|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
24.127.166.64
Amp schematic is shown in triode mode and with two 6V6 per channel. I will use four 6V6 per channel. I am told if a switch is used to apply B+ to the 6V6 screens for tetrode connected, I need a feedback circuit. Any circuit suggestions? Amp has 4/8/16 ohm output.
Follow Ups:
Unless this is a guitar amp, I wouldn't bother switching modes. Both modes can't be optimized with the same output transformer.
--------------------------
Buy Chinese. Bury freedom.
Triode Kingdom. It is a home entertainment amp. Using PPP 6V6GT, I am using Dynaco ST70 audio transformers. I think they are called model 470. These transformers fair very good on tests. Some state 4300 ohm primary, but others state around 5300 ohm primary.
The ST70 O/P "iron" is suited (obviously) to EL34s. PPP 6V6s will be FINE.
Eli D.
With some A470's I would love to do an updated SP20 build. Swap out the front end's first two stages for a single LTP built on something like a 5687 or perhaps 12A4's, and ditch the 6AU6's for EF184's. Tweak the resistor values to maintain the FB levels, or perhaps drop them a wee bit and run fixed bias on all the finals.
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
"Any circuit suggestions?"
Poinz's "Musical Machine" and the "El Cheapo" I'm associated with are closely related. Don't think for 1 second about resistive tail loading, when NFB is applied to the non-inverting triode of the LTP. BTW, Poinz went to CCS tail loading in later iterations of his design.
Eli D.
Eli, do you have the Music Machine later schematic? I rolled a lot of phase splitter tubes & found the USA Raytheon 6CG7 an unbelievable sounding tube.I am using a two RCA 5V4G as parallel rectifiers into a 5hy choke input power supply for high B+.
I also like fixed bias on the 6V6GT tube. Takes cathode to ground without capacitor colorization on cathode biased tubes.
Edits: 12/29/16
Poinz switched to hand matched 6GK5s in the LTP. The triodes in the 6CG7/6FQ7 are the same as those in the venerable 6SN7. Superb linearity, but (most unfortunately) way too little gain for these 2 stage setups. Remember, in these 1 grid driven LTPs 1/2 μ is the theoretical max. gain and 30% of μ is doing well, in the real world.
I too like "fixed" bias in a PPP setup, as exact setting of each tube's "idle" current is important. The downside is that grid to ground resistance limit is lower, when "fixed" bias is employed. Therefore, the NET impedance the driving circuitry is working into drops to a point where both gain and linearity are degraded. That trouble spot is dealt with by buffering the LTP plates. The high gm/low R P types we've been discussing will have no problem in driving DC coupled IRFBC20 source followers. Put a protective 15 V. Zener diode between the plate side of the gate stopper resistor and the FET's source electrode.
Eli D.
"I too like "fixed" bias in a PPP setup, as exact setting of each tube's 'idle' current is important. The downside is that grid to ground resistance limit is lower, when "fixed" bias is employed."
Eli, I've converted a number of amps from self- to fixed-bias without changing the grid resistor. It's never created a problem that was discernible either with measurements or by ear. What's the risk of using a higher value grid resistor than recommended by the manufacturer? Note that I'm not talking about values so high they might create a grid leak effect.
--------------------------
Buy Chinese. Bury freedom.
It depends on the tubes in use. Too "tall" a grid to ground resistance value can result in runaway. :> ((
Current production 7591s are a glaring example of the issue. The tubes fully meet the originally published specifications, but they are INTOLERANT of the liberties taken by Fisher and some other manufacturers.
IMO/IME heeding the warnings contained in the published specifications is a very good idea. Call me a belt and suspenders guy.
Eli D.
I've heard about that with the reissue 7591s, but I haven't worked with those. Is this a grid emission problem? The grid drifts positive when the resistor is too large?
Doing this "by the book" creates a couple other issues besides the need for greater drive current. The coupling cap needs to be made considerably larger to maintain identical frequency response. And there's more risk with a smaller grid resistor that the signal will "see" the decoupling cap in the bias circuit.
