AB to A?

[a91gti]

Hallo all,

I'll see about getting the circuit uploaded so as to get a definitive answer but for now can we theorise?

My amp has four EL34 running in pp for AB, I'd rather they were giving me A.

How might such a thing be done if indeed it can?

[Richard]

How do you know it's running in AB?

[a91gti]

Its a Maplin Millenium so I'm making that assumption. I'm quite prepared to be wrong if that's the case.

I did say "theorise".

[Richard]

Ok, well that holds the key to A or AB. If the EL34s are able to swing so far that one shuts off whilst the other still conducts the stage moves from A into B. Control of this could be for reason of the output stage operating conditions or the driver stage.

If HT is low and current high you can have a situation where as one valve is driven to max output and starts to clip the other is still conducting. Thus neither switch off and so it runs in class A all the time. If it was designed to be a class A amp this is how it would be held in A.

It could though also be due to a stage that would otherwise go into B being held back by lack of driver signal voltage; the driver clips before the output stage is driven to clipping. Thus an otherwise AB amp is kept in A.

My question was prompted by your amp using an ECC82 as splitter/driver rather than an ECC83. If that will not give enough voltage into the EL34s then they may not be driven into class B at their peak output.

It would be like a 5-20/4-20/Millenium but without the class B output at maximum output if you follow me. It would be class A simply because it couldn't be driven into B, the driver stage would clip first and that of course is class A as it's single ended.

That wouldn't make it sound better, just less powerful on peaks, as it wasn't designed to make the most of running in class A. It would just be an AB amp without the B bit! You could get the same from any AB amp by not driving it into B - by simply keeping the input signal level low.

In fact it may sound worse as, if it is only the ECC83 which have been changed to ECC82, they will not be working as a balanced splitter.

So the answer is to draw out the circuit complete with component values and quiescent voltages. That will tell us if it should have ECC83 in there and those voltages may let you see if it's likely to stay in class A or maybe you will need to drive it to clipping and check the stage voltages at that point.

[a91gti]

I'm going to have to read that a few times to get it to sink in Richard but I get the gist of it and thank you.

I have the circuit on paper and will seek to get it uploaded asap.

[Richard]

If you can upload the circuit it's much easier to see what's what rather than trying to explain in a few words all cases of A and AB that might be theoretically possible

With its double mono construction and Danbury TXs it certainly looks like a Millenium. I don't know who drew the circuit you have but if it came with the amp I wouldn't take it for granted that it's correct. In particular follow the wires from the ECC82 anodes and check what value load resistors are attached which will tell us if it should have ECC82 there or ECC83.

This is one of the best pieces I found on the web, well worth a read,
http://www.aikenamps.com/ClassA.htm

Bruce Rozenblit's book is a great read and gives an overview without getting into heavy detail,
http://www.amazon.com/Beginners-Guid.../dp/1882580133

[a91gti]

Success! I've uploaded the circuit as drawn by the guy who added an extra pair of bias pots. The drawing doesn't feature them or the now present volume control but I'm happy that it is accurate.

[Richard]

Oh yes, you'll have fun with that

It's not 5-20 or Millenium. The EF86 is running triode into ECC82 splitter/driver into UL EL34 with global feedback. The ECC82 and EL34 are fixed bias but there are no voltages shown anywhere and you need a voltmeter to balance the EL34s.

I would take a full set of voltages first. Bias is fixed voltage, just adjustable for balance, so check what it is then you will be able to find how the EL34 stage is run by checking the HT voltage.

[a91gti]

Oh christ! More gobbledegook, I get every second or third word so am going away to decide if I'm continuing with this.

[Richard]

Well 5-20 would have been easier

You mention a couple of extra bias pots, they'll probably be for adjusting the amount of bias so you can change bias and balance the valves, A or AB at the twiddle of a pot, or maybe just hot or cold bias, happy days if you know what you're on with

This isn't all bad news. It does look like overkill but presumably it's working at the moment so just fit it with good valves and it will make beautiful music. The only time you would need to adjust it is if the valves were too far out of spec and you'd hear them humming or glowing red.

Really there should be a book explaining where/how to check bias voltage and how much to set. That said, there are lethal voltages inside and it's not a job for the inexperienced. This really is for an enthusiast or dealer servicing. We never had domestic amps like this in the UK market, they were more a US or, recently, Chinese thing. The likes of Quad, Leak, Mullard etc were always cathode automatic biased, the user just changed valves and carried on listening.

It's a bit chicken and egg. You can work it out but first you need to fire it up and check HT voltage at least. Do you have a meter and have you done this sort of thing before?

Taking a different route you could convert, or get someone to convert, it to 5-20 circuit quite easily. The advantage to the user is that all biasing would then be automatic and all you need do is change valves and play.