Within a 12AX7

[jeff]

Is there interplay between the two triodes in a 12AX7 that could cause problems?

Think of a champ. I added another gain stage and cathode follower between the first and second triodes(after the first before the tone stack). Now my first gain stage and last gain stage are within in the same tube. Now the two triodes are in phase where before they were out. Could this cause a problem?

I've pushed wires around and the layout of the amp has the wires/components of V1A going up(phisically located above) and V1B going down so I don't think they're too close could something be going on within the tube.

Would rewiring so that V1A was the first two triodes and V2B the second two be different that having V1 be the first and last gain stage?
Know what I mean? I know having the wrong wires too close to each other can cause problems but can having the two triodes, first and last gain stage(now in phase), in the same glass envelope cause a promlem or the two triodes pretty much electronically isolated(capacitance between the two triodes causing problems???)?

[RicharD]

Yes there can be acoustic coupling between stages, especially in high gain cascading amplifiers.  Slap it into a combo and you could end up with a squealer.  More than once DL and I have built circuits on the bread board (all nice -n- spread out) but it howls in a tight chassis.  One of our first builds worked great on the bench but oscillated in the combo cabinet.  Layout is important.  If you have 2 stages cross-talking within a 12AX7, you probably have other layout problems contributing to the situation. 

[jeff]

OK I just thought of something easy to test. I pulled V2. If there was some kind of cross talk I'd get signal to the power tube and I don't so that's not it.

Here's what I did do. I disconnected the volume from the tone stack and I am getting sound out of the amp so something's up. Probably my layout just can't nail it down.

 When I have the master full up and turn the gain up I get "beeeeeeooooooooop" as I turn it up, then it goes away as I turn it up further. If I turn the gain full up and turn the master, same thing. "beeeeeooop" then as I turn more it goes away. So It seems that there is a "sweeet spot", where the right amount of gain causes problems. If both are full up no problem but if you turn one or the other down both a little you get the problem. If you turn the gain up past the noise, until it just goes away, then turning the master down makes it come back. It's like a balancing act between the two and only the right amount of gain will do it. I would have thought that it would get worse and worse as you turn it up but that's not the case.

Does this how an oscillating amp normaly reacts.

[overtone]

have you tried having that last stage on a different triode bottle?
The beeeeeeeeeooop reminds me of when I had a V1 picking up signal from a cable running to the reverb transformer, you could also kind of dial it in and out. It also came in at different levels depending on the "master". In that case it was lead dress near V1.

[RicharD]

It does sound like a layout/oscillation problem to me.  Overtone gave 1 good example.  Proximity of high level stuff such at the OT or it's primary wires in relationship to early preamp stages also comes to mind.  It's also possible that the amp is oscillating the entire time but typically ultrasonically.  Hard to know for certain w/o a scope..... or a well trained doggie.  :P

[HotBluePlates]

I know others have done it with no problems, but if I were building an amp from scratch, I wouldn't build it the way you described. I would organize the gain stages in the same order as they appear on a schematic. That is, Your first two gain stages (schematic-wise) in one bottle, then the cathode follower and last gain stage in a second bottle.

That's just a good rule of thumb to follow. It lends itself to a layout such that large-signal wires don't get near small-signal wires.

Of course, you might build the amp the way you originally described 10 times, and not have oscillation issues in 9 out of 10. Solving the issue in #10, though, will drive you nuts.

Also, in theory, you shouldn't need an added filter stage, however, you may wish to consider adding one to help with decoupling the stages.

Lastly (maybe it should be firstly), it could just be the individual tube in question.

[kagliostro]

I've the same problem and still thinking how to solve it

---

there are some tubes like russian 6n1p that have a shield between the two triodes



I think that if they decided to put it on they had a reason

Kagliostro

[overtone]

the OP first question: yes, you basically have a later stage running right next your first stage inside the tube.
You may just get away with running a much, much later stage through the same bottle, but usually it ends in tears and increasing hair loss.
 
The 6n1p is an interesting thing but I am not familiar with it.
Suggestion: a quick and dirty test for your set up would be to rig up an extra socket and tube for that final stage and take it from there.

Best, tony

[jeff]

"I know others have done it with no problems, but if I were building an amp from scratch, I wouldn't build it the way you described. I would organize the gain stages in the same order as they appear on a schematic. That is, Your first two gain stages (schematic-wise) in one bottle, then the cathode follower and last gain stage in a second bottle."

I thought about that but the resaon I desiced to go with this was

1) I thought it made more sense if I wanted wanted to try out different tube types 12AU(T,Y)7s
   
2) If something happened to V1, like the heater went, wouldn't that be bad for V2? Since there's no current draw through the 100K plate resistor of the third stage, the plate and grid of the cathode follower would be at the same voltage? Lose two tubes if one failed. Maybe not but that's what I was thinking.

"It's also possible that the amp is oscillating the entire time but typically ultrasonically.  Hard to know for certain w/o a scope..... or a well trained doggie.  :P"

Not sure how to test this, I do have an Oscope. What do I do, just hook to a dummy load and look for waves? Please tell me more.