New DSL40CR challenges

[TheKman76]

Afternoon all from Aus.

I'm a new DSL40CR owner, never owned a proper valve amp before, though I'm very familiar with the workings and have an education in Electronic Engineering.

Now, while I accept that the DSL40CR fits into the budget-friendly end of the Marshall spectrum of valve driven gear, I'm experiencing a couple challenges which feel like quality problems. Here's hoping I'm just missing something.

The first is biasing the power valves. I understand that the accepted normal bias current for the factory EL34 valves is in the 30-40mA range, centred around 35mA being ideal. Having looked at the schematic I see two 1 Ohm resistors used for measuring bias current, so measure 35mV and your golden. Simple.
However, I measured the bias resistors at 0.61 and 0.64 Ohms, not even close to 1 Ohm. Looking over the schematic again I can see no reason for this, so I ran with 21.4mV and 22.4mV respectively. The problem I face is that one valve is adjusted as cold as it will allow while the other is right in the middle of it's range. I estimate they're biased at around 37mA because I ran out of adjustment on one valve.
I realise a more precise bias measuring device is probably a good idea and I'll look into that too. I also note that the schematic I'm working from is for a DSL40C which may be slightly different. But, am I missing something, or is there a quality problem?

Second issue is with MIDI switching the Master volumes. I've read that this is challenging at the best of times and many simply give up on the notion completely. I spent a good two hours yesterday and I simply cannot understand under what circumstances this works as expected. Switching channels is fine, toggling the effects loop also fine, these I have configured on a dozen different patches in the GT-1000. However, I can only make the MV switch seemingly randomly. Even attempting to replicate an exact sequence of events doesn't seem to produce consistent results. I've literally been through the same sequence ten times and found no way to predict what the MV switching outcome will be.
I understand that it will remember the master chosen for a given channel so I thought perhaps the MIDI toggle is only to either enable or disable this option, but no. There appears to be no consistency.
Am I missing something? Can someone point me to a resource which explains what CC#14 is supposed to do? Does this feature work with the Marshall 6-button controller? Am I better off just going completely insane and be done with it?

Finally there's the odd changing of channels when the unit is powered on. This is more of a curiosity than a problem, but it's very curious all the same.
About 30% of the time, if the last channel selected was anything other than 2, at power on it will change to channel 2 after about 5-10 seconds.
To be clear, this is with no MIDI attached, no footswitch, nothing connected to it in any way. Curiously, after warming the amp up and leaving it in 40W mode rather than standby as you normally would it can be powered on without spontaneously changing channels at all.
I did this test about 20 times in each of standby and 40W mode and it only changes channel at power up in standby.

So far it seems my teenage offspring are more predictable that this particular DSL40CR. I'd be vary grateful to have some input.

Thanks you all for reading my lengthy first post.

Dave.

 

[johan.b]

You're probably measuring the bias path as well as the bias resistors. Modern resistors are usually 1-5% tolerance. Unless they look burnt, I'd trust they are what they're supposed to be.

The midi problem is outside of my universe.. 🤔

J

 

[Jon Snell]

Check your meter, most are innacurate on Ohms at less than 10 Ohms due to the test leads. You need a four probe meter for an accurate reading, two probes send current and the other two measure the voltage across the test resistor.

 

[TheKman76]

Thanks gents.

You make a valid point about the bias path, I didn't remove the chassis to test. At this level of resistance it matters.

Regarding the accuracy, I have quite a bit of confidence in my meter at these impedances. The probe leads are labelled 200mOhms and regularly measure 210mOhms which I subtract from my measurements. I have a set of reference resistors at 0.5% tolerance, 1 Ohm and 2.2 Ohm which read to within 2% plus probe resistance. So, while it's not sophisticated measuring gear I do have high confidence that it's reasonably close. Certainly much closer than what the resistors read versus the schematic.

Will open up the chassis to take a closer look and take some better measurements though.

 

[SkyMonkey]

I run my DSL40CR in MIDI 4CM to a Boss GT-100.

