dumb question on maximizing the sound of12A_7 tubes

[Tone Junkie]

Ok I have 50 watt plexi I built it with an extra gain stage . i have been trying differant tubes in the extra gain stage I know how to maximize a 12ax7 what I need to know is the peramiters on the 12a_7 types. Ive been reading on other guitar forums trying to get the info without asking but. I just cant find enough. on one fender forum. they said to use a 12at7 I should set my plate resister to 220k but I know on a 12ax7 that is my max for ultimate distiortion without going way over the top. I want the sweet spot on my 12at7 not max distortion for that tube , then for 12au7 on a differant site they said to run your plate rester at 27k and your cathose rester at 1.5k to bias the tube properly Im confused. What Im looking for and trying to understand is the upper and lower ranges of bias and plate resters for guitar amplifiers . 12at7 ,12ay7, 12av7 and 12au7 and I just cant seem to find that info no matter how much I read and whats for good hifi guys isnt the same for guitar guys. please help. Thanks Bill

PS. for my 12at7 that of a plate resister did sound good but when I plugged in a 12au7 to try it it sounded like shit probably way to high.   

[tubesornothing]

There are a few more parameters you must consider.  What is the B+ voltage?  How much will it overdrive the following gain stage?  What size of signal is coming in? Where is the extra gain stage in the amp? Are you wanting cut-off or saturation type of distortion?

I expect why you didn't get a hard and fast answer is that there is no hard and fast answer.  To begin with, your type of tone is most likely different than others.  Add to that the amp you are using, and where you are putting it in the amp, and there just becomes too many variables.

So what do you do?  Buy a couple of decade boxes and experiment.  Decide what you like and works for your amp.

[Dave]

You could always install a 250k pot as the plate load resistor and a 10k pot as the cathode resistor and adjust on the fly.

Dave

[Tone Junkie]

Thanks guys both are really great idea,s But in my reading I found what my problem was and it was easier than I thought . In reality I was driving my next gain stage to hard and it wasnt a tube gain problem or having to us a differant tube.
what I have is the marshall 1987 circuit with an added gain stage.
and on the return the they have a 470k to ground by changing that to 220k a 12ax7 works fine and I can dial from good to insane without blocking distortion,  which to me ear sounded like bad or to much distortion from that tube.
 Man it sucks to be a newbie. But the lesson I just learned will help me tailer a couple screamin amps in the near future. Thanks Bill

[jjasilli]

Forums are terrific -- but IMHO mostly for answers to specific questions, after you have consulted expert reference books.  There are chapters on this topic, and no forum can take the place of that.  That's the really best place to look.  You can search this forum for recommended reference material.

[Tone Junkie]

JJasilli You are so right Im a top down thinker i think if I can find the peramiters on both sides of something i have it licked, the problem with this question was the peramiters change with voltage to the plates and probably 3 or 4 other things that i dont understand at all.
 Luckily the reason I asked the question got solved the right way to begin with so I didnt have to try and go down that rabbit hole with the very limited knowledge that i have. I have a few more books on there way. Ive read so much in the last year at least I know enough to ask a real dumb question.
 Its kind of like music the more I learn the more I know I have a lot to learn . Thanks for putting up with me gracefully.  Bill

[RicharD]

The 12AX7 and 12AU7 are very different animals.  There's a lot more to tubes than just gain.  What might be optimal for an Ax probably won't be optimal for an AU.  Let's do a little comparison.  I'll use your given values of plate resistor = 220k, load resistor = 220k, assume the cathode resistor = 2k2 w/ bypass cap, and simply pick a B+ of 350V.  Look at the first attachment.  We're getting a big gain of almost 64, current is .72mA, input threshold is 2V, Vtube is 190V, output is pretty dang  symmetrical at 160V, and the operating point is sitting right smack dab in the middle of the load line.    Notice that the load line is pretty much perpendicular to the tube curves. That's a very optimal condition.  Now let's slap in a 12AU7 using the same values.  Look at the 2nd attachment.  Now gain is only 16ish, current has increased to 1.25mA, input threshold is way up at 16V, Vtube is way down at 72V, output is not very symmetrical and it can exceed Vtube at 120V, and the load line is way down in the curvy part.  Far from optimal. 

