Hi Brains Trust,
I have recently made a discovery that has really surprised me, and wanted to find out if others have had a similar experience.
I have been supplying 6N2P valves for an Amp School that I run here in Sydney and numbers of these amps have been built successfully this year. However, there has been a niggling issue that until now I have been unable to lay to rest. The issue concerns the 6N2P valves used in the amps that students build. For those who are not familiar with this valve, it is effectively a 12AX7 with a 6.3v heater.
A few of the constructed amps have had a weird distortion when the preamp is overdriven - the preamp sounds great when run clean, but when overdriven has had a really nasty sounding distortion that I was able to "fix" by changing a few preamp valves. However, I haven't been really satisfied with this situation, and have sine wave tested a few sample amps and the ugly distortion is clearly evident when the preamp is overloaded. The odd thing is that the valves bias perfectly, and when a 6N2P-EV is substituted, the distortion seems to go away. So I have had a growing pile of "faulty" 6N2P valves that stare at me in the workshop and stir me to think about this issue on a regular basis.
As mentioned, the valves bias fine, however when driven by a sine wave, the signal triangulates (looks like slew rate limiting) and when driven further there will be a breakthrough on the peaks of the signal - no wonder it sounds so nasty
Yesterday, I thought maybe I had a bunch of valves that had low emission so I cranked up the heater volts to get an indication of heater current and performance at higher heater volts - made no difference
Anyway, in desperation, I remembered that these valves have a shield between the two triodes - in fact this in common with a heap of similar twin triodes (ECC88, E88CC, 6DJ8 etc) and this shield is brought out to pin 9. Many valves have internal connections brought out to valve pins and normal practice is to leave them unconnected which is what we have done in the Amp School amps. Surely this shield could not have anything to do with this issue?????????????????
Well to cut a very long story short, to my amazement when I shorted the unconnected pin 9 to ground this nasty distortion disappeared!!!!!!!!!!!!!!! And it gets weirder!! When I removed the ground from pin 9 the distortion is still gone! The only way to get the distortion back is to replace the valve with another "faulty" valve from the bin and sine wave drive it again - distortion. Ground pin 9 - clean full sine waves!!
So, has anyone here had similar experiences?? And who is brave to suggest the mechanism by which this could occur?? The best idea that I can come up with is that with all the electrons flying around inside the valve envelope, the shield develops a static voltage buildup and that this somehow affects things. But how on earth can that affect the electron flow within the triode envelope itself???
Your thoughts???
I have recently made a discovery that has really surprised me, and wanted to find out if others have had a similar experience.
I have been supplying 6N2P valves for an Amp School that I run here in Sydney and numbers of these amps have been built successfully this year. However, there has been a niggling issue that until now I have been unable to lay to rest. The issue concerns the 6N2P valves used in the amps that students build. For those who are not familiar with this valve, it is effectively a 12AX7 with a 6.3v heater.
A few of the constructed amps have had a weird distortion when the preamp is overdriven - the preamp sounds great when run clean, but when overdriven has had a really nasty sounding distortion that I was able to "fix" by changing a few preamp valves. However, I haven't been really satisfied with this situation, and have sine wave tested a few sample amps and the ugly distortion is clearly evident when the preamp is overloaded. The odd thing is that the valves bias perfectly, and when a 6N2P-EV is substituted, the distortion seems to go away. So I have had a growing pile of "faulty" 6N2P valves that stare at me in the workshop and stir me to think about this issue on a regular basis.
As mentioned, the valves bias fine, however when driven by a sine wave, the signal triangulates (looks like slew rate limiting) and when driven further there will be a breakthrough on the peaks of the signal - no wonder it sounds so nasty
Yesterday, I thought maybe I had a bunch of valves that had low emission so I cranked up the heater volts to get an indication of heater current and performance at higher heater volts - made no difference
Anyway, in desperation, I remembered that these valves have a shield between the two triodes - in fact this in common with a heap of similar twin triodes (ECC88, E88CC, 6DJ8 etc) and this shield is brought out to pin 9. Many valves have internal connections brought out to valve pins and normal practice is to leave them unconnected which is what we have done in the Amp School amps. Surely this shield could not have anything to do with this issue?????????????????
Well to cut a very long story short, to my amazement when I shorted the unconnected pin 9 to ground this nasty distortion disappeared!!!!!!!!!!!!!!! And it gets weirder!! When I removed the ground from pin 9 the distortion is still gone! The only way to get the distortion back is to replace the valve with another "faulty" valve from the bin and sine wave drive it again - distortion. Ground pin 9 - clean full sine waves!!
