|
Success Stories If you have successfully repaired or restored a piece of equipment, why not write up what you did and post details here. Particularly if it was interesting, unusual or challenging. PLEASE DO NOT POST REQUESTS FOR HELP HERE! |
|
Thread Tools |
19th Jun 2020, 3:15 pm | #1 |
Nonode
Join Date: Aug 2010
Location: Resolven, Wales; and Bristol, England
Posts: 2,614
|
1969 Linsley Hood success & questions
Yesterday I just about completed my rendition of a classic audio amplifier - the Linsley Hood design from 1969 with his modular pre-amp from Wireless World in the same year. It sounds exactly as I hoped it would - transparent, and capable of driving my Keesonic KBM loudspeakers.
The power amp boards were kits purchased from Bang-good and they are as supplied except that I replaced the output capacitor with something more suitable. They use TIP41C power transistors and set up really easily. The pre-amp is built on PCBs I designed and uses the JLH design pretty much intact. There's a couple of problems I need to resolve which you might be able to help with: If I connect the balance control (100Ω) I get a loud earth hum on both channels although it does alter the balance. I've used a single-point earth scheme throughout but altering the earth point for the centre of the balance pot doesn't make a blind bit of difference. The output transistors run hot. Alright I know class A runs hot but these are running 30V at only 1A per channel and they are too hot. The top of the heatsink measures at 66º. My heatsinks may be too small or inefficient, I've read Rod Elliott's treatise on heatsink design but I think I've failed to understand something. Any comments or ideas will be very welcome!
__________________
Richard Index: recursive loop: see recursive loop |
19th Jun 2020, 4:11 pm | #2 |
Octode
Join Date: Dec 2017
Location: Morden, Surrey, UK.
Posts: 1,561
|
Re: 1969 Linsley Hood success & questions
Your heatsinks look too small for a JLH though I haven`t done the calculations. TIP 41s seem a bit marginal as well, though probably within their ratings.
You are dissipating 30 watts on each heatsink, that`s not trivial. Do you have a clear path for airflow through the heatsink "chimney"? As a general rule of thumb transistor reliability halves for every 10 degree rise in junction temperature above 20 C so cooler is a lot better. edit:- A quick fag packet calculation suggests that your measured 66 C is about what you might expect, though rather more than is desirable. Last edited by barrymagrec; 19th Jun 2020 at 4:22 pm. |
19th Jun 2020, 4:55 pm | #3 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,902
|
Re: 1969 Linsley Hood success & questions
It will get worse if you put a cover on it because you'll lose convection on the inside fins.
For a class-A design of that power you've got about half the minimum size of heatsinking advised for good transistor temperatures. Be careful also that the plastic packages don't have the power ratings and the thermal resistances of the old TO-3 metal cased ones - the single mounting hole allows the metal tab directly under the die to come away from the heatsink a little, greatly reducing the contact pressure. In my early days of building amplifiers, I had the preamps on a board with wires going to the pots and switches on the front panel. THis gave me endless trouble trying to get rid of noise and even to keep the thing stable. One of the worst things I did was to have the mode selector switch also switch the feedback networks for the first section, so that it gave different flat gains and switched in an RIAA network. Keeping it stable was a nightmare, and it tends to give big pops and bangs on switching, if you don't add high value resistors to equalise the DC on sections not in use. In the end, I learned to build an RIAA stage as a full time separate thing, to do switching either with passive mode J-fets or miniature relays and to keep tone controls etc close by the board. Transistors are cheap. Fighting problems created in minimising the transistor count as we all used to do is expensive in mental wear and tear. Fans are noisy. I'd consider mounting the power amps separately out of the way, and design the case for each to include a tall chimney. Much quieter than fans and surprisingly effective. If you're throwing away energy in the shape of heat, it's elegant to make it pay for its living by creating its own cooling air flow. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
19th Jun 2020, 4:57 pm | #4 |
Guest
Posts: n/a
|
Re: 1969 Linsley Hood success & questions
66C on the heatsink is probably 80 or 90 at the transistors junction, well within spec (normally 125C). The life will be reduced from perhaps 1000's of years to 100's and a new TIP41 is a lot less expensive than a larger heatsink. Also the capacitor coupled output will protect the 'speakers if one fails.
I would leave it as is. |
19th Jun 2020, 7:44 pm | #5 |
Octode
Join Date: Sep 2017
Location: Bristol, UK.
Posts: 1,659
|
Re: 1969 Linsley Hood success & questions
In his original article, Linsley Hood is very non-committal about thermal design, "suggesting" rather vaguely "a pair of 5 inch x 4 inch finned heat sinks" per channel to dissipate the 17W of heat. He then also provides a (very poor) photograph of these enclosed in a metal box which would severely obstruct any efficient airflow.
I have never known what to make of that, but the OP's arrangements do at least seem vastly better than JLH's! It's difficult to gauge thermal resistance (just measuring surface temperature with any precision is fraught with problems), but my wild guess would be that JLH's heatsinks are no better than about 1.0-1.5 degr. centigrade per watt each - in free air - which they aren't. Mike |
19th Jun 2020, 8:35 pm | #6 |
Octode
Join Date: Dec 2017
Location: Morden, Surrey, UK.
Posts: 1,561
|
Re: 1969 Linsley Hood success & questions
The salient feature of JLHs original heatsinks is that each output transistor has it`s own heatsink - this obviously halves the effective temperature rise, also TIP 41s have a much poorer thermal characteristic compared with the original TO3 types.
|
19th Jun 2020, 9:09 pm | #7 |
Nonode
Join Date: Aug 2010
Location: Resolven, Wales; and Bristol, England
Posts: 2,614
|
Re: 1969 Linsley Hood success & questions
Bearing in mind that the build was precipitated by the purchase of a couple of low-cost JLH class A amplifier kits and that I wanted to make the whole using as many junk box parts as possible...
The TIP41C transistors are operating within their SOA curve, agreed without much margin of error. I should be able to run the transistors at 85º case temperature without catastrophic problems and the projected enclosure is intended to provide unobstructed airflow to the heatsinks. Perhaps I should paint them matt black, I've heard that helps? I'm sure I can keep it within the bounds of reasonable temperature degradation and since it pleases my ears I think I'll keep it.
__________________
Richard Index: recursive loop: see recursive loop |
20th Jun 2020, 12:08 am | #8 |
Dekatron
Join Date: Nov 2003
Location: Heckmondwike, West Yorkshire, UK.
Posts: 9,643
|
Re: 1969 Linsley Hood success & questions
Scrap sky boxes contain a nice quiet little 12v fan.
|