So what happened?  The problem arises from the differences between the 2 tubes internal plate resistance. The AX has an Rp of 62500 and the AU has a RP of only 7700.  Counting on thumbs, that's a 10x difference.  So let's make the AU happy.  Look at the 3rd attachment.  I selected 15k at the plate resistor which is roughly 2x Rp.  Once you have B+ and a plate resistor selected, you can plot a load line.  Yummy, nice and perpendicular to the curves.  Now you basically toss a dart right in the middle to calculate your cathode resistor.  7V/8mA=875 ohms, then tweak it up or down to the nearest standard value.  So we lost a little gain 11, but oh lordy now we're drawing 8.7mA and Vtube is over 200V.  Input threshold is at 11V and output swing is almost 200V. 

So you can see these 2 tubes are quite different by design.  The AX is high gain & high impedance.  The AU is low gain and lower impedance.  You can smack an AU much harder and drive a much lower impedance.  The trade off is gain and current consumption.  This is why you don't often see AU's in guitar amps.

[Tone Junkie]

You see thats what Im talking about you just taught me enough that if I wanted to try it out I could at least change enough things around to plug the tube and see it. And you gave me enough information to then studie and learn more in other words a starting point all I new is that you just cant plug that tube in aand have it sound good like you can with a 12at7. Thank Bill

[phsyconoodler]

You shouldn't be using a 12AT7 anyway,because it is quite a different tube than a 12AX7.They are usually used for reverb driver or phase inverter positions.The bias R/C is different as well,so stick with tubes like 12AX7;5751,12AY7.

  There are lots of ways to make them work for you.You learned one.Don't try and re-invent the wheel here.Look at successful guitar amps with multiple gain stages and start there.

By the way,it's spelled parameter.

[RicharD]

I have found this forum to be the most polite and informative group on the web regardless of topic.  I got better information here concerning a Hammond organ repair than I received form an actual Hammond organ forum.  When I joined, I barely knew the difference between a plate and a cathode. These guys put up with a lot of clueless questions from me.  I'm glad to be able to give back to the forum although honestly I still miss the boat all the time.


I'd have to agree with everything phsyconoodler said.  Plagiarizing the classics is pretty much the best way to go.  There's a million and 1 boutique amp builders copying the 5F6a and sticking their own name on it.  That's fine and dandy but doesn't really teach you much in the way of design.  12 AX, AY, & BZ (for you gain nuts) 7's are the best tubes for preamps.  If you need less gain, there are better ways to achieve this than by selecting a different tube.  Simply not capacitor bypassing the cathode will cut gain by roughly half.  Playing with different cathode bypass caps has a dramatic effect on tone.  Fender used a 25uF which gives you almost full bass response.  That's maybe not what you want for guitar.  Marshall used a .68uF cap which shaves off a ton of low frequencies.  I like to use a 2.2uF most of the time.  Of course another way to decrease gain is to put in a pot or a fixed resistive voltage divider.  This in no way harms tone.

I will say that the 12AT7 is a very unique and IMO under-rated vacuum tube.  It has a fairly high amplification factor but a fairly low internal impedance.  For the sake of argument, it's somewhere in between the AX and the AU.  Take a look at the tube data sheet.
http://www.mif.pg.gda.pl/homepages/frank/sheets/093/1/12AT7.pdf
Under the typical operating characteristics, notice that the approximate plate resistance is closer to the AU7(use the 250V charts).  This right here tells you that simply slapping an AT in place of an AX probably isn't going to make for an optimal condition.  Notice that when voltage decreases, plate resistance increases.  Alright... let's design w/o software this time.

