Ahh well, I tried to explain it...but it ain't easy...
Murphy struck in a manner I did not expect, as usualBird_dog you did well to get it sorted out.
Now, what we have been waiting for, is, how do the Hammonds sound compared to the originals?
Yes, I saw than you explained it to me after I did the modifications. It was well described...but an image worth a thousand words! I'll do some listening today and I will compare my two Meng, the original and the modified. I'll give you my impressions as soon as it's done.
Daniel
6N1-EV's in pre
Hi Gary,
Haven't finished reading the entire saga yet, but looks like you did like swapping in 6N1-EV's for the original pre-amp tubes, right?
__________________
Hi Gary,
Haven't finished reading the entire saga yet, but looks like you did like swapping in 6N1-EV's for the original pre-amp tubes, right?
__________________
Yes, but I think it wasn't the tube per se. I suspect that the lower gain was actually where the improvement came. With the 6N2 and the low B+ voltage that this amp runs, I suspect that the preamp stage was running into clipping distortion very quickly. The lower gain of the 6N1P (and the fact that the negative feedback loop is now disconnected) is the real improvement. That and the ditching of the paraphase topology in favor of cathodyne. I suspect the manufacturer had to use too much feedback to try and cope with the extra gain.
If you were to put the amp through a test of frequency response and distortion now, it probably looks worse, but it sounds a lot better.
I began to listen to some music this afternoon and I noticed than one channel was playing louder than the other. I tried switching speakers and the problem swith side, so the speakers were not the problem. I tried with a new set of tubes, problem still there. I finally tried a sine wave on each input and I discover than whatever input I choose, left or right, I hear the tone in both speakers... I'll try to figure something out to see if I did'nt invert something somewhere but I'm asking you if something can have burn when my original output tranfo burned?
Daniel
Daniel
Rotaspec, the extra gain was probably used to allow near 20db of GNFB. With the 6N1P the mu is 33. The 6N2P has a mu of 100 so much more gain is available to compensate for the GNFB.
I can only get about 6db of GNFB with the 6N1P front end and 6N23P cathodyne phase splitter.
I may try a LTP with current source in the tail to get the gain up for GNFB testing.
Daniel, check to make sure you haven't cross coupled your transformers such that they are both connected to both channels. That is the only thing I can think of off hand that would account for the tone in both channels.
I can only get about 6db of GNFB with the 6N1P front end and 6N23P cathodyne phase splitter.
I may try a LTP with current source in the tail to get the gain up for GNFB testing.
Daniel, check to make sure you haven't cross coupled your transformers such that they are both connected to both channels. That is the only thing I can think of off hand that would account for the tone in both channels.
Daniel, did you happen to disconnect and reconnect the feedback wires? Thats the red, brown, white and black wires in the white circle in this photo . They connect from the speaker terminals to the PCB. If in doubt, you could temporarily disconnect them and see if things come good. I can't think what else may cause the tone to come out both speakers, at this stage.
I could understand one channel being louder than the other if, when the left OPT failed, it blew the 10K feedback resistor open cct, effectively disabling the global feedback circuit.
Maybe, maybe not. The 10K resistors I think are under the PCB, so you won't see them, but you might smell them. It's a long shot though. Even then, disconnecting those four feedback wires would help in troubleshooting.
Unfortunately the builders did not use lug terminals to connect wires to the speaker terminals, you need a heavy duty soldering iron to disconnect and reconnect the wires, something like 80 watts soldering iron.
Maybe, maybe not. The 10K resistors I think are under the PCB, so you won't see them, but you might smell them. It's a long shot though. Even then, disconnecting those four feedback wires would help in troubleshooting.
Unfortunately the builders did not use lug terminals to connect wires to the speaker terminals, you need a heavy duty soldering iron to disconnect and reconnect the wires, something like 80 watts soldering iron.
Check that you didn't accidentally cross the output wires from the trafo's to the binding posts. With both speakers connected, they would end up in series with one trafo effectively shorted out via the common wires in the feedback loop. That's the only way I can see of you getting signal from one input onto both speakers unless you have really crossed up the primary side of the trafo and have one half of the winding going to one channel and one to the other channel.