If you are running the 40CR in 4CM to a GT-1000 then I would use the SOLO parameter in each patch to change volume, rather than 40CR Master Volume switching.
It's just far simpler, and you can change the range of volume increase for each individual patch.
With MV's it's fixed (unless you walk back up to the amp and start fiddling).
I know that sucks, but this is DSL MIDI, not JVM MIDI.
Either way, you end up using a CTL button on the GT-1000.

Also, I use the GT(-100) Patch Level setting to fix the Clean/Crunch volume discrepancy in my 40CR, not MV1/2 assignment.
So, my 40CR is effectively a 1 MV amp.
Plus, the 6-Button footswitch will work alongside MIDI on a DSL40/100, unlike the 2-Button.
So if you really want to use the MV switching, that may be the way to go (I don't bother).
It would allow you to change channel within a patch too, but I find results are mixed when trying this.

My 40CR has one fault that is, thankfully, only mildly annoying, but may be related to your start-up situation.
At start-up the FX Loop light is ON (as I set it for each channel), but the loop is actually OFF.
So, every start-up I cycle the FX Loop button on the 6-Button F/s to switch the Loop OFF and ON again.
This fixes the problem until next start-up.
It's the 'Switching Chip', and only Marshall can change it out for a new one.

Apart from that, my DSL40CR is a good 'un, so I can't really help in other areas.
If you feel you have a bad 'un, and you bought it new, get the shop to swap it out.

They are really good amps, so it's a shame when rare faults spoil the enjoyment.
And faults on these amps are rare. It may sometimes seem that every owner has a fault.
But they are the 'Squeaky Wheels' that make the most noise.
The majority that have good ones don't shout, and go under the RADAR.

IS THE DSL40CR THE BEST DAMN MARSHALL AMP

I’ve had this amp now 3 years with a pedal board going through effects loop and a speaker swap and some new tubes and bias and I honestly think it’s one of the best amps I’ve ever heard. If my house was on fire this is what I would grab on my way out the door. I do have another amp to compare...

www.marshallforum.com

Hope you find resolution @TheKman76.

 

[TheKman76]

@SkyMonkey Thanks for your input.
You're dead right about using the GT rather than relying on the MV switching. Frankly this was more about not letting the amp 'beat' me.
Haven't seen your issue with the loop switching but I did note that sometimes it doesn't recognise the button press. Having owned it such a short time I concluded on those occasions it was me not pressing the button hard enough, but it's happened far too many times to be just me ai think. I don't recall exactly but I suspect it was the first press of any given button after power up. Related? Hard to know with the loop switch as it's silent, but the channel change and MV are all audible.

Agree, the amp is a keeper. Clean and crunch are fantastic, OD channels could use some clarity but I've already got a plan for that. MV switching I can live without and the biasing issue can be fixed with some deft soldering skills. Keeper.

 

[PelliX]

to the MF, new MF'er and congrats on your Marshall amp! HNAD!

However, I measured the bias resistors at 0.61 and 0.64 Ohms, not even close to 1 Ohm. Looking over the schematic again I can see no reason for this, so I ran with 21.4mV and 22.4mV respectively. The problem I face is that one valve is adjusted as cold as it will allow while the other is right in the middle of it's range. I estimate they're biased at around 37mA because I ran out of adjustment on one valve.
I realise a more precise bias measuring device is probably a good idea and I'll look into that too. I also note that the schematic I'm working from is for a DSL40C which may be slightly different. But, am I missing something, or is there a quality problem?

A couple of thoughts; first as @Jon Snell pointed out you want to ensure your meter(ing method) is up to scratch, but as you're an EE I'll assume you have that covered for now. Next up, the 40C and 40CR are in fact two different animals. I've attached schematics for the 40CR to this post (there are a few revisions of the amp, but they're very similar). Note that the 40C/HR and 100HR share the same boards, they're just stocked differently.

Finally there's the odd changing of channels when the unit is powered on. This is more of a curiosity than a problem, but it's very curious all the same.
About 30% of the time, if the last channel selected was anything other than 2, at power on it will change to channel 2 after about 5-10 seconds.
To be clear, this is with no MIDI attached, no footswitch, nothing connected to it in any way. Curiously, after warming the amp up and leaving it in 40W mode rather than standby as you normally would it can be powered on without spontaneously changing channels at all.
I did this test about 20 times in each of standby and 40W mode and it only changes channel at power up in standby.