So, has anyone here had similar experiences?? And who is brave to suggest the mechanism by which this could occur?? The best idea that I can come up with is that with all the electrons flying around inside the valve envelope, the shield develops a static voltage buildup and that this somehow affects things. But how on earth can that affect the electron flow within the triode envelope itself???
Your thoughts???
Not necessary. It it was a trace of some residue between anodes that are on approximately equal potential it could evaporate after applying the full voltage in the middle.
What you described was looking like some kind of parasitic feedback. However, I could not open your images, though...
Edit: oh, never mind: that images were kind of residue from Australian guitar forum from which you copied the message...
Last edited:
If the pin is left unconnected, the distortion might return after hours, or maybe days -- the time it takes for that static charge to rebuild. Something to try out.
Interesting,
In the Aikido pre circuit it talkes about AC grounding with a capacitor. Other audio dual triodes like ECC... do not have a shield. I assume due to the operating frequency. Even stranger that you have batches of "problem" tubes.
gcwills can you please repost the images we can't get to see them! Thank's
Regards
M. Gregg
In the Aikido pre circuit it talkes about AC grounding with a capacitor. Other audio dual triodes like ECC... do not have a shield. I assume due to the operating frequency. Even stranger that you have batches of "problem" tubes.
gcwills can you please repost the images we can't get to see them! Thank's
Regards
M. Gregg
Last edited:
Hi,
I am reading this information:
Wapedia - Wiki: Vacuum tube
It has some information that anyone interested in tubes will find interesting! If anyone finds something in this info please comment!
Regards
M. Gregg
I am reading this information:
Wapedia - Wiki: Vacuum tube
It has some information that anyone interested in tubes will find interesting! If anyone finds something in this info please comment!
Regards
M. Gregg
Last edited:
Sorry, should have mentioned I only ever used -EV versions. But then, the hi-fi people don't normally push any stage of the amp into hearable distortion either, cause then we back off. If it was a guitar amp, then I guess turn it up to 11 and then some more.
To investigate what is happening, connect a DVM from pin9 to ground and see what happens to the voltage on the screen when it is left floating (apart from the DVM input impedance 10M?). If it is collecting electrons (from anode holes?) then it will go negative. On the other hand, it could go positive if each electron arriving kicks out more than one secondary electron - it could get very positive!
If it can receive electrons from the triode structure then it can affect the triode structure. Does the valve have anode holes facing the screen? This could be the difference between those which show the effect and those which don't. Or it could be the exact metal or coating of the screen affecting secondary electron emission.
If it can receive electrons from the triode structure then it can affect the triode structure. Does the valve have anode holes facing the screen? This could be the difference between those which show the effect and those which don't. Or it could be the exact metal or coating of the screen affecting secondary electron emission.
The "images" are just smilies from the Australian forum where Grant first posted his conundrum. They appear correctly on my screen - probably because I frequent that particular forum, and the smilies will be in my temporary internet files.
Looking at Grant's schematic, there are no grid stoppers on the first two triode sections. Perhaps one or both triode sections are on the brink of oscillation, and a large input signal tips the balance. Earthing the shield may be just enough to keep them stable. This doesn't really explain the ongoing stability once the shield is again disconnected though.
If the valve manufacturers went to the extra trouble to include a shield between the triodes, perhaps we should just make use of it.
It would be "nice" to know if it is linked to Miller effect!
The problem not showing after removing the earth is a mystery!
What can remain "magnetic effects" "static/ electrostatic" would discharge after a time!
Or as Wavebourne says contamination being "burnt off". Interesting it dosen't burn off until the shield is connected to earth!
So the shield was not connected to earth at testing?
The problem not showing after removing the earth is a mystery!
What can remain "magnetic effects" "static/ electrostatic" would discharge after a time!
Or as Wavebourne says contamination being "burnt off". Interesting it dosen't burn off until the shield is connected to earth!
So the shield was not connected to earth at testing?
Last edited:
Well if we ground the shield we are changing the capacitance within the tube. Its just a thought!
The Valve Wizard
Things like this do make the hobby interesting!Grounding the screen will mainly affect the capacitance between the two anodes. Mller effect arises from the capacitance between grid and anode. However, if the two halves of the triode are cascaded then one anode will be coupled to the other grid, so there could be an effect but all it would do is change the HF frequency response a little. Unlikely to introduce serious distortion, and it would be repeatable i.e. remove the ground and the problem comes back straight away.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.