"Marshall" sim

[petemoore]

  Brehumbaggertz.
  I wired about half of the 'plurality' filters...all day basically, up to the 5khz peak filter. 
  i shudda made the pot wires shorter
  I Should Have Gotten A PCB...maybe, seems kinda hummy.
  These filters are slow going as far as something to wire up and try out...I just wanted to get some of them going, then try and see if it works, it works but I have ALOT more work to Do.
  Alot of work and the hum is bothering me. I wonder how much of this noise I can get rid of and how much more hum that'd be than a commercial unit, not boxed but still I'm concerned with the hum...seems pretty strong.
  One slow process, building many of these resistors, digging up all these caps...it took me pretty much all dern day [including trip for parts, after 'longsearches'] to just wire up 3 dual OA's, for a total of 6 of the 13 'plurality' of MFB Stages.
  Having no real clue how 'critical' for instance the '202k' resistor value is, I made all the 'funny' values, and all values of R very close if not exactly 'on' value...something like .5 % tolerance...because why not...I figured all that work to make a 'comb' [probably wrong term but] that has 'bent teeth'...no, I decided to make the filters as close to the 'target' freqs as possible.
 

[RDV]

Are ya using 2 batteries, or did you build a dual supply? If it's the latter some extra filtering might help.

RDV

[petemoore]

  Re-update
  This noise hum stuff is the guitar, I turned the vol down on that [   ] and the noise dropped at least 90% right there, didn't have loudness going on but the noise with guitar off is super low.
  I've thus far included the 8Khz peak filter, thinking strongly about leaving off the 11khz as there is plenty of [what sounds like very high end] hash for all without the highest freq peak filter in...Sounds pretty good...
  After the 3rd or 4rth filter speed of building increased alot, following the value charts, schematic and board took a bit of getting used to, hitting target resistor values takes time also.
  I decided to use some brownie caps [in sockets of course lol], for some of the filters, to get more AMP MOJO...2x/lol.
  Its a long process, but I'm building of this beast, piece by piece, I still have the speaker damping and bass boost to go also, that alone will certainly change the Freq Response.
  I'm using 2x longboard [rectangular] with 'end stock' so that I can screw 'sub boards' right to that and connect PS and in/out wiring.
  It's looking like the 'box' will end up being a pan.

[stm]

  Re-update
  This noise hum stuff is the guitar, I turned the vol down on that [   ] and the noise dropped at least 90% right there, didn't have loudness going on but the noise with guitar off is super low.

Pete, 95% of the times I've experienced hum in the way you describe is because I've connected the input or output jack with the wires crossed (i.e. shield as signal and vice-versa).

[tomwarrior]

i built the preamp section of this omiting the tonestack and opamp 8. the thing that gets me is it is way too bright going into the preamp of the clean channel on my amp(valvestate) and way too bassy going directly to the poweramp. it sounds awsome going into my npn boosted ruby amp with the amp set for a clean sound. can anybody explain why? thanks in advance.---tom

[Doug_H]

i built the preamp section of this omiting the tonestack and opamp 8. the thing that gets me is it is way too bright going into the preamp of the clean channel on my amp(valvestate) and way too bassy going directly to the poweramp. it sounds awsome going into my npn boosted ruby amp with the amp set for a clean sound. can anybody explain why? thanks in advance.---tom

Yeah, you left too much of it out...

Seriously... You guys are building bits & pieces of the thing, which was designed to function as a whole in a complete system, and then coming back and asking why it doesn't sound right. (?!?) LXH2 has some pretty detailed explanations on his site as to how it works, what ea stage/section contributes, and etc. Go there for clues.

As to why it sounds good with the npn/ruby, you would have to put it on a scope to see what those other pieces are contributing.

Doug

[tomwarrior]

this is the complete preamp schematic;
http://img257.imageshack.us/my.php?image=marshprb1hi.png
i got it from the LXH2 websight.
looks like a bunch of hipass filtering with no lowpass filtering if im correct.

