SE EL34 Plexi-verb

[JazzGuitarGimp]

Hi All,

In preparing the schematic for a PCB layout, I discovered another error in my original design: I had the inputs (pins 5 and 6) of U2B swapped.

While I was at it, I've added two 1N4148 diodes at the input of U1A to protect it from large incoming AC signals.

Cheers,
Lou

[xtian]

Reverb functioning yet?

I got two BTDR-3H modules and misc parts. Just waiting on my PSU today.

[Phil_S]

Patience please! I hope you understand that work and family get in the way of this on a pretty regular basis. OTOH, I suppose folks here really can't understand that.

I finally received new prototyping boards in the mail today. This week is rather packed with things to do and the weekend is totally committed to be anywhere but at the bench. Let's see if a miracle will happen to give me a window of free time before Friday. If not, it will be well into next week ;-{

Lou! I've only got 1N4007 on-hand. Are those going to be OK?

[JazzGuitarGimp]

Yes, 4007's are fine. They're overrated, but better to go over than under.

[Phil_S]

I have a wiring question. Against good advice (to be fair, received after drilling the chassis) I've got 3 pots on the front panel. One pot receives the dry signal from the 470KΩ and another pot sends the wet signal to the other 470KΩ, both of which are on the turret board. I believe both need shielded wire. The connections are in close proximity to each other.

Can I use shielded twisted pair? This would solve a couple of practical problems. I had used single RG-59U before, and it is tough stuff to use. I is springy and the core is both small and hard to solder well, especially in a tight space. Also, the twisted pair will leave me with only one shield to ground instead of two.

In case you are wondering, the 3rd pot is the double 10KB depth pot which is connected to a 6 conductor shielded cable. No questions about that one ;-}

I'm wanting to wire the pots before I install them in the front panel because working space in the amp is tight.

Thanks.

[martin manning]

I've got 3 pots on the front panel. One pot receives the dry signal from the 470KΩ and another pot sends the wet signal to the other 470KΩ, both of which are on the turret board. I believe both need shielded wire. The connections are in close proximity to each other. Can I use shielded twisted pair?

I don't see any problem there.

[Phil_S]

Thanks Marin. That's one less complication. Meanwhile, I really have to give some thought to the brick board. In going over it I see that I've conflated the VGND and 15V rails ;-( Arghhh! It's going to be a while 'til I can get it reworked. All that tiny stuff is a PITA.

[Phil_S]

OK...I'm getting closer to the finish line. The amp is still on the bench. I haven't had the courage to plug in the BDTR brick. Everything seems to be working OK. Voltages are reasonable. I moved the ground connection for the reverb circuit and I believe I solved the hum problem, but I won't really know until I plug in a proper speaker cab and guitar. I've learned the bench speaker isn't always a good indicator.

There is one more thing I'd like to get done. I'm not sure it's worth doing. I would like to see preamp voltages closer to the typical Plexi. That would be 160-180V at the plates of V1, and 250V on the plate of the CF tube. Given what I'm working with, 250V is not a reasonable expectation, but 220V might be. I can try to tweak the B+ ladder, but I'm basically capped at the 280V screen node for the EL34.

Here are the questions.

For the cathode follower, is this more about the voltage differential Va-Vk or is there something about the absolute voltage that contributes to the Marshall tone? Presently the stage before the CF is sending 81V to the grid, plate voltage is 176, and cathode voltage is 80. Va-Vk = 96. The published chart for the 50W plexi makes this, Va-Vk, 250-140=110. I'm rambling...I think you get what I'm driving at. Is this worth pursuing?

If I fiddle with the B+ ladder, all that I have on-hand that fits the bill are 1K 4W resistors. Is 1K adequate for decoupling between stages? (Using 2 in series will be ugly but possible, for 2K.)

Finally, at the first tube, plates are 99 and 120. Will I get more clean headroom (something I'd like to have) by increasing the voltage at the supply rail? I am thinking I can squeeze out another 20-30 volts at most. Is this worth going after?

Thanks.

[JazzGuitarGimp]

Sorry Phil, I'm not a Marshall devotee. Hopefully someone will chime in and answer your questions.

[martin manning]

Hi Phil, no reason not to try a lower dropper before your PI node to increase preamp voltages. I'd try a 5k there to hear what happens, and you can get that by paralleling another 10k if you have one. From your schematic you are dropping over 50V with 10k, so that should buy you another 25V. Decoupling should still be adequate... 1/(2PiRC)=1.4Hz

On the CF, it's about the bias point. For lower voltages try decreasing the Rk. I'd put another 100k in parallel with the one that's there to see what 50k sounds like. If you like it, use a 56k 1W.

[Phil_S]

Decoupling should still be adequate... 1/(2PiRC)=1.4Hz

Thanks for the help, Martin. I have tried to research this but find that I am out of my depth.

What frequency do you think is too high? With 2K it is 3.6Hz and 1K it doubles to 7.2Hz.

Do I understand correctly, the problem is to decouple sufficiently to avoid motor boating, which is really a form of oscillation?

All of these frequencies are well below the audible spectrum which is supposed to start at 20Hz.

[John_P_WI]

Phil,

As a rule of thumb, I always shoot for 10 times lower than the lowest note on a guitar to be amplified, which would approximately be 8 hz for an open e or 7 hz for drop d.

Remember that there will be an octave lower harmonic around 40 and 35 hz in respect to the open e and d mentioned above.

[Phil_S]

OK then! I ended up with this B+ ladder: FW SS rectifier > Choke > 150Ω > 5K > 1.65K (as metered) > 2K > 10K. 1.65K and 2K were previously 10K. Voltages are pretty good now in the preamp. Not at Marshall numbers, but not so far off. I'll post a revised schematic with changes and voltages after a while.

Took the amp upstairs to plug it into a real speaker cab and a guitar. Hum is negligible (yay!) and it has wide tonal pallet. The difference between the dark channel and the bright channel is considerable and you can have both of them on at the same time. There are too many knobs on the front and I kept forgetting what's what. I made myself a map for next time. It's a good amp. I did not play it long enough to really evaluate it; life keeps getting in the way. I was surprised at how much clean headroom there is and there is a decent amount of volume.

One disappointment is the reverb. I've obviously got something wrong. The signal is only the slightest bit wet. Twiddling all the knobs doesn't change that very much. It will take some effort to get that sorted. I'm not sure when that will happen, but I'll keep at it.

Many thanks for all the help I got along the way. It was considerable and much appreciated. I broke a lot of new ground. I'm very please with the result overall. If the reverb never works (it will), this will still be an excellent amp and different from others I've built.

[Theashe]

If you're not already near rail-to-rail on the output of the recovery op-amp, you could try increasing that 10k resistor or reducing the 39k resistor to get more gain from it.

Failing that, I'm not sure where you mixed your reverb back into the signal path (you might have said, but I've forgotten), but you could try connecting it somewhere earlier in the amp where the guitar signal is smaller and take advantage of a triode for gain.

I tell you, when I attempt this in my next amp in a few weeks, I'm looking forward to a "depth" knob. The spring reverb in my last one was 16" and long decay. The difference between "I can't hear the reverb" and "wow that's a lot of reverb" is like 10 degrees of knob turn. I'm ordering a smaller medium decay tank with the BTDR-3 I get.

[JazzGuitarGimp]

Hi Phil,

Do you have a composite schematic of the amp with reverb as it's now wired? If so, point me to it and I will see if I can help with the low reverb issue.

Cheers,
Lou