a different kind of low pass filter

[khutch]

It is interesting to hear what you all are doing. I've given some thought to building some kind of filter to try out these ideas but currently I am trying to do this another way entirely. Analog Devices makes some single chip audio processors that have an A/D converter, a DSP engine, and a D/A convertor all on one chip. Doing all of this digitally appeals to me as a lark if nothing else so I am in the midst of trying to get a cheap ADAU1701 board from a Chinese company. They make one that will supposedly connect to Analog Device's SigmaStudio software to let you build all kinds of digital filters on the chip and try them out in various combinations. In theory they can be controlled by external pots and switches. The SigmaStudio software lets you configure the filters and control signals graphically on your computer screen -- no DSP programming skill required! So, it might make a neat toy to play with or even embed in a bass guitar some day if it all works out. Right now I am in the throes of trying to get the device actually ordered from the company (Sure Electronics). Something went wrong when I did the web order and it is slow as molasses trying to get that sorted out via email. Hopefully soon....

 

[Passinwind]

It is interesting to hear what you all are doing. I've given some thought to building some kind of filter to try out these ideas but currently I am trying to do this another way entirely. Analog Devices makes some single chip audio processors that have an A/D converter, a DSP engine, and a D/A convertor all on one chip. Doing all of this digitally appeals to me as a lark if nothing else so I am in the midst of trying to get a cheap ADAU1701 board from a Chinese company. They make one that will supposedly connect to Analog Device's SigmaStudio software to let you build all kinds of digital filters on the chip and try them out in various combinations. In theory they can be controlled by external pots and switches. The SigmaStudio software lets you configure the filters and control signals graphically on your computer screen -- no DSP programming skill required! So, it might make a neat toy to play with or even embed in a bass guitar some day if it all works out. Right now I am in the throes of trying to get the device actually ordered from the company (Sure Electronics). Something went wrong when I did the web order and it is slow as molasses trying to get that sorted out via email. Hopefully soon....

Cool! I was looking at some of the AD evaluation boards for that series a few weeks ago, which are not exactly cheap.

Another TB DIY'er sent me a schematic recently for some synth filters using the LM13600 transconductance amp. I haven't had time to vet that part yet, but the schematics seemed very straightforward at first glance.

 

[SteveCS]

I'm looking for a different kind of onboard bass guitar LPF... The regular tone circuits operate by steadily decreasing (or also increasing, in the case of active) the slope of the filter at a fixed knee point, thereby cutting/boosting high end. I'm looking a tone circuit, preferably passive, that has fixed slope, but that slides a fixed cut (downward) slope of 12 or 18dB/octave upward and downward in the frequency range.

Anyone know of such a thing?

How many pickups on your bass? The reason I ask is that with two pickups in a VVT (Jazz-Bass wiring) configuration, the cut-off frequency of the tone control falls off, i.e. does what you want, as you roll back the volume controls. You only need to roll them back a little (99-95%) to dramatically shift the cut-off frequency. If the cap is small enough to start with the cut-off (with the tone control rolled back) can be quite high (1-2k for a 0.033u cap depending on pickup impedance and DC resistance). With 1 volume almost off, rolling the other back by just a couple of % can reduce the cut-off to the 100-200Hz range before you lose too much in overall volume. Roll the other pickup back up a bit and the cut-off starts to rise again (albeit slowly). Classic J-bass interaction and one of the reasons I prefer VVT over BVT. Of course there is no real boost, just a resonant peak, so the overall effect is quite subtle compared to an active filter, but it is there, and its passive which is what you are asking for.
As always, YMMV...

 

[khutch]

Yeah, AD seems to have no interest in serving the hobby/starving musician market for the 1701 chip! Their eval board is around $600, but wait, there is a low cost version for $220. This is for a chip that costs under $10.... The latest email from Sure indicates that they have corrected my order and it should ship today/tomorrow so I am cautiously optimistic that I will get one of their $30 ADAU1701 boards. The DSP board is under $20 but if you want the audio connector board too it drives the price all the way up to $30! And I think shipping is another $5. They really make it seem like ADI just isn't trying, eh?

Now I do have to get a USB to I2C interface for the board. ADI sells one for $80!, but I have ordered one from DigiKey for under $10. I'm not going to say any more about that unless someone here says they want to know and until I have had a chance to hook it all up and verify that it will work. There is always a chance that only the ADI board will work....

There are a couple of other mid priced options if someone wants to play around with an ADAU1701. Elektor magazine sells a board for $107 (with a USB interface board) that is compatible with the ADI SigmaStudio graphical programming tool. Invalid Link Removed but it is NOT compatible with either SigmaStudio or with real time control via pots and switches. So you could certainly play around with it and have a lot of fun/prototype a lot of concepts quickly but if you actually wanted to put it in a bass and have real time control over it you would need to have a Windows computer attached to the bass at all times.... Their business target is the DIY home theater market where real time control is not a required feature and they seem unwilling to move off that position. But if having a computer on stage with you as you play works for you then it certainly is an option.

