Ripple voltage problem, or not?

[event horizon]

Hello good people, here is a question for you all

I'm going to be building a few very low voltage but high current class A power amplifiers. Now these are going to be split rail @ about +3.7V & -3.7V with regards to ground. The thing is with a quiescent current of up to 10.5A the ripple voltage on the capacitors i happen to have will be pretty big.

I reckon that i should have enough BHC Aerovox 15,000uf 10V caps to give me 45,000uf per rail, so 6 caps total - 3 from ground to the + rail & the same from ground to the - rail. However this isn't enough to give me the ripple voltage i want.

So, what would happen if i all of a sudden place a 120,000uf accross from the + to - rail? This would certainly give me more than enough capacitance to get the low ripple voltage i need, but will it work or are there going to be any odd side effects?

Please assume that any rectifiers i use will cope with the current pulses, i just need to know what this will do (having never contemplated this previously).

Thanks in advance

Mark.

[richie00boy]

It will reduce ripple, but you will not be able to draw any current from it as it's connected common mode and your load is differential mode.

[peranders]

Excuse me but +- 3.7 V? What sort of load did you have mind? I'll suspect that you have got the class A thing wrong.

If you want to use a 8 ohms speaker the power will be (3.7*3.7)/16 = 0,856 W only!

[richie00boy]

Hint P-A... 10A quiescent current. More like 0.5 ohm load I'd guess.

[event horizon]

Originally Posted by peranders

Excuse me but +- 3.7 V? What sort of load did you have mind? I'll suspect that you have got the class A thing wrong.

No, i'm aiming to do what i have done previously but at higher power

A push pull class A output stage driven by a class B or AB similar to a certain technics design. The actual output will be about 250W RMS into 8 ohms & 500W into 4 ohms. Allowing for a 3 ohm impedance you'll find you need 10.5A quiescent current

I'm sorry if i didn't make things clear, however i could have been driving a ribbon

By the way, cheers richie00boy It doesn't matter about anything other than getting that ripple voltage low.

[richie00boy]

OK good luck with the project, but I do believe the class-a driving class-b idea is fundamentally flawed and will only ever give class-b results.

[peranders]

10 A, OK but what supply voltage did you have in mind?

[event horizon]

Originally Posted by richie00boy

OK good luck with the project, but I do believe the class-a driving class-b idea is fundamentally flawed and will only ever give class-b results.

Thanks, i'd like to ask you though if you have ever listened to the original Technics amps that used the real class A output stage? I know i haven't as they cost many thousands back in the day.

However, i have listened to a few of my own of both push pull & single ended variety. Likewise with what you'd call a proper class A amp with the full rail voltage accross the output transistors with 100W RMS per channel @ 8 ohms.

I'm not trying to save the earth but seeing as i get the same quality but by using less electricity (& therefore paying a smaller bill) as well as being able to run it in the summer

Besides, there are going to be 6 of them - so every little helps

[event horizon]

Originally Posted by peranders

10 A, OK but what supply voltage did you have in mind?

+/- 38V with a balanced amplifier, or if you like a bridge So allowing for component losses i should have no problem generating 44V RMS.

Another reason to keep the dissipation low, now that makes 12 Class A output stages