Page 1 of 4
#1
So, one of the things that I mentioned before in my kit build tutorial thread (see my sig) was that I'd consider a "design an amp from scratch and build it up" thread. Well, an idea of sorts has popped into my head for an amp that I'd like to build, so I'm going to design it and build it. And document the whole process for anyone who's interested.

Disclaimer: I've designed about 10 amps. Some of them have been awesome, some have been total busts. It's the nature of the beast. I have an inkling that this amp would rule, but I don't know for sure. That's the beauty of this.

Disclaimer 2: Honestly, this thing might suck...but it might be awesome.

Disclaimer 3: I am working on a (very) limited budget. Namely, exactly zero dollars. I'm going to build this with parts laying around. Fortunately, I have a LOT of parts laying around. So much so, the biggest problem will most likely be finding them all.

Disclaimer 4 (getting annoyed yet? ): This is most likely going to take a long time. Not because of anything remarkably difficult about it (my record for designing/laying out/building an amp from scratch is about 16 hours), but because I have a wife who needs a lot of attention (pregnant and sick). In fact, I have no idea if this will ever get done. I hope it will. If you have a problem with this,


This will be a design of an amp, from the schematic level...that is, when this is all said and done, this will be an amp that noone has ever heard before. SCORE! The amp idea itself:

-15-20 watt 2xEL84 push pull amp
-"2 channel"...I put that in quotes because it's going to be a MOSFET booster up front of a crunchy preamp, so it'll go from clean to insane or crunch to insane depending on where various controls are. That will be footswitchable.
-MOSFET for boost: LND150 from Supertex...high voltage MOSFET with gain of 250+ and can swing outputs from almost rail to ground.
-Preamp tubes: two russian tubes that I have that are equivalents to 12AX7s
-poweramp tubes: two russian tubes that are equivalent to EL84s

The tubes I bought on eBay from a guy from the Ukraine, they have actual russian military markings on them. 8 12AX7s and 4 EL84s. Going to use some of them for this amp, some for another amp.


As kind of a table of contents for all (and for myself so I remember), I'm going to try and touch on these things:

1) Simulation using LTSPICE (electronics simulation package - FREE!!) that allows you to simulate tubes, see what the response looks like from a clipping, FFT, frequency response point of view. This isn't an "all knowing" simulation, it's just something to do beforehand so you know that you're on the right track. Ear tweaking will be needed.

2) General layout techniques. I'll show a really lovely program called DIYLC (do it yourself layout creator), which is a free electronics layout software package from bancika, who posts on AX84.com amongst other places.

3) Construction of amp (basically, redo of the kit build tutorial, only for this amp). This will include making/etching PCBs, making a turretboard, setting up the footswitch controller, etc.

4) Tuning/tweaking

I don't know if anyone here will be interested in this. I hope someone will be. This is really fun stuff, very involved too, and certainly not for everyone...but yeah, I hope someone out there gets something out of it...I sure will.

Other background:

Assuming I get this far, it's going to go into a 1x12 combo that I have. It used to be a Marshall SS cab...a friend gave it to me for free because he got it for free from someone who retolexed it in that red glittery crap that was on barstools back in the '50s...some people hate it, I like it. I also like the whole idea of putting russian/soviet tubes in a red amp. :shrugs: I'll post a pic of the cab in all its glory sometime.

General amp idea:

LND150 boost with a "boost level" adjust in front of it. Gain to be set pretty high. Most likely a switch that changes frequency response from "top end" to "full" > parallel 12AX7 section (great input stage for the non-boosted condition, will work well for boosted too) > single 12AX7 stage > 12AX7 cathodyne phase inverter > 2xEL84 push pull power amp Gotta fit a tone stack in there somewhere too, will probably decide where during the simulation stage...might have to add another triode to make it happen though...

In the non-boosted form, it should overdrive the single 12AX7 stage well for a clean to classic rock tone. Boosted it should be waaay up there in gain.


If you managed to read this whole thing, congrads! I'll try to keep things a bit smaller, in "bite size" pieces from here on out. This post was mostly stream of consciousness so that I don't forget anything.

If you don't understand everything as I'm going, a) that's OK and b) ask I can answer most things, and there are a lot of people around here who can answer most questions too. You will hopefully come out of this with enough background so that you can at least understand what is going on inside of amps, even if you can't design your own from the ground up immediately.

And no, the recording/potential tweaking of my SEL kit build will not suffer from this, mostly in a holding pattern waiting to get in some recording time anyways. So, yeah,
Quote by kcdakrt
DLrocket89 makes my ug experience better!


Member of the official GB&C "Who to Listen to" list

Kit Amp Building Tutorial
#2
Kay, here's the paralleled 12AX7 stage. This is a screenshot from LTSPICE. LT spice is a simulator available from a LOT of places. Google "LTSPICE download"...the version I'm using is SwitcherCAD IV available from Linear Technologies.

