Mark V 90W combo speaker change

[MarkV User]

Thank you! It was a little more than I thought I'd have to do, but it has been fun.

I had no idea the Lonestar Special was such a great amp. It sounds amazing in that video.

I played the Mark V through two different cabinets with no other changes except another 12AT7 preamp tube swap at V2. That was too much gain reduction if I want to maintain the character if the amp. With a balanced 12AU7 in the V7 phase splitter position and 12AT7s in V4 and V1 the depth and tonal flexibility of the amp is greatly enhanced, and the high gain tones just sound better. The volume knobs are usable for more of their range. And so between master, volume and gain there is a lot more control over the tone. Going one step further and knocking V2 down to a 12AT7 puts the amp in a very beautiful sounding classic rock place. You get lots of clean and crunchy tones from all three channels, and classic medium gain from channel 3. The IIC+ mode is particularly nice. But the really high gain tones are no longer available and the amp won't get as loud if that is important.

But with any of these gain reductions I'm able to get good amp dynamics that I thought the Mark V couldn't do. The V1 change helped for that, and I will experiment more with which stages benefit the most from reducing gain later.

I get why lowering the gain improved the distortion and harmonics of the amp but I don't yet understand how it seems to improve the dynamics. My best guess is that early gain stages have a higher impedance power rail, and so maybe they contribute more tube compression and have a longer gain recovery time (like the release time knob on a rack compressor). If a later stage saturates then all of those dynamics are lost. If a later stage is on the verge of saturating or is partially distorting and adding harmonics then the dynamics introduced by the early stage will modulate the distortion in a complex and interesting way.

I'll be able to finish my changes and button up the project soon. I'll document what I did a little more. Thank you again for all your help and also to everyone who contributed. I'll reread this thread as I'm finishing up and make sure I've implemented as many of the tips as possible. I didn't really believe it would happen but my amp already sounds amazing.

 

[bandit2013]

The schematic reveals what is going on with each of the triodes. Each will have a letter to designate a net name of the voltage, A, B, C, D, E. A few of these voltage nets change with respect to channel selected. There are some relays that connect other resistor voltage dividers or disconnect them to alter the triode plate voltages. A and B and C will remain constant, D and E are the two that get altered based on channel bus selected. This is in the power supply area of the schematic. There are also JFETs that may pull in more cathode bypass capacitance. V1A is the only stage that has a modified plate resistor for CH1 and CH2 (edge increases the plate resistance by 47k ohms). There is also a relay that alters the cathode resistor for crunch mode on V1A as well. I doubt the length of copper runs on the PCB will be sufficient to result in impedance changes. It is all based on stage definition of the various relays and other circuits. These changes in plate voltage, cathode resistors, plate resistors, and bypass capacitors will alter the characteristics of each gain stage based on the channel used and what voicing mode is active for that channel.

 

[MarkV User]

That all makes sense. I pulled up the schematics to follow along, and meanwhile I have an update. First here's the update:

I've been playing the Mark V in four different speaker setups in advance of making more changes to the circuit:

• Mesa Boogie MC90 in Mark V combo (8-ohm)
• Celestion BN12-300S in Mark V combo (8-ohm)
• Marshall 1960A with 4x12 G12T-75's (4-ohm wiring)
• Compact Marshall 4x12 with older Vox-branded G12H-30's (8-ohm wiring)
• Bogner horizontal 2x12 cab with one V30 and one BN12-300S (16-ohm cab with an 8-ohm load in parallel for testing)

My modified Mark V amp (Mark III only in the feedback loop, no other circuit changes, 5881 and 6L6GC power tubes, many different preamp tube configurations have been tried) sounds great everywhere I've tried it now. It was excessively bright with the MC90 and sounds best with the old G12H-30s. It feels ear safe and my ears no longer ring after I play the amp. I do usually play with the presence turned down to 9 o'clock, but it doesn't get the horrible high frequency singing when I turn it up now. I feel the amp is great in its current configuration. But it is also a matter of taste. It nails the dead-sounding Nirvana tones for example, and from preamp tube selection I have it tuned so the Mark I, Mark IV and Extreme modes will give me the highest gain I want to use if I turn the gain knob up all the way. I have other amps for pretty sounds. The distortion palette is huge and I can make it do so many different things. I feel very happy to have this amp, which is the opposite of how I felt when I started with it.

The BN12-300S is *awesome* paired with a V30 in the 2x12 cab and not quite present enough on its own in the Mark V combo. It is fine for now, and at some point I'll try either an EVM 12L or the Celestion Redback. And I know the Redback is the safe match as the amp is currently voiced, but I like the clarity and flatness of BN12-300S, and the EVM 12L is like that. If the EVM 12L were too present I could fix that easily with a circuit change at the risk of affecting the sound in through speaker cabs where it is currently good.

The treble and presence sound more "rational" than they sounded before the modification. I spent a lot of time looking at the FFT spectra and scope traces of the speaker output voltages before the change and only a little bit after the change, and I probably won't be able to understand why the amp sounds so much better and no longer hurts my ears with the instrumentation I'm using. In some ways the amp sounded brighter after the change. I think there was a resonant peak somewhere in the 4-5K range before change, but my records and memory of all the tests are not good enough to be 100% sure.

Now after considerable time playing the amp in its intermedate state, I think the priorities are something like this:

1. Improve the master and channel volume knobs. Add large-value resistors across all the volume and gain pots to improve the resolution at the low end of the adjustment range. The current behavior is that each of those audio pots wastes a lot of its low-slope range of motion in the regime where no sound comes out. I'm going to use a decade box for this.
2. I'm going to add a small capacitor across the phase splitter to attenuate frequencies higher than the speaker can reproduce. These high frequencies don't belong in the feedback loop because they won't be reproduced by the speaker, or that is my current reasoning. The range of capacitances that might make sense are 75pF-500pF, but something in the 100pF-250pF range is the most likely choice based on what I see in the Mark IV. In earlier testing I found that 500pF in that position would round off all the ear-hurting presence but it also made the amp sound darker than I wanted.
3. Increase the main series resistor in the feedback shunt to normalize the presence knobs to be neutral in the middle position. I'll do this with the scope and the decade box.
4. I am sorely tempted to convert the presence feedback loop to look like the Soldano SLO-100 instead of a Mark III series amp. At least I might want to try it out and hear what the difference sounds like. I'm pretty happy with the current voicing, and still I have a feeling the SLO/Marshall/Bassman feedback circuit would give more natural sounding overtones. If I don't make that change then I should at least try different values for the negative feedback bypass cap.

For pF scale capacitors I have a set of C0G/NP0 1206 chip caps and a bunch of random leaded caps. I think it is unnecessary, but to be sure of getting it right I've ordered a selection of leaded silver mica caps as well and will build a small switch box so I can try out various capacitor values and types at each position. My best understanding is that the C0G/NP0 caps will perform as well or better than silver mica, but I want to be sure.

I'll follow along with you and look at the supply voltages in the schematic in my next message.

 

[MarkV User]

OK, I'm looking at the amp and preamp supply voltages. Thank you for prompting me to look at this further.

I'm going to use the 1959 Mk. II Plexi, SLO-100, and HiWatt DR103 as references because they are all good sounding amps with very different dynamics. I feel like the DR103 is not so far from the MarkV in character and might be a good target to aspire to.

The first part of the MarkV volume dynamics are C83 and C84, which give a total of 110uF between supply A (center tap of the output transformer) and Ground. This is very stiff compared to the 1959 Mk. II Plexi where we have only 50uF, and it affects all subsequent stages. This also bears thinking more about. If we turn the Plexi up loud then all premp stages will see some voltage sag, which looks like lower headroom and lower output from all the preceeding preamp stages. The choke causes a short delay before the effect takes place, just like the attack time of a rack compressor. It also limits the effect to low frequencies, so the effect is more dynamics and less in the audio spectum. I know from experience with the 50W version of that Plexi that the amp has to be turned up quite loud to get this kind of effect, and so to get the same effect from a 100W Plexi would be louder by a factor of 2 in power, and for the MarkV:90 it would be a further factor of 2, but in voltage. So does that add up to +9dB louder than a 50W Plexi before tube dynamics from the power stage kick in? I know I won't hear that effect without ear protection and bothering the neighbors.

Now by contrast the Soldano SLO has twice as much capacitance for the output transformer center tap, at 200uF, so there is really very little compression effect at that first stage. That is one reason the SLO-30 sounds almost identical to the SLO-100, just at lower volume. The HiWatt DR103 (100W) has the same 110uF as the Mark V.

