Solid State Guitar Amp Forum | DIY Guitar Amplifiers

Solid State Amplifiers => Amplifier Discussion => Topic started by: Dino Boreanaz on March 16, 2022, 12:36:03 AM

Title: Reducing Power Output on Marshall 3520
Post by: Dino Boreanaz on March 16, 2022, 12:36:03 AM
I've got a Marshall 3520 Integrated Bass System 200W head that I really like, but I'm finding that it is too loud for home use.  The volume controls (separate treble and bass volumes) are barely above zero and the slightest movement takes it from silent to way-too-loud.

Can I reduce the output power in half by eliminating two of the four output MOSFETs in order to make the power amp like that of the 100W 3510 model?  I've attached power amp schematics of these two models and encircled the differences between the 3510 and the 3520.

If so, would this be as simple as desoldering one leg of R17 and R22 while leaving everything else in place and undisturbed?
Title: Re: Reducing Power Output on Marshall 3520
Post by: Jazz P Bass on March 16, 2022, 01:53:44 AM
"Can I reduce the output power in half by eliminating two of the four output MOSFETs in order to make the power amp like that of the 100W 3510 model"
Simple answer: NO.
The mosfets are handling the power of the amplifier.
ie: The voltage rails.
Not too sure how to go about it but the key is how hard the preamp is driving the power amp.
Title: Re: Reducing Power Output on Marshall 3520
Post by: phatt on March 16, 2022, 03:10:58 AM
If it has an FX loop just insert a volume pedal in that loop.
That then becomes the master volume for the power amp.

As Mr Jazz Bass said, Don't touch the Amp circuit as it could lead to tears. :-[
Phil.
Title: Re: Reducing Power Output on Marshall 3520
Post by: willpirkle on March 16, 2022, 08:48:33 AM
Agree with Phil - simple attenuator in the FX loop is the least intrusive option.
Title: Re: Reducing Power Output on Marshall 3520
Post by: Dino Boreanaz on March 16, 2022, 09:10:42 AM
Thanks for the replies all!  This is the first amp I've owned with an effects loop ... never even occurred to me to try a volume pedal in the loop.
Title: Re: Reducing Power Output on Marshall 3520
Post by: Dino Boreanaz on March 16, 2022, 09:58:43 AM
I'm thinking about building a simple DIY volume control to put in the effects loop rather than buying a more complex (and expensive) volume pedal.  Is there an ideal potentiometer value if I go with a simple input->pot->output arrangement?  Any drawbacks to such a simple circuit?

Thanks again for talking me down earlier!
Title: Re: Reducing Power Output on Marshall 3520
Post by: Enzo on March 16, 2022, 01:21:44 PM
A pot in a box?  Sure.  After all, that is what a volume pedal is.  The amp has no idea what the control is housed in.  I had a pot/box on my bench for years.  You never know when you might want to turn down a signal.  250k worked well for me. 

You neglected to circle the other difference between the two models..The power transformer.  The lower power unit uses a lower voltage power supply.   And that is where the power comes from in the first place.   You want to convert 200 watts to 100 watts?  Swap out the transformer.   And half power will only be a 3db reduction in volume.  Power isn't volume.

But it occurs to me, this is a solid state amp, you have no plans to overdrive the power amp, so why not just turn down the Master volume.  Or it might be called Post volume.   That is what your FX loop control would be doing.
Title: Re: Reducing Power Output on Marshall 3520
Post by: Dino Boreanaz on March 16, 2022, 06:48:03 PM
Yup, my dumb mistake.  I realized after reading the first few replies that the transformer is in fact different between the 100W and 200W models.

Today I tried the "pot-in-a-box" in the effects loop, but it is quite finicky and also affects the way the boost control responds and hence the tone.  This amp doesn't have a conventional single master volume control.  It has separate treble volume and bass volume with a variable crossover.

I've attached the preamp schematic, and it seems there are three gain stages and their corresponding controls (boost, treble volume, and bass volume) after the effects loop.  Would it be correct to say that the placement of the effects loop within this circuit limits the effectiveness of the "pot-in-a-box" to control the overall output volume and causes it to affect other aspects of the amp's tone?

I also noticed that the treble and bass volume pots (VR10 and VR11) are linear taper rather than audio taper.  Would this tend to make them more sensitive at the lower end of their rotation as compared to audio taper pots?  If so, would it be worthwhile swapping these for audio taper pots?  I suppose this thread has become more about managing the quiet end of the amp's volume range rather than reducing the amp's power output.

Thanks again to everyone for your input.
Title: Re: Reducing Power Output on Marshall 3520
Post by: willpirkle on March 16, 2022, 11:17:29 PM
You could replace VR10/11 with log pots if you want to - I'd probably leave them alone.

To get back to the original issue, The input preamp has a max gain of (47/2.2) + 1 = 22.36 = 27dB

If you want to lower the gain of the preamp which is the first op amp after the input, you have to either lower the 47k feedback resistor R4 or raise the 2.2k shunt resistor R5. To experiment, you could replace the 2.2k resistor with say a 10k tweaker-pot and raise the resistance above 2.2k while playing through it till you get what you want.

If still too much gain, change to a 50k tweaker-pot. When you get the gain where you want it, either leave the tweaker pot in place, or measure it and replace with a fixed value.

With 6.8k, you get ~18dB max gain. Note that raising this resistor value also lowers the minimum gain, when the 22k gain pot is at max resistance. So changing that one resistor will alter the overall gain range as well as the max gain.

Will
Title: Re: Reducing Power Output on Marshall 3520
Post by: Dino Boreanaz on March 17, 2022, 08:18:08 AM
Thanks Will!

I probably should have clarified in my original post, that while I would like to reduce the volume (or have better control over the volume at the quiet end), I DO want the preamp to be pushed into clipping.  I like the tone and overdrive grit that I'm getting with the first gain control at maximum and most of the EQ knobs boosted to about 3 o'clock.  So, while lowering the gain of the first stage would certainly reduce the overall volume, it would also not overdrive as much and that's less desirable to me than the too-sensitive volume controls.

I think I'm going to order the log pots for the volume controls and see how that works out.  I'll post back here if/when I try it.
Title: Re: Reducing Power Output on Marshall 3520
Post by: joecool85 on March 17, 2022, 03:06:02 PM
Quote from: Dino Boreanaz on March 17, 2022, 08:18:08 AM
Thanks Will!

I probably should have clarified in my original post, that while I would like to reduce the volume (or have better control over the volume at the quiet end), I DO want the preamp to be pushed into clipping.  I like the tone and overdrive grit that I'm getting with the first gain control at maximum and most of the EQ knobs boosted to about 3 o'clock.  So, while lowering the gain of the first stage would certainly reduce the overall volume, it would also not overdrive as much and that's less desirable to me than the too-sensitive volume controls.

I think I'm going to order the log pots for the volume controls and see how that works out.  I'll post back here if/when I try it.

You could consider using an L-Pad: https://www.parts-express.com/speaker-components/crossover-components/speaker-l-pads

That way you turn that down to half, let's say, and then you can turn up the volume on the amp and make it distort as you wanted without it being ear-bleeding loud.
Title: Re: Reducing Power Output on Marshall 3520
Post by: Tassieviking on March 17, 2022, 10:52:24 PM
Instead if a volume pedal you could make a simple patch lead with resistors in it.
Just solder the resistors into one of the plug.
You have to experiment with the resistor values until you like the volume.
Maybe some one here can work out the best resistor value for you.
Title: Re: Reducing Power Output on Marshall 3520
Post by: Dino Boreanaz on March 18, 2022, 12:04:18 AM
Thanks Tassie, that's a nice solution in a single cable.  If the volume control in the effects loop had worked out, this would have made a very neat and tidy installation.  Unfortunately, since the boost and volume controls are after the effects loop in this circuit, this didn't work out as well as it would have in a circuit where the effects loop is located after all the preamp controls.