Let's stick with a B+ of 350V.  2x plate resistance... call it 22k (standard resistor value).  Now divide 350V/22k = 16ma.  Now we can plot a load line (red).  Now let's pick an operating.  We'll toss our dart right at -2V bias (blue).  Draw a horizontal line to see what the bias current is which = 6.5mA  Now we can calculate our cathode resistor.  2V/6.5mA = 307 ohms  Not a standard value but 280 and 330 are.  Either one will be fine, it'll simply change your bias voltage a negligible amount.  Now draw a line straight down from the operating point.  This is Vtube, call it 207V.  Now lets guestimate gain.  Place a mark on the load line + & - 1 volt above & below the operating point (green).  Draw vertical lines.  You get 242V & 165V which is a difference of 77V, so for 2V input, we get 77V out, therefore AV= 77/2 = 38.5  That's absolute maximum gain with no load.  Seems silly to build an amp w/o a load, so your gain will decrease as load increases.  Bank on actual gain being 10 to 15% less, and 40 to 50% less if no cathode bypass cap is used.  Now to figure maximum swing out, you draw a vertical line straight down from 0 bias and make a mark at B+ (purple).  The swing is measured from Vtube which = -87V to +143V  What else do we need to know?  Input threshold is 2x Vbias = 4V.  Zout is approximately Rp in parallel with the plate resistor & in parallel with the load resistor.  Call it 8k-ish.  Ideally the load resistor should be >5x Rp so something larger than a 100K pot in this case.

This is not an exacting science, but it gets you very very close.  It is hard to find 2 tubes that are EXACTLY the same, but this does work.  It's fun and pretty easy with a little practice.

-Richard

[Tone Junkie]

See now once again both of you have taught me something i didnt not understand  first I didnt know that a 12ay7 is closer to the 12ax7  than a 12at7 of course thats the one tube I dont have but I will fix that soon enough. sorry about the bad spelling my fingers work faster than my brain sucks gettin old and my spell checker is broke.
And Richard I thank for showing how to through around a load line plote for that tube i love math. Bill

[RicharD]

No problem.  Some folks play golf, I play with tubes.  Bookmark this site:
http://www.mif.pg.gda.pl/homepages/frank/vs.html
It has pretty much every tube data sheet there is.  Certainly all the tubes you'll ever need for amplification.  Another really fun tube to play with is the 12BZ7.  Like the 12AX7, it has a AV of 100 but double (oops!  I meant) half the approximate plate resistance.  It's quite macho and I'm surprised it never caught on.

corrected

[tubesornothing]

If you check out Merlin's site (http://www.freewebs.com/valvewizard/) he has a downloadable spreadsheet if which you can plug in various numbers and see the graphs.  For common preamp tubes.

[SoundmasterG]

Buttery,

You remarked about not seeing the 12AU7 often in guitar amps.....about the only ones that use a few that I know of are the Vox AC100 and AC50. The AC100 stock is a very loud, clean amp, and the AC50 has a little more grit but is still loud and clean. They both do have a very sweet and harmonically rich sound though. I wonder how much of that is due to the 12AU7?

The 12AV7 seems like a combination of the AU and AY to a certain extent. A gain of roughly 40 yet a current pig like the AU.

Greg

P.S. Didn't you mean half the plate resistance for the 12BZ7 as compared to the 12AX7? That would make it well suited to drive a tone stack or as a phase inverter I should think?

[RicharD]

DOH!!!!  Yes, 1/2 the plate resistance.

I think those Vox amps are the only ones I've seen with an AU in the V1 position.  I've seen them in old PA amps all the time.  IIRC, Sunn used an AU as the PI in a few amps which it's well suited for.  So far in my examples, I've been shooting for "optimal", where optimal means linear performance.  Once again, this may not be what you want in a geetar amp.  Thermionics is like Leggos.  You can stick them together all different ways, but it might not hold it's shape.

[Tone Junkie]

So a 12bz7 what would change in say a standard ab763 for instance to use the bz7 instead of the ax7 what value would need to change and where . Sorry Im tired and thinking oddly, except for the fact that a whole lot of nos tubes of that type go down the road real cheap and I love the sound of great tubes that come cheap. Bill

[HotBluePlates]

So a 12bz7 what would change in say a standard ab763 for instance to use the bz7 instead of the ax7 what value would need to change and where.

For what you want to do, you would change nothing and simply pop the 12BZ7 in the socket.

There are 3 main tube parameters, and for triodes the guys have already pointed out the most important ones for preamp work (amplification factor and internal plate resistance). 12BZ7 has the same amp. factor (100) and half the internal resistance of a 12AX7.

All triodes actually deliver less gain than their amp. factor when they work into a load of less than infinity, and they almost always do that. But the bigger the plate load resistor and following stage's input impedance, the triode's exhibited gain gets closer and closer to its amp. factor.