Sounds like there's at least a residual charge in the amp. Is it by chance not earthed or earthed correctly?

It's the 'Switching Chip', and only Marshall can change it out for a new one.

Not *quite*. I have the binary dump, so anyone can flash a new ATtiny chip and pop it in. Those chips and a flasher are dirtcheap, it's just a bit of work to get it all done. Shout out if you're interested and I'll give you the file. This only addresses an issue with the image on the IC itself, though - I've often found that actually it's unstable Vcc or poor grounding causing it to malfunction.

 

[Len]

You can’t measure that low of a resistance in circuit.

 

[TheKman76]

@PelliX You're a scholar and a gentleman, sir! Many thanks for the valuable contribution, I wasn't able to find the schematic on my own.
I was just inside the chassis and found reference to the 100 with asterisks against parts which aren't populated on the 40. With this I *may* have found the schematic on my own, but you done me good service and saved me the bother.

As for the spontaneous channel switching, you may be right about residual charge, I wasn't thorough about the test process as it was merely a curiosity. I checked the main earths and grounds while I was in there, all seems well.

I hadn't looked hard enough to know what was responsible for all of the logic, but I did wonder if an AT was in there somewhere. I've played with this stuff in the past so I'm very interested in any insights you have. Any recommended reference material on the particular implementation, perhaps?

@Len I can and I have, with respect. If you're familiar with the schematic it's clear this is entirely feasible.

I actually found and fixed the fault with the bias circuit and I'm sorry to say this was definitely a quality problem. Read on if you're interested...

After opening the chassis and locating the bias measuring resistors I measured them at 0.62 and 0.63 Ohms in circuit. They're clearly labelled 1 Ohm so I considered lifting them to measure out of circuit, but decided to probe around a bit first. The alarm bell rang when the resistance between the two power valve cathodes measured 215mOhms (425 minus probes)... hrmmm. Too high to be a direct short of heavy tracks on the board, but a lot less than the 2 Ohms you'd expect from the series resistance of the two in series. Seems like a short in non-high-current tracks or something.

So I measure across the two external bias probe points (the two red ones) and get 205mOhms. Clearly this isn't right and it occurs to me this is why my attempts to bias the valves didn't go well the first time. After pulling the little board with the probe points out and disconnecting the 10-way cable from the main PCB I measure across the two probe points again expecting to find a problem with this little board, but no, the short is gone! OK, it must be on the main PCB, right? Nope. Now with the probe board disconnected the in-circuit measurements of the resistors are spot on; 0.99 and 1.01 Ohms.

It took about 30 minutes to map out the connectors and figure out exactly what was connected to what, but in short I couldn't fault the boards or the connectors. This is the bit when I thought I'd gone mad, as the fault only presented when all of the components were connected. Remove either end of the cable and the problem disappears! I has to be the cable, right? Probing every single pin looking for a short circuit in the cable connectors reveals nothing. Interestingly, each of the probe points has two pins on the cable:

Pin 1 - Valve 1 bias
Pin 2 - Valve 1 bias
Pin 3 - Valve 2 bias
Pin 4 - Valve 2 bias
Pin 5 - Common ground
Pin 6 - Common ground
Pins 7-10 are for the adjustment pots

So I reassemble the circuit for the 96th time and start probing for problems again when it finally occurs to me.
This isn't a ribbon cable, it's ten individual wires fitted to two connectors. Does anyone see where this is going?
I finally untwisted the cable and flattened it out. It looks like a standard straight-through arrangement, except, there it is! Pins 2 and 3 are still twisted over, connected to pins 3 and 2 respectively on the other end. Now I finally understand why the short doesn't present without the cable connected.

Because pin pairs 1&2, 3&4, 5&6 are all connected directly at the solder joints on the boards, crossing over pins 2 and 3 on the cable creates a relatively high impedance short circuit across the probe points. This is a quality problem in manufacture, though it's fair to say this would be a very hard one to catch at the factory.

Of course, after re-pinning the cable correctly it all now works exactly as you'd expect. I do worry there might be other niggles which will arise in the future.

 

[PelliX]

@PelliX You're a scholar and a gentleman, sir! Many thanks for the valuable contribution, I wasn't able to find the schematic on my own.

