High Gain 2204 preamp

[Burnt22]

Hello all, new to the forum! I've recently been building my own custom all tube preamps that I run into the effects loop of a DSL 40C. The first I built was based a JMP 2204 (I used mustard caps, Vishay CMF60 metal film resistors, Vishay 470pf in the treble peaker spots and EQ with a 33k slope resistor) so to me it sounds like a typical 2204 (bright, not kverly gainy and cuts through well).

However, I decided I wanted something similar to a 2204 with more gain, little more bottom end but still retain the 2204's ability to cut through. I used Mallory 150's (0.022uf's throughout, Vishay CMF60 MF resistors, 2200Pf treble peaker going to the gain pot, 470pF on 470k/470k divider) but I added and 4th gain stage by adding a 12ax7 after the first 10k cold clipper. 220k/220k voltage divider, a 100k plate load, .022uF coupling cap, no treble peaker, 10k cathode before the original (3rd stage in original 2204). Also, for reference this 2nd build uses a 36k slope resistor, 560pF treble cap, and typical Marshall .022uF mid and bass caps. All Pots are typical 2204 value (1M gain, 220k Treble, 22k Mid, 1M Bass, 1M Master volume)

The first issue I ran into was some high frequency oscillation which I added a 470pF on V1A and a 220pF on V1B to fix. I usually run the master on 3-4 (I made another voltage divider using the master volume 470k/1M Gain pot to cut volume) before the effects loop return of my DSL.

However, the issue I'm having is the amp is still super bright. I know 220k plate load on V1A increased the gain at 10khZ from the stock marshall of 35.5dB to around 37dB. But, it also raised gain at 60hZ from 28dB to about 39dB, so it increased both the highs and lows. My question is would it be better to increase my cathode bypass on V1A to get more lows or increase the coupling cap at V1A from .022 to .047. to increase lows? Both would achieve my end result, as a larger cathode bypass would let more frequencies through, and a larger coupling cap would as well. However, would one way be more beneficial then the other? I know I could just try it, but I still like learning and hearing others approaches as I learn more and more.

Now, for the most important part, I do practice safety first. I discharge my filter caps (50uF, 33uF, 33uF on each preamp I build) prior to anything. Speaking of that, would increasing filter cap size also be beneficial to getting mofe lows? I know the filter caps in a 50-Watter vs 100 seem to play a role in bass.

 

[Pete Farrington]

It’s hard for me (and possibly others) to follow text descriptions of a circuit design.
Why not sketch the schematic?
Take a photo of it, upload to a hosting site, and link into a post.

My question is would it be better to increase my cathode bypass on V1A to get more lows or increase the coupling cap at V1A from .022 to .047. to increase lows? Both would achieve my end result, as a larger cathode bypass would let more frequencies through, and a larger coupling cap would as well

Online calculators are available for working out the effects of the above changes.
If you’d like links to them, please ask.
I suspect that the coupling cap change will only affect the deep subsonics, a decade below low E.

Typically, it benefits signal to noise ratio (eg reduces hiss / rumble / hum / buzz) to fully bypass early stage cathodes.

 

[FleshOnGear]

It might be worthwhile to add a shelved low pass between the cathode follower and the tone stack, like in the 5150iii. I drew up this schematic of just the tone section:

This will let you keep the lows out of the early preamp stages for tight response, but “enhance” the lows after all the gain to add beef.

 

[neikeel]

From your description I get the idea but a proper schem and pics of build would help.
As a principle you trim off (shelve) unwanted lows early in the circuit to avoid flub but preserve them in the output stage Iso use 0.047-0.1uF output couplers but trim the first stage coupler to 0.0047uF or 0.01uF.
I use the 2200pF cap over the 470k treble peaker to but only use 100-150pF cap on the volume pot.
Presume you are using shileded wire on first stage grids?
Also presume you have grid stoppers on your output valves. What NFB are you using?