A 12AX7 stage will have values selected to allow it to operate well. If you pop in the 12BZ7 with its lower internal resistance, the resistors already in the circuit will divide down the signal less than they did with a 12AX7.

That's a lot of "how/why" and not the "what" that you asked for, but when you truly get a handle on that concept you will answer your own questions on this subject in the future. I'd suggest a web search for "how a volume pot works" and "voltage divider" (you want a simple explanation using 2 resistors in the voltage divider).

So on to "what": the result will be similar to taking a 12AX7 stage and swapping a 220k plate resistor in place of an existing 100k plate resistor. Or you could say the output impedance of the 12BZ7 will be lower, which drives low impedance loads like tone stacks easier and with less signal loss.

Putting numbers to the words:
Assume a 12AX7 with a 100k plate load resistor and a following stage with a 1M input impedance. The total load looks to the 12AX7 like 100k in parallel with 1M, or 90.9k. Assume the 12AX7 has an internal plate resistance of 65k and the 12BZ7 has an internal plate resistance of 32k (the actual number changes with tube current and operating point, but these are representative numbers). Use the equation for a voltage divider to see how much of the total amplification factor actually gets used as apparent gain.

100 * [90.9k / (90.9k + 65k)] = 58.3 for the 12AX7
100 * [90.9k / (90.9k + 32k)] = 73.9 for the 12BZ7

You could repeat the calculations for the 12AX7 with a 220k load resistor in parallel with a 1M input resistance for the following stage, and prove for yourself that the resulting gain of the 12BZ7 is similar to using twice as big a plate load resistor on a 12AX7.

[Tone Junkie]

Hot blue plates thank you You explained that in a way that even I got it and sometimes with this stuff that is not an easy thing to do.  Bill

[Tone Junkie]

So here is the question in my ever increasing search for tone. Is the 12bz7 usefull in the circuit in other words does it sound good. could I use half the plate resistance and not damage the circuit. Like 47k.
I did order one last night to try out there not to bad price wise . But I can get a premium NOS rca 12ax7 for 30 bucks so Im going to save my real krinkle to by all those I can before there gone. Bill

[tubeswell]

FWIW here's an 'article' from OTM on rp and mu of different pre-amp tubes. The charts at the end provide a helpful comparative context on different tube capabilities

[RicharD]

>Is the 12bz7 usefull in the circuit in other words does it sound good.

Which circuit?  "Sounding good" is subjective.  I like the sound of the BZ.  Some folks love the sound of Silvertones, I personally can't get them off my bench fast enough.

You can slap a BZ in place of an AX w/o fear.  There will be no earth shattering kaboom.  It might sound great, it might not.  It might squeal.... chances are it'll sound a little darker and crunchier.

A more drastic example was the use of a 12AU7 in the V1 position of the Vox AC-50.  I got curious so I graphed it.  The tube is sitting in a very Fenderish 12AX7 configuration: 270V, 100k/1k5.  The tube is not biased even close to optimal.  It's actually in a starved condition.  Vtube is very low (80V) and the output can exceed the voltage across the tube given enough input.   Output looks to be very linear up to about a 3 or 4V input which is more than a guitar will deliver.  This configuration might be frowned upon in a hifi amp but might be a big part of the AC-100 sound.  Slapping a 12AX7 in place of V1 probably wakes it right up.

[HotBluePlates]

Is the 12bz7 usefull in the circuit in other words does it sound good. could I use half the plate resistance and not damage the circuit. Like 47k.


I've never had a 12BZ7, so I can't report on tone. I know there is a guy that is on here at times who goes by "12BZ7" so I'd imagine he likes them. 

I personally wouldn't use them in a 12AX7 circuit and halve the plate resistance, because it seems like you'd wind up with performance that could be gotten with a 12AX7. Of course, the output impedance of that circuit would then be lower, so maybe the end performance is not exactly the same. I'd either use the original plate resistance for the gain boost, or use a different tube or circuit for reduced output impedance.

But that's just me. Maybe I should really try one before making any definite statements about their best use.