Last edited:
Steve, it would be interesting to hear a comparison between LTP and cathodyne with the 6P1P outputs. Do you think you would get enough gain to try this type of feedback as used by Yves and in the Baby Huey and others? In the Baby Huey it was eventually found the GFB was not needed.
For the Mengyue, I think the LTP would be an easier mod than cathodyne, but I have no idea of what the sonic results would be, hence my interest in your experiments with your Svenn amp.
Edit: Glad ya still here Gary
For the Mengyue, I think the LTP would be an easier mod than cathodyne, but I have no idea of what the sonic results would be, hence my interest in your experiments with your Svenn amp.
Edit: Glad ya still here Gary
Yeah. When I heard that Raoul Island managed a 1.5m wave, it sounded like we were in for some action, but it hasn't eventuated. I live near the coast, and about 3m above sea level....
I'm interested in the LTP as well, but by all accounts it is tricky to get sounding right unless it is a VERY long LTP. You may have more luck with your higher B+.
I found a transformer for the heaters in the 812-A valves (each tube needs 4A current). Just got to find some sockets for them. You should just about be able to tell when I fire them up from where you are, Ian, if you look into the night sky
I would use a 6N2P for the LTP. With a mu of 100 there should be plenty of gain for both the local and global feedback. 6P1P for 6AQ5 is a no-brainer in my opinion with the circuit Yves has. It should work like a hose.
I've been thinking of trying an LTP with current source in the tail any way. I will take it one step at a time and try the local feedback after I get the LTP working with GNFB.
I think I have the layout about ready and will order parts Monday or so. I'll post updates under Sven PP.
I've been thinking of trying an LTP with current source in the tail any way. I will take it one step at a time and try the local feedback after I get the LTP working with GNFB.
I think I have the layout about ready and will order parts Monday or so. I'll post updates under Sven PP.
OK, I checked the 10K resistors....mine are 6.8 by the way and they test good. Fortunatly those two resistors were on top of the PCB. I disconnected the feedback wires as you told me and it worked. Both channels are fully independant now. Is there a way to connect them back to get them work. They certainly are not there without purpose?
Daniel
Daniel
It does the trick (disconnecting the feedback wires). Is there any danger running the amp without those wires connected?
You should be able to trace it out pretty easily. Go one wire at a time and listen to the changes in sound.
Looking at the schematic posted in the beginning of this thread, there are two wires coming off the output of the transformer to provide feedback. One attaches to ground, the other provides current mode feedback to the cathode of the first stage via the 10K resistor. Work with one channel at a time, and drive only that channel with an input. If you hear audio out of the other channel, you have cross wired something. Remove your last change and try again. Go one step at a time.
Looking at the schematic posted in the beginning of this thread, there are two wires coming off the output of the transformer to provide feedback. One attaches to ground, the other provides current mode feedback to the cathode of the first stage via the 10K resistor. Work with one channel at a time, and drive only that channel with an input. If you hear audio out of the other channel, you have cross wired something. Remove your last change and try again. Go one step at a time.
Gary, I'll keep an eye out for fireworks in the south-eastern skies, let me know when you are ready
Now just some more ramblings and a couple of questions, trying to learn the effect of different things - I have been intrigued by the operation of pentodes, and by the effect of the screen grid on the operation of the valve, how the voltage on the screen grid effectively changes the position of the grid curves on the static anode graphs (Va vs Ia). My amp is still running straight pentode mode, and no global feedback. I was wondering what the effect would be on the sound if all the grid1 to cathode voltages were set to the same voltage difference, and all the cathode currents were set to the same value by adjusting the screen grid voltage. This would hopefully force all the tubes to operate in the same area, and maybe work together better by sharing the load more evenly, well, that was what I was thinking. So I built 4 x simple LR8N voltage regulators, one for each screen grid. Thanks Gary for letting me know about the LR8N regs. I powered up and no smoke, and set all the grid to cathode voltages to -11V and cathode current to 35mA. The screen grid voltages were adjusted to voltages between 222V and 240V to get all four cathode currents to 35mA. So did it sound any different to before? Not really...it certainly isn't any worse, but I can't say for sure that there is any noticeably good improvement either. Any ideas on this? Do differences in output tubes' performance only affect 2nd harmonics which get cancelled in the OPT?
