Multiple touch electrodes connected to any SNSnK can be
used, for example, to create control surfaces on both sides of
an object.
It is important to limit the amount of stray capacitance on the
SNS terminals, for example by minimizing trace lengths and
widths to allow for higher gain without requiring higher values
of Cs. Under heavy delta-Cx loading of one key, cross
coupling to another key’s trace can cause the other key to
trigger. Therefore, electrode traces from adjacent keys
should not be run close to each other over long runs in order
to minimize cross-coupling if large values of delta-Cx are
expected, for example when an electrode is directly touched.
This is not a problem when the electrodes are working
through a plastic panel with normal touch sensitivity.
1.4 Sensitivity
1.4.1 Introduction
Sensitivity can be altered to suit various applications and
situations on a channel-by-channel basis. The easiest and
most direct way to impact sensitivity is to alter the value of
each Cs; more Cs yields higher sensitivity. Each channel has
its own Cs value and can therefore be independently
adjusted.
1.4.2 Alternative Ways to Increase
Sensitivity
Sensitivity can also be increased by using bigger
electrode areas, reducing panel thickness, or using
a panel material with a higher dielectric constant.
1.4.3 Decreasing Sensitivity
In some cases the circuit may be too sensitive.
Gain can be lowered further by a number of
strategies: a) making the electrode smaller, b)
making the electrode into a sparse mesh using a
high space-to-conductor ratio, or c) by decreasing
the Cs capacitors.
1.4.4 Key Balance
A number of factors can cause sensitivity
imbalances. Notably, SNS wiring to electrodes can
have differing stray amounts of capacitance to
ground. Increasing load capacitance will cause a
decrease in gain. Key size differences, and
proximity to other metal surfaces can also impact
gain.
The four keys may thus require ‘balancing’ to
achieve similar sensitivity levels. This can be best
accomplished by trimming the values of the four
Cs capacitors to achieve equilibrium. The four Rs
resistors have no effect on sensitivity and should
not be altered. Load capacitances can also be added to
overly sensitive channels to ground, to reduce their gains.
These should be in the order of a few picofarads.
2 QT240 Specifics
2.1 Signal Processing
2.1.1 Introduction
These devices process all signals using 16 bit math , using a
number of algorithms pioneered by Quantum. These
algorithms are specifically designed to provide for high
survivability in the face of adverse environmental changes.
2.1.2 Drift Compensation
Signal drift can occur because of changes in Cx, Cs, and
Vdd over time. If a low grade Cs capacitor is chosen, the
signal can drift greatly with temperature. If keys are subject
to extremes of temperature or humidity, the signal can also
drift. It is crucial that drift be compensated, else false
detections, nondetections, and sensitivity shifts will follow.
Drift compensation (Figure 2.1) is a method that makes the
reference level track the raw signal at a slow rate, only while
no detection is in effect. The rate of reference
adjustment must be performed slowly else
legitimate detections can also be ignored. The IC
drift compensates each channel independently
using a slew-rate limited change to the reference
level; the threshold and hysteresis values are
slaved to this reference.
Once an object is sensed, the drift compensation
mechanism ceases since the signal is legitimately
high, and therefore should not cause the reference
level to change.
lQ
4 QT240R R1.11/1006
Figure 1.2 Fast, Spread-spectrum Circuit
S1
S3
S2
VDD
KEY 2
KEY 3
OUT4
OUT1
OUT2
OUT3
10nF
CS1
RS1
2.2K
KEY 1
C1
22nF
VDD
10nF
CS3
10nF
CS4
RS3
2.2K
KEY 4
VDD
VDD
1M
R2
22K
RSNS1
360K
R5
22K
RSNS4
RS4
2.2K
OPT1
10nF
CS2
RS2
2.2K
1M
R1
R6
180K
VDD
62K R4
1M
R3
10 second
timeout shown
22K
RSNS2
SPEED
OPT
OPT2
22K
RSNS3
QT240_ISS
Figure 2.1 Drift Compensation
Threshold
Signal
Hysteresis
Reference
Output
The signal drift compensation is 'asymmetric'; the reference
level drift-compensates in one direction faster than it does in
the other. Specifically, it compensates faster for decreasing
signals than for increasing signals. Increasing signals should
not be compensated for quickly, since an approaching finger
could be compensated for partially or entirely before even
approaching the sense electrode. However, an obstruction
over the sense pad, for which the sensor has already made
full allowance, could suddenly be removed leaving the
sensor with an artificially elevated reference level and thus
become insensitive to touch. In this latter case, the sensor
will compensate for the object's removal very quickly, usually
in only a few seconds.
