Fix treble control to be centered on 1.6 kHz.

[Kazinator]

The ADA MP-1 manual claims that the treble control is centered on 1.6 kHz.  This is reiterated in some written annotations on the EQ schematic, where "1.6k" is written next to the Treble line.

However, this is actually incorrect. The filter circuit for that band has a peak at 2.5 kHz.

It is tantalizing: did they intend to have it at 1.6 kHz? Was it to match some reference amplifier?

Also, if you look at just the two input resistors and capacitors in the circuit, ignoring the feedback, by themselves they do form a band-pass that has a 1.6 kHz center! Just that's now how the overall op-amp filter circuit works out. For the Mid control, it does work out: the RCRC network by itself has around a 600 Hz peak.

Was it just a mistake due to wrong calculations/assumptions? Was the MP-1 supposed to have a 1.6 kHz treble?

I'm motivated to "fix" it mine because I don't use the treble control at all: it is set to zero (at least on dirty presets). On my external equalizer, I have an upper mid scoop which is centered on ... guess what ... 1.6 kHz. A slight cut at 1 kHz and 2.5 kHz, more at 1.2 and 2.0 bottoming out at 1.6.

I recently tweaked the bass control, so I don't need to do anything with bass on my external EQ; it is flat up to 1 kHz. 

If I could use the MP-1 treble to get the cut centered on 1.6k, perhaps I could have a flat EQ all the way out to 4 kHz, after which it's just used for cutting high end.

According to simulation, this can be achieved by changing the two capacitor values from 0.01 uF (10 nF) to 15 nF: a 50% increase.

[Kazinator]

BTW, I love the type who is always simulating. That's why my favorite Spice Girl is LTSpice.

[Kazinator]

OK, so I did this 1.6 kHz mod this morning. While there were some good things about it, it was still cutting too much high end in the 3 kHz to 5 kHz range. This is a very sensitive area where even 1 dB makes a difference.

I took out the 15 nF resistors and put in 22 nF ones. This knocks the center frequency to around 1 kHz.

So with this, we have two midrange controls. The original mid that is centered at around 560 Hz and this new one.

I have Bass at 4 (modded to 195 Hz rolloff), Mid at -8 (stock), Treble at -12 (1 kHz), Presence at 0.

The external EQ is completely flat up to 2.5 kHz. Yay!

I there is a significant boost of 3.1 kHz, and a little bit at 4 kHz.

Then after that, EQ is only used for rolling off fizz.

I think all I need now is to tweak the Presence band to be more useful. Stock is centered on 5.7 kHz, from about 3 kHz to around 11 kHz, if I recall my simulations a bunch of months ago. This is not that useful; it boosts too many ice picky frequencies. If it was from, say 2 to 5 kHz centered on 3.5, it would be more usable. Or 2.5 to 5 centered on 3.7 or something. The high highs are best controlled with power amp presence controls, external EQ, choice of speakers, ...

[Kazinator]

So I did a Presence tweak, changing 390 pF C83 to 1 nF (1000 pF). The high end of the range not extend anywhere near as much as before. Center freq is around 3.5 kHz.

It almost does what I want. On my external EQ, the 3.1 kHz and 4 kHz are no longer boosted, which means it is flat right to 4k. Presence on the MP-1 is boosted to +4, from the previous 0 before the mod.

It's a usable tone. There is just a bit of "tubbiness" to it, likely because it extends into the upper mids. Next, I will try changing R167 to 3.1 kΩ. That will move the center up to 4.6 kHz.

[Kazinator]

There is just a bit of "tubbiness" to it, likely because it extends into the upper mids. Next, I will try changing R167 to 3.1 kΩ. That will move the center up to 4.6 kHz.

I did this yesterday and was quite pleased with the result! I jammed on this for a bunch of hours.

I can hardly believe what is coming out of the rig, with the equalizer flat to 4 kHz.

Overall, my impression is that the repurposed Presence isn't focused enough on its core frequencies, so there is an issue between the tone being a bit dark, or else a bit brittle, without the point in between being a great sweet spot. The fizz doesn't grate on the ears or anything. But speaking of which, with the two midrange cuts, there is a lot of remaining high end past 5 kHz that I'm rolling off externally.

