Force Sensitive Resistors, or FSRs, do exactly what their name suggests: they vary the resistance between their 2 pins based upon the force applied – the resistance goes down as the force increases. FSRs are one of the easiest sensors to use, and are found quite frequently in musical instruments. Monophonic aftertouch on most synthesizers uses an FSR tape similar to the one shown in Figure 1, and the Nintendo Power-Glove used the Abrams Gentile bend sensor for finger flex detection.
Despite the FSR allowing for an easy and expressive interface, it has a few disadvantages such as drift and inaccuracy. We will show how the FSR works, explain the reasons for its good and bad attributes, and give examples of how to best implement an FSR in your project.
Figure 1 – A variety of FSRs. From left to right: 1. Interlink FSR tape (coiled), 2. CUI 5mm square FSR, 3. Abrams Gentile bend sensor, 4. FlexiForce 1/4″ round FSR, 5. Interlink 1/4″ round FSR, 6. Interlink 1.5″ square FSR.
Although all of the above sensors look very different, they have a number of things in common. First, they have a body consisting of an insulator and 2 conductors. In the case of the CUI sensor, this is a kapton strip with copper conductors, the same thing used for flexible circuit boards. For the remainder, they use a thin plastic with printed traces (usually Indium-Tin-Oxide (ITO) or silver ink). The next thing they have in common is a pair of electrodes. These can either be discs on either side of the FSR material, as shown on the FlexiForce FSR above, or with alternating traces on the same side, which the Interlink sensors use. The final piece is the FSR material itself, which is a resistive polymer in most commercial sensors. A cross-sectional view of both of these styles is shown below.
Figure 2. – FSR electrode placement: Left – Opposing sides (FlexiForce), Right – Same side (Interlink).
The Interlink FSRs have a unique feature: their resistive polymer is attached with double sided tape on the edges (the glue/spacer shown in Figure 3). This means that the FSR material does not contact the electrodes when no force is applied, and therefore the resistance is infinite. This allows for easy contact/no-contact detection, and a consistent off-state. It also lets you take the material off easily, if you want to make your own electrodes and custom FSRs. An exploded view of an Interlink FSR is shown below.
Figure 3 – Exploded view of 1/4″ round Interlink FSR.
So how does this all work?
The principle behind FSRs is quite old, and was actually employed in the first decent microphone – invented by Alexander Graham Bell for the telephone, and is still used today in older corded telephones. These were called carbon microphones, because of the carbon granules inside of them which had a slight resistance. As a person spoke into the microphone, a diaphragm would compress the granules inside, increasing the amount of contact surface area between granules. With more contact area, there were more efficient routes for electricity to flow, and the resistance dropped accordingly.
Figure 4 – FSR material being compressed, and creating more conduction paths (lowering resistance).
Modern FSRs use a resistive polymer to obtain the same effect, but because of the small size of the “granules” in the polymer, they achieve a much more uniform resistance change with pressure. You can build your own FSR from conductive foam – the type that integrated circuits get shipped in for anti-static protection. A piece of this foam is shown below, and you can see all of the voids in the foam when it isn’t compressed. If you place an electrode on either side (or two on one side), you can can make the resistance drop by compressing all the voids out of the foam. With the foam shown below, the resistance went from infinite with no pressure, to 100kohms when pressed very hard.
Figure 5 – Conductive Foam can be used as FSR material.
FSR pro’s and con’s:
FSRs are just resistors, so they are easy to work into circuits, and sometimes don’t require any support circuitry. They are also relatively inexpensive and readily available, even Digikey has started carrying the Interlink FSRs. But, unfortunately, that is where the pro’s end. The main drawback to FSRs is a result of its central component – the spongey resistive material in the middle. Just like a sponge, it takes a while to re-inflate after its been compressed. And worse yet, this re-inflation time is a function of how long and how hard you pressed on it. After you release, the FSR value will come back to 95% of its initial value almost instantly, and then drift that final 5% over the next 10 seconds.
The inner material is also very sensitive to how it is pressed. It has a non-linear pressure response which varies with time, temperature, humidity, and even between parts of the same production batch. This makes the FSR a poor choice where accuracy and repeatability are a concern, especially across many units. Fortunately, a lot of musical applications have a human in the feedback loop who is able to just press harder or softer to get the desired effect, making this inaccuracy of less concern.
