TPL0501-100: Can we have a quarter-cycle sine taper digital pot?

Don Baker

Part Number: TPL0501-100

Tool/software:

I've also posted this in the MSP432 forum (misspelled TSP, sorry). I had not filed it here before, because the only "audio" forum choice allowed for a TPL0501 is "Data converters forum" for this device. I'd like to hear directly from TI product engineers working on digital pots about a different application.

If you have two independent signals, put one at +1 on the x-axis and the other at +1 on the y- axis, then it seems to me that the optimal blend-balance curve is a quarter-circle in the (+,+) quadrant. Am I wrong? If you extend one of the signals to -1 on its axis, and call them A and B, then a half-circle covers the range from A to A+B to B to A-B to -B. If the human ear cannot hear a simple phase inversions, then this covers all the phase combinations, with only B and -B as duplicate timbres. I've been filing and getting patents for electric guitar circuits that treat humbucking pickup signals this way, including with combinations of matched single-coil pickups. Three matched single-coil pickups can get you 6 distinct humbucking circuits and 6 non-humbucking circuits - as compared to the usual 5-way switch with 2 nominally humbucking outputs.

But if you linearly mix humbucking signals from pairs of matched single-coil pickups, you can get not only all the possible mechanically switched combinations of circuits, you can get all the continuous variations of timbre in between them. I have a 2020 Springer-Nature textbook on the subject (not many sales), Sensor circuits and switching for stringed instruments, which covers the basics.

Please have a look at Fig. 17 in U.S. Patent 10,380,986; Fig. 16 in 10,810,986; Figs. 10, 11 & 21 in 11,087,731; Fig. 53 in 12,020,674; and Figs. 18, 26, 30 & 31 in US2023/0013236 A1. I've been designing these circuits in mind for TI micro-power op-amps and TSP430 and 432 micro-controllers, to run for more than 100 hours on AA batteries. I can see a large place for a micro-power digital pot with a quarter-cycle sine taper, to use in the linear mixing of humbucking pair string signals. I think that two in combination would also make a good blend-balance pot.

I have more patents in mind. But I'm limited to filing and prosecuting patents. I don't have the resources for anything else. Plus age and medications have removed my software skills. I need partners.

Any bites?

over 1 year ago

0 Paul_Frost over 1 year ago

TI__Guru 53773 points

Hi Don,

I am not 100% I follow the design you are considering here - do you have a schematic snippet you can share? Is there a feature wish list you have for the TPL product family? We create these DPOTs for many industrial, medical, and personal electronic applications, so generally we keep them pretty simple to make them more universal. I was not able to find the exact images from the patents you describe.

In regard to patents, I am affraid that is where I cannot help :) We are limited to customer support activities, as TI generally does not want to limit our customers or potential customers from using our devices. Our reference designs are open and we have a hope and desire that they are copied or leveraged as a starting point in design.

0 Don Baker over 1 year ago in reply to Paul_Frost

Prodigy 30 points

Hi Paul,

Thanks for responding. You can find good copies of the patents and application on patents.google.com, with images. I write them like engineering journal papers with plenty of images. I also have most of the granted patents at: Click Here. I hope that worked. If it didn't, go to tulsasoundguitars.com and click on "New page with U.S. Patents" under "Recent Posts". Try also: https://patents.google.com/patent/US20230013236A1/. I wrote that three years ago, and recently rewrote a lot of it, with a replacement Specification and Claims, but you can get the idea.

Here's one: and another:

P1cos has a center-tap input. The plus signs on the guitar pickup coils (A, B, etc) show the same phase of exterior, interfering hum signal, which cancels in subtraction in the differential amplifier. It's possible to do this with mechanical pots or digital, but works better if the tapers are half-cycle (sine) or piecewise (cosine) sinusoids. A bit like the imperfect f(x) and g(x) in Fig. 19:

RadErr is the difference from a polar vector length of 1. RotErr is the difference between x and the translated angle of an ideal polar vector on a half-circle. This is not my best algorithm, just an available Figure from the application.

I don't use Pi or angle in my plots, having translated the curves to a pot taper space. If the digital pots are linear, I claimed a numerical algorithm that uses only range compare, add, subtract and multiply in the math processing unit to get sinusoids. No cosine, no sine, no square root or divide. So it could run on a T.I. MSP430 or 432 u-controller.

For a blend-balance two gang pot, look at the curves for a Bourns Pro Audio PDB182-GTRB. Now imagine those crossing curves as two quarter-cycle sinusoids, which I think would keep the polar vector of the combined signals on a quarter-circle in 2-space. That's a nice general application, which I have just publicly disclosed and cannot now claim in a patent application. You're covered on that. And if you make them, I can incorporate them into the diagrams above. That is, if I ever get the resources.

Does that help? Would T.I. like to make some advanced electric guitars? I have the patents to license.

0 Katlynne Jones over 1 year ago in reply to Don Baker

TI__Mastermind 41573 points

Hi Don,

Thanks for your more detailed explanation. I'm still unclear on your asks from us. We wouldn't be able to help you develop the MSP430 algorithm for this application for you. Do you have any specific questions about the TPL? That's pretty much the extent of the support we can offer you.

Best,

Katlynne Jones