Thanks, don't mention it, glad I could be of some assistance. I've suggested a section or sticky for schematics on the forum before so that others can benefit from a collective sort of library, but there doesn't seem to be that much animo yet.

I was just inside the chassis and found reference to the 100 with asterisks against parts which aren't populated on the 40. With this I *may* have found the schematic on my own, but you done me good service and saved me the bother.

Anytime, mate. Note there are multiple revisions of the C/HR boards, but the general gist remained the same. You can search in the forum and with a little luck you'll find an exact match.

As for the spontaneous channel switching, you may be right about residual charge, I wasn't thorough about the test process as it was merely a curiosity. I checked the main earths and grounds while I was in there, all seems well.

OK, how about the Vcc for the ATtiny, is that stable and within +/- 0.5V of spec? Those chips are very, very reliable in general, but they don't take well to unstable voltage.

I hadn't looked hard enough to know what was responsible for all of the logic, but I did wonder if an AT was in there somewhere. I've played with this stuff in the past so I'm very interested in any insights you have. Any recommended reference material on the particular implementation, perhaps?

The "logic" is pretty much all the AT, indeed. It switches the relays (channels, muting) and deals with the input from the footswitch, basically. The only material of interest here would be the schematic for the amp (check) and the specsheet for the AT, which is available online from ATmel themselves. In my experience, intermittent fauls are generally not caused by the IC itself but the peripherals including its supplied power. When those IC's fail, they *generally* do it consistently. That's a rule of thumb, of course.

Because pin pairs 1&2, 3&4, 5&6 are all connected directly at the solder joints on the boards, crossing over pins 2 and 3 on the cable creates a relatively high impedance short circuit across the probe points. This is a quality problem in manufacture, though it's fair to say this would be a very hard one to catch at the factory.

Good catch! Not sure I entirely agree on the QC aspect - checking the bias should have revealed "an" issue, albeit a tricky one to track down - as you noticed.

Nice work on getting that fixed!

 

[TheKman76]

Good catch! Not sure I entirely agree on the QC aspect - checking the bias should have revealed "an" issue, albeit a tricky one to track down - as you noticed.

Nice work on getting that fixed!

Thanks. Persistence through perseverance is my moto.

Certainly it's a tricky one from a QC standpoint. As an amp tech you take absolutely nothing for granted, expect every single pair of hands that has touched an amp has tried desperately to destroy it... assume nothing, check everything as you go, be absolutely pedantic about every problem you encounter. While in a manufacturing process everything any given worker does is based on assumptions from those who came before you. You follow your instructions and if nothing explodes, move along to the next set of hands. Even the QC processes have limited scope and assumptions to work with, so I can accept that even a really tight QC system would let a fault like this slip through.

However, a fault is still a fault and it points to a flaw in design allowing failure modes to creep into manufacture. Strictly speaking I think the separate board for the bias test points is a good choice, but the use of this particular connector arrangement is a poor one. These are all relatively low voltage and low current circuits which are well suited to a hard soldered ribbon cable. There are plenty of reasons to not use the ribbon system, but none of them are evident in this design and plenty of designs from Fender, Boss, Boogie and countless others use them to good effect.

Sorry, is my Engineer showing?

So, not really a QC problem, not even really a manufacturing problem, to contradict myself. Like so many things it's a design problem, albeit one probably tied to cost. It is what it is, this is not a super-premium hand-wired amp.

The "logic" is pretty much all the AT, indeed. It switches the relays (channels, muting) and deals with the input from the footswitch, basically.

This all seems to be missing from the schematic. CN203 deals with foot switching and MIDI, though it's not clear how this interacts with the relays. MISO and MOSI both interface with "DFX1" what ever that is. Probably requires effort in working much of this out for myself, I suspect.

EDIT: DFX1 is clearly the reverb. Having never really looked at these it seems odd it should have logic to manage channel switching. *shrugs*

 

[PelliX]

Sorry, is my Engineer showing?

Yes!

There are plenty of reasons to not use the ribbon system, but none of them are evident in this design and plenty of designs from Fender, Boss, Boogie and countless others use them to good effect.

