any problems with increasing filter cap size on marshall 2204?

Discussion in 'Shock Brother's DIY Amps' started by owlexifry, Jul 21, 2021.

  1. owlexifry

    owlexifry TDPRI Member

    Posts:
    51
    Joined:
    Feb 10, 2021
    Location:
    Adelaide, Australia
    previously i’ve tried a few things to improve the bass response/low end from my 2204 clone kit build.
    - resonance control installed (4700pf cap).
    - phase inverter output coupling caps increased to 47nf, then 100nf (didn’t make much difference from 47)
    - changed EL34 to KT77 (very pleased with this change)
    - changed 1st bright cap (goes to gain knob) from 470pf to 2200pf (not sure if this affects bass response much, but thought i should mention it anyway)

    (also done a few other things as well, jose mv, zener diode clipping, 330k plate resistor at 1st gain stage, V2a cathode bypass cap-0.68uf added, NFB resistor changed from 100K to 39K - from 4ohm pin).

    these have been welcome and fun changes, but i’m still not satisfied with the low end response.
    especially after i picked up a peavey windsor (100w EL34 marshally knock-off thing) the other day for super cheap, and compared it to my 50w 2204/800 clone, and its bottom end response and massive tone just stomped my 50w and i found this quite unsettling.
    i understand that im comparing 50w to 100w, and of course they’ll sound different. but i just couldn’t believe how big the difference was.

    this has lead me to reading and comparing various schematics and pondering where it is my 50w clone could be lacking.
    well tonight it just occurred to me that there is a significant difference in the first filter cap stage that directly feeds the OT centre tap (50uf+50uf - 50w; 100+100 - 100w)
    it also occurred to me that my other 40W SLO-inspired build has larger filter capacitance at this stage (80+80) and has a noticeably better low end response than my 50w.
    also noticed another peavey amp i like (5150) has enormous filter capacitance at this point too (330uf).

    so this has me thinking, maybe i just need to increase the filter cap size on that first filter stage and it might just give me what i’m looking for...

    could i just change that first 50+50 filter cap can to a 100+100 cap can? would this cause any issues?
    should my 50w heyboer power transformer handle it ok?
    i’m super keen to give it a try, doesn’t really cost that much to get hold of one..
    746D86C9-6298-4BFE-899F-FB1B4C9DEC04.jpeg
     
    Last edited: Jul 21, 2021
  2. Guitarteach

    Guitarteach Doctor of Teleocity

    Posts:
    11,279
    Joined:
    Aug 6, 2014
    Location:
    UK
    I added a resonance control on the NFB on my ‘79 JMP Marshall. Tried different values/pot? I also have a bypass on the treble cap as it is very bright stock.. that warms it up across the mid ranges. Surprised yours is not too bright.

    The cab gives the bass thump to mine. What you playing through? Closed cab?
     
    owlexifry likes this.
  3. owlexifry

    owlexifry TDPRI Member

    Posts:
    51
    Joined:
    Feb 10, 2021
    Location:
    Adelaide, Australia
    yep i’ve added resonance control. tried 0.0022uf, currently 0.0047uf cap. gonna try 0.0068 next.

    i use a power attenuator, tends to cut a bit of the harshness, so i don’t get that issue with the high end.

    using a mesa recto slant cab (4xV30 closed)

    i guess the main thing i’m looking to confirm is if i can increase that first filter cap can from 50uf+50uf to 100uf+100uf without causing any issues
     
  4. bebopbrain

    bebopbrain Tele-Meister

    Posts:
    139
    Joined:
    Feb 5, 2021
    Location:
    New York City
    I wouldn't do it.

    The current into the cap is in bursts. The larger the capacitance, the more jagged the bursts. This gives a bad power factor, causes noise, and causes heating. Heating is I*I*R, so you can see why pulses of high current could be tough on the power transformer.
     
  5. Tom Kamphuys

    Tom Kamphuys Tele-Holic

    Age:
    41
    Posts:
    836
    Joined:
    Sep 18, 2018
    Location:
    The Netherlands
    I think you could try it, if your clone also has solid state rectifier(s).
     
    owlexifry likes this.
  6. Gijs

    Gijs TDPRI Member

    Age:
    49
    Posts:
    27
    Joined:
    Apr 14, 2021
    Location:
    Normandy
    The output transformer could be the most important limiting factor here (core size ?)
     