I have a tutorial written up somewhere on how to use SWCAD, I'll find it and post later.



In any case, the things to understand...

V1 is the DC power supply (300V)
V2 is the guitar signal

They (and everything else) is referenced to ground, which is the arrow at the bottom.

The two circles are the tubes (the grid on the second one should've been connected too. Oops)

For info on how tubes work, google is your friend, or try goodling "freewebs merlin guitar amp"...that should take you to Merlin's website, which is GREAT, and he has a book out too which is great as well. He has a 37 page article on how to design a single triode gain stage. Good info there.

NEways, R3 is the anode/plate resistor. The typical value for a 12AX7/ECC83 is 100K. This gets cut in half because we have double the current flowing (two tubes). They don't really make 50K resistors, so we use a 51K.

R1 and R2 are the cathode resistors. Frequently with a parallel stage like this, you going with one cathode resistor. I'm going with two because we can then play with the biasing of each tube individually...if we want, we can bias one hot and one cold, and hopefully have a very complex and "yummy" sound.

C1 and C2 are bypass resistors. Basically, they make it so that the AC signal doesn't have to go through the cathode resistor, so you get more gain. They don't amplify all frequencies evenly...below a certian point, they stop working. That "certain point" is determined by their value. The lower the value, the higher the cutoff point. 22u bypasses pretty much all frequencies, so it'll give a "full" sound. .68u bypasses pretty much just the treble range. So, what we have is a composite state that has two different bias states that each have their own frequency responses. Makes for a LOT of tweaking and work getting it right, but it lends itself to great possibilities for great tone. The goal I have here is a respectable full response with "sparkle" in it, hence the boosted treble on one and full on the other.

We may drop the 22u cap to something like a 1u cap One of the top rules of designing a high gain amp: to avoid mud, control the bass early and often.


EDIT: C3 is a coupling capacitor, blocks DC to the next stage but allows AC (the signal). .022u is a common "starting point". We're going to start there and see what happens. The resistor after it is just to apply a load. The resistor between the signal and the grids going to ground is a ground reference resistor and is necessary to keep the tube grid at 0V (where it belongs).
Quote by kcdakrt
DLrocket89 makes my ug experience better!


Member of the official GB&C "Who to Listen to" list

Kit Amp Building Tutorial
Last edited by DLrocket89 at Sep 4, 2009,
#3
Cool dude.
Jackson RR5 ivory w/ EMG 81/85
Jackson DX6 w/ SD Distortion & Dimarzio Super Distortion
Fender Starcaster Sunburst
Mesa/Boogie DC-3
Johnson JT50 Mirage
Ibanez TS-9
Morley Bad Horsie 2
Boss CE-5

ISP Decimator
Boss DD-6
Korg Pitchblack
#5
dlrocket, i love you. i want to design an amp that wont blow up in my face for once
Gibson SG Faded
Epi VJ Stack


Quote by Øttər
Whenever I clean my guitars, my family wonders why it smells so good; I say that I exude a fresh citrus scent from hidden orifices.
They stopped asking
#6
Thanks for the comments everyone. People actually caring keeps me wanting to do this. If I'm just talking to noone, I'm probably not going to spend the time doing it.

So, I finished out some of the input stage. Coupled the second grid in, added two capacitors. One is in parallel with the plate resistor (200pF) and one is in parallel with the ground reference resistor (50pF). These are there to round off some of the high end frequencies, helping to prevent oscillation. The numbers chosen are there as something of placeholders...they'll get the job done, but they shouldn't alter the tone really. Full up input stage:




Now, what is the frequency response of the stage? One of the cool things about SPICE is that you can simulate the frequency response of a stage using something called a Bode Plot (wikipedia is your friend here). In a nutshell, it shows gain vs frequency. I changed both coupling capacitors to 22uF, so there is max gain at all frequencies, to show what the stage does "full on":



The bottom view shows the input stage as configured, the top is the bode plot. Don't worry about the lighter green line. The heavier green line is the frequency response...the gain in dB is on the left, the frequency is on the bottom. Human hearing ranges generally from 20Hz to 20kHz (20,000 Hz). What I take to be "low end" in a guitar is 50-400, mid is 400 to 1k, highs are 1K and up. This varies by style of music, amp, preference, etc, so please so nitpick on this point.

So, as you can see, we have a stage that amplifies all frequencies quite evenly. That's really not good for a few reasons. First, you don't want things to be even...the human ear hears low end better than high end, so you want the low end to be a bit lower electrically to make it sound more balanced in reality. Secondly, you don't want a lot of low end early on in the preamp, as it will lead to mud and mush later down the line.

