luthifer

Sounds great Andy! Yes, pulling the filter pot goes to a higher peak boost. Both the low and high settings are adjustable. Set to 3dB/10dB by default, can go up to 13dB.

Distortion is very low across the board. I figure most folks will use an amp/pedal for that, the big advantage being that you can turn it off! I've been meaning to make a distortion module for the preamp, but can't seem to get around to it yet.

Yes, the new generation FT filter is the same tuning as Wal. I also have multicoil buffers now, and the Q levels for push and pull on the filters can be adjusted (0-8 dB for push, 0-13 dB for pull), which is of interest for a Wal-like setup because you can dial in a higher Q for the bridge. The adjustments are done with trimmers.

Yes, I figure with the cutoffs being around 2k, it's not hiss. I'll bet with the amp cranked up all the way there is quite a loud "rumble" though.

Oh! Thanks!!

I wonder that, need to look into slew rate calculations, because I noticed how incredibly low it is!

So here is the response of the bridge filter. That's all I have time for, for now (I'm in the process of moving, ugh.). It's the response given that the op amp could achieve it, which I doubt the TAB1043 can do, much less a 9V supply. (The TAB datasheet doesn't list its common-mode input range, so I would have to measure one). Those huge Q's should be distorting a lot. Maybe this old op amp is bad enough that everything is rounded off, which makes it sound good (again, I'd love to actually measure one on the bench)?

Wow! That is a seriously crappy op-amp, by modern standards! I'll bet it IS coloring the sound. Not to mention hissing like an angry snake... Anyone have any of them? I'll see if I can't find an equivalent replacement when I get some more time to work on this.

Are you using 9 or 18V? I'm testing a new op amp right now that allows a full input swing. So far it looks great, works with an MFD pickup in series, the gain trimmers up at max, and full Q. With the gain trimmers down, the signal is clean down to less 5V on the battery. I'm excited about it, there would be no need for 18V in any situation. The downside: It uses a little more power.

As you get near those boundaries, the distortion starts increasing smoothly (all the harmonics start coming up) and it sounds good. But to make use of that you would want to only use batteries that are say 7.6 to 7.9 volts (that's an arbitrary range, depends on the actual circuit). So it's not practical to make use of distortion in the op amps. A distortion circuit with diodes or whatever after the preamp, and adjustable, works way better. Plus you have the option of turning it off!

The op amps have pretty hard limits to how far the voltage can swing. You can easily calculate the input signal voltage where it will start to clip for a given battery voltage, and it tests out like clockwork. Most op amps are asymmetrical (i.e. the signal can go from say .1 to 6 volts when the battery is at 9V.) I've found that the clipping doesn't sound good at all for high Qs if a lot of the sound is near the peak (certain notes). You can hear a faint ringing. So I like to avoid any clipping if I can.

Interesting. It could indeed be distorting at 9. I joined this thread late, I hope we're not covering old territory.

Yeah I wish I had one to put on the bench and run some scans, so I could look at distortion and other variables. Or if I knew what the op amps are then I could figure out pretty closely what the distortion is going to look like. I'm skeptical that distortion is much of the sound, though, since distortion in the circuit is going to be hugely dependent on battery voltage. Distortion loud enough to hear at 8V or 9V is going to overwhelm the signal at 6V or 7V (What voltage does the preamp fuzz out at? That would give me a clue...). The circuit doesn't have any intentional distortion elements, and the reference voltage op amp (bottom left) eliminates a major source of distortion that you might get using a voltage divider circuit.

This morning I modeled the above circuit into neck and filter circuits for simulation. I also laid out a board that will plug into the Underhill PD4 and allow me to test and listen to the actual circuits. Honestly, I don't have high hopes for these filters as a general purpose filter preamp. The filtered feedback loop appears to be used to taper the Q, and it shifts the response to the bass end as the frequency rolls down. I don't completely have my head around the feedback circuit; maybe a real EE can explain it? The C19/R32 and C26/R44 high pass stages seem to be to cut out some of the sub bass boost from the feedback response (as well as set the impedance after the cap). It all seems like a bit of a kludge to work with the peculiarities of the Wal bass. The Q, of the bridge filter in particular, gets very high, I suspect because the response of the single coil pickups (the ones I have measured are very flat, and even have a dip in the 3-4k area). This is probably going to sound nasty with normal pickups--the high Q (>12) filters I have made tend to create "wolf-tones" at different spots on the neck, and are generally too honky and nasal in the areas your guitarist will give you dirty looks for! Anyway, it seems to work well for a Wal, but it doesn't look like something that will turn any bass into one (you will still need the same neck, pickups, and pickup placement to get that!). I haven't had time to make up a plot of the bridge yet, I'll post it when I do, but in the case of the neck, you can get the same response with an FV filter by just turning down the Q a little as you roll the frequency down, and you won't have the sub-bass boost problem or the inflexible Q slope. (It would need to be re-tuned it to the Wal's ~2.4kHz upper and ~95Hz lower frequencies.) Lastly, the circuit seems like it would be very noisy in several areas. I don't know what op-amps are used, so I don't have a complete understanding, but i.e. you could use a 10k for R32 and R44 and get 20dB reduction in noise at that point. The input summing circuit seems like it would be very noisy as well. I learn a lot more after I get the test boards back...

