phase inverter question

[daz]

In a JCM 800 style PI would it give optimal results to change the tail resistor from 10k to say 47k is swapping from a 12AX7 to a 12AT7?

[Chuck H]

What is the goal? "Optimal results" is very subjective WRT guitar amps. I would go so far as to say the optimal circuit values for a JCM800 are already in the circuit. Otherwise it's not a JCM800 PI and therefor not optimized to be like a JCM800!?! But if by optimal you mean something like more balanced, better frequency response, lower noise, lower distortion, more swing voltage, higher headroom, etc. etc. Then we really do need to know what it is you want to change about the stock circuits performance.

Changing from a 12ax7 to a 12at7 and upping the tail to 47k with no other circuit value changes would actually create less gain on the inverted side and greater imbalance between the two halves. This, according to an on line calculator I just used. And this may or may not be optimal for what you want to achieve. I don't know. These changes may also affect the NFB circuit. The calculator I used doesn't provide this information. But because the voltage division between the bias and tail resistor would be changed I would expect this to be the case. You might expect that since the 12at7 is a lower mu tube that the output impedance would go down. But in fact I do think the higher tail resistor, being in series with the current flow, might increase output impedance.

You may be able to optimize (depending on what that means) for a 12at7 tube by changing other values in the PI as well. But, FWIW, Fender tried this with the SF amps and those don't win any popularity contests. With most players opting to "blackface" their silverface amps by changing to the unoptimized circuit values. Much of this involves changes to the silverface "optimized" PI circuit.

[JoeM]

....
... This, according to an on line calculator I just used.

...

Chuck, Where is that online calculator?

[Chuck H]

This site has several tube amp spicey, cady type stuff that is really good for those of us that haven't invested the time to learn the available software. Here's the PI balance/gain calculator. But do investigate the rest of the site. Good stuff.

Balancing Long-Tailed-Pair Phase Inverter Gains

[JoeM]

Thanks Chuck.

[daz]

Well, i just figured the circuit was designed around a X7 so i assumed that a T7 might put something out of balance. Maybe for example it wouldn't give the best results as far as dynamics, which is actually why i was using a T7 in the first place....i read that they give better dynamics. What i really want out of the PI it optimal dynamics and clean gain. Thats not to say i want pristine or anything, but i certainly don't want the PI to distort very easily. The main reason i'm using it however isn't because i found the dynamics to be better, at least not to a noticeable degree at low volumes, (i suspect at stage vol it would be noticeable tho) but because i find that turning down the guitar to clean up yields less nastiness in the very high end....smoother highs.

I guess what i would say to the question what do you want would be best dynamics possible and clean and articulate. My amp gets all of it's drive in the preamp, and i really like it's tone w/o adding anything after that point. So a clean and dynamic PI is desirable.

[Chuck H]

Gotcha. Most Marshall amps run the PI bias a little hot. You can cool that down by increasing the 470 ohm resistor that is right on the PI cathodes to 820 ohms. That'll give a little more headroom on the positive swing. Since the power tubes are in cutoff during the negative swing, the positive swing is all we care about. It should also help to raise the tail resistor a little. Maybe up to 22k. This may increase headroom a little. Not certain on this one. But it does raise the standing voltage and that, to me, means that there's less likelyhood of interaction due to the AC signal. But raise it too much and gain may drop. So it seems like a balancing act to me on this value. 22k seems like a safe bet. It certainly worked for the Twin Reverb. Which is notorious for big clean tones. According to the calculator this will start to imbalance the inverting side low. Replace the 82k plate resistor on the inverting side with a 90.1k or a 91k (whatever standard value you can find). This should get the balance pretty darned close. Next would be the plate voltage. You can probably increase the plate voltage to gain headroom and dynamics. Decrease the value of the HV rail resistor feeding the PI plates. Usually 10k or two 10k's in series in most Marshall amps. Maybe try a 5k (to replace a single 10k) or a jumper over one of the 10k's (for series 10k's). This would also increase the preamp plate volts. Which you may not want since it will change the tone of the preamp. If you have an amp where the PI and cathode follower HV shares the same node (like the 1987) you may want to skip this. But if your amp has another dropping resistor before the preamp this can be increased by roughly the amount of the decrease in the PI feed resistor. That will normalize the preamp volts and you'll get the higher PI volts for better headroom.

I hope this doesn't come across too complicated because of my windy description. Because it's really not.

Oh... Go ahead and try the AT7 too. Not much difference according to the calculator but they do always sound a little cleaner to me. And on that note, since your going to some trouble to balance the PI it makes sense to try several AT7's if you have them. Since the two triodes won't be perfectly balanced anyway, you will find one that seems to sound best. Whether it's because of better balance or because of imbalance is another discussion. But do ear test audition several tubes for tone.

[daz]

Oh... Go ahead and try the AT7 too. Not much difference according to the calculator but they do always sound a little cleaner to me.

I think you may have misunderstood, or maybe i'm misunderstanding. But let me clarify in any case. I am using a AT7 now. I posted to ask whether i should change values with the AT7 in order to be sure i am getting the better dynamics and cleaner tone that they are known for. I just figured since this circuit is designed around the AX7 it may not be optimal for the AT7. From that last sentence in the quote about i got the impression you were suggesting those values to get more clean headroom from a 12AX7. But if you WERE suggesting those changes when using a AT7 i will definitely try them, thanks.

[Chuck H]

The at7 and ax7 operate very similar as voltage amps. When you really need gain the ax7 wins. When you need current the at7 wins. But as voltage amps the two are much more compatible swapped for each other than the au7, ay7, etc. All the things I mentioned above apply to the at7 as well. There may be some additional benefit lowering the value of the plate loads too as I think the at7 is data'd for a lower load. Though this may benefit the distortion characteristics it will also sacrifice some gain. Fender also used lower grid resistors. I'm sure they had a good reason but I don't know what it is.

[daz]

The at7 and ax7 operate very similar as voltage amps. When you really need gain the ax7 wins. When you need current the at7 wins. But as voltage amps the two are much more compatible swapped for each other than the au7, ay7, etc. All the things I mentioned above apply to the at7 as well. There may be some additional benefit lowering the value of the plate loads too as I think the at7 is data'd for a lower load. Though this may benefit the distortion characteristics it will also sacrifice some gain. Fender also used lower grid resistors. I'm sure they had a good reason but I don't know what it is.

Thanks. I'll give an 820 and 22k a shot and if that seems to sound/feel better i'll then consider lower plates. I guess a 82k where the 100k was and a 68k where the 82k was might be a good choice for a start.

[deci belle]

That's a nice windy description of exactly what I'm tweaking in my Showman, Chuck… thanks, Man!

[J M Fahey]

.i read that they give better dynamics.

Don't believe everything you read at Forums.
Even less if taken out of context.
Were they specifically discussing Phase Inverters?
...............
I thought so.