[SoundmasterG]

Regarding the Vox AC50 and AC100.....the 50 uses just the 12AU7 at the input, and guys back in the day would put an 12AX7 in place of it to wake them up as you mentioned. The true sound of those amps is with the 12AU7 though, and they can sound just fantastic with that setup. They are loud and clean with a little grit on the top....not quite AC30 territory, but a very appealing sound with lots of tonal complexity, and a more evenly balanced sound than the AC30. (AC30 being more treble freq emphasis) The 12AX7 in place gives more gain, but you lose some tone too when you do that. The 12AU7 just has a sweeter sound overall. The AC100 actually has a 12AU7 on the input, AND one as the phase inverter. These amps are very loud and cleaner than the AC50 when dimed, but they have tons of bottom (probably why they were used as a bass amp too) and are very touch responsive and chimey. I love the sound myself....listen to The Beatles "I Feel Fine" for an example. One of the best amps around for loud and clean, but harmonically rich and touch responsive too. My project one is now making around 85 watts RMS instead of the 105-115 a vintage one is supposed to get, but I think that is because my supply voltage is what the schematic says, and the real Voxes are running higher these days due to the higher AC. My PT is a Sovtek MIG 100U one made for today's wall voltages.

Buttery, would you consider graphing the phase inverter of the AC100 too with that 12AU7 in there? I'm curious what you would come up with? Keep in mind that the common schematic around for the AC100 is the later, fixed biased one that is using 47k plate resistors and a higher supply voltage. The earlier, cathode biased 80/100 amps that the Beatles used were using 220k plate loads and a lower supply voltage. I can get you the specific info if you want it....I'll have to go look it up though. The Jim Elyea book about JMI Voxes has all the info in it that you could ever want about Voxes, and has great schematics of all the amps too. My project amp is using the 220k values and lower supply voltages at this time.

In my latest Bogen project amp, I'm using a cathodyne phase inverter and I had a 12AV7 in there. I didn't have enough swing, so I changed to 100k plate loads to see what happened. With that tube. I got even less swing. While I had the 100k's in there, I swapped in a 12BZ7, thinking that the higher gain would help, and would match the 100k's better, and the lower impedance would provide more swing than the 12AX7 if I tried one. Sure enough that is the case. I get much more signal swing than with the 12AV7 and 56k loads, and the BZ swings more that the 12AX7 does too. If you used one before a tone stack, you would get less loss, and/or could use lower value parts in the tone stack to match up to the lower impedance I should think. I've read that the BZ's are often microphonic and/or defective...as in they have a higher failure rate than a 12AX7, but that they sound thicker. I'm thinking that a phase inverter spot is a good place for them because they will be less microphonic, and can drive the power tubes more, but I guess time will tell. There are a bunch of people at AX84 that have used them in projects and like them though, so there must be something to them besides the low price. :)


Greg

[jjasilli]

Ditto to SoundmasterG. In my Ampeg Reverberocker RI I use 12AT7's for V1 & V2, to tame excessive "gaininess".  But then the amp has trouble reaching overdrive.  So I use a 12BZ7 for the PI.  Overdrive problem solved + rich harmonics from the 12BZ7, but with some microphonics.  Tube swapping is one solution, especially when faced with a diy-unfriendly PCB.

[RicharD]

>Buttery, would you consider graphing the phase inverter of the AC100 too with that 12AU7 in there? I'm curious what you would come up with? Keep in mind that the common schematic around for the AC100 is the later, fixed biased one that is using 47k plate resistors and a higher supply voltage. The earlier, cathode biased 80/100 amps that the Beatles used were using 220k plate loads and a lower supply voltage. I can get you the specific info if you want it....I'll have to go look it up though.

Schematic Heaven only has the later fixed bias version.  Shoot us schematic to digest and I'll slice it up.

-Richard

[SoundmasterG]

>Buttery, would you consider graphing the phase inverter of the AC100 too with that 12AU7 in there? I'm curious what you would come up with? Keep in mind that the common schematic around for the AC100 is the later, fixed biased one that is using 47k plate resistors and a higher supply voltage. The earlier, cathode biased 80/100 amps that the Beatles used were using 220k plate loads and a lower supply voltage. I can get you the specific info if you want it....I'll have to go look it up though.

Schematic Heaven only has the later fixed bias version.  Shoot us schematic to digest and I'll slice it up.

