Separate names with a comma.
Discussion in 'Shock Brother's DIY Amps' started by voskarp, Feb 15, 2020.
It's a bigger cab, so the reverb might fit. You could always use quite small springs.
Errors corrected in Layout:
The 330ohm B+ dropping resistors look too big. If you're aiming for the 270V B+ of your original AC4 drawing then I think you'd need about half that, maybe 120ohm. (assuming about a 50ma draw from those three tubes). The dissipation would be around 1w per resistor so 5w is a bit overkill but whatever, never hurts to be safe. Especially if you end up dropping more than you wanted, that would up the dissipation a bit.
edit: To be more specific - this is tricky to estimate because the transformer B+ winding has resistance of it's own that your dropping resistors will be adding to, and the current is in the form of sharp spikes so resistance here drops more voltage than you'd expect. I'm guessing the transformer's B+ winding has at least 100ohms per side, so 120ohm external resistors would bring the total up to 220ohms per side.
Thanks for your input. I’ve got some different values for those resistors, and my first guestimate was 220R. I figure I just have to build the thing and try some ‘till I get it right.
Board almost ready for soldering. Ruined some eyelets in the process of seating them, so I have to order some more...
The board pretty much fill up the backside of the chassis (it's a generic 5e3 chassis that I have just cut down to 290 mm to fit the cab).
Awesome. So glad to see someone building something that is not a Fender amp. I love Vox amps but I can never make sense of their circuits. They seem to use 10 components to do what Fender does with five, if that makes sense.
Yes, my version is a bit too much with the second AC4TV preamp and some added stuff to the EF86 channel. The original AC4 circuit is quite simple though.
I love the Tweed tone, but for a more crystal clear clean tone I prefer VOXes over Fender BF and SF. It's probably something with the low mids that does it for me.
Will do some soldering this evening, maybe I'll get to finish it.
Put the amp together yesterday evening:
So now I just have to get some blue and/or cream Tolex...
It sounds great, different from the Top Boost circuit that was in it before. The switch selecting different caps for the EF86 screen grid works fine, cutting the bass just enough when switched to the smaller cap.
The different channels have a distinctly different sound, but the problem is that the first channel (EF86) has way less gain than the second (ECC83), maybe something like 10dB less output when maxed. And with my Tele bridge pickup (7,1k A5) the first channel won't go into overdrive even when dimed, but the second channel starts overdriving with the volume about halfway up.
When I hit the first channel with a clean boost pedal I get the most lovely, aggressive and biting overdrive out of it, so it can get there, just too little gain from start.
The EF86 has an amplification factor of 185, and the ECC83 at 100, for each side... so the difference is not very strange perhaps.
Here are some voltage measurements (voltages from original/other schematics in parenthesis):
An idea I've had is to remove the R1 resistor to increase the gain, maybe that would do some difference?
Another idea is lowering the value of the R4 and even more so the R5 resistor...
Removing R1 doesn't buy you much, 2%. The biggest loss is the 220k mixing resistor going straight into a 220k grid leak and a 440k to ac ground. You're throwing away more than 50% of the EF86's signal there. Raising the voltages overall will increase the EF86 gain a little, but it still wouldn't be enough to overdrive the power tube with the guitar straight in with that mixer loss (it would be without the loss). All the same, I'd try your other dropping resistor values now to get the voltages higher and maybe try reducing the EF86 screen resistor. 27V is a very low screen voltage. I wonder why they used a screen resistor so much larger than the book values?
Maybe a channel switch so you can do away with the 220k? Or you could move the mix point to the second stage of the ax7?
Thanks for the input.
I've thought about losing the R28 mixing resistor, and by that also making the P2, channel 1 volume effectively a master volume for channel 2 (and P1 a "master" top cut). It wouldn't be that bad really. Some people actually prefer having a master volume.
I think reducing R5 (the AC15 has a 1M resistor here, and the same 220k for the anode) would be a good thing too.
I think I'll keep the R29, since that second channel dimed really makes the output tube boil as it is.
Maybe I should also change the primary reducing resistors on the ac side to 150Ω, and the El84 screen resistor to 1k.
(As a note: the primary reducing resistors on the ac side are actually 220Ω, not 330Ω as in the schematic.)
1meg is a lot more typical for the ef86 screen, sounds good. Did you get some 120ohm resistors? That was my earlier estimate but you may need even less if this is the voltages with 220.
You know you could actually try just try bypassing R28. Won't hurt anything and you'd get a lot of gain back. Only downside I can see is it would make P2 control the level of the ax7 channel a little bit, kinda like a 5e3, but that might actually be useful if that channel is a fire breather?
I'll try bypassing R28 and changing the screen resistor to start with.
What would be the harm of removing R1? (Even though it won't do very much to increase the gain. I think of it as having the guitar volume rolled down 1/10, does a little difference anyway.)
Went back to the drawing board and decided to go with this advice, an alternative that has been in the back of my mind from start.
But I wanted to try making channel 1 as close to the original as possible. After trying that, and realising that the difference in gain between the channels was too much, I think that this will be much better.
It will make channel 1 more like the AC15 (with the V2b being one half of the phase inverter), as the second channel is much like the normal channel in an AC30.
I also made some other changes and corrections.
As a side note, why this arrangement with the caps and resistors?
Yeah good question. One thing for sure is it's going to set the grid of that stage a couple volts above ground, which will need to be considered when mixing there. Possible solutions are a second blocking cap after the ef86 volume (which is how they handled it in the ax7 channel) or maybe figuring out why they did this and if you can just put it back to normal.
Ouch! Of course it wouldn’t go that easy...
So I either remove R12 and C8 (just bypass them) or ad a cap between R28 and and P2 (or maybe the cap should be after the R28/29 resistors?).
The easiest would be to ad the cap, and if I choose a 47n, like the one before P2 it wouldn't change the tone, right?
I simmed that weird cathode tail in spice and still can't see any purpose it could have. Same frequency and phase response as the 1k5 10u pair has by itself.
Putting a second resistor below that pair is commonly done to provide a tie in point for negative feedback from the speaker (like a blackface champ for example) but this model has no nfb. And the grid is referenced above the 3K9, not to ground, so the tube doesn't care what's happening below that anyways. Very odd.
I'd try simply bypassing that 3k9 with a little bit of wire and playing it. It should sound the same as before and then you don't have to worry about extra caps or anything, your drawing above would work fine as is.
I will try that, thanks for all the help!