Another option entirely is to use Link Removed and similar. They are not as flexible as DSP filters are but it would be easy enough to make a LPF filter, for instance, where you could vary the cutoff frequency by varying the clock frequency to the chip and vary the Q of the filter with a pot. Making a variable frequency clock oscillator that you could control with another pot is something I will leave to the students! They will also do bandpass, highpass, and notch filters so this class of chips could be interesting to play with.

 

[SteveCS]

...but the other thing I just remembered is that you wanted 12dB per octave whereas the the standard tone control, being first order, is only 6. :-( you could add a second stage identical to the first but then you do start to lose output level.

 

[Passinwind]

Last year I pared back my design to just a single knob that actually adjusts the resonance and frequency at the same time. And similar to you it only sweeps the treble region. Low Q / low freq to high Q / high freq, across about two octaves. It is very like sweeping between series and parallel.

I put the filter in a couple of my basses and have been gigging and touring with it for about a year. It's brilliant. I call it a "treble voicing filter". It is especially cool with a stingray pickup.

Good luck with your design!

I installed the first test board in a stompbox today to facilitate doing some measurements and whatever debugging and modding might be necessary. Correlation to my Spice model is very good. It sounds quite nice, and I could easily see your one pot solution working great for me, or maybe just adding a pull pot switch allowing two resonance modes, i.e. boost/cut. I'll build up a second one in the next day or two, put that in my Travis Beam fretless, and see how the front end interacts with passive pickups, keeping your damping adjustment recommendation in mind. Thanks again for sharing your experiences with this topology.

 

[Spacecase]

I installed the first test board in a stompbox today to facilitate doing some measurements and whatever debugging and modding might be necessary. Correlation to my Spice model is very good. It sounds quite nice, and I could easily see your one pot solution working great for me, or maybe just adding a pull pot switch allowing two resonance modes, i.e. boost/cut. I'll build up a second one in the next day or two, put that in my Travis Beam fretless, and see how the front end interacts with passive pickups, keeping your damping adjustment recommendation in mind. Thanks again for sharing your experiences with this topology.

Nice report, Charlie. How well did the linear pots scale with the change in tone?

 

[Passinwind]

Nice report, Charlie. How well did the linear pots scale with the change in tone?

I'm totally happy with the sweep feel so far. 10-2 o'clock for the freq covers 1.8K-4KHz, and the resonance control gives 10: unity, noon: +1dB 2: +3dB, full up: 6dB. All of that is exactly what LTspice predicted, and in practice there's nothing at all goofy about the feel. Putting it in a bass and feeding it with passive pickups will be another test, but for your bass with active p'ups it would probably work pretty danged well.

 

[Passinwind]

I'm totally happy with the sweep feel so far. 10-2 o'clock for the freq covers 1.8K-4KHz, and the resonance control gives 10: unity, noon: +1dB 2: +3dB, full up: 6dB. All of that is exactly what LTspice predicted, and in practice there's nothing at all goofy about the feel. Putting it in a bass and feeding it with passive pickups will be another test, but for your bass with active p'ups it would probably work pretty danged well.

I ended up rewiring my Travis Bean to a straight passive config to test through my temp stompbox:

I have a new case arriving tomorrow with a little more room to allow for a wider variety of switches and configs, but as of right now I don't think the Bean has ever sounded better in the 39 years I've owned it. V2 is on my workbench and should be ready to install in the bass by the weekend. I broke down and bought a stacked pot and knob set for the Bass and Mid controls, so my 5-hole config will be V-V-B-M-T freq-T res.

 

[Spacecase]

Dayum. You do nice work.

 

[Passinwind]

Dayum. You do nice work.

Thanks Bro. Here's the new board I just finished stuffing:

And the newer stompbox I referred to upthread

 

[Spacecase]

Ya see what I mean?

 

[RobbieK]

I've had a handful of pm questions etc about this filter, so thanks for the interest, and here's an update!

The circuits I've posted so far in this thread are just part of the discussion and not really fully fleshed and tested bass preamps.

The dual stage (with 4-gang pot!) and hugh sweep (100-10k, give or take) was directly aimed at the OP. This is pretty much what he/she asked for.

The last circuit I posted (post 83) works fine but since last year I've been running a couple of my basses with filters derived from this discussion. And the filter has had a few tweaks.

So here's a fully functioning circuit that can be used as a bass pre in its own right.

As is, it is designed to have the output of a blend pot (unearthed) connected straight to C1 and have a 10K or 25K volume pot wired at the output.

If you have a single pickup bass, just run it's output straight into C1. You may want to reduce R1 for this. If it's a typical stingray pickup wired in parallel, I'd recommend 100k.

