|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
76.90.39.24
In Reply to: RE: What value should an EL34 Cathode bypass capacitor be in a CCS? posted by JKT on May 07, 2016 at 12:23:55
"The addition of a current source to the power tube's cathode converts a conventional PP amplifier to an all-differential, all-Class A amplifier. In fact, the current source actually prevents the circuit from entering into Class AB."
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Follow Ups:
I removed the bypass caps and the amp went into motorboating when a signal was present. I have been incrementally making slight changes to this amp and when i switched from resistor setting bias to CCS or actually CVS with the LM317. i also removed the resistor on the input grid of the input 6sn7 tube. So i used a straight piece of signal wire from the input rca jack instead of a 1K resistor before the grid. Before i removed the bypass caps as i was testing the new CVS on the amp there were a few putt putts when the amp was first powered up. I am guessing the power tube section is unstable before they warmed up sufficiently. Just several putt putts before disappearing and not reappearing again. But now with the cathode bypass capacitors disconnected the amp is oscillating continuously.
Any suggestions as to what i should do to alleviate this problem? Besides putting everything back to how it was. I am thinking if i should reduce the amount of feedback from the 16ohm tap to the input tube. Should i go further down the rabbit hole or get out when i can? cheers, Dak
Edits: 05/08/16 05/08/16
Have you wired the LM317 as a voltage regulator, or as a current source? I'm asking because your first post on this said you were using it to set bias voltage. Maybe you could sketch a schematic and post it?
--------------------------
Buy Chinese. Bury freedom.
The voltage is set with a 19.6 ohm resistor. Input is at the cathode. out is connected to the 19.6 ohm resistor which goes to ground. Common is set to ground. regards, Dak
I don't understand your nomenclature. The LM317 has three terminals: Vin, Vout and ADJ. In any event, wiring it as a voltage regulator won't work without a parallel cap. You need a current source to force Class A operation.
I'll just add that the LM317 is not a good choice for this job. It's a noisy device, and many people who have used it for this purpose report poor quality sonics.
--------------------------
Buy Chinese. Bury freedom.
Edits: 05/09/16
My phono pre, line stage, and mono-blocks all consist of LTPs with CCS "tails". Over the years that I've been perfecting these most of the sonic improvements have come from "improved" CCSs. To quote The Donald: They make for a "huuuge" difference. There are better alternatives out there for "tail" current sources (current sinks in this case) than the LM317. One place to start reading would be Walt Jung's Current Source 101 articles.
"It is better to remain silent and thought a fool, then speak and remove all doubt." A. Lincoln
I should add Morgan Jones 4th edition to the list.
"It is better to remain silent and thought a fool, then speak and remove all doubt." A. Lincoln
Most of the signal goes through the cap . If these devices are bypassed with a cap , the quality of the cap is more important than the quality of the CCS . I would do it by ear with some 22uF 63V film caps and add in parallel for best sound .
Al
I agree but what's the point of a CCS if it's bypassed with a cap?
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
It's useful for maintaining balance in a pushpull amp , easy to modify a cathode bias network this way .
I would prefer fixed bias and a smart centre reading meter
Al
"It's useful for maintaining balance in a pushpull amp...."
DC balance? I don't see how with only one shared CCS.
Without a bypass cap the impedance of a CCS is high giving a large error feedback signal that maintains dynamic AC balance. At any given moment that error signal is negative FB for one tube and positive FB for the other, forcing AC balance.
With a bypass cap there's no corrective signal.
In terms of DC, the total current will match the set current but why would that current necessarily be evenly divided between the two output tubes?
I don't see the mechanism to accomplish that.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
"In terms of DC, the total current will match the set current but why would that current necessarily be evenly divided between the two output tubes?"
You can use two CCS, one for each half and AC couple the cathodes with a capacitor. The trade-off is a capacitor in the signal path. Something I've wanted to try just to see how it would affect the sound. Matching tubes for my PPP mono-blocks can be a bitch.
"It is better to remain silent and thought a fool, then speak and remove all doubt." A. Lincoln
OK, nomenclature can be confusing.
If you look at the actual TO220 device from the front, the side with the label. The first leg goes to ground, the center leg has the voltage setting resistor to ground and the right leg is at the cathode.
OK, that sounds like you have it wired as a current source. The problem is, 1.25/19.6 is only 64 mA. You need more than that to center the EL34s in Class A. Does it motorboat even at low levels (without the cap), or just above a certain threshold? How much voltage is on the cathodes at idle? What mode is this amp - pentode, UL or triode? Are there grid stoppers on the output tubes?
--------------------------
Buy Chinese. Bury freedom.
Hi, JKt, yes, in its cathode bias w/resistor, the amp was running 70ma at the cathodes. I thought that was too high so i dialed it back a little. The amp is running UL mode but there are no grid stoppers on the suppressor grids. There are 1K resistors on the input grids. As soon as the amp received a signal, it started motorboating.
The LM317 shouldn't cause that. It's likely there's another issue in the mix. See JKT's post below. If you have a scope, you can look at the stages ahead of the outputs to see if unwanted signal is leaking back through the B+ lines.
--------------------------
Buy Chinese. Bury freedom.
Yeah, i did not think it was the lm317 but removing the bypass cap exposed some other situation which has now come to light. I will check the the voltages of the front end tubes and see if something jumps out there. cheers, dak
He'll need to disconnect the GNFB first.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
With motorboating I would also look for the possibility of low frequency coupling within the B+ supplies to the different stages in the amp. Relying on the caps in PS RC filters to shunt noise and signals to ground thus isolating different stages becomes more problematic the lower the frequency of the offending signal.
"It is better to remain silent and thought a fool, then speak and remove all doubt." A. Lincoln
The only way to learn is to jump into the hole.
I hope you still have a resistor from the grid to ground.
How much feedback are you using?
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
The fb resistor is a 1w Takman the colors are either brown/brown/black/red/grey or grey/red/black/brown/brown
the color is faded some so the brown could be gold or grey could be silver. I have not worked on this amp for awhile so i dont remember. For PP el34 a fb resistor of at least 5K would be the usual, so, i am guessing around 7K. I wanted to play with the LM317 so i did this mod to the amp. Truthfully, the amp sounded better with the ccs and i would not want to go back to the resistor biasing. So, having bypass caps is not a deal breaker for me.
Well, as has been said, without the cap you will force Class A operation.
That will limit the maximum output power but should be cleaner (less distortion) up to the Class A limit.
You might try disconnecting the FB all together. If that stops it from oscillating you could use a pot wired as a rheostat and slowly adjust it to see how much FB you can use before it starts causing problems.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Yeah, i guess , l will play with the fb level and see how that affects the amp's stability. cheers, Dak
What I said.
"It is better to remain silent and thought a fool, then speak and remove all doubt." A. Lincoln
.
Have Fun and Enjoy the Music
"Still Working the Problem"
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: