Pickup info

Pickups, Diy Electronics, Stompboxes, Guitar wiring...
ed packard
Posts: 76
Joined: Tue Apr 19, 2011 10:54 am
Location: SHOW LOW AZ USA

Hardware status and details

Post by ed packard »

Pickups and sensors:

Because there seems to be some interest in the non contact sensor for ped/lev position sensing, that also works as an individual string pickup system (also all strings like the standard pickups), I thought that I would try and explain the approach.

The basic sensor element is a Honeywell HMC1501. It is a magnetic field direction sensor that consists of a magnetic Wheatstone bridge. It can be used with an external magnet, or just an external magnetically conductive material, like a screw or a string.

Image

It is followed by a gain stage in the form of an “inamp”…instrumentation amp. My choice is an Analog Devices AD627, or an AD8223…a matter of taste and cost.

Image

The combined elements may be powered by a single supply, or a dual supply system. I am using a Radio Shack available AA battery holder that holds two AA batteries = 3VDC. I add a wire to it from the junction of the two batteries so that it gives +1.5VDC and +3VDC = either single +3VDC source, or +/- 1.5 VDC source.

The reason for the choice is that subsequent circuitry may want a signal that works from “ground” up, or centered about “ground”.

Image

There are two ways to set up the interface between the HMC1501 and the ADxx…AC or DC coupled. If DC coupling is used, the offset in the HMC1501 will be amplified along as any externally sensed signal…if AC the offset is ignored. DC is good for the PED/LEV position sensor, AC is best for the pickup. The board can be laid out so that how it is populated and powered can serve both/either the offset and the “grounding” issues.

The boards that I have now do not have the AC coupling as they were aimed at the PED/LEV sensing, I.E. meant to deliver a ramp voltage as a function of cross shaft rotation. The reason for using the cross shaft instead of the ped/lev is familiar appearance for the first approach.

The boards were laid out to have 6 channels that can be split apart via perforations. They can be used as either pickups, or position sensors. The next generation will have the choice of AC or DC coupling in the front end, depending upon how the board is populated.

The sensors are spaced for 11/32” string separation = most PSGs. Two boards gets you 12 strings, or 10 strings and some other function like maybe a volume sensor etc.

Image

An alternative would have been to use 3/8” string spacing as found on standard guitars and slant the board for use on PSGs.

Connection to the board, or the channels is by 100 mil centered connectors (6 pins for the channels). The power source is common for all channels if used together as a pickup, or has connectors for each channel is split apart. A 50 mil centered connector (10 pins) is added to handle all power and signals for the pickup approach.

A secondary split off section is included for added coupling flexibility. If populated and split, the 50 mil centered connectors will plug together to route power and signals to and from the pickup board to the outside world. The signals in this case are thru resistors that can be ganged at the output to make the individual string pickup function into a combined function as well as individual.
Image

The output signal is from pin 6 on the ADxx, and the signal “ground” is from the “REF” pin 5 on the ADxx. To switch between single supply mode and dual supply mode just interchange the 0VDC and +1.5VDC input wires on the power in connection.

External magnet(s) may be used to bias the offset in the HMC1501 into, or out of saturation. This will interact with the gain setting of the ADxx to provide some interesting overdrive effects.

For PSG ped/lev sensing, an adjustable spring system will be needed to allow adjusting for “feel” to suit the user. If using the cross shafts as the sensor location, the ped to shaft rods will make the unit look “conventional. If the ped sensors are located at the ped bar, the rods will not be needed. In either case the tuning rods and bell cranks (and associated changer mechanism will not be needed….this assumes the DSP pitch(spectral) shift functions are attained. Till then, the position sensing ramps can be used to control other mechanical string tensioning devices, and the pickup used as the common wound magnet pickups are used.

There appears to be about 50dB of separation between strings at the 11/32” string spacing when used in the individual string mode…without tinkering to optimize.

Individual string outputs look/sound quite like those of the common wound magnet units.

