Ways to get more voltage to V1?

[12AX7]

My amp has a marshall 800 style PT from mag comp and the voltage at the PA plates is around 440. Amp has a 20H (don't laugh, i love it) choke and a 100/100uf main can then 20uf node at all 3 preamp tubes. with 10k dropping resistors at every node, V1 with 100k plate resistors gets around 190-200v at each side. I like more voltage at V1 and i was hoping theres a way to get it besides how i'm doing it now, which is to tap directly from the screen node thru a 10k to V1. It still doesn't get me as much as i'd like. Or THINK i would like....i really just want to experiment, but i do much prefer the 240v (if i recall correctly) i get tapping from the screen node. So i'd like to get more, plus i was wondering if theres a way thats more orthodox/correct than the way i'm doing it.

[TubeGeek]

What about experimenting with plate resistor values on v1?

[12AX7]

My amp has a marshall 800 style PT from mag comp and the voltage at the PA plates is around 440. Amp has a 20H (don't laugh, i love it) choke and a 100/100uf main can then 20uf node at all 3 preamp tubes. with 10k dropping resistors at every node, V1 with 100k plate resistors gets around 190-200v at each side. I like more voltage at V1 and i was hoping theres a way to get it besides how i'm doing it now, which is to tap directly from the screen node thru a 10k to V1. It still doesn't get me as much as i'd like. Or THINK i would like....i really just want to experiment, but i do much prefer the 240v (if i recall correctly) i get tapping from the screen node. So i'd like to get more, plus i was wondering if theres a way thats more orthodox/correct than the way i'm doing it.

that changes the tone and gain. I like the 100k there. I spent a long time coming to the conclusion i really like the fisrt 2 gain stages with 1.5k/100k. Adding voltage as i did changes the punch factor but the tone stays the same. So i just want more voltage

[TubeGeek]

I see.  The only other way that i see is to change the value of the 10k resistor feeding the plates but sounds like you're already trying that.

[12AX7]

More or less, actually parallel node vs changing the values. But in effect yes. I guess it's the only way but i had to ask since others here might know of some tricks to do it.

[TubeGeek]

There are smart guys on here...maybe they'll chime in

[HotBluePlates]

Make the 10k smaller, either the original circuit resistor or your new 10k.

The current of all stages connected to the low-voltage side of the 10k resistor is flowing though the resistor, and that creates a voltage drop (voltage drop = current * 10kΩ). If you want less voltage drop, you attack this equation by making the 10kΩ smaller. A byproduct is that ripple will be somewhat higher (may not be a problem in practice) and there will be less isolation of that filter cap stage from the previous one.

I would suggest not going below 1kΩ when lowering the 10kΩ; higher R is better when it comes to the two byproducts I mentioned.

If the voltage on the higher-voltage side (not connected to the plate loads of the preamp stage you're looking at) of the 10kΩ is not a high enough voltage for your needs, then you have to do exactly as you've done: connect that 10k dropping resistor to a higher-voltage point in the power supply. And then of course adjust the 10kΩ value up/down to land on your desired supply voltage for the stages in question.

[12AX7]

Try tapping into the B rail upstream of one of the 10K resistors.  You'll gain voltage, but lose ripple filtering.  (i.e., there will be more noise).  Once you find your voltage add the necessary filters, less the resistors.  with a little luck, you can find a can with the caps

WARNING WILL ROBINSON, YOU ARE DEALING WITH HIGH VOLTAGES THAT CAN AND HAVE BEEN LETHAL

And what HBP said

Thats what i was saying i already did. I was just looking for an alternative method. By HPB basically told me to do it as i already did, so i guess there is not other way.  I'll drop the 10k too to get max V.  i have some 2.2k 3 watters, so i'll try one of those. And yes, i know bout the danger. I've been working with the power rail a lot in this amp especially for years and i learned even long before that how dangerous it is. Haven't thrown a tool thru the wall for decades.

[sluckey]

You should also put a filter cap on the downstream side of that 10K dropping resistor.

[12AX7]

You should also put a filter cap on the downstream side of that 10K dropping resistor.

It does. It was same PSU as a 800 but i simply took the 10k and connected it to the screen supply instead of the preamp node next to it. So yes, it has a filter cap.

