Lower b+ plexi build

[Lindz]

I've read posts here over the past few years of people praising the virtues of a lower b+ plexi - Reeltarded, Jana and others come to mind

I've started a 50 watt build where I was thinking I might play around with this a bit. I am thinking low voltage b+, low filtering, dual switchable ppimv's and a few other popular tweaks. Make a chewy EVHish build with switchable masters so I can have a solo boost sort of thing. Not thinking 90v Variac wall voltage as much just lower b+ and low filtering internally ala Mod 5, Bray and other "brownish" mods. 32uf main filtering, 16uf pi, 32uf v1, v2 preamp is what I am thinking - Mod 5 esque, Bray switchable masters.

Knobs right to left in photo are gain/traditional master in 2 unused input jacks, treble, mid, bass, presence, dual ppimv's, resonance. 3 position power switch off/standby/on and pilot light in normal standby switch hole. Anyone see any potential issues with resonance or PPIMV's that close to PT - noise or oscillation? I will shield where necessary and could shuffle positions if it would be less problematic. I am still debating whether to use relay board or just MacGyver it and epoxy relay to PPIMV ala SDM's dual masters layout from Metro - I think that would be pretty clean, keep relay wire run reasonably short and keep coil away from tone stack etc. Bray does this with his builds I have seen but rear mounts the PPIMV's by v2 and does not even shield the leads, he twists and his amps are pretty quiet. comments?

I am working with a JTM 45 chassis with laydown pt cutout

The Classictone 40-18054 laydown transformer has high and low voltage taps ~450v and ~412v ss recitified - Was thinking I could use a zener string to drop the 412v further to say 380 as Jana/Reel suggest in some posts and I may make b+ switchable so I can a/b/c the tone changes

my first question is this and I know this is subjective - what difference should I expect sonically as I drop voltage? How do people bias power section with lower b+? hot, warm, cool? and why

For preamp voltages do you adjust the dropping string to compensate or leave values as is in typical layouts and drop the preamp voltages proportionally? What are people's favorite voltages in v1, v2 and v3 and why?

What are your favorite new production power and preamp tubes in this sort of build?

I have searched and read lots and know this is a very broad topic likely to have differing opinions so I thought I would start a thread to see who might chime in.


thanks!

[Jana]

I would keep things simple and just use the taps to get the 412 volts. It's a bit higher than 380 but it is still in the sweet range. I wouldn't even bother with the higher voltage taps.

I just use 20K, 10K, 10K for the dropping string of the B+ (the 20K is the node from the screens to the phase inverter).

I bias to about 65 to 70%.

New tubes?--let me know if you find any good ones! I'm fortunate to have a stash of older tubes that will last me the rest of my life (given that I hardly play anymore).

[Reeltarded]

My built amps are 380 and 417. BEST ever.

There are no acceptable preamp tubes, but the EH 6CA7s are killer.

[David Root]

This is a timely thread for me as I'm starting on a low voltage version of Steve Luckey's Dual 50, which is a Plexi with a parallel footswitched 2204 pre.

Jana's low voltage build is what first got me into it so I dug around a bit to learn more and found the Dual 50.

Driving this is I have an older Mercury 4550JT-G4 PT which is a 380VDC nominal based on 117VAC. So maybe 390V on 120VAC.

I too was wondering about EL34 bias; so 65% Pd is about 43mA per tube. quiescent. Using Jana's 20K on the PI dropper would take the preamp V1 plate voltages down well below 150V I think. Not starvation level but getting close. Please correct me if I'm all wet on this, or is that the general idea?

I looked up the old voltage charts page and the older 50W chart (EL34s, 390 B+1) shows B+ on the CF driver/CF as 250V. So V1 Plexi would be ~230V or so and V1 2204 would be ~235 or so. Is this where you went, Jana? I would be feeding the two first stage nodes in parallel off the CF driver/CF node with 10K dropper on each.

If I put in 8K2 or maybe 7K5 in as the PI dropper I get about 320V B+ on the PI. Then the subsequent 10Ks work out. That B+ puts the PI plates in the 210/220V region indicated in the older voltage chart.

Miles, thanx for the tip on the EH6CA7. I have a pair of them, some years old now but hardly used. Did you ever try GE6CA7?

[martin manning]

Nice voltage chart David, thanks for posting. I have a 50W built to Jana's schematic, and it's the go-to amp around my house. I've been curious about the EH 6CA7, so I just ordered a pair to try them out. It'll take a week or so, but I'll report back.

[fusionbear]

I built an amp like this for a client. I definitely did not like the 20K PI drop. I also settled for 8K2. Made the amp sound clearer and richer. Using the 20k made the amp sound too mushy to my ears....

[David Root]

Thanx fb. I was just trying to match the voltage chart. What B+ and plate voltages did you get on the PI?

Also I forgot to bring up the question of filter caps. Steve Luckey's Dual 50 is set up with 100uF on the plates and 50 on each following node, which is 350uF total, way too high for a 50W I think. Even though he designed for higher voltages I think.

I came up with 40 plates, 20 screen, 40 PI, 20, 20 20 on the preamp. I discussed this with Dave Cohrs (Colossal), who has built lots of Marshall circuits and he thought that would work fine at these lower voltages. I want to do it in two can caps, so a JJ 40-20-20-20 and an AES Mallory 20-20-20 fit that bill nicely and the grounding separation will be OK. (I am using a 100W chassis so the two spare power tube holes are for the cans).