My take on this is that the Mark V is ok. Yes, I can imagine that halving the first stage capacitance might improve the amp when cranked, but the strong dependence of dynamics on speaker volume is not that desirable. It would be cool to have a selector switch, but there is no room to fit that kind of modification on this amp.​

Between the center tap (plates) supply and grid supply we have a choke (only) on the Plexi with 50uF bypass capacitance, then a 1K resistor at the grid of each tube. The Mark V has its choke and either 470 Ohm or 1K resistors at each grid, with a 30uF bypass cap. The SLO has a choke and another 100uF of bypass capacitance at the grid supply and 475 Ohm grid resistors--again, very stiff! And on the HiWatt things are a bit more complicated with a 100 Ohm resistor instead of a choke between plates and grids supply, and 40uF bypass to ground, but organized so that the grids and phase splitter preamp supply together each have 50uF to a common point and then that common point has another 220uF to ground. The supply sag compression is a first-class part of this design.

The Mark V with a slightly lower bypass cap here should provide a little bit of power stage compression from the sag of the bias supply at higher volumes, all other things being equal. The differences in choke inductance between the amps also plays into this though, and I don't know those values without doing more research.​

Between the grids and phase splitter the Mark V has--as you said, 5.6K on channels 1 and 2 and 2.5K on channel 3, with 33uF bypass capacitor for the phase splitter plates. The plexi has a 20K resistor and then 100uF of bypass capacitance for the phase splitter plates, with 82K/100K for each of the two plates. The Soldano has 10K and 40uF. The HiWatt has that same funny arrangement with 50uF/50uF for screens and phase splitter plates both into 220uF to ground. And there is just a 1K resistor between the grids and shared plate supply for the phase splitter on the DR103, so we see the grids and phase splitter plates coupled pretty closely together.

The Mark V is closest to the HiWatt again here.

Past this point, the HiWatt really diverges from everything else. Because the plate of the master volume and its companion triode are tied to the input plates, which is higher impedance, at the end of the resistor chain. What a cool design this is, because whatever the gain and overhead reduction from the voltage sag at the input will be applied again--effectively squared--in this master volume stage that drives the phase splitter. Then the other triode in that tube is dedicated to sensing the supply sag from the input stage and further enhance it by lowering the bias of the phase inverter grids when the voltage sags. There is a 10K resistor from the phase splitter plates to the gain stage / tone stack driver. Then another 10K to the guitar signal input tubes.

What I see in the Boogie preamp stages is that we don't have the same intentional compression design in most places. It looks like the design is more like a conventional power supply, designed for isolation. Except in one place. Supply D has 47K and is used only for the effects return / master volume driver. The Soldano and Marshall have the familiar chain of resistors and various bypass caps from the power stages back to the preamp stages.

There is an interesting opportunity here maybe. Look at R4 on the page Preamp1 page of the MarkV schematic. This could be disconnected from E and connected to D instead. That way the stiffness of Channel 3 would remain the same but the input gain would be reduced in reaction to saturating the effects loop and the dynamics of channels 1 and 2 would be enhanced. This is not as far as HiWatt goes, but it is one of the two main tricks they use and it would be interesting to try.​

 

[KRichter]

Are you in France? I may have found some of your playing on YouTube, and if that is yours then your amp sounds really good. If I can get a sound similar to that I'll be happy.

You are one of several people who told me that what worked for them was an oversized 2x12 with one smooth and fairly flat response speaker (the shadow or EVM 12L and one speaker with a pronounced midrange).

I'm going to try an oversized Bogner 2x12 with a V30 (because I have it already) and a Celestion BN12-300S bass speaker because it has a flat response, works good into the lower presence and rolls off the upper presence. That speaker sounds good for guitar clean and dirty and heavily distorted. The V30 is also not too bright but will carry the presence better. So maybe they will be a good match.

In the combo I will use the BN12-300 by itself. I think that is going to sound a little dry, which might be the tone to go for with this amp. If I think I can get away with it later I'm also interested in trying the EVM 12L... when I'm sure I've made the presence circuit work.

Sorry not in France, Cleveland Ohio, States. I'll tell ya, some of that depends on the guitar as well, The cab I'm using now is a Genz-Benz, deep cab, George Lynch 2/12, (You can catch them cheap around 250usd) Super heavy duty, marine ply, One Shadow and one Emenice Texas Heat. I have a loaded custom Tele, Ernie Ball Petrucci and and 04 Les Paul, each are a different animal.. Usually it a matter of kicking up or down the presence according to the room...