Another thing I tried yesterday (again) was tie each pair of LED arrays together to form one higher current array under the cathodes of each pair of output tubes, instead of an individual array under each output tube. Again I didn't like the sound, it seemed good and cleaner, but more sterile and less life in the sound. I have measured 0.5V rms at the cathodes with individual arrays at near clipping, does this give some local feedback to each output tube, and then when the arrays are wired in parallel the output tube pairs miss out on that individual feedback?
This amp is now a real mess inside, here is another angle, nothing here to be proud of but it is serving its purpose well (as a test bed to learn on in this case) and it sounds great (to my ears). There are four different voltages B+ 250V, 350V to the drivers, +/-50V to the mosfet source followers, plus the 4 regs for screen grid voltages. That might explain some of the flying wires.
Now just some more ramblings and a couple of questions, trying to learn the effect of different things - I have been intrigued by the operation of pentodes, and by the effect of the screen grid on the operation of the valve, how the voltage on the screen grid effectively changes the position of the grid curves on the static anode graphs (Va vs Ia). My amp is still running straight pentode mode, and no global feedback. I was wondering what the effect would be on the sound if all the grid1 to cathode voltages were set to the same voltage difference, and all the cathode currents were set to the same value by adjusting the screen grid voltage. This would hopefully force all the tubes to operate in the same area, and maybe work together better by sharing the load more evenly, well, that was what I was thinking. So I built 4 x simple LR8N voltage regulators, one for each screen grid. Thanks Gary for letting me know about the LR8N regs. I powered up and no smoke, and set all the grid to cathode voltages to -11V and cathode current to 35mA. The screen grid voltages were adjusted to voltages between 222V and 240V to get all four cathode currents to 35mA. So did it sound any different to before? Not really...it certainly isn't any worse, but I can't say for sure that there is any noticeably good improvement either. Any ideas on this? Do differences in output tubes' performance only affect 2nd harmonics which get cancelled in the OPT?
Another thing I tried yesterday (again) was tie each pair of LED arrays together to form one higher current array under the cathodes of each pair of output tubes, instead of an individual array under each output tube. Again I didn't like the sound, it seemed good and cleaner, but more sterile and less life in the sound. I have measured 0.5V rms at the cathodes with individual arrays at near clipping, does this give some local feedback to each output tube, and then when the arrays are wired in parallel the output tube pairs miss out on that individual feedback?
This amp is now a real mess inside, here is another angle, nothing here to be proud of but it is serving its purpose well (as a test bed to learn on in this case) and it sounds great (to my ears). There are four different voltages B+ 250V, 350V to the drivers, +/-50V to the mosfet source followers, plus the 4 regs for screen grid voltages. That might explain some of the flying wires.
6N2 to 6N1P(6N1P)?
Hi Gary,
I'm strategizing about what I might inexpensively try when I get my Meng X1. Looks like changing those 6N2's might be the first thing. So...couple of things: 1) Are the "6N1P" and the "6N1P-EV" interchangeable?, 2) if so, can just replace the stock 6N2's with them, without changing components?
I read this on a hifi forum: "You can swap the stock chinese 6N2 with a 6DJ8/ECC88 and 6922 types."
Is this true?
And I got this response from a guy selling the 6N1P-EV's when I asked him if I could substitute those for the 6N2: "You can technically, the 6N1's just have lower Gm and more filament current than the 6922."
Maybe this is saying approximately what you did...?
I'm thinking I'll give mine a listen as is when it comes, then if I don't like what I hear I'll first try changing the pre-amp tubes, and next the power tubes.
--David
Hi Gary,
I'm strategizing about what I might inexpensively try when I get my Meng X1. Looks like changing those 6N2's might be the first thing. So...couple of things: 1) Are the "6N1P" and the "6N1P-EV" interchangeable?, 2) if so, can just replace the stock 6N2's with them, without changing components?
I read this on a hifi forum: "You can swap the stock chinese 6N2 with a 6DJ8/ECC88 and 6922 types."
Is this true?
And I got this response from a guy selling the 6N1P-EV's when I asked him if I could substitute those for the 6N2: "You can technically, the 6N1's just have lower Gm and more filament current than the 6922."
Maybe this is saying approximately what you did...?
I'm thinking I'll give mine a listen as is when it comes, then if I don't like what I hear I'll first try changing the pre-amp tubes, and next the power tubes.
--David
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.