So, let's play with one of the coupling caps. We'll leave the one on 22uF for a full bypass (and max gain through that tube), but let's play with the other one and see what happens:




The above image has one of the caps changed to .68uF. As you can see, the frequency response how has a hump. If you look at the hump, it's about the same overall gain at higher frequencies as before. Below the hump, the gain drops. This is because those frequencies are only fully amplified in one of the triodes, so the response is lower there. The frequency of the roll off point can be changed by changing the capacitor values.


What happens if we remove the 22uF cap? That would make the low end really not be boosted a whole lot, so you end up with this (me changing it to .000001pF is just my way of not deleteing it while making it schematically irrelavent):



Pretty dramatic, huh?


Let's put that 22uF back in, and change the other cap to a smaller value, which will make the boost slighter higher in frequency:



To me at least, that looks like a reasonably good input stage response. *I should have mentioned this before* The tone goal of this amp is something like a hotrodded JCM800 or so, but with more gain. JCMs are reasonably bright, so it would make sense to have this higher frequency response.

Now, the removal of that 22uF before was pretty powerful. If you were to put that on a switch, you could go from the result above to this:



In actuality, Fender does something similar with their "fat" switch on the Blues Jr (I'm sure other amps do this too). If you cut that 22uF out (which is the "normal" response), it'll be brighter. Flip on the fat switch and the 22uF is in the circuit and the lowend is boosted some, making it sound fatter.

I think I'll actually remember that and put a fat switch on this thing.

That's it for now, I have some other ideas I'm going to work on that should be pretty sweet. Bootstrapped cathode follower anyone?
Quote by kcdakrt
DLrocket89 makes my ug experience better!


Member of the official GB&C "Who to Listen to" list

Kit Amp Building Tutorial
#7
Hi,

This is an excellent contribution to the forum. It's awesome you are taking the time to explain this.

For someone without amp building experience but an OK but fairly basic understanding of schematics and electronics, how difficult would it be to follow your method and produce an amp?
#8
Quote by a_man
Hi,

This is an excellent contribution to the forum. It's awesome you are taking the time to explain this.

For someone without amp building experience but an OK but fairly basic understanding of schematics and electronics, how difficult would it be to follow your method and produce an amp?


Thanks. Quite hard is my guess. There's a lot of peculiars with tubes that i'm kindof glossing over and not really mentioning right now that would totally screw something up if you tried something wrong. I'm hoping that by explaining this stuff in detail, people will understand existing amps more and be able to mod them. You need to build and mod a couple of kits before knowing what to try, look for, etc.
Quote by kcdakrt
DLrocket89 makes my ug experience better!


Member of the official GB&C "Who to Listen to" list

Kit Amp Building Tutorial
#9
just read the bit about finding somewhere for the tonestack...

my suggestion (you can and probably should ignore this, its your amp):
have the tonestack follow the cathode of V1b, marshall style? (though marshalls have it after v2, but in this case, v2 is already the phase inverter)
Gibson SG Faded
Epi VJ Stack


Quote by Øttər
Whenever I clean my guitars, my family wonders why it smells so good; I say that I exude a fresh citrus scent from hidden orifices.
They stopped asking
#10
Quote by DLrocket89
Thanks. Quite hard is my guess. There's a lot of peculiars with tubes that i'm kindof glossing over and not really mentioning right now that would totally screw something up if you tried something wrong. I'm hoping that by explaining this stuff in detail, people will understand existing amps more and be able to mod them. You need to build and mod a couple of kits before knowing what to try, look for, etc.


Ahh, I would of thought so. If some money comes my way I'll have to give the kits a shot. Thanks.
#11
Bootstrapped cathode follower. Yummy.

Many guitars use cathode followers to "drive" tonestacks. Tonestacks (passive ones at least) work by sending various frequencies to ground. This makes them "bog down" the tube before them, reducing gain. So, you can use a cathode follower (don't need to mind you) to drive it. In this case, I'm using something called a "bootstrapped" cathode follower. Essentially, the consists of using the cathode follower's output as a postiive feedback to the gain stage before it, helping to increase the gain. It looks like so:



The cathode follower is the tube on the right. It is a low-gain stage (less than 1, meaning the signal is reduced in amplitude). The bootstrap is C3, which goes back to the plate resistor (which in this case needs to be split in half...33K and 33K, fo the "actual plate resistor is now 66K).

That capacitor can be sized to only allow certain frequencies through (same as the other coupling cap, smaller = higher frequencies only).

Using a .022uF cap:




Using a 1uF cap:




As you can see, another 5dB or so of gain, for free. Woot! The bootstrapped CF has the added benefit of making a smoother distortion, so when we put that boost up front and start beating the crap out of it, the distortion should be pretty smooth.

