More newbe stupidity.
Re: Meng splitter, 6N2's.
I measured 108 V on the plate, .66 on the cathode. So current is .66 mA.
(227 V on the top of the 180K plate resistors. )
Looking at the load lines, this puppy is barely turned on and in serious non-linear state, if I draw line between .66mA and 108V.
It looks like to get anywhere into the slightly more linear range, one would need considerably more voltage across the tube and at least double the current. Am I reading this correctly?
If every little bit helps, then I could just replace the 22K in the B+ wiht a diode ( oft recommended anyway). Then reduce the 180 plate resistors and increase the cathode resistors. 1.5 ma looks like a better start.
Re: Meng splitter, 6N2's.
I measured 108 V on the plate, .66 on the cathode. So current is .66 mA.
(227 V on the top of the 180K plate resistors. )
Looking at the load lines, this puppy is barely turned on and in serious non-linear state, if I draw line between .66mA and 108V.
It looks like to get anywhere into the slightly more linear range, one would need considerably more voltage across the tube and at least double the current. Am I reading this correctly?
If every little bit helps, then I could just replace the 22K in the B+ wiht a diode ( oft recommended anyway). Then reduce the 180 plate resistors and increase the cathode resistors. 1.5 ma looks like a better start.
If I can read Chinese correctly....which I can't..
It looks like 275 Volts at 1.75 mA will swing thru some symmetrical loadlines.
Looks like -1.0 Volts for bias...........The loadlines I found are really fuzzy & hard to read.from what I can tell is yes, your set-up is way down there & needs tweeking.
300V Max, 10mA, Max, 1W dissipation.
So??? Your Anode resistor is?? With 300V & 1.75 mA???
_______________________________________________________Rick.....
It looks like 275 Volts at 1.75 mA will swing thru some symmetrical loadlines.
Looks like -1.0 Volts for bias...........The loadlines I found are really fuzzy & hard to read.from what I can tell is yes, your set-up is way down there & needs tweeking.
300V Max, 10mA, Max, 1W dissipation.
So??? Your Anode resistor is?? With 300V & 1.75 mA???
_______________________________________________________Rick.....
Well, it does not want more than 250V, so leaving the 227, line from 1.75 ma to 225 V and I pick about 1V because it looks like about the center of the liner range. That looks like about 120V across the tube. So with 227 available, then the anode resistor would be 61K. Tweak the cathde to get the 1V. Did I do that correctly?
Here's my beginner's take on load lines for the 6N2 in the Meng.
227v B+ with plate at 0v (maximum possible current flow) is 1.25mA with a 180k on the plate. So that is the Y point for your load line on the characteristic curve graph.
227v B+ is (as Ash said) what appears on the plate with no current flow, so that is your X point.
Now we don't want positive grid current (that's bad), so our grid can't go above 0v ref cathode. With the bias voltage you measured, that means a signal on the grid can't more than 0.66v peak (0.466v RMS) before cutoff (clip) on one peak, and positive grid current on the other peak. If you are driving this amp from a CD player, you can get peaks greater than this.
This is why I think that stage sounds not as good as it should. A re-bias will help by giving more headroom, but the whole paraphase idea is messy to my eyes, not to mention hard for a beginner to understand when the signal output from one stage affects the bias on the other (hence the balance pot).
If any of the gurus want to chime in to correct me then feel free.
More thoughts to come when my tube nube brain gets to it. 😀
Gary
The Russian 6N2P will cope with a max of 300v Va, but I don't know about the Chinese one. I wouldn't try it unless I had spare tubes.
My experience of the Chinese 6P1 output tube is that bad things happen over 250v, whereas the Russian 6P1-EV coped well over its rated 250v limit. I didn't bother to try that with the Chinese 6N2.
Remember, bumping B+ up overall will affect the output stage as well.
I had thought about picking the driver stage plate supply from before the original B+ dropper, via a new dropper. That would give (in my case) about 310v max that I could drop from. This is something along the lines of what Ian has done with his extra voltage multiplier stage tacked on the existing doubler.
Whether this sort of fiddling will materially help the paraphase stage is unknown. I still think the cathodyne that Ian used is a better option, but he (and I concur) that it didn't sound much better when used with the 6N2 in this circuit. Happier with 6N1P, happier still with 6N1P and more B+ and idle current.
I would like to try 6N3P but it has a radically different pinout which makes life difficult on the PCB in the Meng.
My experience of the Chinese 6P1 output tube is that bad things happen over 250v, whereas the Russian 6P1-EV coped well over its rated 250v limit. I didn't bother to try that with the Chinese 6N2.
Remember, bumping B+ up overall will affect the output stage as well.
