12AX7 to 12AY7 Question

[Lostfollicles]

Hi Guy's I do hope this question has not been asked too many times before!

With the high prices of 12AY7's and no real cheap alternatives in the market place for a lower gain tube and the many relatively cheap 12AX7 options, as I said they are cheap and readily available, could anyone throw some light on my dilemma?

Firstly I have a Crate Vintage club 30 VC-3112 I hear you say! But this thing sings is built like a tank and is cheap compared to the many alternatives!
The clean channel sounds fantastic and I dont want to change that but the Dirty channel lets itself down by sounding like a wet flabby fart in a tunnel!

Ah! Ha! I thought! I will just fit a lower gain valve in V2 and viola and that will tame her down a bit and promptly fitted the only lower gain valve I had which is a 12AT7! Well actually!! It doesn't sound as bad as you might think but from the little investigations I have made it seems that the 12AY7 would be a better choice not just because it has similar plate loading characterisics to the 12AX7 but it will further help to tame that horrible farting dirty channel from a 100 to 45 gain!

Now being a tight arse and not wanting to spend the high prices on a 12AY7 I thought "Why not just decrease the gain on the cheaper to obtain 12AX7 from 100 to 45 like the 12AY7" By adding a few cheap passive components!

Looking at the circuit diagram (Attached) and from what other people have said, it looks like the main culprit for all that horrible distortion using a 12AX7 is the V2a input providing far too much distortion before it gets to the volume control making it virtually unusable! I mean! It farts at minimum volume WTF? That is so unusable in my view!

So, my question/questions to you guys who know these things is: -

1: To achieve a 45 gain characteristic on a 12AX7 V2a what size should the plate resistor and cathode resistors be?
2: Would it be better to change the plate and cathode resistors also on V2b? So as to tame the whole stage??
3: To help bring in the distortion in a more user friendly way, would it be better to change values of a few other components as well? i.e. Volume pot??

I would also like to increase bass control on the dirty channel as it seems to be a little too middley fore my tastes!? Change C9 perhaps to 1uf???

Thank you all in advance for any help or suggestions you may have! Please "Do Not" ask me to throw this little baby in the bin

Regards
Lostfollicles

[nickb]

Perhaps you are making it too complicated?

(1) Instead of all the mods, try simply turning down the volume on your guitar. Instant taming!
(2) If you really want to mod, change R59 from 470K to something lower say 220k. You can do this by simply soldering another 470k resistor in parallel with R59. This will reduce the gain.

To add bass, or perhaps rather tame the mids and/or highs, there is a tone shaping network after V2B. It's basically a like tone stack with fixed values. Comparing it with the same tone stack in the same amp will give you big clues as to how it works and what to change.

[Lostfollicles]

Hi Guy's

After posting my questions I saw the similar threads columb these sort of answer my questions but I would still like specific to my schematic answers if possible!

This is a great link! Adam's Amplifiers: Triode Calculator

[Lostfollicles]

Wow! Thank you nickb if it really is that simple I will give it try!

Regards

Lostfollicles

[Pedro Vecino]

I know well that model and you can do the following (not complete modification as it is very complex, but a slight approach):
- Cut C51 (10uF). Do not need the extra gain that produces
- Install over R11 a 220K paralleled resistor to slightly reduce the gain
- Install a bypass capacitor in the gain control to clarify the mid/lower gain settings. This will give linearity in the transition. As access to the potentiometer is impossible without removing the PCB, you can take these benchmarks to install (also to test different values if required)



- If you keep the original Sovtek 12AX7, use at least one standard Chinese 12AX7 in the third socket. You get a more controlled overdrive texture in high gain settings

There are four discrete formulas together reduce the gain slightly and put the overdrive capability in a more typical form. If one does not suit you is easy to restore the original state

[cbarrow7625]

Pedro's suggestions sound reasonable and should help but I think there are a couple of other things that you should look at in addition, or maybe instead of.

From what I see, the thing you need to address first is the inter-stage voltage divider between V1 & V2 in the dirty channel (R67 & R59). Two things bother me about that: 1.) like you said, the lack of attenuation between the first two stages is going to make it sound farty no matter what else you do and 2.) a 1M resistor (R67) in series with the signal + the miller capacitance of the tube grid (especially that early in the circuit) almost always leads to a dull, farty-ish tone (for my tastes, and in my experience).

I think you can reduce the values of R67 & R59 without affecting the clean channel tone very much, especially since they are coupled thorough just a small .001uF cap. For starters, try changing R67 to 470k (solder another 1M in parallel to it) and reduce the 470k (R59)down to about 100k (solder a 120k-150K in parallel). That will reduce the series resistance (that I am not fond of) and will reduce the voltage at the V2 grid by about half. I think you might also want to try soldering a 250pF to .001uF cap in parallel with R67 in addition to lowering the value. It will brighten up the tone a bit and still probably won't chage the clean channel noticeably; you'll have to be the judge of that.