I'm going to leave it at the .022uF cap for now in the design phase...like I said, control bass early and often.
Quote by kcdakrt
DLrocket89 makes my ug experience better!


Member of the official GB&C "Who to Listen to" list

Kit Amp Building Tutorial
#12
Quote by AngusJimiKeith
just read the bit about finding somewhere for the tonestack...

my suggestion (you can and probably should ignore this, its your amp):
have the tonestack follow the cathode of V1b, marshall style? (though marshalls have it after v2, but in this case, v2 is already the phase inverter)


Yeah, V1b and whatnot goes out the window anyways because the input stage is V1a AND V1b..

As you can see, I just added a cathode follower. My thoughts are to go (paralleled input stage) > CF > tone stack > 12AU7 recovery from tone stack > 12AX7 gain stage > 12AX7 cathodyne PI
Quote by kcdakrt
DLrocket89 makes my ug experience better!


Member of the official GB&C "Who to Listen to" list

Kit Amp Building Tutorial
#13
yep what you did is what i meant, shoulda clarified myself

we're thinking on the same page though. i did mean cathode follower AFTER v1b.... i can haz fail now...

what are the benefits of the 12AU7 in that spot? or is it one of those things that just happens to sound better?
Gibson SG Faded
Epi VJ Stack


Quote by Øttər
Whenever I clean my guitars, my family wonders why it smells so good; I say that I exude a fresh citrus scent from hidden orifices.
They stopped asking
#14
Quote by AngusJimiKeith
yep what you did is what i meant, shoulda clarified myself

we're thinking on the same page though. i did mean cathode follower AFTER v1b.... i can haz fail now...

what are the benefits of the 12AU7 in that spot? or is it one of those things that just happens to sound better?


After talking with Merlin about it some, if you're going to bootstrap a paralleled tube section, you need a lot of current drive. In this case, that equals 12AU7 or 12AT7. There's a slight chance of the stage after the tone stack being overdriven a bit, at which point in time a 12AT7 sounds craptacular, so 12AU7 it is.

That and, I'd like to have the stage after the tone stack be a very low gain stage, basically just enough to make up for the loss of the tone stack (so, a gain of 4 ought to suffice). Too much gain than what you want is a bad thing too because then noise is amplified too and whatnot, so yeah...

*shrugs* I'll do the pinout on that tube so that if I want to, I can put a JJ 12DW7 in that position, have the 12AU7 section for the CF and the 12AX7 section for the recovery after the tone stack.

These are things that will be further decided while simming it. Nice thing about SPICE is that swapping out a 12AX7 for a 12AU7 is that it takes about 15 seconds.
Quote by kcdakrt
DLrocket89 makes my ug experience better!


Member of the official GB&C "Who to Listen to" list

Kit Amp Building Tutorial
#16
i know you said ignore the light green line, but my listening skills arent very good. i know its the phase section of the bode plot, got that under control. basicly im just wondering if the phase makes a huge difference here. like i notice that with your .022uF cap in the last section the phase swings back "upwards" a bit around 180 degrees before dropping back "down" again around 800Hz. Then with the 1uF cap you have a more straight up phase transfer function like you would expect with a more simple circuit. so basicly, does the phase matter much here or am i just being difficult?

also, have you used any other spice packages? ive used PSpice quite a bit, but it doesnt do tubes (my version doesnt anyway). is this one you like a lot, or is it just one that does tubes well or what? im probably going to dl it no matter what, just curious again.
#17
Quote by DLrocket89
As you can see, another 5dB or so of gain, for free. Woot!
Nope. Not free. There are always prices to be paid.
Your clean just got a little dirtier.
Positive feedback accentuates any non-linearities (distortion). Also, it reduces stability.

Will you like the way this sounds or will it sound like ass? Sims won't tell you.
You'll have to find out IRL.


It seems odd to me that you're designing this around a 300v buss for your preamp tubes. Why not start with a voltage you'll actually use in the finished product?
Meadows
Quote by Jackal58
I release my inner liberal every morning when I take a shit.
Quote by SK8RDUDE411
I wont be like those jerks who dedicate their beliefs to logic and reaosn.
Last edited by SomeoneYouKnew at Sep 4, 2009,
#19
Quote by XgamerGt04
Why do you use the diode in that circuit? What purpose does it serve?
It keeps the control grid from ever being able to go more than a few tenths of a volt more positive than the cathode. If the grid is more positive than the cathode, it draws current and unpleasant things happen. imho, a grid-stopper resistor is a better choice than a diode.
Meadows
Quote by Jackal58
I release my inner liberal every morning when I take a shit.
Quote by SK8RDUDE411
I wont be like those jerks who dedicate their beliefs to logic and reaosn.
#20
Quote by jof1029
i know you said ignore the light green line, but my listening skills arent very good. i know its the phase section of the bode plot, got that under control. basicly im just wondering if the phase makes a huge difference here. like i notice that with your .022uF cap in the last section the phase swings back "upwards" a bit around 180 degrees before dropping back "down" again around 800Hz. Then with the 1uF cap you have a more straight up phase transfer function like you would expect with a more simple circuit. so basicly, does the phase matter much here or am i just being difficult?

also, have you used any other spice packages? ive used PSpice quite a bit, but it doesnt do tubes (my version doesnt anyway). is this one you like a lot, or is it just one that does tubes well or what? im probably going to dl it no matter what, just curious again.


