Need advice

[daz]

I decide to try something i've been thinking about for a long time. Add a 4th AX7. Heres why. My preamp sounds great. Problem is, it sounds great till you turn the master up and then the tone changes radically due to the way the master changes tonality as the always do. Mainly because i designed the preamp very bright in order to get the best possible tightness in the lows. The tone i love happens at the master when down around 8:00 because thas where the master with no treble bleed attenuates all that brightness. If i take that attenuated signal and feed it into another amp thats CLEAN, it sounds fantastic at any volume. So i want to attenuate it at the master, then take that fully processed tone and feed it to another clean stage to drive the PI. So stage1>stage 2>stage3>Cathode follower>tone stack>voltage divider>recovery stage>PI>PA.

If i do this and it works i will build a whole new amp with 4 AX7's. Question is, what issues might i run into and what values would you use on that stage to get the most clean headroom possible? PT has plenty of heater current.

[Alan0354]

I am doing very similar to what you are doing, I just don't see much of the problem after lowering the output level as I described in length to you before. I did bring the volume up and it sounded the same up to 2/3. And after that, it change some but it's acceptable. I think everyone tried to give you suggestion already.

If that border you, design a bigger power amp so you never have to push the power amp!!! Why keep playing with the preamp when you are happy and want to change it. It's time for you to increase the power amp, lower the closed loop gain so it never clip.

[daz]

I don't mind that change way up loud because i'll likely never use it there, and if i did thats power stage clipping with isn't a bad thing anyways. And 2/3 is LOUD.

[Alan0354]

I think you can get there by matching the input output dynamic range in the last thread. What you want should be very simple.............Don't over drive the power amp!!!!

Actually it really helped me just by responding to your post as I didn't like the change of the sound to a grunting like sound when I cranked it up. After I reduce the output dynamic range of the preamp, it improved a lot. I am happy camper. I don't know why you still have problem.

[daz]

I think you can get there by matching the input output dynamic range in the last thread. What you want should be very simple.............Don't over drive the power amp!!!!

Actually it really helped me just by responding to your post as I didn't like the change of the sound to a grunting like sound when I cranked it up. After I reduce the output dynamic range of the preamp, it improved a lot. I am happy camper. I don't know why you still have problem.

Thats not my concern (too much signal) in doing this tho. In that other thread it was, but after all that i realized that if i rectify the larger problem, both will be fixed. Like i said, the larger problem is that the master like all volume controls sounds very dark when turned down. Thats where i designed the preamp, and therefore when i turn up it gets very bright due to the changes the pot causes going away as i turn up and the preamp brighness is revealed unlike at low settings. with a gain pot in the preamp or in a guitar we add a treble bleed cap to get the tone to stay relatively the same from low to high. With the master you can't do that or the tone will change radically. I tried. I placed a treble bleed on the master thinking that while it makes the tone complete treble to an unusable degree, if i could cut the highs after that successfully then the tone would remain the same thru the pot's range. I used a cut control with a .022uf (came to that value after trying every one) to get that to happen and it worked. I was able to get the tone to stay the same. Problem is, the tone was too bassy and uncontrolled. No focus like i need. I tried everything imaginable. It just doesn't work. Thats when i decided to try another gain stage for clean gain. It may also be tonally different but the opposite....dark at low settings. However, a treble bleed will work in that scenrio and i should be able to get it loud to low with little difference in tone or feel aside from the obvious hearing issues (munson curve i think it's called) and speaker/PA compression up high.

[cbarrow7625]

I hear what you are saying about the tone changing as the master volume is raised. I've had the same experience. It sounds like you have an opportunity here to improve the circuit to work even better than you have thus far expected.

