I am nearly ready to test my first project, a preamp with a grounded cathode input and a cathode follower output. It has tested well with my test rig.
My question is this, is it advisable to provide an input coupling cap? I am concered that my CD player may see 300VDC at some point. I have not seen any designs that show this but I want to know what the group thinks.
Marc
p.s. this is still a desing in progress, I will post my final working diagram in a few weeks
My question is this, is it advisable to provide an input coupling cap? I am concered that my CD player may see 300VDC at some point. I have not seen any designs that show this but I want to know what the group thinks.
Marc
p.s. this is still a desing in progress, I will post my final working diagram in a few weeks
You absolutely need a cap: when the preamp is on and the tubes are cold, they can't pass any current. So you have the B+ directly to the input throught 10k + 22k resistors, which form a voltage divider. You'll have at least 70% of the B+ voltage on the input.
By the way, what's that 10k resistor from the plate to the grid? negative feedback?
And there are better ways to polarize a cathode follower...
By the way, what's that 10k resistor from the plate to the grid? negative feedback?
And there are better ways to polarize a cathode follower...
It is for now but I have found changing the operating point of the follower may work better to reduce gain
Marc
Marc
Gain won't be changed by a significant amount by tweaking the biasing of the cathode follower.
Maybe this thread would be useful to you: it contains many suggestions on CF biasing.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=69404
Maybe this thread would be useful to you: it contains many suggestions on CF biasing.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=69404
The 10K resistor will not provide fixed negative feedback since it is shunted by the source impedance of whatever is driving the circuit. Assuming there's a volume control, the feedback will vary all over the place. To make things worse, at lower volume settings, the 10K will effectively parallel the first stage's plate resistor, greatly increasing distortion.
What that resistor will also do is set the CF's grid at DC ground. This will insure that the cathode follower will be starved of current and not function well.
It's an unpleasant thing to say, but I'd scrap this design and start over, beginning with an outline of your actual requirements.
What that resistor will also do is set the CF's grid at DC ground. This will insure that the cathode follower will be starved of current and not function well.
It's an unpleasant thing to say, but I'd scrap this design and start over, beginning with an outline of your actual requirements.
An input cap is not needed as there can never be any high voltage at the input, neither at startup or during operation, the grid of U2 can not give any current through R5. A problem is that as the grid of U2 is at 0V there will be very little current in the cathode resistor and the CF will be quite unlinear.
Note that when using this kind of so called series feedback with a resistor coupled from anode to grid the gain is heavily dependant on the source impedance, more so in this case as there is no series connected input resistor. Gain for the first stage can be described by G=R5/Rs where Rs is the source output impedance, this equation is valid as long as G is much less than mu of the tube. In order to make gain less dependant on the source impedance you can connect an input resistor in series withe the input then gain is G=R5/(Rs + Rin).
Regards Hans
Note that when using this kind of so called series feedback with a resistor coupled from anode to grid the gain is heavily dependant on the source impedance, more so in this case as there is no series connected input resistor. Gain for the first stage can be described by G=R5/Rs where Rs is the source output impedance, this equation is valid as long as G is much less than mu of the tube. In order to make gain less dependant on the source impedance you can connect an input resistor in series withe the input then gain is G=R5/(Rs + Rin).
Regards Hans
You are all right, the CF is very unstable! I originaly planed on using a mu follower and had already built an elevated heater supply for it
I now have the attached mu follower on my bread board and it does seem more stable. I have to play around with the operating point a bit but it sounds ok
The real issue is that the volume pot is very sensitive. I hope this improves when I install my alps blue velvet pot. Any thoughts?
Marc
I now have the attached mu follower on my bread board and it does seem more stable. I have to play around with the operating point a bit but it sounds ok
The real issue is that the volume pot is very sensitive. I hope this improves when I install my alps blue velvet pot. Any thoughts?
Marc
- Status
- Not open for further replies.