I don't think that the phase matters a whole lot. If I had to guess, phases off by some odd number (sy, 74 degrees) will perhaps lend itself to slight muddiness. I've never seen it actually be of concern though...so much else going on.

I've used circuitmaker 2000, which is nice for a quick 5 minute "throw it together to see if it works" thing. Want to go a bit more in depth here though.
Quote by kcdakrt
DLrocket89 makes my ug experience better!


Member of the official GB&C "Who to Listen to" list

Kit Amp Building Tutorial
#21
Quote by SomeoneYouKnew
Nope. Not free. There are always prices to be paid.
Your clean just got a little dirtier.
Positive feedback accentuates any non-linearities (distortion). Also, it reduces stability.

Will you like the way this sounds or will it sound like ass? Sims won't tell you.
You'll have to find out IRL.


It seems odd to me that you're designing this around a 300v buss for your preamp tubes. Why not start with a voltage you'll actually use in the finished product?


In this case, I'm calling it "close enough to free" because I'm not really intending this to be a clean amp. It's also...it's easier to think of it in terms of positive feedback (for me at least). In reality, it makes the anode/plate resistor look close to infinitely large, so that you can get as close as possible to mu of the tube. How exactly it does this, I'm not entirely certain...so, I tend to think of it as a positive feedback that is minute enough that it won't become unstable (I know, that's kindof an impossible condition, I've taken a systems modeling class). From Merlin and a couple other friends who have build it, it creates a very slight compression that is apparently great sounding, a very warm/lush/smooth sounding overdrive.

The 300V is in there as a placekeeper. I don't know what the value will be at each tube, but I'll be able to set each one individually (look up "Supertek LR8 regulator"). The actual power supply buss will be around 400VDC and will be regulated down to whatever I choose at each tube...it's something I've been wanting to try out for awhile, so here's a nice opportunity. I'll explain that when I get there. It'll be a power supply unlike any that you've ever seen in any mainstream guitar amp, I promise that.
Quote by kcdakrt
DLrocket89 makes my ug experience better!


Member of the official GB&C "Who to Listen to" list

Kit Amp Building Tutorial
#22
Quote by SomeoneYouKnew
It keeps the control grid from ever being able to go more than a few tenths of a volt more positive than the cathode. If the grid is more positive than the cathode, it draws current and unpleasant things happen. imho, a grid-stopper resistor is a better choice than a diode.



It's there at startup to keep the cathode follower from arcing shorting internally (I think that's what makes Tung Sols go wacko in a CF position)...

Once everything is up and running, the diode is reversed biased and does not conduct unless you OD the CF. The 10K resistor is there to clamp down on any conduction noises should you overdrive the CF and cause the diode to conduct. No noise issues, no arcing, life is good.
Quote by kcdakrt
DLrocket89 makes my ug experience better!


Member of the official GB&C "Who to Listen to" list

Kit Amp Building Tutorial
#23
So for the supertex regulator are you just going to make R2 something like a variable resistor and then use it to adjust the voltage?
#24
Quote by XgamerGt04
So for the supertex regulator are you just going to make R2 something like a variable resistor and then use it to adjust the voltage?


Initially, yes. That way it is complete cake to tweak the voltage on the tube. Once the "good sounding" voltages are dialed in, it'll be hardwired with fixed resistors. No more of this completely chained together where everything intereacts with everything else crap like every other power supply out there. Not to mention, a lot lower noise floor too, any ripples or whatever will be smoooothed out.
Quote by kcdakrt
DLrocket89 makes my ug experience better!


Member of the official GB&C "Who to Listen to" list

Kit Amp Building Tutorial
#25
Quote by DLrocket89
Initially, yes. That way it is complete cake to tweak the voltage on the tube. Once the "good sounding" voltages are dialed in, it'll be hardwired with fixed resistors. No more of this completely chained together where everything intereacts with everything else crap like every other power supply out there. Not to mention, a lot lower noise floor too, any ripples or whatever will be smoooothed out.