[petemoore]

  Seriously... You guys are building bits & pieces of the thing, which was designed to function as a whole in a complete system, and then coming back and asking why it doesn't sound right. (?!?)
 [>I didn't ask why it doesn't sound 'right']. I know why it sounds like it does [sometimes] hopefully by dissecting it or building it piece by piece I'll be able to discover even more of what makes it tick, and possibly even understand it, we pick every other unit apart and dissect the fragments...why not this one...?...I'd guess because a lot of thought went into the design in the first place, and mods are ... more or less 'moot'....pick the Fender or Marshall and go to town !!!
 No need for a pre-boost. No need for a [   ] or EQ. It is complex, widely variable etc. and stands alone quite strongly...I guess...soon maybe I'll hear the whole completed unit's sound.
 I was merely doing it piece by piece, on perf, sampling each section and testing as I go...to see IF it works, and HOW it works.
 As far as mods for more gain, treble, etc. dial the knobs....there is plenty 'o controls, huge gain available.
 The best/quickest/most sure fire way to get a great LHXS sound...buy one. The commercial ones [there Are Commercially produced ones aren't there?] would probly have all the details worked out, and there are lots of them, all the parts work, lots of those, all the wiring is low noise [Probly EZ to get more noise with sloppy layout] and a plethora of other stuff...like a guarantee the thing'll work.
 I've been hackin away on this thing for three days straight about. The scheamtic looks big, but when you add in the filters, it nearly doubles.
 So far I've read and believe that the bass is rolled off early, then 'built back up' later...among other things.
 Doing this thing on perf is a long, meticulous, labor intensive task, I'd get a PCB if I were to do it again....or reconsider and just buy a pre-made unit.
 I added a couple more sections and that threw my Bias voltages off...at least I probably wont have to wade through a debug of 30+ opamps, because it's most likely a problem with the most recently added 'sub-board', containing 3 duals.

[petemoore]

  The first section, past the diodes clipping, is as Doug said, part of a whole system.
  I has serious bass rolloff, and is very bright. Part of the design IIUC is to start off this way, then 'build back up' the bass with later stages [there's a knob and oa's labeled "Bass Boost'].
  The first section is a dirtbox, basically, a good sounding clipping device, similar to a diode to ground clipper [diodes do shunt signal to ground], but it seems and this is mere speculation/theory or guess...that the OA's before the diodes somehow 'control' the voltage that supercedes the diodes clipping...to produce the 'dynamics' it has. Way too trebley by itself IMO, if you were doing this section for use merely as a dirtbox...I put in some HF rolloff components at the end of this circuit for fun and testing, a couple 12k's strung series with SP, and a couple 2n2's or so going from their junctions to ground [word?].
  It took a fairly long short while to do the clipping section, a full day to build half or so of the filters sections...one long build...I felt the need to pre-test and add sections, for debugging simplification, and just to see what they 'do'.

[petemoore]

  Can't say I recommend this method.
  I'm back from point 'Q' to point "A-'.
  Having troubles with the opamp bias points after installing the 'additional' boards, also after de-installing them.
  Since I took the first section from the schematic, it's kinda funny...I do OK up to inserting the last two opamps I have on my baord, at that point the SUpply Voltage drops about 40%...wierd, used to workk...Daft.
  Anyway, Perfing this thing is rediculous.
  All the opamps drop voltage when either or both of the last two are dropped in, usually I'ts just one single side that gets thrown off, much easier to debug that.
  As it is, one may end up necessarily debugging a rather large number of opamps at the same time.
  I don't know what made me think I could or should do this...whatever...I thought it might be fun and rewarding but it's turning out to be a 3 day nightmare, hopefully this time I've learnt my lesson about getting lucky attempting Huge Circuits on perfboard. It's also eatin' juice fast when batteries are connected.
  I guess any interest in modding or messing with this type of thing is also 'undesirable', as it is turning out, there is most likely little that DIY-ing on this could achieve.
  The clipping section was sounding kind of cool for a minute anyway, I'm not looking foreward to finding a problem among 12 opamps and related components though.
  Doing all the regular stuff, isolating and testing V- Gnd and V+, etc.
  Got to be something though, DC Blockers I'd think would allow an OA on 'the other side' of it to bias, perhaps I've somehow got some blown OA's.
   

