Where did I go wrong?

[Cup of coffee]

Hello,

In 2000, I designed and assembled my first amp. Basicaly, I took a common class A transistor amplifier and scaled up the powersupply. I didn't have Internet access and most books talked of Class A theory, and went on to show Class B and AB with circuits. I don't have most of my drawings and math because I used bits of scrap paper and had to re-calc this and that to get what I needed. In the end I had assembled a single low power stage driving a power follower. With a supply of 36 volts at 2.5 amps, I put together a circuit that had a lot of Hum which I tamed with an extra resistor in series with the potential divider and took a 10,000uf cap to gnd. To block Hum on my output I used a 4700uf cap in series with the Loudspeaker, without it, the driver cone was pushed out to the end of its travel, in response to the DC on the output. OUCH
I used big 25watt resistors bolted to an old amplifier heatsink with the Follower transistors bolted to another heatsink. Both heatsinks were capable of handling the 20Wrms P/Ch of Class AB but not for a Class A, probably pushing a watt, I don't know.
When I put 2 together for stereo, I had a radio as sorce and 2 odd loudspeakers to listen with. I tell you, I could here more detail, than my main hifi. Queen's" Under Pressure", was played by a commecial station, and I could here sound I never heard before. I made the circuits safe, and took it from my little broom cupboard to the lounge, and wired in my CD player. WOW!
I had to listen in bursts of 10 minutes, because I was worried about over heating. I tried to cool it with a fan, but it was too noisy.
My partner and I both agreed it sounded way better than anything we'd listened to before. 2 weeks later and Thermal Runaway or similar, killed it. I was gutted.
In 2007, I want to do it again. And get right what I had done wrong.
Please can you give me some feedback. I think a lot of energy was wasted with the hum it suffered, I think the Hum contributed to the heat, let alone the fact its Class A anyway. The 10,000uf was an addition to the smoothers stacked on to the powersupply.
Also, I worry that if I build, say Rod Elliotts Mosfet Follower it won't sound as detailed, because he says his AB amp has less distortion, if I remember right. I was supprised to hear more detail, even through the Hum, which it suffered.
So please, give me your opinions. This circuit supprised me very much and although the circuit doesn't tell much, it was all I could find, my powersupply was poor as was my earth wiring. There was no star earth.Naughty.
Of the simplicity of the design, having only a few parts means I can spend more on better quality parts.

Cheers

Cup of coffee?

[richie00boy]

Maybe it's just you have never heard a decent AB amp?

For a start power supply rejection is almost non-existent (as you have found with the hum problem) and it is woefully prominent in distortion when driving a speaker. Then again maybe that's what you perceived as 'detail'?

[Dzsoni]

Hi,

DC blocking cap is on the wrong side of load resistor.

[Craig405]

Hi,

Thought id chuck in my observations, im not an expert in electronics but your circuit diagram is a bit flawed.

Your using the power stage for gain, and the driver stage as a buffer, thought this would work better the other way round.

The 4700uf cap on o/p is connected straight to the PSU, connect it to the o/p transistors collector!

I think the 10000uf cap would be better placed on the supply rail then the whole circuit can benefit from it.

having so much resistance in the output current path is a bad idea too for a power amplifier as zout is Rc, and the emitter resistor will only reduce your voltage swing headroom, the stage is not really suited to power output.

In fact it all looks a bit dodgy to me, suprised it sounded ok .. the first decoupling cap isnt needed in a follower i dont think as it will only short out the emitter more at higher frequencies.

Im sure some other people can help more, probably correct me a bit lol..

Regards
Craig

[destroyer X]

Reduce your current to a half of that, those transistors do not "really" dissipates 100 watts.....normally they hold 35 watts without problems.

You will succeed..... go ahead please.

Very good work...simple things uses to sound fine.

There are hundreds of forum folks that apreciate those minimalist designs.

Try to produce a separate sub circuit to bias this darlington one... than you will be able to produce a voltage amplifier using the first stage...no problems, you can couple with condenser and the sound will remain nice....diodes will provide you 1.2 volts (2 units) and them you will switch one your darlington and will control it..including the half of supply voltage into the colector.

The power will be small...1 amp and 18 volts into the colector.... 2 clean watts..... will play loud..do not worry, just use sensitive and soft speakers for that.

The first stage will need 5 volts and 2 miliamps., very small gain needed (colector resistance divided by emitter resistance)... easy to produce.... the input will be 750 milivolts to 2 volts maximum..so..you can calculate.

Good idea man...simple...will surprise many folks related sonics.

The current reduction will also help you related hum.

regards,

Carlos

[Cup of coffee]

destroyer X

Thanks for the confidence, I haven't dabbled with an iron for years.

Can you explain the diode thing? I used the potential divider in the original design, to bring about bias.

You wrote

you can couple with condenser and the sound will remain niceyou can couple with condenser and the sound will remain nice

Can you explain condenser, and why did you have a relay and variable resistor added to schematic?

Cup of coffee?

[Cup of coffee]

Sorry about the repeated bit.
I wanted to change colour of text and got asked something to be allowed. Is it Vista or me?

Cup of coffee?

[Cal Weldon]

It was likely neither, it was probably that you were under moderation. You are free to post in whatever colour you like now.

[Cup of coffee]

Do we do coffee?

[Cup of coffee]

I'm reading about the Zen amplifier,
http://sound.westhost.com/project36.htm
The circuit seems loaded with components to a point I wouldn't say was minimalistic, and yet it has popularity. I havn't learnt much about the bias thing yet, but I can't help thinking I should just join the Pass club and make one already designed.

I see a large cost in power supply and heat sinking, and with wanting to use audio grade parts, the minimalistic approach seems to be sensible. Obviously without hearing one, I can't decide. Of course, we all want the best, the less the parts count , the more can be spent. Yes, I could also build an AB, as richieOOboy pointed out perhaps I've never heard a decent one. I've heard several top designs at the 1998 Hifi show in London, okay the acoustic surroundings were different to home, I wasn't blown away. The valve I heard glowing to Pink Floyd's Dark Side of the Moon, left an impact on everything else I heard. It was Class A. Okay it was a valve, if only I could remember its name!Doh, but surely a Single Ended transistor approach has got to sound good? Don't you think?

Cheers

Coffee anyone?