I was reading the datasheet and they do look really cool. It would be very easy to put into a PCB style amp. Filtering out a lot of that ripple would help quite a bit, especially in the preamp.
#26
So, one of the other things I wanted to do with this amp (if this isn't readily apparent by now) is to do things differently. So, I'm going to break up the tone stack. One of the things learned by Carl B and Jukka from the AX84 forums was that having a bass control up front is good (because you can control the bass very well that way) and that the treble control being late is good too (after the phase inverter in this case) because it allows the amp to "sparkle" more. Obviously, some treble attenuation should occur in the amp to keep it from being really bright, but it's an interesting idea that was successful at least once, so we're going to go that route.

Regarding the bass control, here is what I've come up with:



I direct your attention to C4, R4, and C5. R4 is a 250k log pot...C3 is a smaller coupling cap that has no series resistance, so it passes through mid and upper frequencies very well, not so much on the lower end. C4 and C5 pass low end, with the resistance there to vary how much it passes. It varied by about 5dB (as stated earlier). 250k is enough to pretty much control it well, and the log taper is required (I figured this by trying different values of the pot...found that 1, 100, 10k, and 250k scale pretty linearily, so a log pot will be best).

5dB in my mind isn't really enough to control things well...no real justification here, just going on a hunch. So, recall the 22uF pot that I was considering using as a fat switch? Make the 250k pot from above a dual concentric pot, have one going to the series resistance above, and have the element going to series with the 22uF pot, and we have a bass control:




That one shows the bass set low. The one below shows the bass set high:




Frequency response?

Bass Low:



Bass Hi:



Now we're looking at a pretty good 8dB of adjust. Should be enough, I hope...


Now, someone may point out that the bass control is now behaving like a gain control (as it's boosting all frequencies for that triode, not just the bass). To combat that, we bypass the cathode resistor with a smaller cap that has no series resistance. This bypasses the resistor and higher frequencies and leaves the lower frequencies adjustable by the pot, like so (see C9):



Bass response with bass on high will be the same (now it's bypassed with a 23uF cap instead of a 22uF...no biggie there). With the bass turned down, we get:



We've boosted the mid and top gain a bit, so that now when you turn up the bass, the levels at those frequencies are unaffected.

my only question is whether or not it has enough control over the bass to really be effective. Guess we'll find out later...
Quote by kcdakrt
DLrocket89 makes my ug experience better!


Member of the official GB&C "Who to Listen to" list

Kit Amp Building Tutorial
Last edited by DLrocket89 at Sep 5, 2009,
#27
Quote by XgamerGt04
I was reading the datasheet and they do look really cool. It would be very easy to put into a PCB style amp. Filtering out a lot of that ripple would help quite a bit, especially in the preamp.


Yeahup, good things to have on hand, that's for sure. I don't know if you recall the amp I built into a cakepan awhile back...it was an SE amp with a lone 6V6 in it, I had a "mojo" control that was just an LR8 with a variable resistor on it, driving the screen voltage of the 6V6. Fun stuff.
Quote by kcdakrt
DLrocket89 makes my ug experience better!


Member of the official GB&C "Who to Listen to" list

Kit Amp Building Tutorial
#28
Quote by DLrocket89
It's also...it's easier to think of it in terms of positive feedback (for me at least). In reality, it makes the anode/plate resistor look close to infinitely large, so that you can get as close as possible to mu of the tube.
Not infinitely, but it makes it seem larger than it is.

Quote by DLrocket89
How exactly it does this, I'm not entirely certain...
by use of positive feedback. A triode is a (grid) voltage controlled current sink (anode). When the grid of the first triode becomes less negative, the CF is also stealing some of the current from the upper resistor in the split anode resistor stack. That moves the reference point downward (less positive). The same amount of current will flow in the anode circuit of the first triode as the grid is calling for. The voltage drop across the lower anode resistor is the same, but it's referenced to a lower voltage. So the anode is at a lower voltage than it would have been without the feedback.

Quote by DLrocket89
so, I tend to think of it as a positive feedback that is minute enough that it won't become unstable (I know, that's kindof an impossible condition, I've taken a systems modeling class).
Not minute. Just small enough that it doesn't oscillate. It still is less stable than no feedback or negative feedback.

Quote by DLrocket89
From Merlin and a couple other friends who have build it, it creates a very slight compression that is apparently great sounding, a very warm/lush/smooth sounding overdrive.
There are some similarities between what's happening here to the design concepts in the Dunlop Fuzz Face. Quite different, but the basic concept is still the same.