Well, as someone who recently had the joy of servicing a 90's Blues Deluxe I have my opinions on those ribbon cables. While they work, I'd have preferred a detachable variant, grumble.

EDIT: DFX1 is clearly the reverb.

Yup.

Having never really looked at these it seems odd it should have logic to manage channel switching. *shrugs*

Well, there's some "logic" to it, like muting while switching channels (muting is a separate relay, for example) and so on. It's cheaper to implement that way. I've attached a PDF with among other things the pinout for the AT.

 

[SkyMonkey]

Sifting this thread may be helpful.

So, Now That The New Dsl Schematics Are Finally Out There...

... would it be possible for one of you technically oriented guys to tell us how different they are from the previous gen DSLs? To my (completely untrained) eyes the new (HR) preamp circuit seems quite similar to the older (H) one. One glaring difference is that the new one appears not to have...

www.marshallforum.com

 

[TheKman76]

@SkyMonkey I'll definitely get a look through that, thanks.

@PelliX Yeah, I've worked on some Boogies which use these a lot and they can be challenging to disassemble and such. Ideally a good detachable ribbon is great but this is one of those cost items which often doesn't make it to production, the DSL seems like a prime example.

I seems like none of these schematics are complete. The MarkII uses separate logic for the switching rather that GIO from the AT and it's not at all clear where the AT interfaces. Did the MarkI not have a MIDI interface?

 

[pjcallahan2112]

Afternoon all from Aus.

I'm a new DSL40CR owner, never owned a proper valve amp before, though I'm very familiar with the workings and have an education in Electronic Engineering.

Now, while I accept that the DSL40CR fits into the budget-friendly end of the Marshall spectrum of valve driven gear, I'm experiencing a couple challenges which feel like quality problems. Here's hoping I'm just missing something.

The first is biasing the power valves. I understand that the accepted normal bias current for the factory EL34 valves is in the 30-40mA range, centred around 35mA being ideal. Having looked at the schematic I see two 1 Ohm resistors used for measuring bias current, so measure 35mV and your golden. Simple.
However, I measured the bias resistors at 0.61 and 0.64 Ohms, not even close to 1 Ohm. Looking over the schematic again I can see no reason for this, so I ran with 21.4mV and 22.4mV respectively. The problem I face is that one valve is adjusted as cold as it will allow while the other is right in the middle of it's range. I estimate they're biased at around 37mA because I ran out of adjustment on one valve.
I realise a more precise bias measuring device is probably a good idea and I'll look into that too. I also note that the schematic I'm working from is for a DSL40C which may be slightly different. But, am I missing something, or is there a quality problem?

Second issue is with MIDI switching the Master volumes. I've read that this is challenging at the best of times and many simply give up on the notion completely. I spent a good two hours yesterday and I simply cannot understand under what circumstances this works as expected. Switching channels is fine, toggling the effects loop also fine, these I have configured on a dozen different patches in the GT-1000. However, I can only make the MV switch seemingly randomly. Even attempting to replicate an exact sequence of events doesn't seem to produce consistent results. I've literally been through the same sequence ten times and found no way to predict what the MV switching outcome will be.
I understand that it will remember the master chosen for a given channel so I thought perhaps the MIDI toggle is only to either enable or disable this option, but no. There appears to be no consistency.
Am I missing something? Can someone point me to a resource which explains what CC#14 is supposed to do? Does this feature work with the Marshall 6-button controller? Am I better off just going completely insane and be done with it?

Finally there's the odd changing of channels when the unit is powered on. This is more of a curiosity than a problem, but it's very curious all the same.
About 30% of the time, if the last channel selected was anything other than 2, at power on it will change to channel 2 after about 5-10 seconds.
To be clear, this is with no MIDI attached, no footswitch, nothing connected to it in any way. Curiously, after warming the amp up and leaving it in 40W mode rather than standby as you normally would it can be powered on without spontaneously changing channels at all.
I did this test about 20 times in each of standby and 40W mode and it only changes channel at power up in standby.

So far it seems my teenage offspring are more predictable that this particular DSL40CR. I'd be vary grateful to have some input.

Thanks you all for reading my lengthy first post.

Dave.