[fusionbear]

Thanx fb. I was just trying to match the voltage chart. What B+ and plate voltages did you get on the PI?

Also I forgot to bring up the question of filter caps. Steve Luckey's Dual 50 is set up with 100uF on the plates and 50 on each following node, which is 350uF total, way too high for a 50W I think. Even though he designed for higher voltages I think.

I came up with 40 plates, 20 screen, 40 PI, 20, 20 20 on the preamp. I discussed this with Dave Cohrs (Colossal), who has built lots of Marshall circuits and he thought that would work fine at these lower voltages. I want to do it in two can caps, so a JJ 40-20-20-20 and an AES Mallory 20-20-20 fit that bill nicely and the grounding separation will be OK. (I am using a 100W chassis so the two spare power tube holes are for the cans).

I did what you did. Adjusted the dropping resistor until I got it within 5% of the published chart. I left the standard filtering which is 100uf, 50uf, 50uf (PI), 50uf (CF), 50uf. But, at the lower voltages may be OK. Although, by adjusting the resistor, you are bringing the voltages to the proper spec. which use the 50uf (PI), 50uf (CF), 50uf (preamp) on the published schematics... experiment until you are happy...

[Jana]

My voltages for the B+ nodes are:

Plates = 375
Screens = 372
PI = 273
V2 = 241
V1 = 232

[David Root]

Thanx Jana. Pretty close to the chart, except the PI is lower of course with that 20K.

So then the PI plates would be around 180V or so, given your 820/15K and 100K grid bias, assuming around 45V on the cathodes. And a bit better balance than the standard Marshall cathode/tail resistors, but lower headroom and gain.

How does your PI differ tonally/feelwise from the standard PI?

[Jana]

There isn't any difference in the overall power output of the amp--it still puts out 35 to 40 watts clean (which is what all the Marshall "50 watt" amps with these voltages put out.

Feel and tone are highly subjective.
My desire/preference/goal/taste/whatever you want to call it, is a solid sounding amp. I like it to transition from the clean to dirty sound smoothly and to be controllable with the volume pot of the guitar (that is my "channel switching"). I don't like a ratty, fizzy sound. I want the sound, clean or distorted, to remain musical. I am very opinionated and biased (70%, lol).

One of my very strong opinions is that the phase inverter is an integral piece of the "tone" equation. It is the distortion in the phase inverter, again my opinion, that is the distortion tone that most are searching for when they are instead trying to find it in power tubes, preamp arrangements, etc. All those other pieces certainly contribute to it, however.

Having said all that, I feel that it is important that the PI clip before the power tubes go into saturation (meaning the power tube grids are never driven past 0 volts or into the positive region).

I like 100K grid resistors for the power tubes--I like to keep the power tubes on a short leash.

[David Root]

Thanx for your explanation, Jana.
I agree about the PI's importance, that's why I asked the question.

Interestingly, the 100K grid resistors do load the PI more than the 220Ks as well as going easier on the power tubes, and in "another world" ie Dumble circuits, I have found that my favorite Dumble is the 2nd/3rd generation, which also uses 100K power tube grid resistors and is I think the only generation that does.

Somehow I don't think that is a coincidence.

[Lindz]

Great comments everyone.. made more progress soldering yesterday

2 dumb questions

1. as I build my head shell - the small box plexi plans I see floating around show 1/2" birch ply but I see other comments saying 5/8". 1/2" seems thin in terms of the corner rounding etc, not much material left once you round them off. I've not looked closely enough at a vintage cab to tell what the thickness is but I want to do it period correct (reasonably)

2. I have a hoffman 5v power supply board I was thinking I would use for powering the 1 relay for switching the dual ppimv's. http://el34world.com/projects/relay_switch.htm

Its designed to run off of 6.3 vac

I am reluctant to tap off the 6.3 heater supply I am powering the tubes from fearing noise.

I have the unused 5v tap on my pt.. will that generate enough voltage using this board to switch one relay? Would be simple solution but underpowered a bit vs spec. Using bridge rectifier in board, 7805 regulator and the relay I have is Omron g6a-234p 5v also sourced through Hoffman. Not sure enough about the math to know if it will switch reliably powered this way.

I suppose it would also be easy enough to add a small 6.3v transformer instead... thoughts?

[David Root]

Re your item 2, I think you might come up short using the 5V winding. If you changed the rectifier to four Schottky diodes, which have a much smaller forward voltage drop, you might make it.

I hate to say it, but unless the PS rectifier you have now uses Schottky diodes, the better solution for 5V power might be one of Colossal's 5V relay power supply boards; I use them with his relay board which comes with a G5V-2-H relay and I get 4.99+V every time. colossalamps.com. Dave is a TAG member too. I have no business relationship with him other than as a purchaser. His board uses Schottky diodes.

I use the 5V winding with them too. No problem. The PS board cost is about the same as a 6.3V PT and of course takes up less space overall since you don't have to find room for the 6.3V PT.

[Lindz]

Thanks David - I kind of figured I might come up short that way. I saw posts about the Colossal boards before and bookmarked the page on his site but have the Hoffman on hand. I will be picking some up for future projects and support the local talent

But for this one I'll likely drive by Frys and pick up a transformer - 5 min from my house and I can work on it this weekend that way.

Looks like it will fit pretty easily between the board and PT