I had thought about picking the driver stage plate supply from before the original B+ dropper, via a new dropper. That would give (in my case) about 310v max that I could drop from. This is something along the lines of what Ian has done with his extra voltage multiplier stage tacked on the existing doubler.
Whether this sort of fiddling will materially help the paraphase stage is unknown. I still think the cathodyne that Ian used is a better option, but he (and I concur) that it didn't sound much better when used with the 6N2 in this circuit. Happier with 6N1P, happier still with 6N1P and more B+ and idle current.
I would like to try 6N3P but it has a radically different pinout which makes life difficult on the PCB in the Meng.
It clips at .36 V RMS input, which is 5.77 V out into 7 Ohms, or all of a whopping 4.7W. I have not looked to see which stage is failing first.
Were my calculations correct?
Eventually I do intend to do it up pretty close to Ian's design. The whole point is to learn the tube game on the way.
Were my calculations correct?
Eventually I do intend to do it up pretty close to Ian's design. The whole point is to learn the tube game on the way.
Calculations are right as far as I can tell. Should back up the point I mentioned on one of the threads about not trusting the power claims for Meng amps. Remember, they rate this one at 10wpc.
I'm getting close to 5W at clip out of a single EL84 on my homebuilt SE (and it sounds better). The only thing holding me back from a full EL84 conversion on the Meng is some doubt over the capability of the PT to safely supply the extra current (.25A extra heater juice per output tube over and above the increase I already have by changing the 6N2 to 6N1).
Some ventilation in the transformer cover is already being planned even if I don't change the circuit.
I'm unfamiliar with the amp in question. What's the OEM O/P tube type? If you're worried about overstressing the power trafo, think about a "12" W. type other than the EL84, say the 6AQ5/EL90. Unlike the EL84, the 6AQ5 draws only 450 mA. of heater current. O/P "iron" that works with EL84s also works with 6AQ5s.
The amp is the earlier Mengyue Mini with 6P1 beam tetrode outputs in push pull, driven by a 6N2 in paraphase. A visually identical Meng model sold around the same time used EL84's, hence the possibility the iron may be the same.
Interestingly, the top plate of the chassis is inscribed with valve type 6AQ5, but the PCB and chassis cutouts are for 9-pin Noval tubes, and the amp ships with 6P1's. 6AQ5 are 7 pin mini tubes from memory.
Well, considering the translations they put on the WEB, 6AQ5 looks close enough to 6P1J when they approved the silkscreen.
Correct, the 6AQ5 is member of the 6V6 family. The only substantial difference between the 'AQ5 and the prototypical 6V6 lies in the area of screen grid fragility. When triode wiring 'AQ5s, tie g2 to the plate with 1 KOhm resistors. Use 1 KOhm resistors in the lines to the O/P trafo, when UL mode is employed. If full pentode mode is on the table, regulate g2 B+ with LR8 ICs.
The 6P1 data sheet does look like the 6AQ5 data sheet. Adapters can be made from dead Noval tubes and a 7 pin mini sockets. Are you familiar with the string method for cutting glass? Does the expression fire polishing mean anything to you?
A paraphase splitter? No, thank you! "Steal" the 12DW7 front end from the Dyna ST35. The Fisher/Dyna/Scott style of voltage amplifier and DC coupled "concertina" phase splitter will serve you well.
Thanks for the info regarding screen regulators, I'll keep that in mind.
The paraphase is long gone from my amp, but I think TVR has yet to change.
On a PCB it's not quite as easy to do a topology change as point wired but I bit the bullet.
Mine has the Dynaco style driver stages using 6N1P-EV. I'm still playing with currents to see what the 6N1P likes most. At the moment about 1mA but I think it could do with more. I run a 6N3P-EV at 3mA driving an EL84 in SE in my other experimental amp toy, and it sounds good.
Gary
The paraphase is long gone from my amp, but I think TVR has yet to change.
On a PCB it's not quite as easy to do a topology change as point wired but I bit the bullet.
Mine has the Dynaco style driver stages using 6N1P-EV. I'm still playing with currents to see what the 6N1P likes most. At the moment about 1mA but I think it could do with more. I run a 6N3P-EV at 3mA driving an EL84 in SE in my other experimental amp toy, and it sounds good.
Gary
Eli, just looking at the ST-35 schematic and noticed an RC network between the cathodes of the voltage stage and the splitter. Is that feeding "forward" some of the negative feedback from the voltage stage to the splitter?
I'm a real newbie when it comes to NFB circuits, and have no NFB at all in the Meng yet (it's stable fortunately).
No. I'm not familiar with glass cutting with anything other than the right tool or a hammer 🙂 Flame treating the sharp edges makes sense.
I'm a real newbie when it comes to NFB circuits, and have no NFB at all in the Meng yet (it's stable fortunately).
