1987/2204 50W Combo Build

[herbvis]

[mnegrete]

So, with a 47k resistor coming off the bias pot I was only able to get the cathode current up to around 25mA. I had to swap out the 47k resistor with a 33k resistor to get the current up to 35mA.

What's the reason that the value in the original circuit did't get the current where it needs to be? Is it the tubes, or the B+ voltage? or something else?

[danman]

All power tubes are slightly different in the way they bias up. You may see as much as a 25ma variance between different pairs, so sometimes it is necessary to adjust the bias voltage supply to get them to bias properly. I have had to swap resistors myself to get different sets to bias properly.

[shredgd]

Hi everyone,

this is my first post on the forum! I've completed my first tube amp build and I wanted to share my experience and my mods!
It started out as a stock 1987 Plexi lead, then ended up as a 1987/2204(JCM800) combo, that's why I'm resuming this old thread. I didn't use any existing schematic/layout, but ended up with my own, because I wanted to keep a stock Plexi front panel look and I also wanted to keep the interaction of the normal channel volume pot when in Plexi mode (the SMD schematic doesn't allow that).
Also, my layout gives me an additional third mode, which is a higher gain JCM800 mode with a 2.7k||.68uF cathode in place of the 10k cold clipper.

Straight to the mods:
- Having the choice of both a 0.1uF and a 0.68uF presence cap (switchable) is probably the single coolest mod for a Plexi style amp, so I highly recommend it. Using a pushpull pot at the bass pot spot is the perfect place to have it, if you don't want to drill your front panel and you don't want to order an expensive custom 5k pushpull pot to have it right at the presence pot spot. I actually used a switchable 0.47u cap paralleling the existing 0.1u cap (close enough).
- Another "must have" mod, in my opinion, is a switchable 2200p cap (vs the standard 470p) at the first treble peaker of the 2204/JCM800. I just couldn't decide which of the two to keep, so I had to have them both. I took advantage of the unused half of the bass pushpull pot (see above) to have both the .68u presence cap and 2200p cap active when in pulled mode... instant Slash mode! (See the Ceriatone AFD schematic). I actually used a 1n cap paralleling the 470p fixed cap... close enough.
- Having the choice of both a 2.2nF and a 22nF coupling cap for the bright channel in Plexi mode is a nice addition to choose between two different flavors of tone. In my case it was an happy accident because I had to have it switchable when shifting from Plexi to 2204-mode.
- If you have to switch a ground-connected resistor or cap in and out of the circuit, always remember you might take advantage of the input jacks inherent switching capabilities at the sleeve contacts, without adding any supplemental switch to your layout! Of course, you have to be sure you will never need to switch that component in while you're plugged into that channel input.
- I used to hate Bright caps, but if that pot is a gain pot and is usually set past halfway, then a bright cap might transform an ok tone to a great tone. If you think you're getting too much treble, increase its value (yes, increase), so that the cap will pass more mids and balance your tone.
- I left the Plexi bright channel 4.7nF bright cap off as I really like that channel clean tone at low volume settings without it.
- a Plexi style gain stage with volume at about 3 o'clock (equivalent to a 470k/470k divider, more or less, as we're talking about log pots) is probably the highest gain you would ever want from a tube stage. Trying to squeeze as much gain as possible into a couple gain stages doesn't really work well in a tube amp, you really have to spread gain over multiple stages if you want a high gain tone without any boost pedal.
- I didn't like the V2a .68uF cathode cap in almost any of my experiments, so I left it off. The 2204/JCM800 circuit doesn't use it, the Marshall 1987X reissue does the same, and Bogner Ecstasy does it also, so it probably is quite a shared opinion. However, making it switchable in and out of the circuit might be a cool mod.
- A bigger V2a cathode cap (25uF or more) saturates the PI tube too much if you use a PPIMV, in my opinion. Again, a three way switch to choose between .68u, 25u or no cap might be a cool feature, if you don't mind drilling your front panel for a toggle switch.


The whole story:

First of all, I used a Modulus amplification kit, with a very positive experience.

I wanted a "Plexi looking" head, without any additional front drilling, but I also wanted to get some hot rodded tones out of it... so I went for the push-pull pots route! I first put two of them, one for each volume pot, then I added a third one at the bass pot spot (which is a 1M log pot as well), because I needed more options.