Image

Higher supply voltages can be used (say 5 or 6 VDC) to play around with the offset to gain interaction.
richard37066
Posts: 517
Joined: Sat Dec 11, 2010 11:44 am
Location: Gallatin, Tennessee, USA

Re: Pickup info

Post by richard37066 »

Ed -

Yup - you're filling in the gaps in my understanding of this contraption. Many thanks.

Did you assemble the boards yourself or farm it out to a specialty house?

If one were to commit to this unique design, can you supply a ball-park cost figure for, say a 12 or 14 string configuration? Budgets and all of that rot, don't you know.

Keep it coming.

Richard
ed packard
Posts: 76
Joined: Tue Apr 19, 2011 10:54 am
Location: SHOW LOW AZ USA

Re: Pickup info

Post by ed packard »

RD...I do the layout and farm the board out. I send parts (not Rs Cs and connectors) just chips from Digi-key (Google it), and they assemble.

Cost is a function of quantity 3 to 4$ each for the chips in batches of 100 ea. I buy the bare boards in batches of 100. Usually have about 10 populated for experiment and just store the rest.

I don't have the "gear" to populate the boards...too small for these shakey old hands and unfocused eyeballs.

How is the chord hunt coming?

Edp
richard37066
Posts: 517
Joined: Sat Dec 11, 2010 11:44 am
Location: Gallatin, Tennessee, USA

Re: Pickup info

Post by richard37066 »

Ed -

The chord hunt? The fox is still outrunning the hounds. I often think that I'm attempting to create the proverbial blivet - ten pounds of crap in a five pound bag - or of flying to the moon on a bottle rocket! Your approach has given me direction on a couple of fronts and, should I approach my ideal, I'd like to lean on you for a critique of the method. 'Til then, "into the breach" or some such rot.

Thanks for asking.

Richard
ed packard
Posts: 76
Joined: Tue Apr 19, 2011 10:54 am
Location: SHOW LOW AZ USA

Re: Positioner info.

Post by ed packard »

A bit about the non contact position sensor application. The link is to the magnetic Wheatstone bridge chip, and a manufacturer recommended way to use it.

http://www51.honeywell.com/aero/common/ ... 1-1512.pdf

For our purposes, we want the motion in the ped/lev systems to provide a ramp of some sort when the ped/lev is operated. The ramp shape should be adjustable re the ped/lev operation. There are several ways (locations) to sense the ped/lev position. The most simple way is to use the rotation of the cross shafts as they are presently used =let the rotation change the relationship between the sensor and the item being sensed (a magnet for now). If the tuning rods are not used, then the cross shaft will need a return spring that makes the ped/lev feel normal.

The above approach leaves the pedal rods on the PSG so that it looks the same to”traditionalists”. The alternative is to mount the sensor at the pedal board and not use the pedal rods or pedal cross rods…still have the levers to worry about.

As described before…the position sensors are just a single channel of the pickups. The perforations for separating the channels are built into the boards.

These functions could be done with pots as someone has suggested. I don’t care for that approach as a single pot in a volume pedal causes much aggravation…now put pots on 8 pedals and 7 levers!!!
Halldor
Posts: 2
Joined: Wed Jan 16, 2013 8:27 am

Re: Pickup info

Post by Halldor »

Dear sirs.

Thank you for the thread above.

I have a string instrument project in perpetual development, see here:

http://halldorulfarsson.info/halldoroph ... dorophone5

It uses controllable feedback for perpetual sustain and tone color…

I have been using "spot" pickups to isolate the strings on individual channels (which is essential to the premise of the instrument). Coil pickups are problematic for me because they "spill over" from string to string and there is a a problem with coil to coil feedback (pickups to speaker, which cuts out all the nice acoustic stuff inbetween…).

I have recently become aware of the magnetoresistive pickup as a promising solution to these problems. I first saw mention of it here:

http://jap.aip.org/resource/1/japiau/v9 ... horized=no

Which seems to be an feasability report published by Philips corporation researchers, discussing the benefits of such a pickup. (you can download a .pdf of the full article in one of the sidebars)

And then this thread, as the first practical discussion among builders I have seen, again, thanks for that Mr. Packard.