[tubeswell]

To get any noticeable change in voltages from changing the dropping resistors, you have to make the changes early on in the power rail and make them 'drastic', i.e. use a choke between the plate and screen nodes, and at least halve the value of supply resistors. Or do parallel supply nodes

[jjasilli]

Use Ohm's law.  Measure the voltage drop across the 10K dropping resistor to determine the current draw to V1.  Then use Ohm's Law to figure the value of the dropping resistor you need.  You can check your work by temporarily placing another resistor of proper value in parallel with the 10K resistor.

If you want the supply voltage more constant (better regulated) to V1, you can pace a 1N4007 diode in series at the beginning of the parallel B+ supply.  This will inhibit variations in the power tubes' current draw from sucking B+ voltage from the V1 supply.

To reiterate prior posts:  most of the hiss, hum & noise in an amp will originate from the first preamp stage, because any noise there keeps getting re-amplified by subsequent stages.  So, the first gain stage is at the end of the filter stages.  That way it gets the most filtering of the B+ supply.

Presumably you are happy with the noise level in your amp.  If not, you could build a parallel B+ supply from anywhere on the B+ rail to the first gain stage -- with as many filter stages as you want in the new B+ supply line.  

[jojokeo]

I don't see why you can't adjust both the plate & cathode r's to attain the same load line response using a higher voltage if necessary? Changing voltage using the same r's as mentioned before DOES change the load line an operating conditions anyway...

[jojokeo]

The gain & frequency response stays the same but the load lines do not. I used a few arbitrary numbers not knowing exact circuit values but these illustrate my point.

[alerich]

Why is your voltage so low throughout the power rail? I built a 50 watt JCM 800 with Magnetic Components transformers and my voltages pretty much match the ones here:

http://mhuss.com/MyJCM/JCM800_2204.gif

440VDC is low for the plates. Should be about 470VDC. A normal JCM has about 245VDC at V1 pin 6 with the standard power rail but you have to tap off the screen node to get that. Something does not sound right.

[ac427v]

I know you love your 20 H choke but the added resistance is probably dropping the voltage of everything downstream from it. That adds compression (love that sound!) but reduces punch. How do your plate voltages look on the PI and other preamp tubes? If those tubes are seeing a lower voltage they will lose punch just as much as V1. Changing the first 10K dropping resistor to a 5K might compensate. That would bump up everything including your V1.
Craig

[HotBluePlates]

I know you love your 20 H choke but the added resistance ...

The extra resistance of a 20H choke as opposed to a 4-10H choke is probably negligible in this case.

Let's compare 2 Hammond chokes (because they have published DCR): 4H 194A & the 40H 194C.

On average, a guitar amp gain stage will draw 1mA or less, and let's say we have 3 12AX7's for the preamp and phase inverter; that might be 6mA of total current draw for the preamp. Screen current draw varies, but is generally only 5-10mA per output tube for many amps. Total current through the choke for a 50w amp will then be 6mA + 20mA = 26mA.

The DCR of the 4H 194A is 167Ω, while the 40H 194C is 420.2Ω; that's 10x the inductance but only ~2.5x the DCR.

0.026A * 167Ω = 4.34v
0.026A * 420.2Ω = 10.9v

The voltage drop, whether 4 v or 10v, is negligible. Which is, after all, why a choke is generally used instead of a resistor in this part of the power supply... because by comparison it has a voltage drop you can ignore.

By comparison, if you pull 6mA through a 10kΩ resistor, there will be 60v dropped across the resistor.

[alerich]

Why is your voltage so low throughout the power rail? I built a 50 watt JCM 800 with Magnetic Components transformers and my voltages pretty much match the ones here:

http://mhuss.com/MyJCM/JCM800_2204.gif

440VDC is low for the plates. Should be about 470VDC. A normal JCM has about 245VDC at V1 pin 6 with the standard power rail but you have to tap off the screen node to get that. Something does not sound right.

Looking at the Triode site I see they stock two JCM 800 PT models. The laydown version 40-18089 is 625 VAC CT while the standup 40-18023 (the one I used) is 690 VAC CT. That could account for the difference if you are using the laydown PT. If that's the case you're probably just stuck with what you're stuck with. The power rail is working as it is designed. If you want more supply voltage you'll need a hotter PT.

[12AX7]

Oops, sorry. I meant to say it's a Hammond PT. I HAVE and have used the mag comp which does indeed have more voltage, but as i recall the V1 voltages were still not as high as i seem to see in a lot of schematics. I also have a 4 or 5 (i think 4) henry mag comp choke but i also don't recall that making any notable difference. I stopped using the MC PT because it was 150ma while the hammond i first had in this amp was 250, so i went back to it just so i could run all sorts of power tubes safely. Now that i'm certain it's 6L6 for me for now on, maybe i'll put the MC back in.