First, find the tone sweet spot on your master volume and replace the pot with a voltage divider of the same values. I assume that the shunt leg of the master volume is going to end up being a fairly small value; you can replace that resistor with a smaller value pot to act as the new master volume. The smaller pot should give you less tonal change from 0-10. this is all pretty rudimentary; I'm sure you've thought of it already. Here's where I think you have an opportunity:

If the preamp with the master down sounds good through a clean power amp, it makes sense that we might want to do something to make sure that the power amp in your build works better than the average long-tailed pair. Instead of just tacking in another gain stage in front of the existing PI, why not change the PI to a 2-stage PI to get your gain AND reduce distortion while producing a better balanced signal going to the power tube grids? There are some great circuits here: New Phase Splitter/Inverter AMPLIMOS one stage amplifiers In particular, I think the one down the page a ways called "Figure 4 Direct Coupled Long Tail Pair Phase Inverter" would fit the bill perfectly. The 12BH7 tubes in that schematic are very similar to 12AU7's. I'm sure you could make the concept work with 12AT7's or 12AX7's if you wanted to; although you would probably have way more gain than you need. The circuit couldn't be simpler and it will give you the extra gain needed without additional series DC blocking capacitors or cathode capacitors (which may cause additional undesirable frequency and phase shaping in your "perfect tone" circuit.

Another cool thing about that circuit is that you could set it up to allow local negative feedback from the output of the PI back into the first cathode (same place the global negative feedback is feeding) to dial in the exact amount of gain you want. The negative feed back could even be completely "decoupled" from the signal path by using the other 1/2 of the extra tube you are going to install. Picture this: The unused half has it's cathode directly tied to the cathode of the PI input tube, the plate goes directly to the supply (no plate resistor) and the grid is fed directly from the output of the upper PI tube (after the coupling cap). To really get it working optimally, you would probably want to use a voltage divider to set down the feedback voltage and also use a bootstrap arrangement on the input tube to provide a littlemore voltage swing in the cathode so that it doesn't clip. That's the "fancy way" to do it; maybe more trouble than it is worth. It would just be easier to make some portion of the 100k load resistor from the upper PI tube into a pot and use it to adjust the amount of voltage fed back into the cathode. Obviously you wouldn't want to tie the output of that PI stage directly to the 3.3k cathode resistor, the tube wouldn't be happy driving that load. Something like 90K in series with a 10k pot to ground and take the output voltage off of the 10k pot wiper. That would give you a VERY wide range of gain control (especially if using 12AX7 tubes) and still effectively keep the negative feedback "out of" the direct signal path. The feedback from the amp output could still be fed back to that same node as long as the series feedback resistor was fairly high ("normal" ranges for guitar amps).

.....but that'as just one way to handle it!

[daz]

Thanks, that page looks interesting. However, looking at the circuit you mentioned, that shows 3 stages which would require i add a tube anyways. Or am i missing something?

[R.G.]

Problem is, it sounds great till you turn the master up and then the tone changes radically due to the way the master changes tonality as the always do.

I ran into this issue in another way in a volume pedal design. The amount of treble changed with pot setting. This happens because the effective load on the driving stage and the source impedance at the wiper into the driven stage change with pot setting.

With a pot at either extreme, the source impedance driving the wiper output is minimum. At the peak, it's just the driving impedance of the stage driving the pot. I'm guessing you have a 12AX7 plate driving a cap into the "hot" lead of the pot, so you have the signal voltage in series with about 60K driving the hot lead. When the pot is turned up to maximum, the pot is just a resistor to ground in parallel with whatever the wiper drives, probably a grid and maybe a capacitor then a grid.

With the pot at minimum, the source impedance to the driven stage is nearly zero because the wiper is nearly at ground.

In the middle, the impedance seen by the driving stage varies from the pot resistor in parallel with the driven stage up to just the pot resistance as the wiper is at zero. It hits a minimum at exactly half-resistance point. There, the driving stage sees half the pot in series with the lower half in parallel with the driven stage as a load. The driven stage input sees a source impedance of 1/4 of the pot resistance.

With all that impedance wobbling and the capacitors at the top of the pot and possibly the wiper, yep, the amount of treble changes.

The solution I found was to buffer the pot. Buffering going in meant that the change in loading from the pot wiper position changing did not matter to the incoming signal, as the loading (and hence the loading *change*) was insignificant to the driving signal. I buffered the output as well, because I had some feet of cable capacitance to drive. The high input impedance of the buffer at the pot wiper meant that no treble was lost driving the cable, so no treble loss variations happened.

