Thanks SY. But I forgot to mention that the "preferred" circuit (http://www.diyaudio.com/forums/attachment.php?s=&postid=335624&stamp=1077837811) has an ECC88 in it, and I don't have any ECC88 at hand. Only 81,82,83. Is there some way to recalculate the resistance values thus adapting the second stage to accomodate an ECC83? Maybe it would bring up some gain, as you say if it is lacking in this department.
Can you explain it a little bit more? I don't understand. You're saying that there's too little excess gain to use the NFB?
Sorry for so much unexperienced questions...
Can you explain it a little bit more? I don't understand. You're saying that there's too little excess gain to use the NFB?
Sorry for so much unexperienced questions...
Hi Giaime,
Feedback theory is not trivial and can take a while to understand. Questions are good.
The three tube versions use a buffer to drive the feedback network and the output (next device in line). This allows for more gain in the second stage. The extra gain is useful to reduce distortion and correct the filter response.
SY can give a much better description than I can here.
-Chris
Feedback theory is not trivial and can take a while to understand. Questions are good.
The three tube versions use a buffer to drive the feedback network and the output (next device in line). This allows for more gain in the second stage. The extra gain is useful to reduce distortion and correct the filter response.
SY can give a much better description than I can here.
-Chris
It's a mess from the beginning. The gain starts out by being marginal to get to proper line level. Then the second stage has a severely reactive load in the RIAA network. It will presumably also be loaded by some wiring and volume/balance controls, and lose a bit more gain. Ugh. Dynaco recognized this and put some positive feedback between the two cathodes to bump the gain. Having lived with that phono stage, I can attest that the cure is worse than the disease.
Now there's half a fighting chance if you put a buffer after the second stage to drive the RIAA and the output. In too many designs, it's a weedy buffer, but that's better than nothing. At least the second stage can have a reasonably flat high resistance loadline. Now you're just faced with a stage that has very low feedback at low frequencies, with the attendant rise in distortion. And it can have a tendency toward blocking distortion when overloaded by the odd tick, pop, or mistrack. Well, life's not perfect, and that's why really good RIAA stages have lots of glass in them.
Now there's half a fighting chance if you put a buffer after the second stage to drive the RIAA and the output. In too many designs, it's a weedy buffer, but that's better than nothing. At least the second stage can have a reasonably flat high resistance loadline. Now you're just faced with a stage that has very low feedback at low frequencies, with the attendant rise in distortion. And it can have a tendency toward blocking distortion when overloaded by the odd tick, pop, or mistrack. Well, life's not perfect, and that's why really good RIAA stages have lots of glass in them.
Oh I understood well, thanks Chris. And thanks SY, so that circuit cannot properly drive a RIAA network and the following stage.
So let's stick to this circuit:
http://www.diyaudio.com/forums/attachment.php?s=&postid=335624&stamp=1077837811
You have pointed that there's a lack in the gain. So, if the ECC88 has a mu of 33, instead the ECC83 has a mu of 100, if I replace the second stage with an ECC83 stage (of course tweaking the resistor values to obtain the right bias point), I should obtain all the gain I need! Am I wrong?
So, how can I tweak the resistors value to accomodate an ECC83?
I see no cathode resistor in the second stage. Is it grid-leak biased? Is this the purpose of the 10M resistor in the grid circuit? Seeing the plate resistors of the previous stage, I think that those of the second stage should be increased to about 10k/120k (a factor of 10), or not?
Thanks anyway for all the help you're giving to me!
So let's stick to this circuit:
http://www.diyaudio.com/forums/attachment.php?s=&postid=335624&stamp=1077837811
You have pointed that there's a lack in the gain. So, if the ECC88 has a mu of 33, instead the ECC83 has a mu of 100, if I replace the second stage with an ECC83 stage (of course tweaking the resistor values to obtain the right bias point), I should obtain all the gain I need! Am I wrong?
So, how can I tweak the resistors value to accomodate an ECC83?
I see no cathode resistor in the second stage. Is it grid-leak biased? Is this the purpose of the 10M resistor in the grid circuit? Seeing the plate resistors of the previous stage, I think that those of the second stage should be increased to about 10k/120k (a factor of 10), or not?
Thanks anyway for all the help you're giving to me!
So, led in the cathode circuit for biasing. I undestand that even with ECC83 it will work but...
...do I need to change those plate resistors values to match an ECC83? I think so... anyway I'm starting to build it, at least the first stage. So we'll verify by hand the proper value of those resistors.
But what is the way to calculate them? I think I should draw the loadline of that stage, calculate the bias point, the Iq and Vaq, the gain... and put those values in an ECC83 kind of circuit to obtain the same gain... Or I'm wrong?
...do I need to change those plate resistors values to match an ECC83? I think so... anyway I'm starting to build it, at least the first stage. So we'll verify by hand the proper value of those resistors.
But what is the way to calculate them? I think I should draw the loadline of that stage, calculate the bias point, the Iq and Vaq, the gain... and put those values in an ECC83 kind of circuit to obtain the same gain... Or I'm wrong?
anatech said:
Yes I agree. But remember that I'm driving a cathode follower on the output of that circuit, so extra driving capacity isn't needed (at least the minimum needed to drive a 100k potentiometer).
I'm starting now to assemble at least the first stage, which I think can be good. For the plate resistors, there's time to tweak them later.
Still thanks!
Hi Giaime,
That's what I figured you were going to do. Wire for a 12AX7 and use the higher value plate resistor to increase available gain. The CF will drive everything and maintain your gain and curve accuracy.
The ECC88 was originally used I think because there was no CF in the circuit.
-Chris
That's what I figured you were going to do. Wire for a 12AX7 and use the higher value plate resistor to increase available gain. The CF will drive everything and maintain your gain and curve accuracy.
The ECC88 was originally used I think because there was no CF in the circuit.
-Chris
Exactly what I'm going to do. Thanks so much, I'm now selecting the components. Sorry, no audiophile caps, just polyester generic ones... I will report later!
Should I mantain that subdivision of the plate resistor with the capacitor in the middle? So I will increase both resistors!
Thanks, I will post later!
Should I mantain that subdivision of the plate resistor with the capacitor in the middle? So I will increase both resistors!
Thanks, I will post later!
Sorry people, no chance to build it yet. I don't have the proper capacitor values at home for the RIAA: I have ceramics, that aren't good, and some odd polyester, too low voltage (different capacitors for each channel... bad thing)... 
No way now. Maybe when I've got more time I will start with the sockets, resistors etc etc...
Thanks, anyway, for all the help you gave me.
No way now. Maybe when I've got more time I will start with the sockets, resistors etc etc...
Thanks, anyway, for all the help you gave me.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.