Quote by DLrocket89
The 300V is in there as a placekeeper. I don't know what the value will be at each tube, but I'll be able to set each one individually (look up "Supertek LR8 regulator"). The actual power supply buss will be around 400VDC and will be regulated down to whatever I choose at each tube...it's something I've been wanting to try out for awhile, so here's a nice opportunity. I'll explain that when I get there. It'll be a power supply unlike any that you've ever seen in any mainstream guitar amp, I promise that.
your PS is going to be nearly as complex as the amp itself.
Meadows
Quote by Jackal58
I release my inner liberal every morning when I take a shit.
Quote by SK8RDUDE411
I wont be like those jerks who dedicate their beliefs to logic and reaosn.
#29
DLrocket89, would you suggest starting with a simple 386 based amp like a runoffgrove.com style amp and work up from there, or would one of the kits from ax84 be ok to start from? I have a basic understanding of electronics, I can solder like crazy, and would be willing to take my time. Also, do you know of any cheapers kits, or easy tube amps to start with building?
Lefty Strat Copy w/ GFS Lil' killer rails + Crazy wiring
Lefty Kona Thinline Acoustic/Electric
Righty-to-Lefty Silvertone Bass
Righty-to-Lefty Memphis Short Scale Bass
Vox AD30VT-XL

Tascam US-428 USB Interface Mixer
Guitar Rig 3
#30
To SYK:

Agree on all of that. The power supply will be more "busy" yes, but the LR8 thing is so nice that I think it's worth it. Not a whole lot of trouble to use them, and the adjustability is great...
Quote by kcdakrt
DLrocket89 makes my ug experience better!


Member of the official GB&C "Who to Listen to" list

Kit Amp Building Tutorial
#31
Quote by GeToChKn
DLrocket89, would you suggest starting with a simple 386 based amp like a runoffgrove.com style amp and work up from there, or would one of the kits from ax84 be ok to start from? I have a basic understanding of electronics, I can solder like crazy, and would be willing to take my time. Also, do you know of any cheapers kits, or easy tube amps to start with building?



Start with the P1 from AX84.com, it's good for starting out.
Quote by kcdakrt
DLrocket89 makes my ug experience better!


Member of the official GB&C "Who to Listen to" list

Kit Amp Building Tutorial
#32
Quote by DLrocket89
To SYK:

Agree on all of that. The power supply will be more "busy" yes, but the LR8 thing is so nice that I think it's worth it. Not a whole lot of trouble to use them, and the adjustability is great...
Should be fine for your preamp stages, but what are you gonna do for the EL84s?

Also, you can use a pot for your cathode resistor.
Connect the cap to the wiper for your bass control.
Meadows
Quote by Jackal58
I release my inner liberal every morning when I take a shit.
Quote by SK8RDUDE411
I wont be like those jerks who dedicate their beliefs to logic and reaosn.
Last edited by SomeoneYouKnew at Sep 5, 2009,
#33
Quote by SomeoneYouKnew

Also, you can use a pot for your cathode resistor.
Connect the cap to the wiper for your bass control.


DC on pot = scratchy tone control

EDIT: regarding power supply for EL84s...it's going to be pretty basic up front. After the screen tap, it's going to just be a 400VDC bus going to the preamp area, with a cluster of LR8s there to set voltages.

Also, I'm uncertain regarding the voltage...I may end up going to JJ 6V6s so I can run higher voltages. I have a PT sitting here that'll do about 490VDC with a GZ34 rectifier.
Quote by kcdakrt
DLrocket89 makes my ug experience better!


Member of the official GB&C "Who to Listen to" list

Kit Amp Building Tutorial
Last edited by DLrocket89 at Sep 5, 2009,
#34
Quote by DLrocket89
DC on pot = scratchy tone control
mis-applied partial truth.
has more validity when the signal is being sent to the next stage, not with a bypass cap.
if you use pots that don't suck you won't have a problem.


Quote by DLrocket89
EDIT: regarding power supply for EL84s...it's going to be pretty basic up front. After the screen tap, it's going to just be a 400VDC bus going to the preamp area, with a cluster of LR8s there to set voltages.

Also, I'm uncertain regarding the voltage...I may end up going to JJ 6V6s so I can run higher voltages. I have a PT sitting here that'll do about 490VDC with a GZ34 rectifier.
best to go with 6V6s, regardless of JJs @ 490v or other @ 400v.
EL84s won't be very happy @ 400v.
Meadows
Quote by Jackal58
I release my inner liberal every morning when I take a shit.
Quote by SK8RDUDE411
I wont be like those jerks who dedicate their beliefs to logic and reaosn.
#35
General update on where this is going, seeing as how the tone stack is broken up, we now have:

parallel 12AX7 section (with bass control) > bootstraped 12AU7 CF > mid notch filter > 12AU7 gain stage > 12AX7 gain stage > 12AX7 cathodyne

The last set of 12AX7 GS > 12AX7 cathodyne may end up being a 12AU7, we'll see what happens...
Quote by kcdakrt
DLrocket89 makes my ug experience better!