Earlier today I remembered that I had modeled this preamp circuit in Micro-Cap so I took a look at it again.  I used the stepping analysis to model the effect of sweeping both volume controls from 0 to 100% in 10% steps using both linear and logarithmic tapers.  I've attached the plots and it seems the linear taper produces a huge jump in the first 10% of the pot's rotation, where the log taper uses over 70% of the pot's rotation to cover the same range.

I don't have much experience using Micro-Cap, so I'd be happy to share the file with anyone who's more familiar with it and could verify whether my analysis is valid or find any mistakes in my use of the AC analysis tools or the model itself.

I've written to an engineer at Marshall to get their input on whether swapping from linear to logarithmic taper pots would make the low end of the volume controls less sensitive to small movements.  I'll post an update if/when I get a response.
Title: Re: Reducing Power Output on Marshall 3520
Post by: Enzo on March 18, 2022, 02:50:43 AM
FX loop is a concept, not necessary to take it literally.  Your FX loop is before some controls, I see.   But I also see the split highs and lows output jacks over on the right.  It appears using the bass alone jack is still full range, until the highs jack is used.  Don't plug into highs, then the bass jack is full range.  Added bonus you still can use the highs and lows controls to balance your tone.    Does not the power amp have an input jack?   Try the volume control there instead of in the official FX loop.    That is after all controls.   

Bass jack out to power amp in jack.
Title: Re: Reducing Power Output on Marshall 3520
Post by: Tassieviking on March 18, 2022, 03:46:00 AM
My fault, I should have looked at the schematic closer.
You want to reduce the signal between the pre-amp to the power amp.
It looks like the place to insert the patch lead is between jS4 on the pre-amp schematic and JS1 on the amplifier schematic.
I think that means the Master Post socket is the pre-amp out. I think Marshall did not install JS1 on the power amp circuit on the back panel.
If you really wanted to you could remove the shielded wire on the power amp PCB and put the resistors there.
Or you could install JS1 on the power amp schematic on the back yourself, but it means drilling a hole.
If you did drill a hole, a switch would be simpler to add / remove the resistors.
A power amp in jack would make it simple.
If you can provide a photo of the bottom of the PCB the rear jacks are on we could check to make sure, just have to look at the tracks.

Title: Re: Reducing Power Output on Marshall 3520
Post by: phatt on March 20, 2022, 07:56:56 AM
My advice would be to cut the input cable to power amp as per TassieV's idea and mount a 50k~100k pot in the back panel and that becomes the power amp Master volume. Looks like enough room for a pot mount next to the Socket PCB in the back panel.
Phil.
Title: Re: Reducing Power Output on Marshall 3520
Post by: Dino Boreanaz on March 20, 2022, 11:27:49 PM
I very much appreciate everyone's input on this as I'm quite new to the world of modifications.  I have a couple of questions for those with more experience that myself (which is just about everyone!):

As my end goal is to make the volume controls less sensitive at very low settings, is there something preferable about adding resistance (either fixed or a pot) before the power amp versus changing the volume pots from linear taper to log taper?

I have found volume settings that I am quite happy with, but I would really like to have finer control in this area and it seems like log taper pots would accomplish this without adding components or making permanent modifications.

If I may ask another question to further my understanding of this circuit:  It seems like the two volume pots (VR10 and VR11) are not located between their op-amp's input pins and output pins, and so should not affect the gain of these op-amp stages.  Is it correct to say that these two volume pots control the signal passed to the power amp as voltage dividers rather than by altering the gain (and therefore clipping behaviour) of the op-amps?