[MartyMart]

Pete, you're a brave man ..... perf !!
I had no problems at all with the "Pre-amp" section and have only made some
changes due to using the "Condor" as my speaker sim section.
I used a large piece of stripboard, about 4 cm by 8
I recommend this 100% as it's a great result.
I took one look at all those opamps and thought  .... no no !
4 duals and single for the preamp was enough for me and is also the largest
build I've taken on yet.
My de-bug for that lot was only 10 minutes  !!
The pre amp with a few small adjustments works very well with the condor, which
is only 2 more opamps ... !

If anyone wants the details, which myself and Pete came up with, PM me.

Marty.

[petemoore]

  Maybe I'll work on the filter sections some more to blow off steam.
  I'd like to get 'em going and try some Fuzz through them, possibley figure out what they need to drive them...that'll be a day or two's wiring right there, compact and not pots means I'll have a potless, fairly compact board for debugging..
  Stupid me won't let the board bias OA 8, the + input insists on 1/3v, despite all wiring, voltage, continuity, etc. beep mode checking possibilities, the tone stack, with a direct connect to +input from 1/2v at certain settings brings the voltage near half.
  I changed the opamp a few times, got some other wierd voltages, seems to have settled on 1/3v for today at least, even with the opamp out.
  Super dangly potted Gangly board is not DIY Debugging friendly.
  Suffice it to say I'm not liking it. Much too much for perfboard.
  That 1m bias resistor is going directly to 1/2v from +input, not Gnd, V+ or anywhere else on the board, 1/2v only...I think, after clipping the lead there and touching every node/connection on the board. Why it insists on finding itself at 1/3v, with only a direct connection to 1/2v through the tonestack [or 1m connection through the bias resistor], and nowhere else, with no opamp present, is baffling me totally.
  I spent all day, ripping it down to the clipper/TS, debugging, and all I have to show is a rant here.
  Perfing this thing is pretty rediculous.
  Well I figured out I can plug only one dual OA in, otherwise the battery time starts ticking, 9.08, 07, 06, .05 ...Funny thing is these opamps are showing 1/2v on the inputs and outputs, except the defiant one, which didn't change matters when unplugged.
  A real puzzler, and, I'm at a standstill as far as disbugging this one, I suppose I should see if a PS can handle the load better for debugging...or...
  I've been over and over every conductive surface on the board, mainly looking for how the power supply is loaded so much.
 

[MartyMart]

Pete, you're talking about the "preamp" 1M to ground off the tone cap wiper to opamp 8's + input yes ?
All seems right up to that point ..?
Are you powering with 2X batteries on the in/out sockets.
Across both the power lines I have 47uf's and 1N4001's reversed biased to ground rail and
extra 0.1 poly's to ground too for filtering.
All my pin "8's" are 9v and pin "4's" -9  when you say 1/2V you mean 1/2 V is ground right ?
Would seem like a "drain" somewhere around the tonestack or opamp 7 ...
I thought you had this section working fine and just the "Speaker sim" was a problem 
I'm about to try a version as a "9v" circuit, just the preamp section using 9v 4.5v and ground .....
Luck,

Marty.

[krille2]

Uh, troubleshooting isn't always the most rewarding activity, petemoore. I also had problems with the preamp, around op amp 7 or 8. I had to rewire the lot with new components and off we go with a working amp! And man, was I smiling when I first heard the sound of it! The sound is amazing and it was worth every effort I put into it.
There is a lot of discussion regarding the preamp section which I thought of as to be really good. But as I understand it you should plug it into a power amp -> 4x12 if you don't build the cab sim, from there one could judge if it's bright or not. Compared to other units on the market I think this has a really steady, strong and genuine "Marshall" tone much more so than the weak and buzzy units you could buy. The sound when using my Les Paul is so nice, I could dial in for example a ridiculous close Satriani tone just out of the box, or a really convincing AC/DC tone! When using a Stratocaster it is actually a little too bright to my taste (try a real Marshall stack with stock strat singlecoil pickups – it is REALLY ear piercing bright!!!), so I tried the "another-470pF" mod as suggested here and it worked well, I also added a 27k to the GND leg of the 220k treble pot (to get closer to 250k). I have to investigate a little further... I thought of making this mod in the cab sim instead, around the 6db per octave low pass filter to decrease the brightness somewhat, maybe change the 0.1uf to ~0.2uf. Actually a Marshall should be bright and this is a concern only with the Fender as the Gibson (with humbuckers) needs the treble at 10.