So, overall good success for a Round 1, but I'm now thinking about Round 2:

I would like to change the Mid control (which I have not touched) so that it spans the entire range currently occupied by Mid (560 Hz) and Treble (1.0 kHz). It might be centered on, say, 800 Hz, and wider than stock.

Then Treble could be higher than stock (2.5 kHz) rather than lower. Say, boost a bit past 4 kHz. (That one is easy: pair of 5.6n capacitors).

Then Presence could be set higher than stock also, so it can be set negative, and start cutting somewhere past 5 kHz. Well out of the way so that it doesn't take the bite out of the tone.

IDEA: If we change the C-R-R-C band-pass network in Presence to C-R-C-R, we can get a two-pole high pass, so when used for cutting, it will cut -12 dB per octave after the second knee frequency.

[Harley Hexxe]

Kaz,

   I've been following your thread and getting a few chuckles along the way with how you are manipulating the EQs in the MP-1, and I mean that in the nicest possible way.
   It's cool to see someone actually working on fresh mods for the MP-1 

  The only reason I'm chiming in here is because of your statement that you jammed for hours on it yesterday. Doesn't that say you landed in the place you were hoping to be?

Harley

[Kazinator]

It's cool to see someone actually working on fresh mods for the MP-1

I've executed the ideas I was talking about. The 31 band EQ in my rig is now disconnected.

I will probably start another thread on it because it's wandered.

I had to do all the mods at least twice to tweak the frequency response. I'm still not 100% happy with the midrange, but it's difficult (probably next to impossible) to get what I want out of one filter.

For Bass, I have that 195 kHz mod from the "fixed boofy bass" thread. Have not touched that in the recent work.

For Mid, I first had it centered at around 780 Kz, with -3dB points at 400 Hz and 1.4 kHz. That was too tubby sounding.  I widened the filter and centered in on 700, with -3dB points at 280 and 1.8 kHz.

For Treble, I first had a narrow Q configuration at 3.6 kHz, with -12 dB at 2 kHz and 6.5 kHz. (For the narrow one, I look at a -12 dB "box"). This was quite harsh when boosted. I narrowed it an dropped it down. It is centered at 3.15 kHz, with -12 dB at 1.85 kHz and 5.35 kHz. It sounds very similar to a 3.15 kHz band on a 31 band EQ!

For Presence, I changed the CRRC band-pass network to CRCR high pass, cascading two single-pole filters which each having the same cut-off frequency. So it is now a shelving control.  First I had it with -3dB at 5 kHz. That turned out to be taking too much of the good stuff away. I changed that cut-off to 6 kHz, which is working a lot better.

The shelving Presence does a really good job of cutting out fizz. I can get useful sounds out of the right without any other EQ. I still have a variable frequency filter available in my effect mixer. But I can set it to pass through (knob dimed at 20 kHz).

I'm getting a nice tone, similar to my usual one. It's hard to let go of the security blanket of the 31 band EQ. But the fact is, I used it in a particular way, with little deviation, and most of the sliders unused. I'd be better off with a parametric with two or three bands just to have that "get any tone shape" weapon in the arsenal.

Getting good tone without the 31b EQ and being able to set if flat revealed its limitations. It affects the tone even when set flat. It has a sound quality to it, like everything. What surprised me is that there is a certain harshness to it when it is set to pass through. I thought maybe it's just some highs in the tone not being filtered out when the EQ is in bypass, but when I spliced the EQ out of the chain entirely, it went away.

[Harley Hexxe]

Hey Kaz,

   If I'm not mistaken, the EQ in the MP-1 is a parametric EQ, just without the knobs. I guess they had a particular sound in mind when they designed it, after all, it was the 80s.

   What 31-band EQ do you have? I noticed the same thing with the Rane EQs that I have too, they do impart their own sound characteristic in the signal path which is why I don't use them. I suppose a transparent EQ is a hard thing to find. Then again, the same is true with most effect processors as well, they will affect the sound character with their own flavor even when bypassed.

[rnolan]

Transparent EQ = $s, they are not so hard to find but do you want to  spend the money.

I've use lots of different 31 band eqs in various live and studio settings. They all sound different.  This is where my dislike of the "Roland" sound originated.  The TEAK 31 band (of the day) was much more transparent.  Also the Klark Technic (much more expensive) were very good.  Urei also made some decent eqs.

The DBX eqs (dual 15 band) I've been using in our PA are quite good for the money.

[Kazinator]

The EQ is an AB International.