[TIMBO]

Hi guys, I didn't read right through the post so shoot me if this has already been suggested. Thanks to Sluckey this PSU is a treat for those builds that you need the experiment with voltages.

http://www.el34world.com/Forum/index.php?action=dlattach;topic=15795.0;attach=38187;image

[12AX7]

yeah, thats parallel rail vs series, and basically thats what i did, tho only for V1. I just thought there might be a way to get more voltage because with this method i'm still not getting a lot.

[12AX7]

heres a question....someone mentioned less filtering the way i'm doing V1 now, IE: 10k resistor right from the screen supply after the choke to V1 having less filtering because it doesn't have the effect of the filters in the stages after it like the PI and V2. Do those filters actually add up in effect to v1? In other words,  lets say you have if you have 3 12ax7 stages. And for the sake of argument lets take the PA filtering and choke and all that out of the equasion. It's a series PSU just like typical and each stage has say 20uf. Does this mean V1 has the effect of 60UF? I would think not due to the droppers. So is it cumulative like that but with an equasion that would tell you how much filtering it will have?

What i'm getting at here is i'm trying to understand how this works so i can determine if i should put more filtering at V1. I have 60uf there now, but i want to determine if in effect that amount of filtering would need to be altered up or down in order to give V1 the same filtering it had when the rail was all series and all 3 ax7 stages had 20 uf.

[jojokeo]

heres a question....someone mentioned less filtering the way i'm doing V1 now, IE: 10k resistor right from the screen supply after the choke to V1 having less filtering because it doesn't have the effect of the filters in the stages after it like the PI and V2. Do those filters actually add up in effect to v1? In other words,  lets say you have if you have 3 12ax7 stages. And for the sake of argument lets take the PA filtering and choke and all that out of the equasion. It's a series PSU just like typical and each stage has say 20uf. Does this mean V1 has the effect of 60UF?

No, if all of the caps all had a series resistor before the next, etc. V1 ultimately would only have 20uF, they don't add up cumulatively. That's what ISOLATES each stage/node from one another.

BUT - after each filter node, your filtering effectively gets better and better after each node and why the "series of nodes" is a good way to provide the best filtering (aka least amount of ripple) to your most sensitive preamp tube or stage when the signal gets amplified the greatest - at V1. Any ripple or hum or interference at V1 gets amplified right along with the signal and then through each successive stage after that furthering along the menacing noise that it is. You want V1 to ideally have THE best filtering possible for this reason.

This is the easiest & least technical way to explain it I think???

[HotBluePlates]

... someone mentioned less filtering the way i'm doing V1 now ... Do those filters actually add up in effect to v1? ...

Yes, but not like 20uF + 20uF + 20uF.

Power supply ripple is 120Hz. Let's guess 15v of 120Hz ripple at the 1st filter cap at the rectifier. You say you have a 20H choke and (I assume) a 100uF can following the choke.

Inductive reactance of the 20H choke @ 120Hz = Xl = 2 * pi * 120Hz * 20H = 15,079Ω
Capacitive reactance of the 100uF cap @ 120Hz = Xc = 1 / (2*pi*120Hz*0.0001F) = 13.26Ω

These form a voltage divider for a.c. only of 13.26Ω / (13.26Ω+15079Ω) = 0.000879:1  --> 15vac * 0.000879 = ~13.18mV of ripple at the screen node. Yes, the big choke value knocks ripple down a lot; it would have been 65.7mV of ripple with a 4H choke.

Next set of components is a 10kΩ resistor and a 20uF cap. 13.18mV of ripple becomes:

Capacitive reactance of the 20uF cap @ 120Hz = Xc = 1 / (2*pi*120Hz*0.00002F) = 66.31Ω
10kΩ resistor & 20uF cap form a voltage divider for a.c. only of 66.31Ω / (66.31Ω+10000Ω) = 0.00659:1  --> 13.18mV * 0.00659:1 = ~86.86uV ripple

And on and on. Each filter stage provides successive voltage division of the a.c. ripple voltage riding on the clean d.c. With the typical Fender series arrangement of the power supply, the input stage has very clean d.c. But if you make the cap bigger or the series resistor bigger (or raise the Henries of the series choke), each of these knocks down the ripple just from different directions.

