Origin 50 tilt circuit and a shoehorned (fet) CF

[stratomaster]

I picked up a Marshall Origin 50 for next to nothing.  It has an interesting "tilt" control that appears to blend between two fixed but complicated passive filter networks.  It also has no CF to drive the tonestack. 

I'll have to spend some time with spice to get a real feel for what the tilt is panning between.  Any gut feels at first glance at the schematic?  UPDATE: Took a closer look.  It's not affecting the odd passive filter network after all.  It's panning between big-ish cap to ground and smallish (for Marshall) bright cap.  Sounds like it's doing more than that. Still need to figure out that passive network, though.

Merlin and others have advocated the cathode follower as an integral component to the sound.  I've seen others hack up the board to repurpose the effects loop buffer as the tonestack CF.  I think a more elegant solution would be to shoehorn in a small FET board source follower and cut the trace between the plate and slope resistor/treble cap junction. 

My question is, how close to the behavior of a tube CF is the FET facsimile?

My goal is to get this thing to sound more like an old Plexi without resorting to heroic efforts.  It just needs to behave more like a Plexi than my old Blackfaces to scratch my Marshall itch and prevent me from buying or building one.  I think I can get there with some strategic tweaking.

[tubenit]

My first response is I'd try to add a mosfet cathode follower.  Mosfet cathode follower (el34world.com)


And then make everything between V1a and V1b look like a Plexi 50 preamp.


with respect, Tubenit

[Ronquest]

I've hacked up a few of these, 50s and 20s.  Personally I say there are a few small tweaks you can make to reduce noise , cross talk, beefing up a few week components and a bigger tweak (moving the ill placed output transformer).  I tell people, their money is better spent getting the amp they want.  The moment you jack up the gain and reroute the circuit you open up other issues. 

The other question about the mosfet you should be asking is, does the mosfet distort like the tube?  If you're going for "plexi's secret sound" and believe it's in the cathode follower, why then change the cathode follower to a mosfet.  Maybe, use the mosfet as your input stage, as it shouldn't be overdriving anyway and then you could use a real tube stage for the follower. However, Tubenit says he can't tell any difference 

On the bright side if you don't like it, you can gut it for a plexi build and move that OT.  That's my opinion on these amps.

[stratomaster]

I've hacked up a few of these, 50s and 20s.  Personally I say there are a few small tweaks you can make to reduce noise , cross talk, beefing up a few week components and a bigger tweak (moving the ill placed output transformer).  I tell people, their money is better spent getting the amp they want.  The moment you jack up the gain and reroute the circuit you open up other issues. 

The other question about the mosfet you should be asking is, does the mosfet distort like the tube?  If you're going for "plexi's secret sound" and believe it's in the cathode follower, why then change the cathode follower to a mosfet.  Maybe, use the mosfet as your input stage, as it shouldn't be overdriving anyway and then you could use a real tube stage for the follower. However, Tubenit says he can't tell any difference 

On the bright side if you don't like it, you can gut it for a plexi build and move that OT.  That's my opinion on these amps.

I have no real interest in reallocating triode stages.  I like that there's a built in effects loop and power scaling.  I'm just wanting to voice is more like a traditional Marshall than it currently is.

If I spend more than 3 hrs and $20 doing so I will have failed my mission.

Having said that, you bring up an interesting possibility.  Use a FET stage as the normal channel input gain stage, revoice the existing input as the high treble, wire them in parallel and use the tilt to dial in the normal channel.  Depending on the board layout this can be done cleanly without much hackery. 

Combined with the FET CF, I'm firmly in Plexi land in terms of gain staging and voicing. And I get to keep my effects loop. 

I can use strip board for the normal channel and CF daughter boards.

I'll have to give this some thought.

I think the FET CF is an easy enough thing to investigate for any potential value and might be a good first tweak

Again, all I want is for it to be a bit more Marshall than it currently is.  I don't need it to be a Plexi clone. 

[stratomaster]

I finally got around to doing this.  Well at least the addition of a source follower between the plate and the tonestack.  All it required was to splice into one of the board to board jumpers. No traces needed to be cut.

Install was pretty clean and didn't require pulling the main board.  I populated a small daughter board for the FET source follower.  Used a 330 ohm input resistor and a 68k source resistor on an IRF820.  I can't say how much of a difference just the source follower made as I also tweaked the cathode bypass caps, input stage plate resistor, and converted the Master Volume to a cross line post phase inverter style.  I also added a fixed resonance network in the NFB to help with the lack of low end in this amp. The top end is still pretty harsh, so I'm considering either a Cut control or tweaking some values in the passive filter network that resides between the first two stages.