It turns out I must have an older DSL 40Cr because mine has a pentode/triode switch non the back

Afternoon all from Aus.

I'm a new DSL40CR owner, never owned a proper valve amp before, though I'm very familiar with the workings and have an education in Electronic Engineering.

Now, while I accept that the DSL40CR fits into the budget-friendly end of the Marshall spectrum of valve driven gear, I'm experiencing a couple challenges which feel like quality problems. Here's hoping I'm just missing something.

The first is biasing the power valves. I understand that the accepted normal bias current for the factory EL34 valves is in the 30-40mA range, centred around 35mA being ideal. Having looked at the schematic I see two 1 Ohm resistors used for measuring bias current, so measure 35mV and your golden. Simple.
However, I measured the bias resistors at 0.61 and 0.64 Ohms, not even close to 1 Ohm. Looking over the schematic again I can see no reason for this, so I ran with 21.4mV and 22.4mV respectively. The problem I face is that one valve is adjusted as cold as it will allow while the other is right in the middle of it's range. I estimate they're biased at around 37mA because I ran out of adjustment on one valve.
I realise a more precise bias measuring device is probably a good idea and I'll look into that too. I also note that the schematic I'm working from is for a DSL40C which may be slightly different. But, am I missing something, or is there a quality problem?

Second issue is with MIDI switching the Master volumes. I've read that this is challenging at the best of times and many simply give up on the notion completely. I spent a good two hours yesterday and I simply cannot understand under what circumstances this works as expected. Switching channels is fine, toggling the effects loop also fine, these I have configured on a dozen different patches in the GT-1000. However, I can only make the MV switch seemingly randomly. Even attempting to replicate an exact sequence of events doesn't seem to produce consistent results. I've literally been through the same sequence ten times and found no way to predict what the MV switching outcome will be.
I understand that it will remember the master chosen for a given channel so I thought perhaps the MIDI toggle is only to either enable or disable this option, but no. There appears to be no consistency.
Am I missing something? Can someone point me to a resource which explains what CC#14 is supposed to do? Does this feature work with the Marshall 6-button controller? Am I better off just going completely insane and be done with it?

Finally there's the odd changing of channels when the unit is powered on. This is more of a curiosity than a problem, but it's very curious all the same.
About 30% of the time, if the last channel selected was anything other than 2, at power on it will change to channel 2 after about 5-10 seconds.
To be clear, this is with no MIDI attached, no footswitch, nothing connected to it in any way. Curiously, after warming the amp up and leaving it in 40W mode rather than standby as you normally would it can be powered on without spontaneously changing channels at all.
I did this test about 20 times in each of standby and 40W mode and it only changes channel at power up in standby.

So far it seems my teenage offspring are more predictable that this particular DSL40CR. I'd be vary grateful to have some input.

Thanks you all for reading my lengthy first post.

Dave.

Hello Dave! I thought you had been on here for years!😁You've been a great help to me so far as I have just recently joined and boom! The first guy I talk to is an electrical engineer with a dsl40cr! I am afraid we may have some differences in our amps as I was under the impression that I too owned a dsl40cr but I believe what I in fact have is a dsl40c with a celestion creamback instead of the V-type that comes stock with the dsl40cr. There are also a few more differences so starting with the rear panel the dsl40c has no emulated out, audio in, midi in, and is also minus the 1x4ohm/2x8ohm speaker out. It has the f/x loop switch and the half power Pentode/Triode on the rear panel also. The differences on the front panel include a channel select switch between the classic and ultra gain and volume knobs, od1 and od2 are called lead 1 and lead 2 on the dsl40c and I am also minus the master volume section, which would be a nice mod, but that would take some serious surgery. Could you tell me where I can locate the schematic for the dsl40c? When I bought the amp the celestion creamback was an exclusive Sweetwater Music modification also. I see that a new dsl40cr is also $300 more than the dsl40c. I thought my amp was already going up in value!!🤣🤣🤣

 

[TheKman76]

New to valve amps and Marshalls, relatively speaking, but not new to the other stuff.

I think the correct schematic can be found here:


The Mark I has quite a few things different to the Mark II. Tonally they're very similar, but features are different. Yours will still have the C19 bright cap for instance.