 

[MarkV User]

Sorry not in France, Cleveland Ohio, States. I'll tell ya, some of that depends on the guitar as well, The cab I'm using now is a Genz-Benz, deep cab, George Lynch 2/12, (You can catch them cheap around 250usd) Super heavy duty, marine ply, One Shadow and one Emenice Texas Heat. I have a loaded custom Tele, Ernie Ball Petrucci and and 04 Les Paul, each are a different animal.. Usually it a matter of kicking up or down the presence according to the room...

Thanks! I paid more attention this time and listened to some video comparisons of the Texas Heat and Swamp Thang. Those are really nice speakers. I'll look at them more closely.

That cab sounds intriguing and I'll keep an eye out.

You are using an EVM type Black Shadow right, not an MC90 Black Shadow like I have?

I've tamed the shrillness of my amp now, although it is brighter than before my changes. If it doesn't seem to make sense, there was an irrational high frequency peak that would happen when a hard-clipped signal hit the power amp. Now I've reduced that effect with a change to the presence/feedback loop but the amp seems to be brighter now without being so hard on the ears.

I'm going to take the amp out to the guitar store and play it against a Mark VII and a Mark V:30, and also have some other people play it to get a better idea of where I'm at before making more changes.

I've spent a lot of time swapping preamp tubes to try to get the gain structure as rich as possible and I found a combination I like. I might be able to improve it more with a 12AY7 to mellow channel 2 a little. I've seen that Mark I mode can sound massive with lower gain stages. Every other mode sounds great. I am particularly happy with the crunch sounds I can get from channel 1 with gain cranked in the current configuration.

The Celestion BN12-300S in the combo is a little too dark. I might try a Redback, an EVM 12L, or a Swamp Thang there.

In the big Bogner 2x12 the BN12-300S and V30 are really nice. When I got the cab it had a V30 and the Celestion Classic Lead 75, which is a harsh and grainy sounding speaker that compliments the V30 well... But it is just not my sound. It makes a very recognizable 1980s heavy rock solo sound. I have a few ideas but the EVM 12L might be another good match for the BN12-300S there.

 

[KRichter]

Thanks! I paid more attention this time and listened to some video comparisons of the Texas Heat and Swamp Thang. Those are really nice speakers. I'll look at them more closely.

That cab sounds intriguing and I'll keep an eye out.

You are using an EVM type Black Shadow right, not an MC90 Black Shadow like I have?

I've tamed the shrillness of my amp now, although it is brighter than before my changes. If it doesn't seem to make sense, there was an irrational high frequency peak that would happen when a hard-clipped signal hit the power amp. Now I've reduced that effect with a change to the presence/feedback loop but the amp seems to be brighter now without being so hard on the ears.

I'm going to take the amp out to the guitar store and play it against a Mark VII and a Mark V:30, and also have some other people play it to get a better idea of where I'm at before making more changes.

I've spent a lot of time swapping preamp tubes to try to get the gain structure as rich as possible and I found a combination I like. I might be able to improve it more with a 12AY7 to mellow channel 2 a little. I've seen that Mark I mode can sound massive with lower gain stages. Every other mode sounds great. I am particularly happy with the crunch sounds I can get from channel 1 with gain cranked in the current configuration.

The Celestion BN12-300S in the combo is a little too dark. I might try a Redback, an EVM 12L, or a Swamp Thang there.

In the big Bogner 2x12 the BN12-300S and V30 are really nice. When I got the cab it had a V30 and the Celestion Classic Lead 75, which is a harsh and grainy sounding speaker that compliments the V30 well... But it is just not my sound. It makes a very recognizable 1980s heavy rock solo sound. I have a few ideas but the EVM 12L might be another good match for the BN12-300S there.

Yep I have the Shadow C90 in there. One thing for sure, with the MKV, you can sure go down a rabbit hole, I sure did. I have a cool tone after several years and just play into the amp.

 

[MarkV User]

I played it in that configuration I ended up with for these recent months. It is so much better that I haven’t wanted to change with it. It doesn’t hurt my ears anymore. It is a really different amp now, and I can get it to do all kinds of things. It is like the amp I originally thought I was buying. It is probably not able to be a good heavy metal amp, but the range of distortion sounds is a lot bigger than before and I do get interesting focused high gain sounds.

Now I’m going to try swapping the BN12-300S speaker for the 300W EVM 12L. The BN12-300S is a better sounding speaker than the MC-90 I had, more musical and even flatter. The MC-90 had a lot of resonant modes that sounded bad, and would distort unpleasantly at low volumes. I could hear the problems clearly when I drove the amp with a signal generator. The BN12-300S is just heavenly but rolls off too much of the presence. So one of the two EV speakers might be the ticket, and I’m trying the darker one first.