http://www.kjell.com/, thanx for the tip!!

[petemoore]

  Gadzooks hard to debug.
  I finally went door to door, point to point, wire to wire, soliciting for errors, I FInally found a wire on wire nonconnect of 1/2v to the last two dual OA's biasing...Youch that took some time 'n effort. ...following the 'wire' train, I noticed a voltage difference along it...One of those wires that din't get pinned to the wire right and the solder just didn't 'set', sure looked 'connectey' at a glance, had to be one, but it was a blob all over both wires, I surmize only actually physically connected to the top one. Big StrongBear Debug. Took many times the time I'd hoped for.
  The other 'input re-wire' error was much easier to find, being as I wasn't relying on something I'd taken for granted that wasn't there.
  For right now, I'm rewarding my extensive efforts, having a good 'ol time just using the BMF's with 'something' cooking distortion to get the inputs drive up, and a Mosfet Boost for recovery. The fact that all OA's are 'active' with source guitar or even 'thumbuzz' [touching the -input of each stage is cool, each stages 'frequency specific emphasis' can be heard]...I may choose to tune the output levels [EQ] using these sounds as a guide].
  Anyway I'm just peached that I got this part of the circuit working, especially since it is the most foreign to me, even though the difficulties I encountered were not.

[krille2]

It's said: if it isn't broken, don't fix it, which maybe is true. Anyway, I took apart mine yesterday just to get rid of the 470pF cap, and bring it back to original state once more, I swapped the 1N4148 for 1N914 and added a 27k to the GND leg of treble pot 220k to match a 250k more closely and it sounds good too, not really much of a difference actually. Maybe, maybe the distortion is a little smoother with 1N914, but it sure is subtle. I like the brightness of the amp, without the added 470pF which to my ears hides some of the Marshall specific overtones. I think one should attenuate the highs in the cab sim instead if it bothers which it might do if you're using single coils and want a "smoother" sound.

[krille2]

Well, the unit sounds terrific in its original form; I know this from trying out different mods which actually don't seem to bring in any sonic improvements   . I have tried different tone stack configurations, lots of capacitance changes, symmetrical/asymmetrical diode configs (with odd and even number of diodes) and more. But still, the original is the best so far. And yes, it's not the favorite unit to experiment on as petermoore said, new problems might arise if one's doing it on perf 

[MartyMart]

Well, the unit sounds terrific in its original form; I know this from trying out different mods which actually don't seem to bring in any sonic improvements   . I have tried different tone stack configurations, lots of capacitance changes, symmetrical/asymmetrical diode configs (with odd and even number of diodes) and more. But still, the original is the best so far. And yes, it's not the favorite unit to experiment on as petermoore said, new problems might arise if one's doing it on perf 

Krille, I'm sure that you're correct - when used with the specificly voiced "cab sim" that it sounds "right".
I have only adjusted the preamp to use with my condor sim, which I'm now very happy with
I just needed to get rid of some high shrill content, but have not removed TOO much, there's still plenty
with the tone pot at 50% and above ...
My only other change is 2X 1N4148's by 1X 1N4148 for a bit of assym distortion, which I like better.
Great tones now, in particular from a HB pup gtr 
MM.

[krille2]

Would love to here a sound clip (or two) from that one!

[rody82]

If i use a quad opamp, does it matter how do i use the discrete opamps in it?
I mean if i use 2 of 4 for the low level (guitar level) signal (near input) and 2 of 4
for the high gain (near output) signal will this cause any problems?