PI tubes; balanced or unbalanced that's the question

[conger]

Seems to me you aren't going to whip out your resistor collection every time you change a PI tube? so TRYING to balance the PI is an excersise in futility.
Stew

On the other hand you could use a pot as outlined here by BooBird and just adjust it by ear.

[lovedat800]

I guess that depends on the person. For me it's not an issue because preamp tubes last me years. I have some Chinese from the 80's i still use and sound like new. So i can't envision needing to do it once every 8 or 10 years. But even then, when i imbalanced my PI, the imbalance was such that it was more than any tube makes. When i first built my current project i tried a number of tubes in the PI and none made a difference like imbalancing the PI via the resistors. In fact, never in any amp have i noticed much of a change in the tone when switching PI tubes unless i was going to a different brand that just inherently sounds different. Adding even 4.7k in series to the 100k resistor made a difference no swapping would ever do. at least thats what i find. So once you change the balance via the resistors the tonal difference will probably be there no matter what tube you put in it.

[stoo]

I just read the artical on the Aikens sight re; the LTP PI. He says ..if you're using the second i/p on the PI as a mixer for the other channel then the plate resistors should be the same so as to not compound the issue for the other channel. Also that a high value tail resistor mitigates the inbalance to an extent and I'd say 56K is fairely high.
Stew

[zaphod_phil]

merlinb, who has the valve wizard web site, did some tests with PIs and found that after about 15k there was little improvement in balance as you increased the tail resistor value.

Seems to me you aren't going to whip out your resistor collection every time you change a PI tube? so TRYING to balance the PI is an excerise in futility.

Well provided you stay with the same generic type of tube - eg 12AX7 - then the PI will stay approximately (un)balanced to the same extent. Although no two 12AX7s (or 12AT7, 6SL7s, etc) are the same, there isn't enough variation between them to make a significant difference to PI balance in a guitar amp. In a finely knife-edge balanced hi-fi amp it may be different,

[overlock]

I have a couple of questions about this, if anyone can help. The trimpot idea sounds great, but I'm having trouble visualizing how it lays out. I was thinking the logical place for the pot would be across the HT ends of the two load resistors, with the wiper connected to Vc, but that doesn't really fit with the description in this thread. Any chance of a partial schematic?

Also, is 0.5W OK for the carbon comp resistors? By my inexpert estimate, they could be dissipating around 0.45W when one triode is off - does that sound about right?

[zaphod_phil]

I've never tried the trimpot thing myself. 0.45W sounds rather high to me. As I recall, you usually you get less than 0.25W in that position. Don't forget that there's a significant voltage drop across the PI's tail resistor.

[dotfret]

Resistor failures are a matter of taste - when I was a lad, metal film were desirable because they were the most exact value types, combined with low noise.
As time went by, people avoided the MF for a while, because they "went high" if they failed, and usually took out some associated transistors in the process.

Parts you buy these days are a long way forward - they are all "low noise" compared to the old days, the precision is amazingly high, you need to look at long term life to decide.

The objective here is to build amps based on old designs. Remember, when those designs were conceived, the valves and inductances were the precision parts. Resistors came +/- 20% as standard, and capacitances were as bad as 50% off the nominal value - and people succeeded in producing working designs.

That's why some old amps sound better or different to others with the same designations - you get variations.

90% of the specified work is probably OK for any resistor - but don't ever try to exceed the values quoted.
Sometimes capacitances will work at higher voltages than specified, especially PIO and silver/mica types - but don't try pushing it with other components.