bit more.

For messing about you can always point a fan at it.
 
For messing about you can always point a fan at it.
I built a twin-fan platform under my Pioneer M-22, and run them at a low speed. Without them, the amp is almost unbearably (no more than a few seconds) hot to the touch. I was mainly concerned about reducing heat soak and longevity issues in the longer term. It consumes 240W to produce 30 class-A watts per channel. But, oh, such sublime watts they are too.
 
Bob pease used to use old vw cylinder heads as heat sinks.
Second hand CPU coolers can be had for free from your local wee dump.
It would be fun to make something with some unique character.
 
It would be fun to make something with some unique character.
Good plan. I shall build a Stirling engine to dissipate the residual heat and blow a fan across the other components. The hotter the amp gets, the faster the engine will run and the more airflow will be generated for cooling. How neat is that?
 
I fancy building an amp, a Class A design would be interesting. The Nelson Pass designs look good but comments from Arkless of this parish suggest that all Class As operate on similar principles. This has led me to the JLH 10W Class A, which appears very simple. I have even seen prebuilt populated PCBs for under £20 a side, and kits for half that. This would make for an easy build, I can design and build a linear PSU or use a computer/printer type SMPS, play with regulators, etc, and build my own case easily enough. Is this a sensible choice for a first build, bearing in mind I haven't played with DIY electronics since I was in 6th form? I can read a schematic but I don't have a scope or a sig gen.


What speakers do you want these amps to drive?

If you just fancy buiding something I wouldn't go for a class A amp.
I've built a few NP designs and I really don't rate them soundwise.
I built a JLH on home-dremmeled PCBs. Pretty good sound, but not better than many class AB designs. Not really worth the hot/large heatsinks IMO.
Build a Qudos. At least if you don't like it they will be easy to sell on.
For something completely different, how about a KEL84:
http://www.world-designs.co.uk/power amps.htm
 
I fancy building an amp, a Class A design would be interesting. The Nelson Pass designs look good but comments from Arkless of this parish suggest that all Class As operate on similar principles. This has led me to the JLH 10W Class A, which appears very simple. I have even seen prebuilt populated PCBs for under £20 a side, and kits for half that. This would make for an easy build, I can design and build a linear PSU or use a computer/printer type SMPS, play with regulators, etc, and build my own case easily enough. Is this a sensible choice for a first build, bearing in mind I haven't played with DIY electronics since I was in 6th form? I can read a schematic but I don't have a scope or a sig gen.

Absolutely not! I said the JLH and certain Pass designs do, and I was thinking mainly of the PLH... there's also a similar idea on the Elliot Sound site.
 
I heard one driving Quad ESL57s in comparison with Quad IIs and a well regarded amp of the time that I cannot remember (this was 1977 I think). The JLH class A was clearly the best of the three and I told myself I'd build one, one day... that one day was probably about 15 years later, and I knocked one together from scavenged parts, and got it working beautifully. It alternated in my system with my Leak Stereo 20, and probably, at the time my Nytech 602. I got it out again to run the mids of my Isobariks for a time, but it went wrong, and it was easier to just use my Naim power amps that I then had so it never got repaired. I have on occasion toyed with building another, but the environmentalist in my is a bit reluctant. It does chuck out a lot of heat!
 
My first amp, a long time ago! Used it with some Richard Allen home mades (RSDD 3 and RSDD 10 if I remember) and then some Tannoys. Sweet little amp.
 
Thanks all. I still fancy a Class A, I've not had a Class A tranny amp. The KT88 design described above doesn't do anything my EL34 doesn't, and I'm sure the Qudos is great but it's a bit too, well, polished. If I wanted something like that I could but a NAPxxx and build regulators, power supplies and the like, or just buy a completed AB amp. All these are rather "cheque book solutions" and don't inspire me. I want to build something interesting that isn't available off the shelf. I have a very nice heatsink to hand and a case, if I can run it from 2x SMPSs then it will cost loose change. I like the idea of building something competent from "objets trouves" and a couple of boards.

In addition the JLH is a classic. What's not to like? The commercial kits often use 2N3055 output devices, which again are design classics but apparently you can and should do better. I shall look out a version with better devices.
 
I built the "ultimate version" from the TCAAS, with the long tail pair input arrangement, and made several further improvements of my own including using fast Japanese output devices. It was about 20WPC and had fan cooling. Sounded very nice indeed!
 
Even the capacitively coupled output version is still a good sounding amplifier, what it gets right it does well and what it doesn't, lack of power, slight bass roll off isn't annoying. Being class A it can tolerate low spec output devices like the 2N3055 and 2N3773, so long as you select for decent gain. If you want something a bit newer then high current low Ft devices like the MJ802 MJ15003 type devices would be my choice, high Ft devices are not necessary if you are running in class A and you may hit stability problems if layout is not optimal.
 
You do need output devices of moderate gain, dodgy MCP 2N3055s which didn`t meet the gain spec produced poor distortion and low output power. The original Wireless World article gives a table of results for transistors of differing gain.
 


advertisement


Back
Top