At first I tried a simple cascaded style mod (normal channel into bright channel) activated by the normal channel volume pushpull, with a pre-PI master volume. However, I believe the Plexi normal channel is too bass-loaded early in the circuit and I didn't like the result... so I tried to mimick a Friedman style preamp (BE-100), but that amps really take a lot of bass out in the first stage to add it later in the power amp with a resonance circuit. They also shave treble at various spots in the preamp... too many switches needed to trasform a Plexi into a BE-100, some of the treble cutting caps are also across plate resistors, something I didn't want to put on a switch close to the front panel.
So I decided to try some Ceriatone style hot rodded preamps circuit, which are more similar to cascaded Plexi channels. However I couldn't get all the bright caps combinations those amps provide with my pushpull volume pots, so I still wasn't 100% satisfied with the tone (most of them also include a Jose style master volume, which didn't convince me tone-wise in the beginning and soon got off the circuit).
After studying the Bogner Ecstasy preamp, I tried to invert the cascading (bright channel into normal, which by that time got a standard 2.7k||.68u cathode combo), but I couldn't get rid of an oscillation at gain levels over half the gain pots, so I couldn't even really fully test it. That was probably due to layout, because I could cascade the normal into the bright channel and max the volume knobs without any oscillation.
In the end, I decided to give the 2204/JCM800 circuit a try and... I fell in love with it!

I used the normal channel as the first stage to avoid any oscillation... also, compared to the Plexi bright channel, the 470k||470p treble peaker is before the gain pot rather than after it, and it has a 100p cap across anode and cathode, so using the normal channel made me able to avoid major permanent alterations to the bright channel. However, I still had the problem of the cold clipper transformation of the bright channel while in "2204 mode". I solved it by taking advantage of the input jacks inherent "hidden" switching capabilities! So I hard wired a 10k resistor at V1b cathode, and wired a 3.3k||.68uf combo to the switching sleeve contact of the normal channel low input. This way, the V1b cathode sees 10k when I plug my guitar cable into the low normal input (2204-mode), while it sees 10k||3.3k||.68u when I'm plugged into any of the bright channel inputs, which is equivalent to 2.5k||.68u, close enough to the standard bright channel cathode combo... pretty cool, isn't it?
I decided to choose the low input for the 2204-mode just because it has a 68k grid stopper, as in the original 2204/JCM800 circuit... however I later realized that this caused a 68k input to ground (grid leak) with the standard plexi inputs wiring, which obviously lowered input gain a lot, so I did a little modification of the inputs wiring to have the 1M resistor to ground for the lo input rather than for the hi normal channel input.
An happy accident of this "poor man's" switching is that if I plug into the hi input of the normal channel, then I'm getting a 2204 circuit with a higher gain 2.5k||.68u combo at the second stage cathode. So my amp is actually 3 in 1! This is becoming my favorite mode! Having the bright channel volume still active in this mode (I simply max it out when in stock 2204 mode, so it is virtually out of the circuit) allows me to dampen the high total gain of this cascaded configuration, which can be too much. So whenever I'm using this high gain hot rodded-JCM800 mode, I simply lower the bright channel volume pot to 3 o'clock, still using the normal channel volume pot as the gain pot. With or without the "Slash mode" (first treble peaker and presence cap switch engaged, see above), this circuit delivers a musical and versatile high gain tone!

If you do permanent mods to the normal channel of a Plexi and they involve treble peakers/bright caps as the 2204 first stage, as in my case, note that there is one drawback: setting the volume I pot above 0 when plugging into the stock bright channel adds hiss to the tone. The "raise volume I trick" while using the bright channel is a cool way to add some clean bass to your tone, as that pot wiper is wired to the bright channel 470k grid leak resistor, and setting it above 0 raises that 470k value by adding series resistance. At the same time, however, that resistor is the grid stopper of the normal channel, so raising that volume can introduce hiss into the circuit if bright caps are permanently bypassing resistance.
To avoid this annoying drawback (I really like to set volume I to 3-4 when using the bright channel), I wired a 2.2nF cap to ground after the normal channel 22n coupling cap, making it active only when using the bright channel. This shunts treble frequencies, included hiss, to ground. I actually used the bright channel volume pushpull to do this (one half of the DPDT), while the other half is used to raise the 2.2nF coupling cap of the bright channel to 22nF when in 2204-mode. I still can use that pushpull with my bright channel and experiment with the 22n coupling cap, as that bigger cap adds bass, so I would set the normal channel volume to 0 anyway, shunting hiss to ground in another way.

My kit luckily had two extra chassis holes in the back so I only had to drill the holes in the backplate to install a simple "insert style" effects loop. This is very cool to have because it allows you to troubleshoot problems or even just noise in the preamp or power amp section separately, and it is also the perfect place for a pre-PI master volume box. I used a short patch cable modified with a 1M log pot at its half, so now I have my "master volume cable"! I then installed a Lar-Mar PPIMV, so now I have the choice of a pre-PI MV, PPIMV or both! I really like the transparency of the pre-PI MV, which works especially well with the 2204 and the modded-JCM800 circuits, but I also like the added gain of a pushed PI tube when in Plexi mode that you can achieve with a PPIMV, although the reduced negative feedback at low settings is quite annoying... combining the two is probably the perfect compromise!

I hope this information will be helpful for some of you!

Giulio