The source follower added a marginal amount of gain (a 7.45V signal in resulted in 7.54V out), but the real goal was impedance to drive the tone stack. Mission accomplished.

I'm pretty happy with it so far.  This thing definitely has more of a stereotypical Marshall voice now.  But the noise level is rather high--this was true pre-modification, but the noise floor seems higher post mod.  It's not "I must've done something wrong" noisy, just "I know this can be quieter" noisy.  Ronquest mentioned that some quick tweaks that could help with that.  If anyone is privy and savvy to what these are, I'm all ears.  It looks like the first two tubes are already being fed DC for the heaters, so not much can be done to improve that. Usual tricks of shielded grid leads and added gridstoppers are harder to implement on these PCB amps.

Edit: added a hacked up schematic showing what's been done to the preamp.  I think I did a better job with the install than the graphic. I just updated values in red. In practice I paralleled additional caps across cathode resistors.

Edit 2: Adding some LT Spice sims of the passive filter network--didn't bother with the loading effects of the volume pot or the triode.  Looks like the 220pf cap is pretty heinous.  A simple snip might clean up a good bit of the ice pick and hiss.

[stratomaster]

Removing that 220pF cap helped tremendously. To the point where I can probably go back to 1μF for the first stage cathode.  There's still hum in the non-boost mode that goes away in boost mode (which simply changes the V1b input voltage divider ratio and switches in a 2.2μF cathode cap on that same stage). 

This hum could be the transformer hum Ronquest mentioned (especially since it's present immediately upon power up before the tubes warm up). It seems to be shunted to a great degree by the cathode cap--so I'm likely also experiencing some degree of power supply ripple (haven't bothered to measure yet). 

Interesting note, the current limiter glows brighter in the lower power modes.  These modes must introduce massive inefficiencies.

[Dave]

RL1A is essentially a "bright cap". A significant amount of treble will be lost across that 2m2 resistor, so the cap will restore that. Some will probably argue that a resistor is not frequency dependent, but there is more to it than that. It's a "bright cap".


RL1B shorts the 47k resistor between the cathode bypass cap and ground. Essentially, more gain. Although there is another small bypass cap that is treble-heavy. So, probably the idea is that when the switch is engaged, the bypass cap gives more gain across all frequencies where the treble was already boosted. That's why you need the "bright cap", to, sort of balance the frequency response of the gain boost.


Dave

[stratomaster]

Interestingly the LT Spice sim shows the 470pF isn't doing much--though this could be user error on my part.  And it's shorted along with the 2.2M in boost mode, anyway.  The Tilt control blends in a bright cap that gets plenty bright enough on it's own.  This design has 3 bright caps and a partial cathode bypass well into the upper midrange on the non-boost mode, and two relatively small coupling type caps in series.  Combined with the pre PI master volume, no wonder these things sound brash at lower volumes.

In stock form you need to get the master up pretty high and the boost mode engaged to get anything resembling a useable edge of breakup type sound.

After the conversion to post PI and the tweaking of the cathode values and FET CF (the real focus of this thread initially) the only remaining gripes were noise and harsh high end.  Eliminating the 220pF took care of the harshness and some hiss.  Now just the hum remains and the Tilt control still gives me plenty of brightness.

[tremdoctor]

I picked up a Marshall Origin 50 for next to nothing.  It has an interesting "tilt" control that appears to blend between two fixed but complicated passive filter networks.  It also has no CF to drive the tonestack. 

I'll have to spend some time with spice to get a real feel for what the tilt is panning between.  Any gut feels at first glance at the schematic?  UPDATE: Took a closer look.  It's not affecting the odd passive filter network after all.  It's panning between big-ish cap to ground and smallish (for Marshall) bright cap.  Sounds like it's doing more than that. Still need to figure out that passive network, though.

Merlin and others have advocated the cathode follower as an integral component to the sound.  I've seen others hack up the board to repurpose the effects loop buffer as the tonestack CF.  I think a more elegant solution would be to shoehorn in a small FET board source follower and cut the trace between the plate and slope resistor/treble cap junction. 

My question is, how close to the behavior of a tube CF is the FET facsimile?

My goal is to get this thing to sound more like an old Plexi without resorting to heroic efforts.  It just needs to behave more like a Plexi than my old Blackfaces to scratch my Marshall itch and prevent me from buying or building one.  I think I can get there with some strategic tweaking.

