Putting together a Mark III Guide

[jrb32]

So the question is should it be possible to effectively have a purple stripe modded by a local tech into a blue or red stripe for fairly cheap? I already know I would be looking to do the R2 volume mod and the amp might need a general servicing as well, so the stripe conversion would be easy to tack on as long as it was possible and should not cost much. I live in Houston and Mesa has a certified repair center that is nearby, so that was my thought to save on shipping costs into Mesa. If that would be required, then I’ll just continue to wait for the perfect model to pop up. But I the ability to buy now and return if needed has its merits.

My advice would be to buy a purple stripe (or anything else you want) and send it to Mike B to get whatever you want done to it. No-one else should be doing this unless you live on the other side of the planet. Even 'authorised' repair places. Just send it to Mike He'll be able to advise on whatever you need and do the work including the R2 mod/service/anything else. You can also usually just ship the chassis to save on shipping costs rather than the whole amp.

 

[stephen sawall]

Agree with send it to Mesa and buy the stripe you want. I've played on all the stripes and they sound different enough.
Personally I wouldn't do it any other way.

 

[GJgo]

So here's something new to consider. I just acquired this amp, and I can find zero references to the CP1H. Ever heard of it? This is my 8th Mark III and I have to say this one has some feels that the others didn't have, save for maybe the early no stripe III+ SG.

- Blue stripe s/n 2837X, dated 7/99, preamp CP1H

 

[woodbutcher65]

It'll cost me some money but I'd like to eventually have one Mark III of every stripe color. Even better if they're all in the same configuration. (DRG)

 

[CoG]

So here's something new to consider. I just acquired this amp, and I can find zero references to the CP1H. Ever heard of it? This is my 8th Mark III and I have to say this one has some feels that the others didn't have, save for maybe the early no stripe III+ SG.

- Blue stripe s/n 2837X, dated 7/99, preamp CP1H

7/99? On a mark iii? That seems late. I'd call Boogie and see if one of the old guys can explain, you might have a unicorn there

 

[user 41014]

The coli numbers are not correct but GJ has just about the last blue stripe made.

 

[SKULL_KRUSHER]

Just out of curiosity, how much does a blue stripe go for these days? I'm potentially looking to sell mine to fund another amplifier endeavor.

 

[Theresa]

I have one of the first DRG III's if you need help ID'ing anything. This one has the 105 Transformer. SN:15128

 

[GJgo]

Just out of curiosity, how much does a blue stripe go for these days? I'm potentially looking to sell mine to fund another amplifier endeavor.

In the last year or so I've seen them sell in the $1800-$2500 range depending on condition, service history, III+ mod, etc..

 

[quad48uF]

Just picked up a Green Stripe DRG from May 1994 with the CP1H preamp board

 

[knoe8790]

Is this where I should be to get someone to build me a mark iii purple?

 

[knoe8790]

Is this where I should be to get someone to build me a mark iii purple?

 

[user 41014]

Is this where I should be to get someone to build me a mark iii purple?

I don't think the factory is taking any more requests to build a purple stripe.

With a couple thousand sold you are in luck that you don't have to wait for someone to build you one tho.

You could always send Gibson an email every day as a different sender stating how you would pay them insane money if they offer a Mark III mod shop - any stripe, power section, preamp, tolex, dummy light color, and maybe they will force Petaluma to do a re-issue mod shop for 4 times what the amp is worth before they realize you were the only person willing to pay for it.

No?

 

[TheTubeAlchemist]

I made a simulation of the RED, GREEN and BLUE stripes, just for fun, based on the circuit you posted.

I only included in this simulation V1B, the lead circuit with its feedback loop, and the following triode. So the input signal is a 0dB signal that replaces the tone stack output.

Keep in mind that this kind of simulation doesn't account for harmonics introduced by tube clipping (they assume the valve can get any signal without distorting the signal), so what you're seeing here is just the contribution of resistors and capacitors to frequency, while tubes are amplifying the signal "transparently".

What I see from this graph is that the blue has more gain than the red, while the green should have a fatter sound and be slightly less aggressive than the blue.

Do you find any match with what you experience from the true amp? I know that this ignores much of the dynamics of a distorted sound, but should be somewhat relatable.

 

[mark0]

yep this looks dead on
red does best c+ scooped
blue is tuned and can step on any marshall voicing
green is similar to a blue on steroids and tuned even more aggro but w the perfect Mark amp hair

 

[user 41014]

I made a simulation of the RED, GREEN and BLUE stripes, just for fun, based on the circuit you posted.

I only included in this simulation V1B, the lead circuit with its feedback loop, and the following triode. So the input signal is a 0dB signal that replaces the tone stack output.

Keep in mind that this kind of simulation doesn't account for harmonics introduced by tube clipping (they assume the valve can get any signal without distorting the signal), so what you're seeing here is just the contribution of resistors and capacitors to frequency, while tubes are amplifying the signal "transparently".

What I see from this graph is that the blue has more gain than the red, while the green should have a fatter sound and be slightly less aggressive than the blue.

Do you find any match with what you experience from the true amp? I know that this ignores much of the dynamics of a distorted sound, but should be somewhat relatable.

View attachment 3996

Any chance you can add purple to the simulation? I can get you any component values.

 

[TheTubeAlchemist]

yep this looks dead on
red does best c+ scooped
blue is tuned and can step on any marshall voicing
green is similar to a blue on steroids and tuned even more aggro but w the perfect Mark amp hair

I think I'm still not grasping how the "R2 CHANNEL RELAY" works, plus It's just the simulation for a piece of the circuit, so definitely don't take it as gospel. I'll come up with a more sophisticated analysis asap.