-Richard

Richard,

All I have is the paper schematics that came in the JMI Vox book...no easy way to send them your way without copying and mailing, or finding someone with a scanner, etc. But if you look at the fixed bias version at Schematic Heaven, and sub in 220k plate loads in place of the 47k's, and adjust the supply voltage and/or actual plate voltage at the tubes to reflect the load change, then I should think you should be able to run both versions? The 80/100 cathode biased version is using 220k plate loads on the paraphase inverter, and they list the actual voltage at the plate as 89v. The voltage at the cathode is 4.2v. They do not list the supply voltage per se, though you can infer somewhat. They list the screen voltage at 410v, and this is after a 470 ohm screen resistor. The supply before that screen resistor and the choke is listed at 430v. There is a seperate tap off that screen supply before the actual screen resistors that goes to the PI, though through a 2k2 resistor. So I'd guess somewhere south of 430v and upwards of 410v for the PI supply. I'm sure someone could figure it out with the current draw and everything but that is beyond me.

I'd actually be interested in knowing what both versions are like as far as the analysis if possible? I can get you an actual copy of the schematic, but I'll have to go get it copied and scanned or emailed once it is in pdf form. Let me know if you need this or if you can get by with what I posted? Thanks!

Greg

[RicharD]

Paraphase?  The one at Larry's is a LTPI......IIRC.  Gimme a few minutes to sketch something up.

[RicharD]

OK, yeah it's a paraphase, my mistake.

[RicharD]

AU7-1 is with the 47K plate resistors.
AU7-2 is with the 220k plate resistors. 

Scenario 1 has voltages very close to what's shown on the schematic.  Close enough to not split hairs over.  I could not get Vtube down to 89V in scenario 2.  I ran with bias = -4.2V because that was coming up pretty naturally.  Unless it's a different cathode resistor, I suspect that 89V is actually supposed to be 98V.

IMO, the 47k scenario look really good where as the 220k scenario has a fairly limited input and output. 

[SoundmasterG]

Thanks a bunch for running that!

My AC100 project right now is using the 220k's and my power is down a bit from where it should be for an AC100. I've been working on the Bogen lately, but when I get back to the AC100 project I think I'll start fitzing around with those resistors to see if I can get more swing out of the phase inverter. I was aiming mostly for the sound of the early amps and have it in spades right now, but would like to get the power up closer to the 100 it should be before I add power scaling to it. Who knows...it may even sound better too! Incidentally the schematic does show 89v and not 98v at the plates with the 220k plate resistors....and the cathode resistor is a 1k5 just like the later amps. Since my amp has another channel and the current is all different, I just adjusted supply voltages around until I was getting close to the 89v. My amp also does fixed bias too though so I have to shoot for a happy medium between the two.

Greg

[Geezer]

My favorite tube to stick in certain 12ax7 slots is the lowly & misunderstood 12AY7.....not as drastically low gain as the AU7, & sounds very "musical" to my ears. It allows me to crank up the gain & get the rest of the circuit working harder (if that makes any sense). I don't change anything in relation to Rp or Rk/Rc values.

Both of my favorite amps (HoSo56 & Tweed Overdrive Special) have a 12AY7 somewhere in the circuit.....V2 on the HoSo56, & the 1st OD gain stage/parallel triode on the clean channel of the T-ODS.

If you look hard, there are reputable internet sellers where you can get NOS 12AY7's for $10 each.

G

[SoundmasterG]

The 12AY7 is one of my favorite tubes and I'm constantly looking for them at surplus places and for good deals online. They sound fantastic!

Greg

[WiderGates]

Hello,
this Impact should be a exact copy of the AC 80/100 poweramp:

http://www.edgaraudio.se/tubes/impact/Impact%20100.JPG

and a "better" version that don't eat tubes for breakfast:
http://www.edgaraudio.se/tubes/impact/Impact%20100%20rev4.JPG


Dieter

[SoundmasterG]

That impact amp is close to the Vox 80100. Coming out of the phase inverter, the stock 80100 has .1uf coupling caps instead of the .047uf listed there. The Vox also has 470 ohm screen grid resistors. Many of the surviving 80100's have had the cathode bias resistors changed to 330ohm, and that was also a factory change to the later cathode biased amps. The cathode bypass cap is also a 100uf for each power tube on the 80100. The preamp on that Impact amp is completely different, using different tubes too. The phase inverter circuit is close, but again, using a different tube.

Greg