Contact Us

The resolution for the encoder is the number of digital pulses produced
by the encoder over a unit of distance, or the instance of a single pulse as a unit of measure. The line count of an encoder is number of cycles of line and space pairs, over a unit of distance.

If two count channels, A and B, view a cycle (360 electrical degrees), and the two channels are separated by 90 electrical degrees, each channel now produces two counts, and the summation of the two channels will produce four counts per cycle. This is referred to as “quadrature”, or edge detection, and produces 4x more counts than the initial line count code pattern. If the disc or linear scale has 1000 cycles, the resolution of the encoder utilizing quadrature will be 4000 pulses.

Additional resolution maybe achieved via electronic cycle interpolation. A phase resistor bridge may be used to determine additional positions within a cycle that interpolates the two analogue sine waves electronically. ASIC’s or circuits that incorporate DAC’s are also used to determine position by calculating an arc tangent function from the phase position of the sine and cosine wave inside a cycle. This interpolation factor may be as high as 512x before quadrature and 2048x after quadrature (11 bit interpolation). More typical interpolation factors used in optical encoders are 2x, 4x, 10x. If the disc or linear scale has 1000 cycles, and 25x interpolation is employed, the number of cycles increases to 25,000. The resolution of the encoder utilizing quadrature will be now be 100,000 pulses.

A low cost “kit encoder” that mounts to the user’s shaft requires certain mounting limits to be met to operate accurately. Shaft run out, and wobble affect the gap between the disc and reticle and the electrical phase relationship between the count channels. This needs to be controlled to use a kit encoder. Typical specification for lower resolution encoders allow shaft end play up to +/- 002”, and a run-out of .001 TIR

A better option would be flex mount encoders that contain their own bearing and flexure. The gap and disc run-out requirement of the encoder are set internally in the encoder, and the flexure isolates the users shaft misalignment. Position accuracy is still a concern, but the flexure will allow a shaft endplay of +/-.010, and radial run-out of .005 TIR

In most applications it is desirable to determine a “home position” when moving in either direction within the final resolution of the encoder. When using quadrature 4x detection, the final resolution of the encoder is ¼ of a cycle for one count and a full cycle zero reference would represent 4 counts. If you initialize the “home position” when moving in one direction, and then wanted to re-initialize the “home position” when moving in the other direction, the true position would be off 4 counts.

The reason the disc is not normally made with the zero reference already ¼ cycle wide is that this disc may be used in encoders that will use cycle interpolation. Cycle interpolation circuits use digital logic that takes an aligned full cycle zero reference and produces a zero reference count that is 1 interpolated count wide, and can be used bi-directional.

The zero reference pulse for the encoder can be made 1 cycle, ½ cycle or ¼ cycle wide on most encoder models.

Case grounds, cable shields, or drain wires, are for different purposes to help minimize the effects of electrical noise. Case grounds are sometimes available for insuring that a metal encoder has a good ground connection between the earth ground and the encoder. Shields and drain wires are available for shielded cables and are not attached at the encoder end but may be required to be connected to ground at the opposite end.

The maximum frequency response of the encoder is the maximum velocity or rate of rotation that the electronics inside the encoder can operate given a particular line count and RPM. This rate is not based on 4x Quadrature. It is based on the line count. There may also be maximum rotation limits on the mechanical structure, which should also be reviewed, and specified as maximum
rpm or ips.

For example: an electronic circuit rated for 100 KHz frequency response would allow a 2500 line disc to rotate at 2400 RPM. A 9000- line disc may only rotate at 667 RPM to meet the same 100 KHz requirement.

(frequency response x 60) / line count = RPM
RPM / 60 x line count = frequency response

Accuracy error is the difference between the position that the instrument indicates that you are at, and the actual true position.

When stating an accuracy specification on a rotary encoder, the most common specification is +/- arc seconds instrument error. This is the error due to disc master and printing error, disc mounting eccentricity error, bearing run out, and non-repeatability bearing error. Because this is an incremental encoder, the error “closes on itself” once per revolution (returns to zero), unless there is some nonrepetitive bearing error. Typical error specified is the error rated “cycle to any other cycle” or without considering quadrature. Disc master error and printing error are typically <+/- 5 arc seconds. Disc centering error remains the largest contributor to instrument error and may account for +/- 3 minutes of error on some kit models and is dependent on disc run-out and disc diameter. Typical error with electrically centered disc encoders is >+/- 20 arc seconds. This all makes up the total instrument error, which is typically rated +/-30, +/-45 and +/- 60 arc seconds on standard models with bearings. These values are regardless of line count, and are the errors associated with making the components that make up the encoder.

Another accuracy error is quadrature error, which is the 90degree phase difference between the two count channels allowed within one cycle. Typical quadrature error is allowed to be 90 +/- 22 ½ degrees. This means that the position error can be off ¼ of a count. For example, one count of a 2500 line encoder equals 130 arc seconds and can have an instrument error of +/- 30 arc seconds, but have an additional +/- 32 arc second error due to quadrature phase error on top of the instrument error. Quadrature error is line count dependent, a 1000 line disc with the same 22 ½ degree phase error would have a +/- 80 arc sec error.

Resolution is the number of counts over a given distance or the numerical value of a single count or “bit” from the encoder. An encoder with a disc that has 5000 lines / rev may be referred to as 5000 line encoder. The actual resolution that the user would see after quadrature would be 20,000 counts / rev. The actual resolution or 1 bit is .018 degrees. (360 / 20,000 = .018 degrees)

It is very possible to have higher resolution than the accuracy of the encoder. If the above encoder has a 10 x cycle interpolator then the resolution increases to 200,000 counts / rev. or .0018 degrees / bit (6.4 arc sec). This model with a +/- 30 arc sec instrument error and a possible +/- 30 arc sec quadrature error could have 10 times more resolution then total encoder error.

Repeatability is normally stated as “bi-directional”, which is the expected dispersion on each side of the mean when approached from either direction. This is not normally stated in encoder specifications because the factors that effect repeatability in a rotary encoder create less error than the normal digitizing error (quantizing error) of +/- 1 bit.

In a non-contact encoder, bi-directional repeatability is limited to the amount of hysteresis designed into the circuit, or non-repetitive runout of the bearings.

Case grounds, shields and drain wires are used to reduce the likelihood of external noise corrupting the encoder signals causing miscounts and consequent control errors.

British Encoder units generally have the screen and drain wire, present in the cable, connected to the encoder case and 0V line internally. This provides excellent protection against EMI, crosstalk and noise.

Care needs to be taken with avoiding ground loops however, and we recommend not connecting the screen at the receiving equipment.

Most manufacturers of equipment that is to be connected to an encoder will specify the type of output required. They may have the option of using any of the common output types, in which case it is recommended to specify our Universal Line driver circuit type “HV”. These outputs offer the facility to use differential signals to reduce the electrical noise picked up by the cable and allow the use of long cable runs. However, older equipment may need Open Collector and use of our “HV” outputs will function in these cases as well. For more information please consult our Technical Bulletin TB109.

Most controller manufacturers will specify either a single ended TTL signal (A, B, Z) or complementary pairs of signals (A, /A, B, /B, Z, /Z). These may be specified as output from a complementary TTL or differential line driver devices. The user should always try to match up the line driver with the line receiver family although most are specified as RS 422 compatible. The sink and source current requirement may indicate which device to specify. Typical line drivers used are, 3487 and 26LS31, which are 5 Vdc devices. Both are available in CMOS for extended temperature requirements. High voltage line driver options are available and are specified as +5 Vdc to +24 Vdc.