yeah, that could explain all the plate to plate caps and plate to ground caps and so forth in the schemo. that and the fact that it sat under some projector motor may be partly why it originally used 12vdc heaters.
its a "tight" sounding amp, though, and now with a decent PI driver its loud and "clean". surprisingly balanced. i've often considered converting it to 5e3 or some other schematic, but it sounds good as it is - i collect (though no longer seek out) small amps to acquire the pauper's version of a collectors selection, for both gigging and recording. i'm not really looking to change the sound. in fact, that dying 12au7 sounded awesome for a particular range of material - i'm tempted to keep it on hand to plug in when the material calls for it.
thanks again for all the input. i'll update if something about the PI comes up, but otherwise I'll start a new thread if the microphony proves to be a problem.
cheers!
its a "tight" sounding amp, though, and now with a decent PI driver its loud and "clean". surprisingly balanced. i've often considered converting it to 5e3 or some other schematic, but it sounds good as it is - i collect (though no longer seek out) small amps to acquire the pauper's version of a collectors selection, for both gigging and recording. i'm not really looking to change the sound. in fact, that dying 12au7 sounded awesome for a particular range of material - i'm tempted to keep it on hand to plug in when the material calls for it.
thanks again for all the input. i'll update if something about the PI comes up, but otherwise I'll start a new thread if the microphony proves to be a problem.
cheers!Back to the PI -
this second stage has a grid resistor (ground ref) that is not in the "floating paraphase" circuits that I've been reviewing - is this providing the input ground reference? is it also providing ground reference to the output tubes? i'm wondering if it would be better to remove it from pin 7 and put it at the 6v6s pin 5-to ground, at say 470k. or is that ridiculous?
this second stage has a grid resistor (ground ref) that is not in the "floating paraphase" circuits that I've been reviewing - is this providing the input ground reference? is it also providing ground reference to the output tubes? i'm wondering if it would be better to remove it from pin 7 and put it at the 6v6s pin 5-to ground, at say 470k. or is that ridiculous?
It could be the coupling cap (.005) going to the grid of 12AU7, pin 2, has DC leakage, upsetting the bias of the phase inverter circuit. Worth checking. Grid to cathode difference is 3.7v, according to schematic, so it should have enough headroom.
Regarding the phase inverter circuit, (paraphase PI is my favorite sounding circuit, great tone) I prefer to run both sections in equal current so the cathode resistors of 3.3K are not really equal as the first tube's cathode is really parallel with the 10K feedback resistor which makes is really 2.48K. Either parallel another 10K resistor to ground on the second tube's cathode or just use one 2.5K resistor. In this case both plate voltage will be equal. Checking AC balance of outputs from both phases can also reduce distortion by adjusting the value of the 390K input resistor. Bear in mind I am only thinking in terms of audio amp but the mismatch might be responsible for the sound or "tone" in a guitar amp. Try it and see what happens.
You also try to add AC balance this way:
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
thanks DirectDriver!
(edited to add: i just saw your pic after writing this post)
I was wondering about possibly using a pot for that balance (390k/240k) - maybe 1meg with 240k resistors from either lug to the wiper (which then heads to pin 7 of 12au7) - my thinking on the resistors is primarily to ensure ~470k input resistance for the 6v6s.
i'll try swapping out the second stage cathode resistor as you recommended. i'm going through checking figures and finding some typos in this schematic and changing values to compensate (most obvious one being the 12ax7 plate voltages - they're flipped) so I'll also be reevaluating the components in the PI in general
what is your opinion on my previous post - should I "free" the input grid of the 2nd stage of the PI from ground reference ( remove 240k pin 7 to ground) and then give the 6v6s the needed ground reference by adding 470k (pin 5 to ground) for each?
(edited to add: i just saw your pic after writing this post)
I was wondering about possibly using a pot for that balance (390k/240k) - maybe 1meg with 240k resistors from either lug to the wiper (which then heads to pin 7 of 12au7) - my thinking on the resistors is primarily to ensure ~470k input resistance for the 6v6s.
i'll try swapping out the second stage cathode resistor as you recommended. i'm going through checking figures and finding some typos in this schematic and changing values to compensate (most obvious one being the 12ax7 plate voltages - they're flipped) so I'll also be reevaluating the components in the PI in general
what is your opinion on my previous post - should I "free" the input grid of the 2nd stage of the PI from ground reference ( remove 240k pin 7 to ground) and then give the 6v6s the needed ground reference by adding 470k (pin 5 to ground) for each?
Last edited:
Either way will work but having extra set of resistor is not necessary. If you want the PI to see 470K, then just increase the input resistor to ground to 470K and add a pot to get AC balance. The second half of the 12AU7 has to see a portion of the first half 12AU7's signal that's equal to its input voltage, basically both halves of 12AU7 need to see equal voltage at their grids to ensure AC balance. Let's say the mu of the first half of 12AU7 is 15 then the second half of 12AU7 needs to be feed 1/15 of amplified signal to be back to unity. Having a pot at the grid makes adjustment much easier and you can even purposely make it imbalance to get the distortion or lack of distortion you want. Adjust-ability comes in handy.
