Class A power amplifiers

[Reid Malenfant]

I was refering to the circuit itself!
nothing wrong with output transformers when their designed correctly, also they provide excelent isolation for your expensive speakers when things go tits up further down the chain, unlike horrible directly coupled transistor output stages.

Ok Anthony, yes i do agree that transformer isolation is a good thing But an output capacitor will give the same protection in reality.. Not that i'll be using one

These will be over engineed. I have 4.5KW (dissipation) of transistors for each class B amp & 1.8KW on each class A section. They'll run at a cool temperature under normal conditions, i figure the heatsinks will be about 10C above ambient with a 4 ohm load & 11 Amp quiescent. Only 5.5 Amp for 250W 8 ohm allowing for a 6 ohm load..When driven though this will increase obviously, but this is a huge amount of heatsinking even for a 500W RMS class B amp. So going to H class with the rail switching will definately keep things cool

There will of course be some kind of protection for the loudspeaker if things go wrong

[Reid Malenfant]

Got to admit i slightly prefer the sound of Quad '50's (Output Transformer) againt the '303' (Output Transistors)

Aye, the 303 is a class B amplifier though - the same as the 50 With a transformer output it'll tend to filter out a good deal of the upper harmonics of any low level distortion. This might make it sound a little mellower as well as reducing the output damping factor & giving a very slightly fatter bass

[anthonyTD]

Ok Anthony, yes i do agree that transformer isolation is a good thing

But an output capacitor will give the same protection in reality.. Not that i'll be using one

These will be over engineed. I have 4.5KW (dissipation) of transistors for each class B amp & 1.8KW on each class A section. They'll run at a cool temperature under normal conditions, i figure the heatsinks will be about 10C above ambient with a 4 ohm load & 11 Amp quiescent. Only 5.5 Amp for 250W 8 ohm allowing for a 6 ohm load..When driven though this will increase obviously, but this is a huge amount of heatsinking even for a 500W RMS class B amp. So going to H class with the rail switching will definately keep things cool

There will of course be some kind of protection for the loudspeaker if things go wrong

I'm not even going there!!!
but full respect on your design.
look forward to hearing about it when its finished.
Anthony,TD...

[anthonyTD]

Aye, the 303 is a class B amplifier though - the same as the 50

With a transformer output it'll tend to filter out a good deal of the upper harmonics of any low level distortion.

This might make it sound a little mellower as well as reducing the output damping factor & giving a very slightly fatter bass

I agree [partialy]
but if the amp has sufficient feedback then the damping factor should still be sufficiently low enough for the amp to produce a decent bass response.
Anthony,TD...

[Reid Malenfant]

I agree [partialy]
but if the amp has sufficient feedback then the damping factor should still be sufficiently low enough for the amp to produce a decent bass response.
Anthony,TD...

No arguement there Anthony That's why i said "very slightly"

[anthonyTD]

No arguement there Anthony That's why i said "very slightly"

[DSJR]

The only ss amp with transformer coupling I remember hearing was the old AR amp towards the end of its commercial life (1975 or so) and very lovely it was too. The few of these (and even more so the tuner) that come up tend to fetch good money I think.

[Rare Bird]

The only ss amp with transformer coupling I remember hearing was the old AR amp towards the end of its commercial life (1975 or so) and very lovely it was too. The few of these (and even more so the tuner) that come up tend to fetch good money I think.

was once on auction early this year around £80 i think it went for, they are nice your right Dave

[Reid Malenfant]

I agree [partialy]
but if the amp has sufficient feedback then the damping factor should still be sufficiently low enough for the amp to produce a decent bass response.
Anthony,TD...

Actually having thought about this a little more carefully i doubt that many people would be able to detect the difference between an amp with a damping factor of 25 & one of 1000 With a well designed speaker you might be looking at less than 1Db in difference near the speaker bass resonance. It's only when you start going much below about 25 that things might get a bit bouncy in the bass end

The only ss amp with transformer coupling I remember hearing was the old AR amp towards the end of its commercial life (1975 or so) and very lovely it was too. The few of these (and even more so the tuner) that come up tend to fetch good money I think.

Macintosh make some very nice looking (& performing as well apparently) transformer coupled solid state power amps. A little bit on the expensive side for most people though

[Reid Malenfant]

Ok, time to resurrect the dead so to speak

Some of you may recall that I bought a couple of class D amplifier modules a while back, which you can see here I really should get back to that with my findings in actual fact, but for now lets just say that the heatsinking was totally inadequate & the amp would only stay alive with music signals. It certainly wouldn't last with a test tone due to the H bridge switching Mosfet having a totally insulated case & a maximum power dissipation of something like 18W. That'd be with a 25C heatsink to You don't need to do math to know that if the amp can output 200W RMS & is 90% efficient that 20W is being lost somewhere...


The whole idea of this is to make an efficient high power class A amp, a misnomer if ever there was one My original intention was to build a class H rail switching amplifier to increase the efficiency of the class B amplifier, the class A section will be just about as efficient as possible due to it's low rail voltages.

Well time doesn't stand still & it was only a question of time before yours truly spotted something quite interesting There have been a few more developments of that class D amplifier & the latest version uses two very fast standard packaged TO220 Mosfets. These are each capable of up to 100W of dissipation so a crazy idea crept into my head...

Enter the L25D Class D amp..

Like everyone on here I hate spending money unless I absolutely need to, so I sat down & had a good think about what exactly I have here that could make six class A amps a reality As it turns out I realised that I have over twelve 0 - 22, 24, 26V 500VA transformers & a couple of boxes of BHC ALS 10000uf 40V screw terminal capacitors. That'd be the class D amps sorted, as long as I go & run them in a bridge I'll get the kind of output power I'm looking for (200W RMS 8 ohm, 400W RMS 4 ohm). All the heatsinking I need for the class A amps, power transistors, transformers, capacitors etc.

So I make a couple of bridge amps with the class D amplifiers driving the zero volt rails of the class A amps & connect the speaker between the two class A amps... Sounds easy, not really

Though it shouldn't be impossible, there are is one problem that i'll need to overcome. As the frequency increases there will be a gradual phase shift of the output of the class D amps due to the output reconstruction filter. This appears to be centred at 45KHz. Now either I could put some kind of lead network on the class D amps (possibly the best idea), or I can put an all pass delay network onto the feed of the class A amps...


Anyway, I bought six lots of those kits which is twelve amplifiers in all to make six bridge amps. Looks like someone didn't read the application notes of the IRS2092, so it's a good job I realised I'd have to modify them to run at lower voltage & in a bridge I found a few errors which could have compromised the amps, which will be rectified before they get soldered together.

Each of the class D amps will have a 0.8C/W heatsink to keep the power Mosfets cool & with an expected 20W of dissipation they should certainly stay cool. Each half bridge of class A amps will use 4 x 300W power transistors mounted on two 0.5C/W heatsinks. Current needed for 400W RMS class A 4ohm push pull is 7.07A, with +/- 4V rails these will run cool as well...


This is going to take some time, first of all I need to get these class D amps built & tested - all twelve of them At least it'll give me something to do over the winter months....

Should be fun at any rate See you in another year


E2A:- I should say that unlike the L15D the PCBs of these kits come with the IRS2092 pre soldered to the PCB. It's not exactly fun soldering surface mount components without a tiny soldering iron bit. I managed it previously though with little difficulty, but I'm used to that kind of work.. Most are not