Now, V2 itself: It basically has "Modern Marshall" values - not my favorite; that combination tends to add a very crunchy, midrangey quality to the distortion characteristic that I don't care for. As Pedro said, reduce the value of R11 to 100K for sure. If it were my amp, I would also reduce the value of R12 to around 1.5k (put a 3.3k in parallel with it). The cap might be OK at that point, or you might want to make it slightly larger, but I would not go above about 5uF, maybe 10uF at most. If the amp can handle it (not oscillate or squeal, that is), I would also experiment with removing C1 or at least finding the smallest cap that works there to tame the oscillations / squeals. I always try to avoid plate resistor bypass caps at all costs; amps sound much more "open" without them. My tastes are for a very smooth, fluid distortion, not crunchy. You may like it better with the cap in but it is worth trying; you'll be surprised how much it can change the character of the sound. Plate bypass caps basically put a very heavy load on the tube at high frequencies (like 0 ohms, essentially!) and I find that, in general, 12ax7's start sounding very dull & muffled when driving tougher loads. The plate bypass cap is enough to give that character to the higher frequencies.

Interstage attenuation between Stages 2 & 3 looks reasonable. If you put the cap around R67 & clipped out C1, you might want to look at removing C55 (470pF) that bypasses the 475k resistor if things are too bright. In general, treble bypass caps like that have less and less obviously noticeable effect on tone at each successive later stage in the circuit.

The cathode ont he 3rd stage (R14, 4.75k) is going to produce a somewhat crunchy tone (standard Marshall-type value again) but it might still sound OK. It is probably there to reduce the overall gain of the stage so that the whole thing doesn't squeal. Since we increased the attenuation between stages 1 & 2, you might be able to reduce R14 down as low as 1.5k. The lower the value there, the less crunch (but more overall voltage gain) you will get from that stage. Or if you want to go the opposite way, you an reduce the gain of that stage even further and put it more into cold biasing by increasing it to anywhere from 10k up to the SLO value of 39k or even 47k. Increasing probably won't affect the tonal character all that much since the 4.7k (with 100k plate resistor) already puts that tube into a somewhat cold-biased state to begin with; it will just scrub off some gain.

I don't know what's up with that fiixed tone circuit after the 3rd stage. It is basically the equivalent of a Marshall-type tone stack with the treble & midrange knobs turned up to 10 with the bass knob on 0. If that is your kind of thing, leave it; I would be looking to remove it or bypass it. That combination of R15, C12 & R66 are also putting a failry heavy load on R13 (as all Marshall/Fender tone stacks tend to do), which will lead to the dull, farty loaded-down 12ax7 tone that I was describing earlier. It is actually worse in this case becasue the 4.75k cathode resistor (R14) causes the output impedance of that tube to be higher than normal, so the loading effect is heard even more dramatically than with a "normal"-biased stage. I would try clipping out R15 & R16, then replacing C11 with a 100K-470k resistor to get back a bit of the (probably necessary) attenuation. The stage 3 tube will now be seeing a nice, easy to driver 1M ohm load to drive. If more attenuation is needed, just reduce R69 to 100k-200k (with the 470k resistor replacing C11).

There is really no reason for the cathode of stage 4 (R18) to be 1k; again, it is just a typical Marshall-type value they stuck in there. It can add a bit of "tightness" to the sound, but increasing it to 1.5k-2k will reduce the gain at that stage a bit and also reduce the "crunch" factor a bit again. Think about it; that stage is being hit by the biggest signal voltage in the entire channel, yet it has the least grid voltage swing available of all the 4 stages! It is going to distort no matter what, give it a little breathing room for crying out loud! In the old 2 & 3-stage Marshalls a low value was needed there to make sure the right part of the waveform was being "helped" into clipping; here it's clipping both sides of the wave, pretty much all the time. As Pedro said, you might want to remove C51 (10uF cathode bypass cap) if there is too much overall voltage gain still, but it is not really going to do much to alter the character of the distortion. The problem with this circuit is the gain structure early in the circuit, not here. I actually find that a 10uF bypass cap on the stage right before the tone controls is perfect to produce good pick attack definition without getting too muddy. Leaving it out tends to make the tone just a bit "mushier"; which I don't care for. By the way, reducing that cap below 10uF is not a great idea either (unless you are reducing it to less than 1uF for just a higher frequency treble boost), values of 2-5uF in that position end up emphasizing the lower mids and it just sounds odd to me. 10uF seems like the perfect value for a "full range" boost. 15-20uF starts getting into muddy, flabby, boomy territory.

Cathode follwer & tone stack look OK. If things still sound a little dull, experiment with 100pf-.002uF bypass caps around R21 in the master volume attenuation circuit. Usually, a bypass cap in that position sounds way too bright, but it has its place sometimes. Conversely, to reduce some overall treble, bypassing the master volume pot (P8) with a cap to ground can reduce the overall treble without really affecting anything else too much.

So, that circuit has too many WTF? choices in it for me! Looks like it should be pretty easy to improve it with just a few tweaks without having to actually unsolder anything or take the amp apart. Good luck. Let me know what you end up doing and how it sounds!

