02-11-2016, 07:43 AM
I've made a bit of a boo boo when I said the OP of the ECC82 is 25v, it is 25v but that's P-P, so that would be 8.8v rms. The datasheet for the EL84 says a 4v input is needed so I used a 82 after looking at mutual characteristics graphs for the 81/2 and 3. I was looking to get a low linear OP, IE as little THD as poss. Using a ECC83 is overkill, too much gain as the input is 1v. The 82 has high straight slopes so with a 12k anode R and it biased at -4.5v this gives a linear OP, IE 12.5v anode V for a change of .5v input g1.
The OP from the triode voltage amp is clean at the EL84 g1 and IMO has sufficiant gain for the OP stage. As I only wanted a 4v OP, this is why I used a 12k anode resistor. I'll post some load lines later when I get the time and all will become clear hopefully.
Yesterday I tried another OPT. The OPT's I have, have on the whole too low an impedance ratio. In theory for center biased class you find the optimum Zpri by HT^2/Pa which is 250^2/10 = 6250k or a ball park 5k. The EL84 datasheet gives a Ra of 5.25 or 4.5k in similar operating conditions to that which I'm using. So I tried another OPT, I found the Zpri by sticking a sinewave on the primary,( this I established by measuring the DC res, IE high DC R + primary) measuring the OP and using Vin/Vout^2 and times the result by load. I couldn't find one that was suitable, so used what I had, which I think had too low a Z ratio. This was born out after I used a 32R resistor instead of 8R, and got some improvement. I the end I found that with a sec load of 200 ohms I got a 2w clean undistorted OP. this OPT came off a radiogram i think, and had three speakers wired to the secs, one bass and two "tweeters" in series on different windings. As I have no schematic replicating this load is impossible.
I take your point Michael that it would be better to build a P-P amplifier, but I'm not going to build this amp and use it. I'm only doing it for a bit of fun and to try and go deeper into some aspects of amp building that I havn't done before and become familiar with things like open loop gain, NFB etc.
Mark is right about the input stage, for a modern AC source 10k ish is about right. I could have popped a 1meg pot on it but read somewhere that this is too high a value. I used this formula to find C in - 1/(2 x pi x f x R) ... 1 / (6.28 x 1* x 50000) = 3.184713376 -6, therefore a 3u cap is what I used. * Why 1? We want to block DC and pass all frequencies over DC.
I'll re read your posts in more detail later and try some of your suggestions and get back to you. Andy.
The OP from the triode voltage amp is clean at the EL84 g1 and IMO has sufficiant gain for the OP stage. As I only wanted a 4v OP, this is why I used a 12k anode resistor. I'll post some load lines later when I get the time and all will become clear hopefully.
Yesterday I tried another OPT. The OPT's I have, have on the whole too low an impedance ratio. In theory for center biased class you find the optimum Zpri by HT^2/Pa which is 250^2/10 = 6250k or a ball park 5k. The EL84 datasheet gives a Ra of 5.25 or 4.5k in similar operating conditions to that which I'm using. So I tried another OPT, I found the Zpri by sticking a sinewave on the primary,( this I established by measuring the DC res, IE high DC R + primary) measuring the OP and using Vin/Vout^2 and times the result by load. I couldn't find one that was suitable, so used what I had, which I think had too low a Z ratio. This was born out after I used a 32R resistor instead of 8R, and got some improvement. I the end I found that with a sec load of 200 ohms I got a 2w clean undistorted OP. this OPT came off a radiogram i think, and had three speakers wired to the secs, one bass and two "tweeters" in series on different windings. As I have no schematic replicating this load is impossible.
I take your point Michael that it would be better to build a P-P amplifier, but I'm not going to build this amp and use it. I'm only doing it for a bit of fun and to try and go deeper into some aspects of amp building that I havn't done before and become familiar with things like open loop gain, NFB etc.
Mark is right about the input stage, for a modern AC source 10k ish is about right. I could have popped a 1meg pot on it but read somewhere that this is too high a value. I used this formula to find C in - 1/(2 x pi x f x R) ... 1 / (6.28 x 1* x 50000) = 3.184713376 -6, therefore a 3u cap is what I used. * Why 1? We want to block DC and pass all frequencies over DC.
I'll re read your posts in more detail later and try some of your suggestions and get back to you. Andy.