Maybe I've just been lucky making these changes in the past. If you know, how far away from the manufacturer's specs did Fisher push the grid resistor?
--------------------------
Buy Chinese. Bury freedom.
"Is this a grid emission problem?No, not a problem in the sense that the grid emission is excessive; but a situation that can occur if the tube is used with a grid resistance that exceeds the spec.
"The grid drifts positive when the resistor is too large?"
I would say it this way - "The grid drifts less negative when the resistor is too large." The grid doesn't have to drift all the way to a positive voltage for issues to arise.
Edited to correct wording error...
Edits: 12/30/16
So do you believe it's necessary to stay strictly in accordance with the manufacturer's spec on this to operate safely? I've seen 6L6 fixed-bias amps for example with 100K grid resistors, but there are also many examples of designers using 220K and higher. I've personally used 270K without apparent problems at the time. Maybe I should rethink this aspect of my designs?Apologies to the OP, didn't mean to hijack the thread. I realize the discussion is about 6V6s, so I'll keep it short.
--------------------------
Buy Chinese. Bury freedom.
Edits: 12/31/16
100K is the top limit of the 6L6GC spec, at least according to GE and RCA. That is the max for fixed bias so that wouldn't bother me at all. For cathode bias the limits are typically 500K.
There is no such thing as absolute safety, but I do believe you are getting the best odds when you stay at or below spec. Above spec some tubes will be fine, but some others will cause trouble. It seems to me that it is irresponsible to ask the amp owner/user to bear the risk of operating at too high a grid circuit resistance. So I recommend that the spec be adhered to.
I should probably step back and review my design philosophy on this issue. I've been gradually increasing the size of grid resistors over the years, primarily to enable better low frequency response without resorting to larger coupling caps. Some manufacturers have gone as high as 1uF with 100K grid resistors, but PIOs are physically too large for that in most amps.
Thanks for your input Jim, it's valuable advice as always!
--------------------------
Buy Chinese. Bury freedom.
Back to original question. I am now using one section of 6CG7 to drive 6CG7 phase invertor as shown in first schematic. First section 6CG7 tube uses 50K plate resistors, 1.2K cathode resistor (no by pass capacitor) with 180 volt plate voltage.What feedback circuit parts would you use in PPP tetrode from either 8 or 16 ohm audio transformer secondary?
Edits: 01/01/17
I think you should pay attention to the more important parts of circuit design, such as which wire sounds best.
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
The design is completed with best components including solid silver wire. I think 80pF shunted with 80K resistor is a good feedback circuit for 16 ohm audio transformer tap to input tube cathode. I can adjust resistance for best sound.
15K, not 80K used on 16 ohm tap for feedback.
If this project is as important as you claim, why are you working with a global loop of NFB?
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
The feedback is just in the tetrode operating position. Triode connection does not have feedback. Simply want 30 watts for outdoor parties, 15 watts in triode is a little too low power for that.
There also seems to be a cause and effect pointing to a lower grid resistance sounding better. Now it could be a change to the load of the preceding stage, but I have made the attempt to leave the load unchanged in these experiments...LOL
The end being a choke or transformer secondary sounding the best...with gapped, low loss NI sounding best...:)
What ever the limit, putting two in parallel halves the resistance the driver sees...so while these are not 50's or 813's I will hold onto the idea that minimizing the grid resistance is a good thing, even if it requires a rather more solid than is traditional driver. 12AX7 need not apply...even for a pentode-rigged EL84.
cheers,
Douglas
Friend, I would not hurt thee for the world...but thou art standing where I am about to shoot.
Eli, you are right. Not enough gain. I will use 1/2 a Raytheon 6CG7 driving a 6CG7 per channel using the LTP phase splitters. I have enough tube stock.BTW- when rebuilding Fisher 400/500B/500C/800 receivers I replace the power tube grid resistors with 200K.
Edits: 12/30/16 12/30/16
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: