Marshall 3520 bass amp

[MarkusBass]

Does anyone know how to calculate the filter frequency for filters that are in Marshall bass amp 3520: http://www.drtube.com/schematics/marshall/3520pre.gif ?
According to SPICE simulation, the last filter (marked as 5kHz) is actually 2.5kHz. But the manual clearly states that this is 5kHZ filter.
Also the Contour filter does not seem to be shown correctly on the schematic. It seems to me that the wiper should be connected to the ground.

Mark

[Steve Conner]

Just by scaling of the components from the other sections, it makes sense for the final section to be 2.5kHz. (its resistors are the same as the previous section but the caps are a little less than half the size, so the cutoff frequency should be a little more than twice)

So I guess the 5kHz is a typo.

I think the contour control is a bridged-T notch filter, but I'm not sure about the wiring. I've seen such filters that feed signal into the base of the "T" to adjust the notch depth.

[Enzo]

If you ground the wiper of the contour control it just becomes a rolloff knob. it also appears to be part of a feedback loop around IC1b.

[MarkusBass]

Thanks for you answers.
Steve, for me it is also a mistake but, taking into account that Marshall created a whole series a such amps in the late 80's - Integrated Bass System, it's quite a mistake. In the manual it is specified as 5kHz. And, as far as I remember, the text below the knob in the amp was also "5 kHz".
Enzo, I'm not sure whether I understand the term "rolloff knob" (my poor English) but I simulated the corrected version in SPICE and it started working correctly as a contour filter with middle frequency at about 550 Hz (and not 700 Hz as specified in the manual). Appart from this, I'd seen the board some time ago and I remember the wiper being connected to the ground (although I may be wrong). It would be the best to verify my findings with a real amp (they are still popular in UK).

Mark

[Enzo]

I first called it a treble knob, but when I saw the size of the cap associated with it, I realized it went a lot lower than "treble." But if you ground that wiper, then the pot is just a variable resistor in series with a cap, and they are in shunt across the signal path. This would be exactlty the same arrangement used in guitars as a tone pot. And that tone pot does nothing more than roll off the high end. SO I called it a rolloff. I am too lazy to calculate what freqs it would affect.

I am not ready to call it a corrected version, you made a changed version and Spiced it, I don't doubt then that it differed from the published spec.

What do you consider a "contour filter" function? I consider it a sort of wide notch filter.

Did you Spice the schematic version?

[Steve Conner]

Well, what if the component values on the schematics are wrong? Marshall could have changed the frequency from 2.5k to 5k during development, and forgotten to update the schematic. Or even deliberately left the wrong values to make life harder for their competitors. The voicing of these filters probably gets changed a lot during development, and it ends up like a secret tone recipe. In fact, I guess that's what you're after, and Marshall probably would rather you just bought the amp.

I agree, contour is a notch filter that produces a mid scoop, which implies a T type of circuit.

Do you mean the solid-state Marshall bass amps with the big shiny black volume knob? I'm in the UK, and it's a while since I've seen one in the local music stores or on the gig circuit.

[MarkusBass]

..I am not ready to call it a corrected version, you made a changed version and Spiced it, I don't doubt then that it differed from the published spec.
What do you consider a "contour filter" function? I consider it a sort of wide notch filter.
Did you Spice the schematic version?

Enzo,
Your explanations are great as always . Thanks for this.
The filter is marked on the schematic as "contour". As Steve has written: "contour is a notch filter that produces a mid scoop, which implies a T type of circuit". Now I have some doubts regarding my change but I'm (almost) sure that the "old" version does not work correctly in SPICE as the "new" one does. I will check it once again.
Yes, I have the schematic in SPICE. I thought that it makes sense to test the design using such software instead od struggling with not working preamp. In such a program you can clearly see other problems. For example, it seems that in this preamp the capacitor in the feedback loop of the first stage is too high (220p). It causes roll-off of high frequencies (so the frequency plot of this preamp is not linear). If I change it e.g. to 100p, the results are much better. Of course this has to be verified in the "live" version. It can be a mistake on the schematic, or they wanted to prevent high frequency oscillations. I'm going to check it.

Mark

[MarkusBass]

Well, what if the component values on the schematics are wrong? Marshall could have changed the frequency from 2.5k to 5k during development, and forgotten to update the schematic. Or even deliberately left the wrong values to make life harder for their competitors. The voicing of these filters probably gets changed a lot during development, and it ends up like a secret tone recipe. In fact, I guess that's what you're after, and Marshall probably would rather you just bought the amp.
....
Do you mean the solid-state Marshall bass amps with the big shiny black volume knob? I'm in the UK, and it's a while since I've seen one in the local music stores or on the gig circuit.

Steve,

I checked several other schematics by Marshall and I think that the components are OK but they forgot to update the schematic (frequency value). I also play bass guitar and two filters like 5kHz and 10kHz do not make much sense for a bass guitar (especially if there is no tweeter in the combo). This could be a first approach and later they changed the filter to 2.5kHz.
At least currently, with software like SPICE, it is very easy to verify whether the components are correct or they deliberately left the wrong values (but in the 80's it was different).

Yes, I mean exactly the amp you mention but there were several versions: 100W, 200W head, 200W combo, 400W head. I'm talking about 200W version with cross-over in the preamp and MOSFET power amp. I've seen the amp several times in Poland. There is a whole Web site about this and other Marshall amps: http://www.drtube.com/marshall.htm
And here is the manual I mentioned: http://www.drtube.com/schematics/marshall/ibs.pdf
I wonder how difficult it would be to contact the factory directly (Milton Keynes, UK ?? - I was once very close to the factory) and ask them for details.

Mark

[MarkusBass]

Contour filter description

Description of the Contour filter from the manual:
"Contour allows for a variable amount of cut at either 40Hz or 700Hz. At "12 o'clock" position (0), no cut is realized. Moving the control clockwise from 0 increases the amount of cut at 700Hz. This helps provide a cleaner, more rounded sound, especially at low volume settings. Rotating the control counter-clockwise from 0 increases the amount of cut at 40Hz. This will tighten up the sound when the amplifier is being overdriven."

Mark

[MarkusBass]

It looks like I found it; my "corrected" version is OK (wiper to the ground) but two resistors have incorrect values on the schematic: R9 and R10 should be 3k3 (instead of 4k7).
See the attachment for the simulation results. I get exactly what is written in the definition of the contour filter in the manual. And exactly 40Hz and 700Hz. Great - it's time to build the preamp.

Many thanks for your help,
Mark