    Greg70, owlexifry and bebopbrain like this.
  7. Tom Kamphuys

    Tom Kamphuys Tele-Holic

    Age:
    41
    Posts:
    836
    Joined:
    Sep 18, 2018
    Location:
    The Netherlands
    Looking at the schematic, I think the only changes you can do to improve the bass response (if that means just more bass) is to change the treble bleeders (besides the resonance control that isn't present on the original). Either remove them or bypass them with a wire. The gain will be different in both cases, the amount of bass not.

    All coupling caps are large enough to pass the lowest (normal) guitar note without problem, which is probably why you didn't notice a change changing the ltp output coupling caps. The bypass cap on v1b is also large enough to boost all guitar frequencies. And power supplies are generally stiffer in an amp that is designed to be overdriven.

    In principle, changing tubes should not change bass response.

    Having said all that, the way tubes are used in an overdriven amp makes these statements a bit less true. E.g., because of intermodulation distortion new frequencies below the lowest guitar note can be generated. On my RobRob-ish 6V6 JCM800 there is a point on the gain control where I can here the bass 'fatten up'.

    Btw, I agree in principle with @bebopbrain, but for a short test I don't see much of a problem. You might wanna do some simulations with duncan's PSU software if you decide to keep it.

    @Gijs also has a good point.
     
    Last edited: Jul 21, 2021
    owlexifry likes this.
  8. tlp123

    tlp123 Tele-Meister

    Age:
    62
    Posts:
    174
    Joined:
    Jan 8, 2012
    Location:
    Denton, TX
    Go here https://ampgarage.com/forum/viewforum.php?f=9

    You'll get better and more accurate info.

    Mostly appliance operators here. And the few here that actually do have noteworthy technical knowledge get run off by the appliance operators that think they know something.
     
    Chipss36 and owlexifry like this.
  9. Wally

    Wally Telefied Ad Free Member

    Posts:
    40,169
    Joined:
    Mar 17, 2003
    Location:
    Lubbock, TX
    More capacitance in that first stage will yield a punchier, more articulate in the low end, ime. And yes, as one can observe, most solid state rectified amps targeted toward ‘modern’ sonics will have larger capacitance in that stage.
    That said, a small output transformer with less iron will have problems yielding big,firm low end at volume compared to a bigger OT. We see and hear this when we compare say a Fender Super Reverb to a Pro Reverb before the Ultralinear era. The SR has more authority at volume, and the Pro Reverb will saturate much earlier.
    I see no problem in increasing the first stage of filter capacitance with that solid state rectification. If it were tube rectified, then I agree with the concerns about the limits that tube rectifiers have.
     
    24 track, owlexifry and hepular like this.
  10. dan40

    dan40 Friend of Leo's

    Posts:
    2,296
    Joined:
    Aug 19, 2015
    Location:
    Richmond Va
    I had the same issue with my 2204 build. I tried many mods but none of them gave me the tight, strong low end that some other amps exhibit. These amps really need to be played loud to sound decent, and getting the master volume up to a decent level helps quite a bit. It's also important to remember that the circuit in these MV amps is designed to remove a lot of the low end and accentuate the higher frequencies. There are several treble peakers throughout the circuit that help the higher frequencies to pass through easily and much of the low end thrown away to help the amp cut through the mix.

    Some folks will change out the coupling caps at the phase inverter to allow a bit more low end. Try replacing the .022uf caps with .047uf or .1uf and see if you like it. Raising the capacitance at the first filter will also help as the fellas mentioned above.

    This old thread at the Gear Page is also a wealth of info.

    https://www.thegearpage.net/board/index.php?threads/jcm800-tgp-research.701204/
     
    24 track and Wally like this.
  11. Wally

    Wally Telefied Ad Free Member

    Posts:
    40,169
    Joined:
    Mar 17, 2003
    Location:
    Lubbock, TX
    +1 with Dan40….ime, any vintage Marshall amp needs to be running a bit past halfway on the gain controls…whether they are the early circuit or the later MV circuits. In the MV amps, the preamp gain and the MV need to at 6 at a minimum for me, or the amp doesn’t come to life.
     
    dan40 likes this.
  12. owlexifry

    owlexifry TDPRI Member

    Posts:
    51
    Joined:
    Feb 10, 2021
    Location:
    Adelaide, Australia
    thanks, this is the answer i was looking for.

    yep, SS rectifiers.

    i’m gonna give it a go. just put in an order for a 100+100 cap.

    thanks for pointing this out.
    i’m confident the OT should handle this fine, especially since the 40W build w/ 80+80 caps at the same filter stage hasn’t had any issues at all and the OT on that is fairly small (it’s essentially a super reverb 40W replacement OT) where the 50W 2204 OT is comparably much bigger (nearly twice the size).

    fingers crossed, super keen to hear the results
     
    Wally likes this.
  13. Wally

    Wally Telefied Ad Free Member

    Posts:
    40,169
    Joined:
    Mar 17, 2003
    Location:
    Lubbock, TX
    without knowing the specifics of each OT, I will observe that when looking for articulate low end, one would design with a strong OT as a first step, imho. I was figuring you were using something close to the original OT.


    I reread your first post. 2 x 50mfd in parallel = 100mfds. 2 x 100 in parallel = 200mfd. In that Peavey, those 330mfd caps are in series. That Yields a 165mfd cap….very similar to the 2 x 80mfds in parallel that you have been using as that arrangement yields a 160mfd first stage cap.
    Ime and IF your OT is ‘small’, imho you may have realized all that you can realize in this search. I have used 2 x 220mfds in a BF Bassman amp. That amp would physically punch you in the chest. Those 2 x 220s were in series for a 110mfd cap.
     
    24 track, owlexifry and bebopbrain like this.
  14. King Fan

    King Fan Poster Extraordinaire Ad Free Member

    Posts:
    6,855
    Joined:
    Jan 1, 2013
    Location:
    Salt Lake City
    Heh, it's true that Amp Garage and several other sites have a higher density of tech than TDPRI. Amazing this place is so popular... maybe it's the low density of snark?

    And then, of course, there's the excellent ignore function... :D
     
  15. Greg70

    Greg70 Tele-Meister

    Age:
    51
    Posts:
    444
    Joined:
    Feb 21, 2021
    Location:
    St Louis, MO
    Marshall actually reduced the number of filter caps in the 2203 over the years. The early ones had 6, there was a transition year to 5 and then down to 3. I believe the 2204 followed this same pattern.
     
  16. Wally

    Wally Telefied Ad Free Member

    Posts:
    40,169
    Joined:
    Mar 17, 2003
    Location:
    Lubbock, TX
    Interesting. The schematics I find at Schematic Heaven do not show this. The ‘70s schematic shows 6 caps in 3 stages.
    https://schematicheaven.net/marshallamps/jmp_mastervol_100w_2203u.pdf

    the 1988 schematic also shows 6 caps in 3 stages.
    https://schematicheaven.net/marshallamps/jcm800_lead_mstvol_100w_2203.pdf

    the only difference between those two filtering schemes is that the ‘70s scheme has 2 x 100mfds in series for a 50mfd cap in the first stage while the 1988 scheme has 2 x 50mfds in parallel for a 100mfds cap in the first stage.
     
    24 track likes this.
  17. Greg70

    Greg70 Tele-Meister

    Age:
    51
    Posts:
    444
    Joined:
    Feb 21, 2021
    Location:
    St Louis, MO
    I realize that screenshots from vendors isn't necessarily the gospel, but this is what I found to be the case. My 2203 was built in 1984 and has 5 filter caps. It was the first year for the horizontal inputs and I believe it was also the first year for the PCB mounted pots. I have yet to find a Marshall schematic that shows the 5 caps. At some point they went from 5 caps down to 3. Below is my 2203 with the 5 caps. These were the originals that were replaced last year. I added the extra preamp tube a long time ago for an effects loop.




    upload_2021-7-23_10-59-54.png

    upload_2021-7-23_10-58-42.png

    upload_2021-7-23_11-5-41.png
     

    Attached Files:

  18. Wally

    Wally Telefied Ad Free Member

    Posts:
    40,169
    Joined:
    Mar 17, 2003
    Location:
    Lubbock, TX
    Greg, what you are seeing as three caps are three dual section caps with a total of 6 caps there. This is exactly what the schematics show. There are three stages of filtering in both of those schematics I posted. Each stage consists of two capacitors wired to gather. In the older circuit, the first two caps are wired series to establish that first stage while the other two stages consist of two caps in parallel for each stage. In the later circuit, all stages were built with two caps in parallel for each stage.
     