Finally, FSRs can also be very fragile. They are made of thin, laminated plastics, and are frequently placed under high forces. This usually results in them coming apart if they are not mounted correctly. Despite these drawbacks, with careful mounting, the FSR can be a very useful tool in creating expressive interfaces, like the MIDI gamelan instrument below. In this instrument, the FSRs are sandwiched between two plastic plates with a cushioning disc (click on the image for a close-up view).
Figure 6 – A percussive MIDI gamelan instrument implemented with FSRs.
How do I use an FSR in a circuit?
An FSR can easily replace almost any resistor in your circuit and give you instant external control. The resistance varies from several megohms with no pressure, down to a few hundred ohms under very heavy pressure. This is particularly useful for audio circuits, as you often want to modify a frequency or amplitude over large ranges.
If you want to create a control voltage with an FSR, or send its signal to the ADC on a microcontroller, there are a couple of options. The first is shown on the left of Figure 7 below. This is a standard voltage-divider configuration, with the FSR acting as one of the resistors. You can place the FSR on the top of the divider if you want the voltage to increase with pressure, or at the bottom if you want the voltage to decrease. A simple op-amp follower is shown on the right of Figure 7, which will keep the FSR output consistent and buffer the signal for driving smaller loads. You can also put gain in this op-amp if you want to increase the output swing of the FSR signal.
Figure 7 – Typical FSR circuits.
One of the characteristics of an FSR is that its resistance does not change linearly with applied force. The resistance will tend to drop quickly at first, and then more slowly as most of the spaces within the FSR material become compressed. This can be advantageous if you want to cover a large range of force values, as it essentially compresses the range, but it also makes the circuit very sensitive to the initial light touches on the FSR. To compensate for this, and linearize the FSR output, the circuit in Figure 8 can be used. As with the voltage divider circuit, you can get either positive or negative voltage swings by tying the FSR to either the positive or negative voltage supply, and gain can also be used to increase the FSRs effective range.
Figure 8 – FSR linearization and gain circuit.
With any of these circuits, it’s important to remember to keep the current low through the FSR. The more current you send through them, the more they heat up, and the more the resistance changes. With too much current, the FSR material can even be damaged. It’s best to keep the current under 1mA.
How do I attach the FSR?
Although FSRs are relatively easy to use in a circuit, they can be a bit tricky to attach to things, both mechanically and electrically. Most FSRs use metal staples (see Figures 1 and 3) which pierce the plastic film and crimp on to the electrical conductors printed on top. The problem with these staples, is that it can be difficult to solder to them without melting the plastic and hurting the electrical contact within. If you do solder to the contacts, be sure to have the wire you are soldering to pre-tinned, and only heat long enough to melt the solder and make a good joint. Other options for connecting an FSR include screw terminals or SIP sockets. These have the advantage of making sensor replacement quick and easy at the expense of a less robust connection. Most FSRs also use .1″ spacing, so you can easily press them into a protoboard as well.
A good mechanical connection to your FSR is very important, and often over-looked. The mounting system should do two things – provide a stiff, flat surface for contact, and limit the amount of shear force on the sensor. One of the problems with FSRs is that they can give different readings under the same force conditions, depending upon how they are pushed. This is because the sensor averages out the pressure over the whole surface, and a poorly distributed load will apply a lot of pressure in one place, but not another. To counter this effect, always use a relatively stiff material on either side of the FSR, with a thin cushion between, in order distribute the load evenly.
Figure 9 – Delaminated FSR.
Another drawback FSRs have, is that they easily split in half under sideways (shear) forces. They also only work under compression, and tear under tension. This is why you want to make sure that pressure can only go straight down onto the FSR (a delaminated FSR can be seen above). Depending upon how hard you expect to press on the sensor, this may not be an issue. But, if possible, have the actuator that presses against the sensor constrained to move only up and down. This can be accomplished with a small groove, or a cantilevered hinge, as shown below.
Figure 10 – FSR attachment examples.
Which FSR should I use?
The main two manufacturers of FSR sensors are Interlink and FlexiForce. Interlink FSRs come in a wide range of shapes, are relatively inexpensive, and fairly robust. On the downside, they vary greatly between units, have a lot of resistance drift over time, and are a bit slow to respond. The FlexiForce sensors, on the other hand, tend to be more accurate and repeatable, but are more expensive and fragile. In applications where exact forces don’t need to be measured, the Interlink products will probably serve you best. For more repeatable results, the FlexiForce sensors will be better (as they rely on the piezoresistive effect). But, if you really need consistent results, an FSR isn’t the best choice, and a load-cell or strain guage should be used.
For more information, check out the Interlink FSR Guide.