Thank you again for all your input and insight.
Title: Re: Reducing Power Output on Marshall 3520
Post by: phatt on March 21, 2022, 06:27:44 AM
Quote from: Dino Boreanaz on March 20, 2022, 11:27:49 PM
I very much appreciate everyone's input on this as I'm quite new to the world of modifications.  I have a couple of questions for those with more experience that myself (which is just about everyone!):

As my end goal is to make the volume controls less sensitive at very low settings, is there something preferable about adding resistance (either fixed or a pot) before the power amp versus changing the volume pots from linear taper to log taper?

I have found volume settings that I am quite happy with, but I would really like to have finer control in this area and it seems like log taper pots would accomplish this without adding components or making permanent modifications.

If I may ask another question to further my understanding of this circuit:  It seems like the two volume pots (VR10 and VR11) are not located between their op-amp's input pins and output pins, and so should not affect the gain of these op-amp stages.  Is it correct to say that these two volume pots control the signal passed to the power amp as voltage dividers rather than by altering the gain (and therefore clipping behaviour) of the op-amps?

Thank you again for all your input and insight.
Yes A log pot may well alter how early the volume jumps but I think a larger value pot (Log or Lin) would give a more progressive rise in volume. 5k would make it jump up in volume with only little rotation, as you have noted.

I have a hunch 50k pots might resolve the issue far better,   then log or lin would not matter much.
That may resolve this better than working on adding an extra master pot.

Regards to VR10 & VR11.
Yes they are just voltage dividers and do not effect the gain.
Phil.
Title: Re: Reducing Power Output on Marshall 3520
Post by: Dino Boreanaz on March 21, 2022, 09:53:49 PM
Hey Phil,

Thank you so much!  My understanding is improving slowly, but steadily.  If you'll indulge a few more, somewhat rambling, lengthy questions ... if not, I totally understand and thank you again for everything you've shared.

I tried modelling a few different potentiometer values (4.7k, 47k, and 470k) to gain additional understanding and found that the higher value pots produce the same range of sweep when the upper and lower limits are treated as a percentage of pot value.  For example, the total gain range is the same for the 4.7k pot swept from 5 ohms to 4.7k ohms as compared to the 470k pot swept from 500 ohms to 470k ohms.  This is as I would have expected since the voltage divider should operate the same as long as the proportions remain the same regardless of the absolute values.  Am i correct here?

One thing I did not expect is that the higher pot values seemed to have a slightly stronger low-bass response (from about 40 Hz to 200 Hz) as compared to the lower pot values.  See the attached "Overlay 1" where the blue traces are 4.7k pots, the green are 47k pots, and the red are 470k pots.  The plots are essentially identical above 400 Hz, but show that the higher value pots provide more bass extension at corresponding settings.  Is it correct to assume that this is due to the higher resistance to ground of the high-value pots is allowing more low frequency content to pass through rather than bleed off to ground?

Lastly, I tried modelling the same three pot values swept from an initial resistance value of 5 ohms up to their respective maximums (5 ohms to 4.7k, 5 ohms to 47k, and 5 ohms to 470k).  In these plots I did see that the higher pot values indeed provide greater range at the quiet end of their sweep, as shown in the "Overlay 2" attachment.  This again is as I would have expected, but my question is which sweep is more representative of what a real potentiometer would produce?  In other words is the lowest attainable resistance of a potentiometer some particular value regardless of the pot's maximum value or is it a percentage of the pot's maximum value?  Or (as I fear) does it vary from pot to pot and manufacturer to manufacturer meaning that this has all been an exercise with no practical value?!

Thank you again!
Title: Re: Reducing Power Output on Marshall 3520
Post by: phatt on March 22, 2022, 04:20:23 AM

A larger value pot gives you more room to move over the lowest 10% of the rotation,, especially if using an Audio log taper pot.

The plots of log and linear responses are resistance vs, pot position.
The issue is that there are many different log Curves,, even reverse log. :loco

Ideally the closer you get to 10~15% of total resistance at half rotation point the better it will be for what you want in your application.