No. I'm not familiar with glass cutting with anything other than the right tool or a hammer 🙂 Flame treating the sharp edges makes sense.
Yea, only changed to 6ax7's up front. Trying to understand it fully before I pull it apart.
Now this is for learning, not better sound. I am going to replace the 22K resistor with a 1K and a diode, which should slightly up by available voltage to the splitter keeping isolation between stages. Then drop to 125K plate resistors, and twiddle the bias until I get about .8 to 1 ma quiescent. The graph says it should be in a more linear region and more of the available swing. Errr, I think. Then see what the symmetry looks like, and what the output distortion measures on the old spectrum analyzer. After that, rip it out and put in a cathodyne. Eventually I will get to the output stage and cleaning up the PS. I am going slow as I don't have a well stocked parts bin for amps, just speakers. I have to order and wait for everything.
Going back and forth, the 6N2's measure not that far off the charts, the 6AX7's are way the heck off. By the chart, for the voltages and currents seen, I should be at 1V bias, not .66.
Now this is for learning, not better sound. I am going to replace the 22K resistor with a 1K and a diode, which should slightly up by available voltage to the splitter keeping isolation between stages. Then drop to 125K plate resistors, and twiddle the bias until I get about .8 to 1 ma quiescent. The graph says it should be in a more linear region and more of the available swing. Errr, I think. Then see what the symmetry looks like, and what the output distortion measures on the old spectrum analyzer. After that, rip it out and put in a cathodyne. Eventually I will get to the output stage and cleaning up the PS. I am going slow as I don't have a well stocked parts bin for amps, just speakers. I have to order and wait for everything.
Going back and forth, the 6N2's measure not that far off the charts, the 6AX7's are way the heck off. By the chart, for the voltages and currents seen, I should be at 1V bias, not .66.
See my replies to your posts in our other thread for some background on output stage. I would get that area right DC-wise before you play with the input tubes. Less confusing for both of us. We obviously have some rather different plate currents or output iron, and as the driver tubes get their supply from B+, changing that will alter all your voltages and calculations for the drivers.
I was also thinking about the drifting grid you have on one tube. Maybe lack of grid stopper resistors is sending it into oscillation (that can affect your multimeter readings as well). I reinstated all mine (10k) which were originally there but with a shorting wire across them.
I was also thinking about the drifting grid you have on one tube. Maybe lack of grid stopper resistors is sending it into oscillation (that can affect your multimeter readings as well). I reinstated all mine (10k) which were originally there but with a shorting wire across them.
The string method entails tightly wrapping a piece of white kitchen string around the envelope, where the cut is desired. Moisten the loop of string with alcohol and set it on fire. The thermal stresses of combustion cause the glass to crack.
You worked out what fire polishing is. A propane torch, like that used for household plumbing chores, gets the sharp glass hot enough to smooth over.
The 150 KOhm/220 nF. pair in the ST35 schematic combine to form a high pass filter whose F3 is "4.8" Hz. Hafler was doing some freq. resp. shaping. Another "oddity" is the asymmetrical resistors in the "concertina". Hafler was anything but stupid and I will not sell him short.
Fisher used the 12DW7 in an EL84 design too. Hop over to AA and find where I uploaded the schematic fragment, in using modern 7591s inside "vintage" Fisher stuff threads. "Steal" from that too. Verbatim copying need not be necessary. Ideas are the object, not specific details.
You worked out what fire polishing is. A propane torch, like that used for household plumbing chores, gets the sharp glass hot enough to smooth over.
The 150 KOhm/220 nF. pair in the ST35 schematic combine to form a high pass filter whose F3 is "4.8" Hz. Hafler was doing some freq. resp. shaping. Another "oddity" is the asymmetrical resistors in the "concertina". Hafler was anything but stupid and I will not sell him short.
Fisher used the 12DW7 in an EL84 design too. Hop over to AA and find where I uploaded the schematic fragment, in using modern 7591s inside "vintage" Fisher stuff threads. "Steal" from that too. Verbatim copying need not be necessary. Ideas are the object, not specific details.
Something in the back of my foggy brain says something about turntable rumble filters being about that frequency.
One other thing I noticed was the output tubes having their cathodes and heaters connected. Trying to get my head around that. The 95ohm/100uF cathode bypass must present quite an extra load to the heater supply.
I'll check out audioasylum as well thanks.
It would be much easier just to switch from 6P1 chinese tetrodes to 6P1P(-EV) Russian tubes. I'm running 6P1P-EVs in PP and getting 10W with no difficulty.
At 13W I get a slightly red fin on one plate (Have to turn the lights off to see it).
Sound is great.
- Status
- Not open for further replies.