 

[bandit2013]

Hmm, if they are older models of the EV then there will be some differences in tone. Considering the current available EVM12L speakers. Black Label vs the Classic, there is not much of a tone difference. The EV BL is a bit tighter response and not forgiving on frequency, there will be no HF roll-off. The EV Classic is just a tad different, a little softer on the low end, but still retains the upper frequency response as the Black Label.

To date, the only amps I ever appreciated with the EVM12L speaker was the Mark III DRG, JP2C and the Mark VII. It was a no-go for me with the Mark V90. It did not work all that great with the Mark IVB. However, the Mark IVB was just a honky box sound and not as much of an ice pick as the V90. Sort of makes me wonder if Mesa had an alternate supplier for the dc blocking caps or the parts they made use of in the phase inverter. It could also be related to the OT as well. Trying to figure it out is beyond my skill level.

 

[DECEMBER]

I love how bright and detailed my Mark V 90W is. I've never had a clean channel sound so nice. And the high gain on channel 3 is so clear and detailed in the mids yet not harsh and fizzy in the high end. I have to crank the treble and presence on my other amps to get rid of their muddiness, but it makes them so fizzy and harsh, and boosting the mids makes them sound boxy, they just don't compare to the openness and detail of the Mark V.
The Mark I setting on channel 2 is pretty dark and bass-heavy to me. I have to cut the bass to 0 on both EQs and crank the treble and presence to make it usable. I'm also using a parametric EQ in the loop, to fine-tune the mids and low-pass the high treble (usually at somewhere between 5kHz - 10kHz).
I have the head with 6L6GC and two 2x12s both connected. One has 2 V30's, one has 2 Jensen Tornado Stealth 80's. I swapped one of the V30's for a G12K-100 but it got a bit harsh with high gain, so I put the V30 back, but I do like the G12K for cleans. The V30's and Tornadoes have a smoother high end and boosted mids, not quite at the same frequencies, and the V30's have more low end. They're a little dark to me, not my favorite speaker, but I think it would take a really bad speaker to make this amp sound bad.
I'm actually wanting to get a Celestion Black Shadow C90 soon, because I played a Mark VII combo at the local store and it sounds perfect.

 

[bandit2013]

What I found the most interesting was a combination of the V30 and MC90. I had a wide body 112 open back cab that came with the MC90 loaded in it. I paired it up with the Recto Vertical 212 cab and wow. The C90 complemented the V30 quite well. If you are going to get an MC90 black shadow, it is better served in an open back cabinet. V30 will work either way but it excels in a closed back or seal cab.

 

[rarebitusa]

What I found the most interesting was a combination of the V30 and MC90. I had a wide body 112 open back cab that came with the MC90 loaded in it. I paired it up with the Recto Vertical 212 cab and wow. The C90 complemented the V30 quite well. If you are going to get an MC90 black shadow, it is better served in an open back cabinet. V30 will work either way but it excels in a closed back or seal cab.

IIRC that the pairing in the Road King Cab Not sure if you can get them anymore.

Road King 2x12

 

[DECEMBER]

What I found the most interesting was a combination of the V30 and MC90. I had a wide body 112 open back cab that came with the MC90 loaded in it. I paired it up with the Recto Vertical 212 cab and wow. The C90 complemented the V30 quite well. If you are going to get an MC90 black shadow, it is better served in an open back cabinet. V30 will work either way but it excels in a closed back or seal cab.

My 2x12 with V30's is a closed-back, but the back panel is removable. I keep it off because I find it to be more boxy with it on. I also have a plywood 1x10 I assembled myself (from Tube Depot) that also sounds a lot more boxy with the back panel on. So I keep them all open.
I don't play at high volume. I don't know if that has anything to do with it. I can be quite loud for an apartment, much louder than most apartment neighbors would tolerate, but it's still nowhere near half the potential volume of my amps. I get my Mark V master volume around 40% on 45W. My Carvin V3 on 50W can barely go past 1, but I have the Deep knob cranked, and that raises the volume quite a bit.

 

[rarebitusa]

In the FWIW dept... I didn't love the Recto 2x12 in it's original form. For me it was too much v30 mid range hump goodness. Since my buddies 2x12 Road King cab rocked, made some simple mods. Had the MC90 paired with a v30 initially, which was great. Did end up swapping the MC90 out for a Scumback M-75, which was just as good. I prefer the semi open back on stage as it's not so directional.