Loved reading your post, extremely technical but you speak with authority. Interestingly there is a guy from Finland on the Marshall Forum who goes by the handle "Timo V" who posted the same schematic and mods you did. I asked him if he was you, he said no. Anyway great mod! You mentioned HeadFirst, I am doing his JCM800 mod to my Origin 20 only because some of the tone he gets is that raspy Plexi tone while adding lots of distortion, as he repurposes V2 as a gain stage disabling the fx loop.

I like having an fx loop, I use them extensively on my 3 DSL's, but the fx loop on the Origins (20H & 50H) are abysmal when you are using high gain, whether it's using the onboard Gain/MV or an OD out front. I get decent distortion using a 5150, better than using an SD-1, but the only thing that works in the fx loop is my delay, an EQ such as a GE7 which works fabulously in a DSL as a clean boost for shaping endless tones, but it sounds like nails on a chalkboard on an ORIGIN.

You wanted to preserve the fx loop functionality, with high gain (lets say MV & Gain set to 3:00, how are you able to use the fx loop?

Have you heard of anyone who rerouted the fx send so that the preamp gain was not introduced in such a way it distorts the fx loop signal so negatively?

Thanks,
TD Dave

[stratomaster]

Thanks for your compliment. I'm unaware that this mod caught on with at least another Origin owner. I posted a post modification summary at the Marshall forum. Perhaps that's where he got the info?  Haven't checked responses there in a long time.

As for your issue--I haven't noticed any issues with the fx loop. But I don't use very high gain.  I max out at about ZZ Top levels of grit.

I also went back and added the Cut control for use with the two lower power settings in the amp.  I changed the cathode network on the first preamp stage as well since I posted that schematic. I'll have to see if I took updated photos or marked up the schematic again.

I understand FX loops and their placements can be very tricky in higher gain amps.  It may be beneficial to study either the drop in FET loops available from places like Metro or to look a Dumble ODS schematic for his passive loop placement and put in an internally buffered loop at that point on your modified circuit.

By the way, the amp is still pretty noisy for my tastes I'm pretty sure the source is the transformer placement as Ronquest has indicated .

[tremdoctor]

Thanks for the great and timely response. I'm surprised the poster on MF didn't credit YOU for the work you did. Anyway, people are using your mods ;)
You suggested using a FET, I think it's a great idea, in fact I asked Jason Tong about using a FET on an Origin, he said if it fits do it. That was something you suggested in this thread and I did some research what a FET was. You are basically talking to a child here (me) regarding this stuff, I LOVE amps, but am very green with electronic circuitry, my experience was wiring schematics, wiring harasses, relays etc, not circuits so you'd need to really dumb it down for me to grasp it, sorry about that, don't want to waste your time. I like high gain, Metallica, Megadeth and my favorite of all time, EVH. He ran high gain Plexi's right?

I bought this Origin as a mod platform searching for plexi tone, but as you said here it falls far short sounding like a Plexi. The Plexi crunch, the raspy chords, cat purring gain rather than chainsaw sound I am searching for, the chimey sound a Plexi delivers, good edge of breakup, not nasty sounding overdrive that IMHO the Origin has. I get decent tone from a 5150 pedal, but then plug that pedal into anything clean and it sounds good, so calling the Origin a good pedal platform is no complement. To me anyway.

A fellow contributor on MF did your mod recently and loves it, I pointed out it is YOUR mod ;o He says the more he plays it the better it sounds. I may do the same mod on the 50 which is now stock. The Headfirst mod which will be on the 20 is as you say "heroic"? Repurposing V2 and saying goodbye to the fx loop I'd call herioc. That and changing over 20 components, it's intense. But I bought the 20 to Frankenstein, the 50 I want to be more careful.

Since you mention wanting to capture Plexi tone, I wonder if you can hear it in this JCM800 tone like I do. Sharing to see if you hear the same Plexi tones as I do once distorted.

I'd like to hear how yours turned out the way you did it.

Best,
TD

[stratomaster]

I used his schematic as a basis of what's there. My mark-ups and mods are on post 5 here. I don't think anyone is using them as there are no responses at all to my post there.

The final version of my mod is attached here.

The video you posted is on the upper end of what's possible after my mod, but mine is a bit looser in the low end.  By design.

[tremdoctor]

Great information here, now I see the connection. There are a lot of posts on the Origin on the MF, different subjects. People definitely talk about the Timo V mod.
I'll check out what you did and perhaps do that to my O50H. Like you said earlier in this thread, I just want it to sound more like a Marshall than it does. For me it as an inexpensive kind of a Plexi I could massage and get close, at least as close as an SV20H.