Member of the official GB&C "Who to Listen to" list

Kit Amp Building Tutorial
#36
and the treble control goes ... where?
Meadows
Quote by Jackal58
I release my inner liberal every morning when I take a shit.
Quote by SK8RDUDE411
I wont be like those jerks who dedicate their beliefs to logic and reaosn.
#37
Quote by SomeoneYouKnew
mis-applied partial truth.
has more validity when the signal is being sent to the next stage, not with a bypass cap.
if you use pots that don't suck you won't have a problem.


best to go with 6V6s, regardless of JJs @ 490v or other @ 400v.
EL84s won't be very happy @ 400v.



I'm going to be using crappy Alpha pots that I have laying around (unless you want to subsidize?). So...yeah...going the route I have laid out. I also don't like sending DC in any form to the front panel. If there is a short in the tube, you can end up with HV DC in the pot. Yuck. The method I have also lends itself to easy and convenient wiring.

...and yes I know gain pots, etc have high voltage AC on them. I'm saying, there's a potential for a HV standing current through the pot, which is something that I don't like (and the pot probably won't like either).

I have another PT I could use that would give about 350V, but that would kindof constrain what I'm trying to do with the variable power supply/regulators in the preamp, so I want to start with a high voltage. JJ 6V6s sound great with a freaking high voltage on them, so why not?

Another possibility that I have is a pair of JJ 7591s that I have that I bought before I knew that 7591s have different pinouts than normal. They're supposed to be like a cross between 6V6 and 6L6...like, tone of a 6V6 but power handling of a 6L6. That's all hearsay from another website, so I dunno though. worth a shot maybe?
Quote by kcdakrt
DLrocket89 makes my ug experience better!


Member of the official GB&C "Who to Listen to" list

Kit Amp Building Tutorial
#38
Quote by SomeoneYouKnew
and the treble control goes ... where?


after the phase inverter, as a circuit between the two signals...two caps with a variable resistor in between. Works to cancel out the high frequencies, by canceling due to the phase difference between them. sorry for the crap explanation, I'm getting tired.
Quote by kcdakrt
DLrocket89 makes my ug experience better!


Member of the official GB&C "Who to Listen to" list

Kit Amp Building Tutorial
#39
Quote by DLrocket89
I have another PT I could use that would give about 350V, but that would kindof constrain what I'm trying to do with the variable power supply/regulators in the preamp, so I want to start with a high voltage. JJ 6V6s sound great with a freaking high voltage on them, so why not?
350v is still a bit high for EL84s. 6V6, regardless.

Quote by DLrocket89
Another possibility that I have is a pair of JJ 7591s that I have that I bought before I knew that 7591s have different pinouts than normal. They're supposed to be like a cross between 6V6 and 6L6...like, tone of a 6V6 but power handling of a 6L6. That's all hearsay from another website, so I dunno though. worth a shot maybe?
It's a remake of an old tube that was used in a few Gibsons and several Ampeg models. Most notably the Gemini. Great sounding amp. Probably more due to the fact it had a Bax rather than a Fender/Marshall/Vox ToneSuck.

Dunno what the remake is capable of. The originals could do about 40w for a pair, no more.

Quote by DLrocket89
after the phase inverter, as a circuit between the two signals...two caps with a variable resistor in between. Works to cancel out the high frequencies, by canceling due to the phase difference between them. sorry for the crap explanation, I'm getting tired.
np. I read you.
Meadows
Quote by Jackal58
I release my inner liberal every morning when I take a shit.
Quote by SK8RDUDE411
I wont be like those jerks who dedicate their beliefs to logic and reaosn.
#40
Quote by SomeoneYouKnew


It's a remake of an old tube that was used in a few Gibsons and several Ampeg models. Most notably the Gemini. Great sounding amp. Probably more due to the fact it had a Bax rather than a Fender/Marshall/Vox ToneSuck.

Dunno what the remake is capable of. The originals could do about 40w for a pair, no more.



I was reading an article somewhere about the JJ 7591 and they said it was like a beefier 6V6. Which, if it's beefier than a JJ 6V6....daaaaamn. I have a cleanish amp that I built that has a pair of JJ 6V6s running UL with 490V on the plates (that'd be the transformer I'd use for this build, going to gut it) with the idle dissipation at about 120% of max, and they're fine with it, zero replate or screen nuking going on. So...my question is, is JJ using the same plate structure for the two saving $ and the max dissipation of the 7591 is the same as what their 6V6 is really, or is the 7591 similarly overbuilt?

something to think about for the time being. Might go with the 7591 just to be different. And they are supposed to sound good...
Quote by kcdakrt
DLrocket89 makes my ug experience better!


Member of the official GB&C "Who to Listen to" list

Kit Amp Building Tutorial
Page 1 of 4