I have ordered parts and am wiring up a test on a breadboard. I am working with replacement parts as I have been unable to order either the HMC1501 or the KMZ10 (used by the Philips team) here in Iceland (it seems to me they are out of production?).

The magnetoresistive sensor I have is Honeywell´s APS00B

And the opamp I decided to use is TL072 because it´s well documented among the audio electronic DIY crowd.

I wired up a test based on the documentation above and I do get a signal, albeit noisy when I tried listening to a string.

Mr. Packard would you be willing to drop a few lines here on the AC ("pickup") wiring for your circuit to help me along?

Any comments and directions and ideas well appreciated!

greetings from Iceland,
Halldór Úlfarsson
Halldor
Posts: 2
Joined: Wed Jan 16, 2013 8:27 am

Re: Pickup info

Post by Halldor »

Dear sirs,

More on the major topic of the thread above.

We are now working here at EMS in Stockholm, Sweden to create an accessible package for anyone who might want to use a magnetoresistive sensor as a pickup.

We are working with the Honeywell HMC 1512 containing two of the Wheatstone sensing elments (vs. one in the HMC 1501 used by Ed Packard in the examples above).

We have a test circuit wired up

Image

And we are doing some testing

Image

Here is a diagram of the circuit:

Image

The main problem we are running into is a fairly constant noise. We did try changing the input impedance to the opamp from 10k to 1k but it didn't have a noticable effect.

About the circuit:
We have added a pot to control the signal bias, which we have found to be good at around 2.5v. Note that the signal bias is also dependant on the mass of the respective string being sensed... (hence the variable resistor).

And we have added a second pot to control the gain.

We did experiment with the diractionality of the ambient field (a function of supply voltage), which is interesting but perhaps not relevant to the work we are doing (easier to just move / tilt the sensor in relation to the string).

Here is the layout of the circuit:
Image

Is there anybody out there who has experience with instrument amplifiers (op amp) who can give us pointers about noise reduction?

Many greetings from Sweden,
Halldor Ulfarsson
User avatar
Jif
Posts: 86
Joined: Thu Nov 08, 2012 12:54 pm
Location: Scotland

Re: Pickup info

Post by Jif »

Hi Halldor,

Looks really interesting what you are doing here :)

I'm no expert but in my experience with op amp ccts the op amp itself is generally not the cause of the noise but amplifies the noise present in the circuit.

Looking at your circuit and looking at the circuits in Ed's post, the op amp being used is a specific instrumentation amp with gain control through the external resistor. I have no experience of using these, I have just used conventional op amps, however I would guess that the principles are similar.

In the suggested circuit in the diagram below, there is a feedback loop between the op-amp output and the inverting input consisting of a 249k resistor and a .01uf capacitor. The non-inverting input is being biased at 1/2 the supply voltage. As the description says the combination of the 249k resistor in the feedback loop and the 10k resistor on the inverting input sets the gain of the amp to 25, (249k/10k)

The 0.01uf capacitor provides frequency stabilisation and removes noise from the circuit, it is possible to tune this to a frequency range through the combination of the capacitor value and the resistor values to suit.

f=1/2*Pi*R*C

My feeling is that there should be a capacitor(s) in your circuit somewhere to filter out the noise in the circuit. I found this data sheet for the AD627 which seems identical to the AD623 you are using.

http://people.virginia.edu/~cas8m/class ... /AD627.pdf ,

if you look at page 21, chapter - RF INTERFERENCE there is a good description with circuit diagram on how to reduce noise in the circuit. This might be worth a try.

If that doesn't help it might be worth trying a more conventional op amp using a variation of the suggested circuit below, with the RC feedback loop providing the gain factor and noise reduction.

Hope this is of some value to you, good luck ;)

Image
Post Reply