In a tube amp setup like yours, before you go to the work to drill a hole in the chassis for a new 12AX7, it might be interesting for you to - just as a lark, just as an experiment to see what happens - wire in a MOSFET follower on the plate of the driving tube. This might eliminate a lot of the tone variation in the master pot. Maybe - but maybe not. If it doesn't, you have only lost some time, one $1.50 MOSFET, and a resistor. If it helps, depending on how much it helps, you may or may not need to think about a buffer on the wiper. I really suspect that the grid of the following tube may be high enough and not need a buffer.

Just thinking.

[daz]

Thing is RG, you're looking at it from the perspective of tweaking it so theres no treble loss. But my issue is that i WANT that treble loss from the master because thats where it sounds good. But since at that point the level coming out of the pot is so low the amp's volume is at home or bedroom level. Therefore i'm thinking a voltage divider to set the tone like the master set low and an extra tube to boost the level back up so it can amplify that darker tone to full 50 watt amp levels.

[potatofarmer]

Have you tried using a dual-ganged pot for your master, with one deck controlling the master volume and the other wired (backwards) as a treble control to roll off highs as the master is raised?

Or you could just add another treble pot right off the top of the master control. Call it "sparkle" or "fizz" or some such.

[daz]

Have you tried using a dual-ganged pot for your master, with one deck controlling the master volume and the other wired (backwards) as a treble control to roll off highs as the master is raised?

Or you could just add another treble pot right off the top of the master control. Call it "sparkle" or "fizz" or some such.

I've thought about that, but i'm positive that i'd never be able to get a ganged pot where the one used to attenuate highs would be a value and taper that would track exactly the right to cut the highs as needed at every volume setting. Not to mention getting the cap value right and all that. I'm positive theres zero chance of retaining the exact tone exiting the master or even close. i would need a lot of experimenting and a custom pot to the exact specs to even make it somewhat close unless i happened to get very lucky.

[cbarrow7625]

Oh, yeah. You definitely have to add a tube one way or another. I wasn't trying to help you get around that. Unavoidable. I was just suggesting that you may be able to take advantage of that extra tube to get results that are even better than what you originally might have expected from just adding another gain stage ahead of, and decoupled from, the PI.

[daz]

Oh, yeah. You definitely have to add a tube one way or another. I wasn't trying to help you get around that. Unavoidable. I was just suggesting that you may be able to take advantage of that extra tube to get results that are even better than what you originally might have expected from just adding another gain stage ahead of, and decoupled from, the PI.

Ahh, I see. Yeah, i will have an extra stage unused, but thats ok. I've always wanted reverb so maybe if i can find a way to implement it with one tube stage i may do that down the road. I'm planning on trying this tomorrow by just tacking it in temporarily. I think i may draw it up with some values and post it to see what you and others think. I was figuring 100k plate, 1.5k cathode with no bypass and 1M grid. Then a master with treble bleed so the tone doesn't get real dark down low where i use it most of the time.

[daz]

Any advice on this drawing as far as mistakes or changes that would be good?

[cbarrow7625]

I don't think you're going to be too happy with that arrangement. If you are taking the unattenuated output from the tone stack directly into the extra tube stage, you are just going to overdrive it pretty badly. A better arrangement would be to set a voltage divider in place of your current MV at the MV sweet spot and feed the additional tube with that. There is little point in using a 1M MV after the additional tube stage; the changing impedance will interact with the PI and change your tone anyway. As I mentioned in my earlier post, the best arrangement is probably going to be replacing the resistor to ground in your voltage divider with a lower value pot; It will act like your current MV from 0 to the "sweet spot" that you set at maximum. This arrangement will also give you additional headroom in the extra tube stage as you turn the volume down. A 12ax7 probably has WAY more gain than you need to just re-amplify the attenuated MV output unless you sweet spot is very low on the pot rotation. You might want to consider using a 12AU/T/Y7 instead. You probably can just determine by experimenting.