[12AX7]

This is the easiest & least technical way to explain it I think???

yes, actually that made sense to me. So then amps like the matchless chieftain that use parallel nodes are sacrificing cleaner power for MORE power at each stage. For me, being the electronic moron i am, i look at filter caps as more of a tone thing, or more specifically, a feel thing. Doesn't seem like more filtering helps with noise, but it does radically change the feel. But i'm really not sure exactly what ripple causes other than noise.  Is that it?  or if not, what does having too much ripple do to the tone?

I tried a few things tonite out of curiosity, things that were mentioned, one at least i've tried before but thought i'd have another go. First i tried reducing the resistor from the screen supply to 2.2k from 10k.  Then i tried taking the tap from before the choke at the beginning of the power rail. Then i tried more filtering. Even tried the 4H choke again. All at different times tho, never 2 things at the same time. None sounded/felt near as good so i think i already have it pretty much nailed. But the voltage at V2 didn't move much so i still don't know the effect that would have. It's high enough, but i was just very curious. My problem is not that the amp doesn't sound good as it....it's actually so good i really never feel the need to make it  better. But i have this extreme curiosity about amp design and what might cause what. I get ideas in my head and even if i'm totally happy with it i gotta try this new idea.  Crazy, huh?

[HotBluePlates]

... So then amps like the matchless chieftain that use parallel nodes are sacrificing cleaner power for MORE power at each stage. ...

Sort of, but not really.

Most boutique amp makers of the 90's copied tweed Fenders, with a few copying Marshall (who themselves originally copied the tweed Fender Bassman). Matchless was different, and copied Vox.

Vox amps were largely based around EL84's (yes, with the exception of the AC50 and AC100, but how many people do you know in the U.S. that have even seen one of those in person?). Because there was no good reason to have a B+ in excess of about 320vdc with EL84's when using cathode bias (the most economical and safest choice for operating the tubes), you don't have a lot of voltage to throw away when feeding the preamp stages.

Vox did not use series power supply architecture because you get successive d.c. voltage division, along with the a.c. voltage division I described earlier, and the use of 150k-220kΩ plate loads would leave very little plate voltage left on the preamp tubes. The high plate loads were generally chosen to maximize voltage gain in the preamp, because the designer is also capitalizing on the relatively-little drive needed by the EL84's to have the fewest number of tube stages to create the amp (as well as leave a few tubes for special effects and extra channels).

So it is about keeping "more power" at each stage, but only because the amp started off with 320vdc or less in the output section, not 420-450vdc like comparable Fender amps.

... But i'm really not sure exactly what ripple causes other than noise.  Is that it?  or if not, what does having too much ripple do to the tone? ...

1. It could cause hum, because it is a hum voltage.

But to do that, it would have to sneak down from the B+ node through the plate load resistor and out through the coupling cap toward the following tube stage. But triodes have relatively low internal plate resistance, which forms a voltage divider along with the plate load resistor to noise from B+. Therefore, triode gain stages are generally pretty good at rejecting noise/hum from B+, with lower-gain tubes being the best (mainly because their internal plate resistance is generally much lower).

Pentodes are relatively poor at rejecting B+ noise because their internal plate resistance is generally many times larger than any practical plate load resistance, so there is little division of noise voltage from the B+ node. This is a double-whammy for Vox-style amps, as they often use a pentode input gain stage and might also have parallel supply paths with more B+ ripple as a result. The saving grace for vintage Vox designs is there is often no amplification after a pentode gain stage, except what is offered by the phase inverter so hum may be relatively mild.

Of course, if you graft-in a pentode onto your amp with several gain stages afterward and don't account for noise in the power supply or how the pentode reacts with that noise (or how the noise may be amplified by successive stages), well...

[HotBluePlates]

2. You've heard of the famous Vox "sub-harmonic"?

Well, have you? Sometimes this gets blamed on speaker issues, but ripple voltage due to an under-filtered amp trying to keep B+ voltages high while starting with too-little voltage... well, it shows up as a hum-modulation of the output signal, generally when the amp is distorting or making maximum output power.

Along with sagging voltage, heavy current draw when the output section is cranked also increases the size of the ripple voltage. The fist filter cap still charges to the peak of the incoming voltage, but the cap is having current drained by the output stage faster than it can be replaced and the trough of each ripple cycle dips ever-lower, which you observe as a reduced average B+ voltage (i.e., sag).

The higher ripple means more hum potentially appearing at the plate of every stage, and in the case of the Vox AC-15, more ripple at the screens of the output tubes. That cause output tube plate current to bounce up & down at 120Hz, and 120Hz beats against the notes you're playing. Combine that with a stage that is distorting, and you have the chance to have that 120Hz heterodyne with the notes you're playing to create strange subharmonic notes.