Google; *Log pot curves*,, That may help you get your head around it.
And also learn about how we all hear sound;
https://ozvalveamps.org/pots.htm

It may well be cheaper and less hassle to just mount a power amp Master level pot on the back panel. 50k pot should work fine.
Phil.
Title: Re: Reducing Power Output on Marshall 3520
Post by: Tassieviking on March 23, 2022, 04:39:31 AM
Looking at the circuit again, I would definitely go for the Log pot first, that would give you more control over the volume at lower level.
Another option is to put a resistor in series with the pot, but then you cant turn it up to full any more.
I presume it is a 24mm Alpha Taiwan potentiometer you have in there (RV24AF-22 model) since it is a Marshall.

The very last option if it were my amp, I would get a 16mm pot (log A4k7) with a push pull switch on it, and wire that in.
Wire it so if you pull the switch out you put a resistor in series with the pot, when pushed in it is like it is now.
Being 16mm you would have to run short wires to the PCB as the pot won't reach the PCB.
If the pot is too close just turn it sideways so the terminals point sideways.
I don't know if the capacitor C31 will be in the way for this.
Title: Re: Reducing Power Output on Marshall 3520
Post by: Dino Boreanaz on March 23, 2022, 09:56:19 PM
Thanks Tassie ... very much appreciate your input!

I have decided to go the route of trying the log pots for the treble and bass volume controls.  You are correct, the pots are the 24mm Alphas.  I've ordered the three different values that were available in log taper (4.7k stock value, along with 22k, and 100k).  Based on the results of the circuit modelling, it seems there's nothing to loose by trying the 100k pots first.

I love the idea of the push-pull, but I'm going to keep this simple.  I'll be very happy if I can just get some finer control at low volume levels.
Title: Re: Reducing Power Output on Marshall 3520
Post by: Dino Boreanaz on August 18, 2022, 09:18:35 AM
Well it's been quite a while since I first posted this topic and I just wanted to provide an update for the sake of completeness.

I did try a few different values of audio taper pots to replace the stock linear pots and I found that they made the volume controls slightly less sensitive at the low end of their rotation as compared to the stock linear pots.  It was not as dramatic a change as I had hoped it would be, so I also tried using a reverse audio taper pot and this did not work well at all (being more sensitive at the low end of the rotation ... as it should).

Since the different values did not have any noticeable affect on the behaviour of the volume controls, I stuck with the stock value of 4.7k but used the audio taper pots in both the treble volume and bass volume controls.

I should also note that I am running this amp into a 16 ohm load rather than the rated 4 ohm load.  So, according to Marshall's senior service engineer, the amp should be putting out approximately 75 watts.  Even at this power level, I find that I have both volume controls set at (or slightly below) 1 out of 10 on the dial which still makes it quite sensitive to small changes.  But I have found a balance that I'm happy with and have no plans to make further changes at this point.  I will post a new topic if I change my mind and decide to start tweaking things again.

Thanks again to everyone who contributed their time, knowledge, and ideas.
Title: Re: Reducing Power Output on Marshall 3520
Post by: Tassieviking on August 18, 2022, 07:52:36 PM
I can think of one more thing you can try, increase the taper of the pot.
place a resistor between pin 1 and pin 2 of the pot and the audio taper curve will increase.
The taper effect will be more pronounced at low volumes, you can buy audio pots with different taper curves, but the easiest way is just add the resistor.
You can read more here :http://www.geofex.com/article_folders/potsecrets/potscret.htm
It would give you even more control at low volumes.
Title: Re: Reducing Power Output on Marshall 3520
Post by: Dino Boreanaz on August 18, 2022, 09:55:49 PM
Great now I've got more work to do ... just when I thought I was done tinkering!  This actually seems like a fantastic option to make the taper more gradual at the low end.  I will have to try this and I'll reply with my results when I do.