[tubeswell]

Its not just the amplification factor that makes a 12AY7 sound like a 12AY7, its a whole bunch of things:



and you need to make a distinction between the amplification factor ('mu') of a tube and the gain of the stage.

The gain of the stage is set by at number of characteristics which you can't change and a number of things which you can change, so you only have limited ability to manipulate the characteristics of any particular tube type. Amongst the aspects you can't change are; the tube's ('internal') plate resistance (ra), the tube's transconductance ('gm'), and the tube's amplification factor (mu).

The aspects you can manipulate to change the gain of a stage include: the (resistive) load on the plate (Ra), and the bias point of the tube which, in your typical cathode-biased pre-amp tube is set by the cathode resistor (Rk), and the AC load on the stage - which you can change by changing the Ra and any AC-parallel grid load resistance ('Rl').

You can also manipulate the frequency band-pass characteristics to a limited extent. For example, the Rk can be selectively gain-boosted at different frequencies by adding cathode bypass capacitance (Ck), in order to manipulate the Lo-freq 3dB roll-off point of the gain boost (giving the effect of full-bandwidth gain boost or selective mid to high freq gain boost). The intrinsic band-width ('bw') of the tube is something you can only have a limited influence over (because it is affected by the gm and the ra - which you can't change). However you can change the Ra and the method of coupling different gain stages, and any coupling capacitor (C) and any other freq-shaping RC filter networks,as well as the stage's input frequency HF roll-off point, which you can manipulate by changing grid resistor, which interacts with the inter-electrode capacitance ('miller capacitance') of the tube to shape the input frequency roll-off point.

The upshot of all this is that there are many variables that will affect the result, and it will take more than a month of sundays to conduct piecemeal experiments to suit your tastes. I highly recommend reading the first 2 or 3 chapters of Merlin Blencowe's 1st book (preamps), if you want to come to terms with the intricacies of pre-amp gain stage design.

However having said all this, nothing really beats getting a 12AY7 and bunging one in - if a 12AY7 is what floats your boat.

[Lostfollicles]

Phew! Lots to take in!

Hi guy's

Thank you so much for all your input, there is a lot more to this than I thought and I obviously need to have a good chew on all the info you guy's have to offer before I can make a decision as to what I'm going to do!

I'll get back to you soon but for now, cheers guy's I'm really grateful to you, these amps are awesome for the money even though they need to be tweaked!

Respect

Lostfollicles

[HTH]

Why not just buy a 12AY7? - you can get a current production EH one for under $12.00, thats hardly breaking the bank.

My own thoughts would be to bypass the third gain stage altogether - clip the two components at the junction of R69 and C13, then run a wire there from the wiper of the gain pot (P4).

You'd still have a decent amount of gain and a dirty channel in the ballpark of a 2203 Marshall, happy days.

[cbarrow7625]

Heck, if you want to go the direction of bypassing one of the gain stages, I would just bypass the one that is giving you the most trouble, stage 2. Having the gain after it instead of before is pure silliness. Just re-route C41 over to the junction of C10 and R76, then clip C10 out of the circuit (shunt around R59 to ground to ground the grid of that tube and keep it quiet). Drop another resistor and cap in parallel with the cathode resistor, R14, to get the gain up a bit and you're in pretty good shape; similar results to what HTH suggested. I would still get rid of that shaping circuit right after that stage (clip R66 from ground, then just shunt around C11 to bypass the whole thing). At that point it should be a pretty hot sounding and useful lead channel.

[Lostfollicles]

Crate VC-3112 Sorted Many Thanks Guy's

Just a short note to all the guy's that contribuited to my quest for a better sounding VC-31312!

I ended up changing on V2a: -
R11 to 100k
R12 to 1.5k
R59 to 240k-ish (2x470k) in parallel
C9 to 1uf

V2b:-
Removed R15-R16-R66-C11-C12 replacing C11 with 470k resistor.

For bias of 2x EL84's instead of 4x EL84's I found that a 100R resistor in series with R43 worked the best with no red plating I did try 60R but I found that the EL's were red plating a small ammount so stuck with the 100R.
I also tried a 220R resistor in parallel with R52 to bring the spec the same as the 20 Watt model but found this to cause red plating, plus it sounded slightly distorted so I left it as spec 470R.
I have also fitted a switch and wired across the 100R resistor to enable me two change between 2x and 4x EL84's

What does she sound like now?

Awesome guy's just plain awesome, she sings like a bird, with the clean channel giving chimey cleans leading to a nice warm break up at higher volumes and a awesome sounding dirty channel!

Thank you so much for all your help guy's!

I now have an amp where can quickly change my VC-3112 from a 30 Watt to a 15-20 Watt-ish screaming beast! If I ever need to that is!?

Thanks again

Regards and respect

Lostfollices

[cbarrow7625]

Sweet. Good choice to keep the mods as simple as possible. Sounds like you directly addressed the two stand-out "problems" with the original preamp design. Glad you are happy with it now! Go sound clips?