Stock circuit:
Since you like to have 470K input resistor at the grid of 6V6, I just tweaked the values to use commonly available resistors. Adjusting the 100K pot is important, most likely the wiper will be at around 30K to ground and try that as the starting point and adjust accordingly. Make sure to measure the AC balance of both outputs of PI.
An externally hosted image should be here but it was not working when we last tested it.
I'm not hooked on 470k for the 6v6s, but in the original schematic I saw the series resistance as over the 6v6 spec of .5M max grid resistor - the 240k to ground on pin 7 plus the 390k. I wonder if that was another typo in the original schematic. i can't recall if the component was there or not, and have since learned that its a good idea to verify/modify the schematic during deconstruction when looking to maintain an amps sound but update components.
does 100k on the PI plates seem a bit high to you? would the headroom decrease or increase if I went to say 47k? In preamp stages I've built, the higher the plate resistor the more "nasal" and "pinched" the sound seemed to become - well, particularly getting around 220k and higher on a 12ax7.
I have since realized that I do have a ton of gain here, even when quick swapping the 12ax7 with a 12at7. I guess my night audition must've been with the guitar volume at 5, because this amp starts to break up pretty quick beyond that! I may bring the 220k 12ax7 stage down to 100k. I may just put in another input and jump it to the volume pot (skipping the 12ay7) for variety instead.
thanks for your efforts. hopefully i'll get some soldering time today - refinishing cabinets for now.
Jeff
does 100k on the PI plates seem a bit high to you? would the headroom decrease or increase if I went to say 47k? In preamp stages I've built, the higher the plate resistor the more "nasal" and "pinched" the sound seemed to become - well, particularly getting around 220k and higher on a 12ax7.
I have since realized that I do have a ton of gain here, even when quick swapping the 12ax7 with a 12at7. I guess my night audition must've been with the guitar volume at 5, because this amp starts to break up pretty quick beyond that! I may bring the 220k 12ax7 stage down to 100k. I may just put in another input and jump it to the volume pot (skipping the 12ay7) for variety instead.
thanks for your efforts. hopefully i'll get some soldering time today - refinishing cabinets for now.
Jeff
It does seem rather high and only drawing less than 1.5mA of current for a 12AU7, as if it's designed with a 12AX7 or 12AY7 in mind. If you change to 47K plate resistor, you will need to reduce the value of the cathode resistor so plate voltage maintains the same. You're essentially changing the operation point of the tube. I don't think you need to mess with the original circuit too much. Just adding AC balance is a nice touch. The PI circuit does not have a lot of gain and further reduced with global feedback so the preamp section is very important to have the necessary gain. It's rather strange that it would clip so easily to begin with.
Anyway, I am not a guitarist or deal with guitar amp too much so I am not equipped to answer all your questions. Good luck with your project and, most important, have fun.
thanks DD! i don't have much room in this chassis for the balance pot (its a cramped thing) without possibly inducing noise as I'd have to cross from the back of the chassis to the faceplate - though I may try it with shielded wire.
i did drop the 220k to 100k on the 12ax7, and will see how that opens things up.
i did drop the 220k to 100k on the 12ax7, and will see how that opens things up.
Trim pots don't take up space at all. You can place it right next to the other parts, no need to mount on anything.
An externally hosted image should be here but it was not working when we last tested it.
oh yeah, right you are - i had the standard panel knob in mind, but I suppose a trim pot hidden away in there would work
one thing bothers me about the schematic you posted though - with the pot turned all the way toward ground, wouldn't the signal be nil? isn't that equivalent to putting the two resistors (470k & ~470k) to ground and pulling a lead from ground to pin 7 of the PI? I think the PI pin 7 needs a ground reference resistor - which is what the original 240k was doing. I'd be inclined to put the pot on the other side of the 390k resistor you suggest, so that it always sees at least 390k as ground reference. doesn't that make sense?
one thing bothers me about the schematic you posted though - with the pot turned all the way toward ground, wouldn't the signal be nil? isn't that equivalent to putting the two resistors (470k & ~470k) to ground and pulling a lead from ground to pin 7 of the PI? I think the PI pin 7 needs a ground reference resistor - which is what the original 240k was doing. I'd be inclined to put the pot on the other side of the 390k resistor you suggest, so that it always sees at least 390k as ground reference. doesn't that make sense?