Part of the problem in the case of the AC-15 is it's given a good amount of filtering, but the output tube plates and screen are sharing the supply supply node with almost no decoupling of the screens.

... But i have this extreme curiosity about amp design and what might cause what. ...

Amps are frustrating in that regard. Once you know what's going on, you see that if you change 1 thing, 5 other things change in response. A guitar amp is a multi-way balancing act; few things are cut & dried.

[12AX7]

Thanks. Some of that went whoosh, some  i got and makes good sense.

Amps are frustrating in that regard. Once you know what's going on, you see that if you change 1 thing, 5 other things change in response. A guitar amp is a multi-way balancing act; few things are cut & dried.

Man, do I know that ! over time i have seen so many tweaks that were great improvements become just the opposite when something else is changed. long ago i came to the conclusion that the secret to a great amp is perfect balance of all areas the circuit. And the criteria of what that means is hard to explain and a lot harder to achieve ! But i feel i've come close because anymore it seems like i can't change anything w/o hurting the tone. That tells me i have pretty good balance.

[jjasilli]

. . So then amps like the matchless chieftain that use parallel nodes are sacrificing cleaner power for MORE power at each stage. For me, being the electronic moron i am, i look at filter caps as more of a tone thing, or more specifically, a feel thing. Doesn't seem like more filtering helps with noise, but it does radically change the feel. But i'm really not sure exactly what ripple causes other than noise.  Is that it?  or if not, what does having too much ripple do to the tone?  . . . i have this extreme curiosity about amp design and what might cause what. I get ideas in my head and even if i'm totally happy with it i gotta try this new idea.  Crazy, huh?

MORE power at each stage:  Not really.  Changes of 60V or so on the plates have negligible effects on gain.  Pragmatically, even a very low mu tube would have a nominal gain factor of 20:  1VAC of signal in becomes 20VAC of signal out.  This is more than enough to swamp the next stage, so attenuation is necessary.  Essentially the power output of a preamp tube is not crucial.

Ripple:  Too much ripple (not enough filtering) = audible hum.  Excessive filtering lacks sparkle & liveliness; sound SS. (I don't know why.)

curiosity about amp design:  Loadlines; loadlines; loadlines.  Changing B+ supply voltage changes the operating point of the tube and has a major effect on tone, other than gain itself.  The operating point of the tube is affected by supply voltage, plate resistor, and cathode resistor.  E.g., the purpose of a hi-gain preamp stage is not for the gain, but rather to convert the input signal into a square wave to get a "metal" or "shred" tone.

[jojokeo]

curiosity about amp design:  Loadlines; loadlines; loadlines.  Changing B+ supply voltage changes the operating point of the tube and has a major effect on tone, other than gain itself.  The operating point of the tube is affected by supply voltage, plate resistor, and cathode resistor.  E.g., the purpose of a hi-gain preamp stage is not for the gain, but rather to convert the input signal into a square wave to get a "metal" or "shred" tone.

Load lines are VERY important as mentioned earlier in this thread and is what's behind what 12AX7 is experiencing when changing the voltage applied to the tube(s). Real tonal control comes from the choice of topology, frequency shaping and manipulation of overdrive characteristics, and from a complete understanding of the circuit’s functionality.
Understanding a tube's static anode characteristics and transfer characteristics (also known as dynamic & mutual characteristics) enables one to understand the linearity of the tube of choice that's going to be utilized in a circuit to accomplish the tonal goal for that particular design or amp.
This is where the tube's operating point comes into play in setting the load line where it is desireable in creating the characteristics that the designer wants to achieve. We can control the voltage swings, currents, determine where the signal gets cut-off or goes into grid current limiting, allowing for greatest headroom & input sensitivity or is bunched below the line that creates "squashing" (compression), which in turn creates the 2nd order harmonics that we all know and love, etc...there's a whole lot more on the topic to understand fully.
It's not quite a simple and easy as just "converting into a square wave to get metal or shred tone" as there's much more that goes into a design that you might realize?

[Ed_Chambley]

What is it you are wanting to achieve from increasing voltage?  I read through the posts and really did not get a good idea of what you want.

Reason I ask is I have done something similar to a Marshall build, but it also involved a change in the Output Transformer.  I was looking to get a cleaner, tighter response.  There are many ways to achieve similar results of a specific tonal quality you are after.