So I just put together a bass for a guy (really good player) who has owned a lot of Wals, including one of the first double-necks. He confirmed that my treble filter is too low, so I made up a prototype for him that just stays above the filter peaks. I thought it sounded thin until he played it (he exclaimed "there it is..!"). It really adds something without being in your face. So I'm going to update the filter I think. So it confirms that the 6k-ish HPF is the way to go for a Wal sound.

This is what my current FT4 treble boost filter is doing. It has a fairly soft knee (Q=0.5 I think) and the boost is settable with a trimmer. I tried going a little lower, but then the sound comes back to pretty much what you get with no LPF and some bass boost, unless you are only using the LPFs set way down low. It also has a pickup selector microswitch, and there's a big difference: The neck pickup (like Wal) gives a clean sparkle, and the bridge gives a much fuller sound. Setting the gain right is also critical or it gets nasty. One thing I might do is switch to a bandpass filter with a fairly sharp peak, to accentuate the pick sound somewhere around 1.8-2.5k, but eliminate the boosted fret noise. It works pretty well with the high filter on my two-band parametric eq, but that's maybe not sharp enough.

Not bad news, just surprising they want to boost way up there. I spent enormous amounts of time playing with a pick and trying different frequencies to get the same kind of meaty pick sound, and settled on ~2k. 6k seems more like a "fret noise" boost to me...

Wow! That's as I suspected if the caps are 1n. I'm surprised the pick attack is up that high. Yeah I can't hardly hear it either. Nor can my bass amp when set flat!

So I'm pretty sure C24 and C25 are supposed to be 10n. Even with the gain of the inverting "attack" amp all the way up (2x), you would barely hear the boost at frequencies below 7 or 8 kHz. 10n caps there would bring the cutoff frequency down around 500 Hz, which gives you lots of boost around the 1-2k "pick" frequencies and above. Maybe I'm missing something; what happens with the OUTB signal before it hits the final mixing amp is very confusing without a clearer picture.

I don't think Slowburnaz has a scope. Yeah I've measured well over a volt on the scope with slap! Slap is y'ur low battery indicator That's exactly what the EQ-01 does. I don't like the noise at high Q (or the Q control curve), but for super aggressive playing or "one size fits all" it's a great choice.

Hi Charlie! Yes, tradeoffs and design decisions... By the way, folks, Passinwind is the guy who sucked me into this rabbit hole! I don't know whether to love or hate him!

It is very clean. The variable Q filter uses a circuit that is much quieter at full Q (and has very linear Q control - 1/2 way is 1/2 way - and base gain is dead on), but has the downside of clipping at roughly the level of 350mV with 7V supply (and max Q). With a fresh battery, gain trimmers only used to balance pickups, and normal output pickups, it won't clip. But high output pups may need more than 9V. The sample is with the variable Q filter (The switched Q filter has a lot more headroom). I think it may be clipping a bit (hard to tell) but it doesn't sound harsh. Sounds pretty good, actually! The gain trimmers are applied before the filter, so if you wanted some distortion intentionally, you could crank them up and get it for sure. The high gain is available so you can use the preamp by itself as a booster, but it can definitely overdrive the variable filter, which has internal gain above the gain coming out.

In a Wal configuration with two switched filters (one, two, or three Q level switchable) and the pick attack filter, the Underhill draws 3.3 mA. The power drain of all the modules is on the web site.

No comment. Nice to get a look at it though. Thanks Funkle and Passinwind!

I'm curious if the pick attack was on in the Wal scan? If the schematic posted here is correct as far as the attack circuit goes, the cutoff frequency of that filter is around 500Hz (it doesn't get to full volume until about 1k). It's a low Q filter (0.5), so it doesn't have a sharp cutoff. Turning it on would appear to make the Wal-ish closer, though there's that hole around 1k... I just finished my own treble boost filter; I set it a bit higher and crisper (about 1500 Hz and Q=.707). I wasn't big on the concept because of noise, but it's really growing on me, and its actually pretty quiet.