First, you don't ever want to turn it towards ground. You can start at midpoint. The 390K resistor and the wiper to ground forms a voltage divider that sends a signal voltage that should be equivalent to unity or same as the input signal and that depends on the mu or gain factor of the tube. Pin 7 is seeing ground through the pot. Once you find the right value, you can replace the pot with a fixed value resistor if you want. But having a pot there is very convenient for tweaking. Besides, the pot is not for you turn all the time, just there to dial in the AC balance and then you leave it alone.
Again, one end of the pot is connected to ground and that's your ground reference already. There are many ways to do this and since the 6V6's are self biased, this is the easiest and simplest way to do this. If you place the pot on the other side of the 390K, you will not get the correct signal voltage to the grid, it will be too high no matter where the wiper is.
Here's a schematic of the famous Pilot 232 amp (best sounding 6BQ5 amp ever) and perhaps give you an idea of one way of balancing a paraphase circuit and in this case two fixed resistors forming a voltage divider to the grid:
An externally hosted image should be here but it was not working when we last tested it.
Here's a calculator for values of all the parts in a paraphase circuit. But it assumes the circuit uses no feedback and the cathodes are bypassed with big caps so the gain is higher than many circuits, including yours.
Paraphase Inverter Gain and Resistor Calculator
But you'll get the idea.
Paraphase Inverter Gain and Resistor Calculator
But you'll get the idea.
Last edited:
Right on! Thanks for your input and energy. Now that you say not to turn the wiper all the way to ground, it obviously makes more sense. I was considering the pot to cover a range of useful tweakyness, and how to achieve that -
i guess still caught up in the control panel knob mindset
but your inclusion of the schematic helps illustrate things
the one element i'm a bit confused is this - in the original schematic, the 390k and 260k (or whichever, can't remember exact figure) came together at pin 7 - instead of 1 leg (390k, in your suggestion) going to ground. isn't the signal to pin 7 a summation of the signals through these two resistors? it visually appears as though taking one of them off of pin 7 and grounding it would reduce the signal that appears at pin 7.
i'll review your calculation link as well. thanks for that!
i guess still caught up in the control panel knob mindset
but your inclusion of the schematic helps illustrate things
the one element i'm a bit confused is this - in the original schematic, the 390k and 260k (or whichever, can't remember exact figure) came together at pin 7 - instead of 1 leg (390k, in your suggestion) going to ground. isn't the signal to pin 7 a summation of the signals through these two resistors? it visually appears as though taking one of them off of pin 7 and grounding it would reduce the signal that appears at pin 7.
i'll review your calculation link as well. thanks for that!
The original circuit is using a floating paraphase. When the two dissimilar resistors form at one point to grid or pin-7, it also has another 240K resistor to ground. The idea is again to have a portion (unity) of the first tube's gain to the second tube. Regardless of how you do it, the bottom-line is to feed a tiny bit of the out of phase signal to the second tube, which really has no gain at all. Try to think of the paraphase circuit as TWO single ended circuits with opposite phases.
Don't think of it as summation of of two signals. Think of them as subtraction. When two identical level but out of phase signals join together they cancel each other out. That's why the 390K and 240K resistors are not the same value. Regardless of approach, at the end of the day you need a portion of the first tube's signal feeding into the second tube so second tube's gain is unity. Getting this out of phase signal in the right portion is what the pot is for. Some circuits use a fixed value through calculation or measurement but, to me, dialing the pot is the most accurate. The original circuit might not be balanced at all and that might contribute to its sound but having a pot there makes life easier and the best of both worlds as you can purposely make it imbalance to get a more distorted sound. Either way, you have some control.
Thanks again!
Much better now - turns out I had a pretty good imbalance there - almost twice as much signal on one 6v6 than on the other. I followed your advice and now they are very close to each other - it changes slightly with the volume knob position.
I do still have significant distortion with the guitar volume all the way up, and the amp volume above about 30% - I'd like to get a little more range. I also wonder if I'm not hitting the output section a little hard - i get about 60v signal at grid 1 when cranked, and it hits that point pretty fast when turning up the knob (about 30% of the way, as mentioned above). This amp may be a good candidate for a master volume. Or for turning one of the preamp stages into a cathode follower.
But, the correcting the imbalance made the distortion much more musical, which is a step in the right direction!
Thanks again for your help Direct Drive!
Much better now - turns out I had a pretty good imbalance there - almost twice as much signal on one 6v6 than on the other. I followed your advice and now they are very close to each other - it changes slightly with the volume knob position.
I do still have significant distortion with the guitar volume all the way up, and the amp volume above about 30% - I'd like to get a little more range. I also wonder if I'm not hitting the output section a little hard - i get about 60v signal at grid 1 when cranked, and it hits that point pretty fast when turning up the knob (about 30% of the way, as mentioned above). This amp may be a good candidate for a master volume. Or for turning one of the preamp stages into a cathode follower.
But, the correcting the imbalance made the distortion much more musical, which is a step in the right direction!
Thanks again for your help Direct Drive!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.