I don't own a real Mk3, but a Quad, which should be somewhat similar in its L2 channel (but with lots of weird values and redundant resistors possibly related to how they implemented channel switching), so It's the only "reality check" I have at hand.

If I decode the Mk3 schematics correctly, when the relay is active (which mode is which?), R133 and C28 should be grounded, as well as R105 and C47, but it's not what I'm doing here.
This would be also different to what I found in my quad, that is R133 + C28 in parallel with C27 and R106 and R105 // C47 to ground. This would be the classic arrangement that we usually see in modded IIC+ or in the III. Maybe someone more expert than me can chime in.

Now, I'm trying to experiment with many different models using the "classic arrangement" above, and I noticed some funny things from simulating them:

- C47//R105 play an important role on taming high frequencies and interacts highly with the other two components, especially in lead mode when they are in parallel with the lead EQ shelving circuit.

- My 1990 Quad has NO R105 and C47 is 47pF, so it's more akin to a red Mk3. However it's soaring, beautiful and uncontainable, in a very different way compared to the IIC+ on channel 1. So, by looking at the values it is a Red stripe with a different lead output EQing, (that makes it loose a lot of volume) but it's also nothing like a IIC+.
Weird thing is that if I simulate it, I get a crazy ramp up of ultrasonic frequencies, so it could oscillate wildly. May be a problem with SPICE, but I would put a resistor there anyway. They seem to have put a global patch by putting a feedback loop between the tone stack and the input.

- Most importantly, in the BLUE STRIPE, I get that the with C47 at 82pF just eats out most high frequency content, maybe too much. This may be due to the 2.2M resistor swap, but It's late and I don't feel like doing the math now. So, the question mark on the schematics may be more relevant than you think.
I tried fixing it by putting there the classic 680k/47p (like in the purple or IIC+) and I get the most "1KHzish-ear-piercing" frequency response of all the II(I/C+/C++), which can be consistent on how everyone is describing it.
...but then I get a Green Stripe which is annoyingly close to a IIC+.

Any chance you can add purple to the simulation? I can get you any component values.

If you can get any component values than you could double-check the schematics that jrb posted, it would be much more useful for everyone to have everything in a single place. However, yes, I can do it as soon as I have some spare time. However, being the primary difference in how the clean and lead frequencies get mixed, my bet is that I wouldn't get a graph very different from that of a IIC+. After all as this stage is concerned it's just a Green stripe (in triode mode) with the Lead output EQing of a IIC+, so it may stay in that ballpark. After all, in these simulations you can't actually see the combination of clean and distorted signals anyway, which is what may be the real difference.

So, this is what I'm getting now. Again, take it with some grains of salt since there's a lot going on in this circuit and may be entirely wrong.

 

[FedWar]

I think I'm still not grasping how the "R2 CHANNEL RELAY" works, plus It's just the simulation for a piece of the circuit, so definitely don't take it as gospel. I'll come up with a more sophisticated analysis asap.

I don't own a real Mk3, but a Quad, which should be somewhat similar in its L2 channel (but with lots of weird values and redundant resistors possibly related to how they implemented channel switching), so It's the only "reality check" I have at hand.

If I decode the Mk3 schematics correctly, when the relay is active (which mode is which?), R133 and C28 should be grounded, as well as R105 and C47, but it's not what I'm doing here.
This would be also different to what I found in my quad, that is R133 + C28 in parallel with C27 and R106 and R105 // C47 to ground. This would be the classic arrangement that we usually see in modded IIC+ or in the III. Maybe someone more expert than me can chime in.

Now, I'm trying to experiment with many different models using the "classic arrangement" above, and I noticed some funny things from simulating them:

- C47//R105 play an important role on taming high frequencies and interacts highly with the other two components, especially in lead mode when they are in parallel with the lead EQ shelving circuit.

- My 1990 Quad has NO R105 and C47 is 47pF, so it's more akin to a red Mk3. However it's soaring, beautiful and uncontainable, in a very different way compared to the IIC+ on channel 1. So, by looking at the values it is a Red stripe with a different lead output EQing, (that makes it loose a lot of volume) but it's also nothing like a IIC+.
Weird thing is that if I simulate it, I get a crazy ramp up of ultrasonic frequencies, so it could oscillate wildly. May be a problem with SPICE, but I would put a resistor there anyway. They seem to have put a global patch by putting a feedback loop between the tone stack and the input.

- Most importantly, in the BLUE STRIPE, I get that the with C47 at 82pF just eats out most high frequency content, maybe too much. This may be due to the 2.2M resistor swap, but It's late and I don't feel like doing the math now. So, the question mark on the schematics may be more relevant than you think.
I tried fixing it by putting there the classic 680k/47p (like in the purple or IIC+) and I get the most "1KHzish-ear-piercing" frequency response of all the II(I/C+/C++), which can be consistent on how everyone is describing it.
...but then I get a Green Stripe which is annoyingly close to a IIC+.


If you can get any component values than you could double-check the schematics that jrb posted, it would be much more useful for everyone to have everything in a single place. However, yes, I can do it as soon as I have some spare time. However, being the primary difference in how the clean and lead frequencies get mixed, my bet is that I wouldn't get a graph very different from that of a IIC+. After all as this stage is concerned it's just a Green stripe (in triode mode) with the Lead output EQing of a IIC+, so it may stay in that ballpark. After all, in these simulations you can't actually see the combination of clean and distorted signals anyway, which is what may be the real difference.

So, this is what I'm getting now. Again, take it with some grains of salt since there's a lot going on in this circuit and may be entirely wrong.

Super nice.
May I ask what is the name of the program you are using?
Thanks