    Greg70 likes this.
  19. owlexifry

    owlexifry TDPRI Member

    Posts:
    51
    Joined:
    Feb 10, 2021
    Location:
    Adelaide, Australia
    update: great success!

    done a few things now.
    very pleased with the changes made today+yesterday.
    many thanks for all your comments and input.


    1)
    - took out the 50+50, replaced with 100+100.

    2D36C96B-8646-4E8A-92B2-B2BAA16590B0.jpeg 406C1BCC-3D37-47E6-BB5F-D4DD16DF31A5.jpeg EE9F92DE-065E-4CAE-8548-D3F04A6D97BA.jpeg F6BF2D9C-722B-4C0D-8E31-7269CDEC4F4D.jpeg

    yes. this did almost exactly what was promised / expected.

    tighter everything. more punch. more article low end. ‘bigger tone’.


    then, given the comments about treble peaker/bright bypass caps, i thought i should have a look at these (bright caps at the gain control pot), cos the tone was a bit overly ‘spanky’.

    [previously, the stock value 470pf (silver mica) cap (bypasses 470k resistor) was replaced with 2200pf (ceramic), and the cap that ‘bypasses’ the pot, 1000pf, was changed from stock film type to ceramic (same value)]

    2)
    - so i decided to get rid of that 1000pf cap, because i don’t like how the tone changes as the gain knob goes up and down (by nature of its bypass location).

    - then i compared 1000pf, 470pf, and 2200pf, where it bypasses 470k resistor going to gain pot).

    2200pf was better than the others to my ears.

    with the 1000pf removed, tone thickened up a little bit, it was a bit duller, acceptable, but could do with a little more zing.

    so i had an idea - replace the 470pf that bypasses the 470k between V1 and V2...

    3) - so i did and put a 2200pf in place of the 470pf. this was good to my ears, added that zing i was looking for. (although, after what i did next, i think this will need adjustment again)

    4)
    - then i did a mod to the NFB that i’d been meaning to do for a while (inspired by a specific layout/schematic)

    - previous setup:

    4ohm tap—resonance control(1Mlogpot,4.7nf cap)—0.1uf cap(intended to block DC, probably didn’t need to be there**)—39K—presence pot

    **got the idea from the SLO schematic, it didn’t appear to do anything detrimental to the tone (but why even have this? would there even be any DC voltage coming from any of the 4,8,16 ohm taps?)

    - new/current setup:

    8ohm tap—resonance control(1Mpot,4.7nf+2.2nf in parallel for 6.9nf equivalent)—27K— NFB control[250K Linear pot(220K across)]—presence pot

    5CF9ABF3-5F14-40AC-96C4-33085640BF12.png

    B795AA73-4F95-40BC-8277-470FAB3A7ABF.jpeg F02380FF-7FB7-4F2C-922D-6BC5FF906C30.jpeg

    And wow.
    this NFB mod (and the depth control cap change) made a huge difference.
    i’m sure the bigger filter caps are helping, but this change in the NFB loop has made a huge amount of punch/balls/‘bark’ available with these controls that wasn’t possible before.

    A/B ‘d with the stock 100W EL34 peavey windsor, and it now appears to have an equivalent amount of low end response/punch. stoked. i’ll call this a win.


    only gripe is, with the NFB mod, the tone can be a little bit overly ‘barky’. somewhere in the high mids is a bit too much (and i can’t seem to dial it out without killing the tone where i like it).
    i’m thinking i’ll try some different values for that bypass cap between V1 and V2 that i mentioned above.

    - another thing i wanna try is to insert a 820pf-1000pf bypass cap across a plate load resistor somewhere..
    maybe the second gain stage? (cold clipper, currently stock-10k cathode, 100k plate) or somewhere further down maybe....

    - also keen to try 120pf plate load bypass caps on the PI (V3).

    would be interested to hear opinions/experiences on these mods/approaches
     
    Last edited: Jul 25, 2021
    Wally, Gijs and 24 track like this.
  20. owlexifry

    owlexifry TDPRI Member

    Posts:
    51
    Joined:
    Feb 10, 2021
    Location:
    Adelaide, Australia

    tbh i find this place to be the most laid back, but also the most helpful. coming to this place feels like coming to ‘old joe’s parts shop’, vs. places like tgp, where it often seems like a blur of incoherence..
     
    zeppelinofled69 and King Fan like this.
IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.