 

[Burnt22]

(htts:f//ibb.co/T11cJ9m remove the () and f in the link. As I can't post links yet.

I'm using a Marshall DSL40C as the power amp and speaker. I'm running my preamps through the Effects return ( my preamp goes through the Phase inverter of the DSL, Power tubes and then output transformer to the speaker)

Hopefully that image uploads. That is the exact preamp I've built. B+ voltage is 280v.


I'm using a traditional Marshall tone stack. But with a 36k slope resistor, and 560pF treble cap. Treble pot is 250k, Bass is 1M and Mids is 50k.

It's basically a standard JCM 800 preamp, but with an additional coupling cap (.022uF) voltage divider (220k/220k) plate resistor (100k), cathode resistor (10k) and the changes to the tone stack I previously mentioned.

 

[Burnt22]

It might be worthwhile to add a shelved low pass between the cathode follower and the tone stack, like in the 5150iii. I drew up this schematic of just the tone section:

This will let you keep the lows out of the early preamp stages for tight response, but “enhance” the lows after all the gain to add beef.

I actually looked over the 5150 III schematic as you recommended, and noticed that it has a .001uF bypass cap over the 100k plate load feeding the 10k cold clipper. My fourth stage is an additional 10k cold clipper but, I have it being fed off a 220k/220k divider with a 150k plate load and .022uF coupling cap.

I wasn't having an issue with not enough bass, it was I had too much treble from the previous treble peakers 470k/2200pF and 470k/470pF. The 4th stage I added didn't have the plate bypassed at all so the treble already being bypassed by the previous stages made it treble focused. I added the .001uF on the plate of my additional 4th stage and that fixed all issues I had. Sometimes it's a simple solution!

As is, my stages currently are
V1A 100k Plate/1.8k Cathode
V1B 100k Plate/10k Cathode
V2A 150k Plate/10k Cathode
V2B 100k Plate/820-Ohm Cathode

 

[GlideOn]

Your post is quite a bit of info to digest and interpret based on reading comprehension alone without a visual circuit diagram for us 'get off our lawn' types, but I will do my best to suggest some tailoring that can at least help with the "brightness."

In order from most suble effect to greater effect, imho:

33k Slope resistor will shift the eq more towards the low mids and definitely relax the treble forward nature especially with your brighter 560pf peaker cap. 33k is more modern Marshall for sure.

You can snub the V2 100k plate, use 390pf to 500pf, even more for audible cut in the harsher treble freq.

Larger PI stability snubber - try instead 120pf to 500pf for audible cut, I believe a favorite for EVH fans to 'brown' the tone in Marshall style preamps.

Mesa, Rivera amps I've seen snub both the 82k or 100k PI with 120pf, 250pf values. I hear a slight smoothing change here, not too dramatic.

If you have the traditional JCM800 6.8k PI tail resistor, you can try 10k for more compression, more old-school Marshall value. Relaxes the punch a bit in the extreme highs, more of a era/voicing change than a treble change.

Larger grid stopper to the pin 1 power tubes help too, try up to 5k, for audible cut in the highs, slight voicing change as it tightens the power tube overdrive.

Make sure to use the higher voltage 500v caps for plates and at least 1w for the pin 5 resistors for longevity.




Conservatively speaking, just being a different place in the room away from the beaming of the center cones of the speakers can too, alone make the biggest difference. Speakers are also bright and inconsistent in their projection, so manipulating your body in relation to the amp requires no hard-wire fixes at all.


Hope this helps!

 

[TheKman76]

Maybe my eyes are going, but I could use a better pic of that schematic.

That said, I'd have thought this shouldn't be that bright. Though the plate snubbers early in the circuit will reduce noise, the lack of them later in the circuit may be allowing too much of the clipping harmonics through.

Also, what speaker is in the DSL? The V-type in my DSL40CR was way too bright and has been swapped for a cream back, now it loves the brightness I've added to the circuit.

 

[jagrenet]

Great post - This is the kind of stuff I like to read.