OTOH, if you have a really hi z pickup that is wired in series, I'd recommend bumping up R1 to 470k.

I haven't had a chance to try this in my P, but when I do I'm planning to run the pickup in parallel (with R1 as 100k), and slightly redesign the first opamp for 6dB of gain.

The pot is still a dual gang C50k. As far as dual gang reverse-audio taper pots go, 50K seemed the most common last year! I found them on ebay, amazon etc. Now I'm struggling. FWIW, any dual gang reverse audio pot from 5K to 100K can be made to work. You just have to scale the other parts in the filter appropriately. eg, with a 10K pot, you divide the resistors by 5 (1k2), and multiply the caps by 5 (22n and 5n6).

Stuff I've changed:

The sweep is now between 1K and 8K. IME, sweeping through those three octaves is about right as a useful/musical resonant treble control.

The hi res is now 6dB, for a slightly more obvious difference between the switch setings.

The low res setting always seems a bit bland when you switch back to it, but it gives a rolloff that is more "passive" sounding.

On the schematic I've laid out SW1a&b to resemble the lugs on a push-pull switch.

The reason the high pair of caps now switches in series is to avoid (notoriously inaccurate) ceramic caps. I could have used say, a pair of 680p's in parallel but green caps and mkts are far more common in sizes above 1nF. C4 and C5 are at Vref now to avoid switch noise. (Technically, they should have always been there ).

The opamp can be just about any low to medium current audio opamp. TL 062, LF442, NE5532, TL082, LF352, LM833 etc etc, there are thousands.

Here are some sims:
Hi res...

Low res...

And finally, I've had a few requests for stripboard layouts etc. I'm afraid I don't have a heap of time for this at the moment. Plus I don't really have a program that I can easily create good looking ones with. TBH, with a little circuit like this, I'd cut a piece of strip board, slice the tracks in a line, place the opamp socket to straddle this line, then just go from there. Unless you are really short on space, it shouldn't take too much forethought.

 

[Passinwind]

And finally, I've had a few requests for stripboard layouts etc. I'm afraid I don't have a heap of time for this at the moment. Plus I don't really have a program that I can easily create good looking ones with. TBH, with a little circuit like this, I'd cut a piece of strip board, slice the tracks in a line, place the opamp socket to straddle this line, then just go from there. Unless you are really short on space, it shouldn't take too much forethought.

I'm contemplating eventually making my design fully open source, and sharing the PCB fabrication files and a Mouser BOM. But everything about my design is more complicated and expensive, and will probably evolve to be even more so. Big props to to you for putting something out there that people could build today!

 

[RobbieK]

Big props

Cheers mate. There are such huge discussions on TB about which pots and cap to use in passive basses. Really, with a very simple circuit like this, you get far more from a tone pot. If you use a low current opamp, like the 062, a battery will probably last several years with normal use. Or if you don't have room and don't want to route, well I think you'd get pretty good life from a little 12V battery as well. Like an A23.

If you can confidently wire a point to point passive jazz bass circuit with a series-parallel switch or something, a project like this is really not a million miles away.

 

[BassLife77]

Thanks Bro. Here's the new board I just finished stuffing:

I use those same LT chips in my builds. they are so underrated. Can you post the layout of this build?

And the newer stompbox I referred to upthread

 

[Passinwind]

I use those same LT chips in my builds. they are so underrated. Can you post the layout of this build?

I plan on posting the schematic once it's finalized, which is likely to take at least a few more weeks I think. But if you mean a strip board layout (?), I don't build things that way much at all these days, and in this case I just did my first two test builds on a PCB, and may or may not do one more with the remaining first gen board. It'll take ~ two weeks for the revised boards to come in, assuming I order them in the next day or two.

 

[RoadRanger]

I plan on posting the schematic once it's finalized, which is likely to take at least a few more weeks I think. But if you mean a strip board layout (?), I don't build things that way much at all these days, and in this case I just did my first two test build on a PCB, and may or may not do one more with the remaining first gen board. It'll take ~ two weeks for the revised boards to come in, assuming I order them in the next day or two.

Any chance of running some extras for those interested?

 

[Passinwind]

Any chance of running some extras for those interested?

Absolutely, just want to keep things non-commercial on my end at this time though. The idea of doing a board share would be that anyone could order a short run (3 bare boards from a US vendor, = approx. $16 shipped, including overseas destinations) at their leisure, and the middleman (me) would be pretty much cut out of the loop. I already have a beta test builder lined up, so hopefully I can fast track the whole process. So far it's going better than expected, if anything.

 

[BunchyMutt]

Sub'd. Love seeing tidy electronics and schematics like above.


Also, now that you've figured out an onboard LPF with variable state cutoff and adjustable Q, time to get to the HPF version