I threw it into the effects loop now. I still need it for comparison. I'm looking to mod that Mid control more, and I use the EQ for getting a rough idea.

I saw a post in another thread about EQs. In the effect loop and a light bulb went on right away. With that, your MP-1 preset can bypass it. It solves the problem of the weird audio when the AB International is on bypass. Moreover, I can make a preset where the MP-1's, say, Mid is on 0, and the external EQ is in with a mid cut, and an almost identical preset where the EQ is out, but the MP-1's Mid is cut. Then quickly A/B between them.

[Harley Hexxe]

Hey Kaz,

That's the cool thing about a programmable preamp, you can save all kinds of different settings and call them up at a touch of a button.

@ Richard,
Putting outboard gear through a PA doesn't mean it's going to sound good in the MP-1 loop. Case in point: you know how much I love the old ADA rack gear that was way before the MP-1 was developed, and I have a huge rack worth of that stuff. Well, regardless how much I like them, they don't match up with the bandwidth of the preamps. They are only at 12,000Khz instead of 20,000Khz where all the high-end processors live. That means in the effects loop of any of my preamps, these will cut the level and darken the sound of the guitar signal. That's why these work in mixing desks better than in the amplifier, (meaning preamp>,power amp>, speaker cabs), signal path.
I know this is true because when I do use them in any of the effects loops, I have to boost the signal and trim the EQ a bit to get my signal back to where it was before the processing.

I use my Rane ME-15, (Dual 15-band EQ) in my PA system only. That is also where I use the Aphex Type C, and Carver C-9 Sonic Hologram projector. (Originally, I had these in the effects chain of my original MP-1).

[rnolan]

Hey Harley, I wasn't commenting on how any eq would be in the MP-1 loop, just that I've used a bunch of 31 bands and they all sound different.  Moreover the very expensive ones were much more transparent. The PA and studio references were purely because that's when/where I noticed the different sounds and colourations. And I was not so pleasantly surprised by the differences.

What would probably sound good in a MP-1 loop (if you need/desire) is something like the Boogie eq thing where the target frequencies are guitar oriented (as are the eqs in the preamp).  31 band eq's are not guitar oriented, they are 3d Oct centered around ISO standard frequencies.  A 4 band parametric eq is probably the way to go where you can control the 3 main parameters of 4 bands:
Boost/Cut
Centre Freq
Width

[Kazinator]

So I think I'm done with the EQ tweaks. For the past bunch of days I had just been playing, and made one more change.

It became obvious I had made the Treble control way too narrow: Q factor of around 3.45! Cut offs at around 2.7 and 3.6 kHz, centered on 3.1.

And at the same time, too wide ranging; big change from 0 to +2 and to +4.

I fixed both issues with a resistor change which knocked down the narrow peak. The peak being a bunch of dB's down makes the EQ steps finer, and the Q dropped to 2.0, widening the bandwidth a bit wider than 2.5 - 4.0 kHz.

You might not think it's a big change, but the ear is very sensitive. It feels dialed in just right to me.

You can see the experimental process in the image. I made five copies of that circuit alone with different part values.

I implemented three or four of these. The cyan curve is the second-to-last which I tweaked to the red curve.

I experimented multiple times like this with the Mid control and Presence as well. Only the Bass took just two tries.

Kind of time consuming! Plus with multiple trips to the resistor and capacitor grocery halfway across town.

The search for tone takes you to strange places you'd never think you'd go ...

I just have to document all this stuff. Sound clips would be nice.

[rnolan]

Hey Kaz, it's been an interesting journey .

( I just have to document all this stuff. Sound clips would be nice. ) +1 good idea

[Kazinator]

Oh man; I was looking at some Marshall tone stack frequency responses on a website. And I noticed that a bunch of the mid scoop profiles are exactly like what I dialed in for the Mid control: -3dB half-power points at about 320 Hz and 1.6 kHz.

I have Marshall-conditioned ears!

It had to be that way. Too far below 1.6, and the tone is too nasaly/telephoney. Too far above, the tone just disappears. Same with the opposite end: cut too far below 300 and you lose body. Cut too far above, and it's too tubby.

I played with it this way and that, and in the end, what sounds  good, like what I want to hear when I'm changing the FM dial in the car around rock stations, turns out to be the Marshall scoop.

It's kind a sad in a way.