Dsl1 (2018 Vietnam Version) Mod's

[grappo-soda]

Hi all!
I’m a bit confused on several opinions that there is too much treble response from this amp.
Making some guesstimate calculation (using spec sheet on ECC82 and the reported OT winding earlier in this thread), one would expect the transformed source impedance to be ~4ohm, ie less than 16ohm nominal load.
That would imply reduced response for higher frequencies (and also around speaker/cabinet bass resonance) by the typical impedance curve.
Perhaps designer did consider this, but over-compensate the EQ to be overly bright?

 

[acevh00]

Hi all!
I’m a bit confused on several opinions that there is too much treble response from this amp.
Making some guesstimate calculation (using spec sheet on ECC82 and the reported OT winding earlier in this thread), one would expect the transformed source impedance to be ~4ohm, ie less than 16ohm nominal load.
That would imply reduced response for higher frequencies (and also around speaker/cabinet bass resonance) by the typical impedance curve.
Perhaps designer did consider this, but over-compensate the EQ to be overly bright?

Could be a right argument....
I just sold mine (sixth Marshall sold, without being really satisfied). Clean channel very bright, but useful with pedals, and gain channel (despite new high-priced tubes that anyway damped the problem a bit) too dark and dull.

 

[grappo-soda]

Comparing to other available schematics for DSL amps (ie DSL20) the preamp is very much the same, other than a few decouple caps and some tweaks to resistor values. Have not analyzed/simulated differences though. Power stages (after fx loop) are completely different of course.

The extreme red channel is another interesting topic. I understand some swap to lower gain triode (ECC82) in first stage of overdrive circuit. I wonder if lowering the gain in second stage would be worth trying instead, as ECC832 tubes are more available than suggested ECC823. Perhaps a better idea if cold bias is also moved to first OD stage?

 

[gdvsbp]

Comparing to other available schematics for DSL amps (ie DSL20) the preamp is very much the same, other than a few decouple caps and some tweaks to resistor values. Have not analyzed/simulated differences though. Power stages (after fx loop) are completely different of course.

The extreme red channel is another interesting topic. I understand some swap to lower gain triode (ECC82) in first stage of overdrive circuit. I wonder if lowering the gain in second stage would be worth trying instead, as ECC832 tubes are more available than suggested ECC823. Perhaps a better idea if cold bias is also moved to first OD stage?

I was going to ask something about this today, and here I think you’ve about confirmed it for me. And now just to be really sure, here’s the question: I like the clean channel as-is, but would like to seriously reduce gain in the red channel. Would an ECC823 in V1 and an ECC832 in V2 reduce gain in 50% of the red channel’s gain stages while leaving the green’s gain as-is? I think that lines up with what I read elsewhere in this thread but would like to confirm before trying to source both tubes.

Thanks!

 

[fitz288]

I was going to ask something about this today, and here I think you’ve about confirmed it for me. And now just to be really sure, here’s the question: I like the clean channel as-is, but would like to seriously reduce gain in the red channel. Would an ECC823 in V1 and an ECC832 in V2 reduce gain in 50% of the red channel’s gain stages while leaving the green’s gain as-is? I think that lines up with what I read elsewhere in this thread but would like to confirm before trying to source both tubes.

Thanks!

I can say that the JJ ECC823 in V1 works at reducing the gain on red and leaving the green as is.
There's still plenty of gain when cranked, it's just more controllable in between 1 & 10
I've never heard of putting an 832 in V2 - not saying it won't do something, I've just never tried it.

 

[gdvsbp]

I can say that the JJ ECC823 in V1 works at reducing the gain on red and leaving the green as is.
There's still plenty of gain when cranked, it's just more controllable in between 1 & 10
I've never heard of putting an 832 in V2 - not saying it won't do something, I've just never tried it.

Excellent; that confirms the theory then, based on the signal path stated on page 4. I’m not looking for a lot of gain at all. (Why a DSL then? 3-band EQ, effects loop, master volume, and 0.1 watt power scaling doesn’t really come in any other packages.)

I’ve got a few 12AU7s sitting around, so I’ll swap those into V1 and V2 to see if it’s doing what I want to the red channel. If so, I’ll source the JJ ECC823/32 combo.

 

[grappo-soda]

The schematics were posted earlier in this thread (much obliged!) and from this it is clear the green/clean channel use 2nd triode of first tube and 1st triode of second tube. The OD circuit is then put in between these two stages. That means if you want to mod the red channel but keep green channel as is, you will need to use these odd mixed triode tubes such ECC823 in V1 or ECC832 in V2.

Simplistic calculation would predict you get 11.5/26=44% (let’s say abt half) relative gain in first stage of OD circuit by using a ECC823 in V1. Also putting a ECC832 in V2 would reduce total gain even further (by factor 6.7/10=67%), but I think in practice it’s a question of either-or if you want some OD gain remaining on red channel.
Also, as 2nd OD stage is a cold clipper i think most of that character would go away if you swap that triode?

 

[grappo-soda]

I’ve got a few 12AU7s sitting around, so I’ll swap those into V1 and V2 to see if it’s doing what I want to the red channel. If so, I’ll source the JJ ECC823/32 combo.

That wont work, unfortunately, as the red channel effectively uses all stages in both tubes. Putting eg ECC82 in V1 will affect both first two gain stages in red channel, not just one.

 

[gdvsbp]

That wont work, unfortunately, as the red channel effectively uses all stages in both tubes. Putting eg ECC82 in V1 will affect both first two gain stages in red channel, not just one.

Oh duh, that makes sense. Thanks for pointing that out.

 

[Mavrim]

What if the DSL1 fits the old Lead12 cab? And 12” speakers can fit in Lead 12. Not suggesting anyone should mod anything but Lead 12 does look a lot like DSL1.

I had purchased a broken tattoo JVM1 head. After repairing it, I sold the gigantic small box cabinet and trimmed the chassis to fit into a lead 12 cabinet. Worked out great! I unfortunately sold it because I wasn’t using it much. But the JVM1 is a great amp.

 

[Mavrim]

I just came across this thread that I started out regarding the extra speaker tap. I ended up putting the amp in the closet and moved on to a DSL20H that I absolutely love!

So cool that we finally have a schematic! I had called marshall about a problem on the JVM1 and they emailed a schematic for it. I posted that a while back!

One thing I did on my DSL20H to match the older 1x12 cabinet was too make the shiny tolex look more like the elephant tolex on UK Marshall’s. I removed all plastic corners that are screwed on. I scuffed the tolex lightly with scotchbrite and then used armour all to add shine. I used gold marshall rivets to replace the screws. Now head is very close to matching the order tolex. I hate that fake shiny tolex they are using!

 

[grappo-soda]

Did anyone measure HT1-5 voltages?

 

[Harbard]

.

 

[Rocketman]

Happy Thanksgiving to everyone!!!

 

[Rem223]

This is a great thread. I really like my little DSL only wish the red channel had less compression and was more open sounding. Seems no mods have been posted to address this but possibly by using a ECC823 in V1 it might help.

Also, does anyone have the full original Marshall schematic for this yet that they could post?

thanks!

 

[fitz288]

using a ECC823 in V1 it might help

to the forum.
Give the 823 a try, it does make a difference.

 

[WiderGates]

So here it is:

 

[gregr]

Now we need the DFX board.

 

[Max Gahne]

The schematic @WiderGates posted has matched up to the DSL1 I bought in Nov 2020 although in recently doing some tinkering I noticed R91 on mine is 100K. Just pointing out that it looks like there's been that change since that schematic and there might be more.

 

[Rem223]

So here it is:

Thanks!

I’ll check it out against my amp and see what we got!