With large values of Cs and small values of Cx, drift
compensation will appear to operate more slowly than with
the converse.
Drift Compensation in Slow Mode: Drift compensation
rates in Slow mode are preserved if there is no Sync signal,
and the rates are derived from the ~90ms Sleep interval.
However if there is a Sync signal, then drift compensation
rates are derived from an assumption that the Sync
periodicity is ~18ms (which is corresponds to 55.5Hz). Thus,
drift compensation timings in Sync mode are correct for an
~18ms Sync period but different (slower or faster) for other
Sync periods. For example a Sync period of 36ms would
halve the expected drift compensation rates.
2.1.3 Threshold Level
The internal threshold level is fixed at 12 counts for all four
channels. The hysteresis is fixed at 2 counts (17%).
2.1.4 Max On-Duration
If a sufficiently large object contacts a key for a prolonged
duration, the signal will trigger a detection output preventing
further normal operation. To cure such ‘stuck key’ conditions ,
the sensor includes a timer on each channel to monitor
detection duration. If a detection exceeds the maximum timer
setting, the timer causes the sensor to perform a full
recalibration (if not set for infinite). This is known as the Max
On-Duration feature.
After the Max On-Duration interval, the sensor channel will
once again function normally, even if partially or fully
obstructed, to the best of its ability, given electrode
conditions. There are three timeout durations available via
strap option: 10s, 60s, and infinite (Table 2.2).
Max On-Duration works independently per channel; a
timeout on one channel has no effect on another channel.
Note also that the timings in Table 2.2 are dependent on the
oscillator frequency in Fast mode. Doubling the
recommended frequency will halve the timeouts. This is not
true in Slow mode.
Infinite timeout is useful in applications where a prolonged
detection can occur and where the output must reflect the
detection no matter how long. In infinite timeout mode, the
designer should take care to ensure that drift in Cs, Cx, and
Vdd do not cause the device to ‘stick on’ inadvertently even
when the target object is removed from the sense field.
Timeouts are approximate and can vary substantially over
Vdd and temperature, and should not be relied upon for
critical functions. Timeouts are also dependent on operating
frequency in Fast mode.
Max On-Duration in Slow Mode: When Sync mode is used
in Slow mode, the Max On-Duration timings are derived from
the Sync period. The device assumes the Sync periodicity is
18ms (midway between 50Hz and 60Hz sync timings). Thus,
Max On-Duration timings in Sync mode are correct for an
18ms Sync period but different (shorter or longer) for other
Sync periods. For example a Sync period of 36ms would
double all expected Max On-Duration timings.
2.1.5 Detection Integrator
It is desirable to suppress false detections due to electrical
noise or from quick brushes with an object. To this end,
these devices incorporate a per-key ‘Detection Integrator’
counter that increments with each signal detection exceeding
the signal threshold (Figure 2.1) until a limit count is reached,
after which an Out pin becomes active. If a ‘no detect’ is
sensed even once prior to the limit, this counter is reset to
zero and no detect output is generated. The required limit
count is 6.
The Detection Integrator can also be viewed as a
'consensus' vote requiring a detection in successive samples
to trigger an active output.
In Slow mode, the detect integrator forces the device to
operate faster to confirm a detection. The six successive
acquisitions required to affirm a detection are done without
benefit of a low power sleep mode between bursts.
2.1.6 Forced Sensor Recalibration
Pin 13 is a Reset pin, active-low, which in cases where
power is clean can be simply tied to Vdd. On power-up, the
device will automatically recalibrate all channels of sensing.
Pin 13 can also be controlled by logic or a microcontroller to
force the chip to recalibrate, by toggling it low for 10µs or
more, then raising it high again.
2.1.7 Fast Positive Recalibration
If the sensed capacitance becomes lower by 5 counts than
the reference level for 2 seconds, the sensor will consider
this to be an error condition and will force a recalibration on
the affected channel.
2.2 Options
These devices are designed for maximum flexibility and can
accommodate most popular sensing requirements via option
pins.
The option pins are read on power-up and about once every
10 seconds while the device is not detecting touch on any
channel. Options are set using high value resistors
connected to certain SNS pins, to either Vdd or Vss. These
options are read 25 times over 250µs to ensure that they are
not influenced by noise pulses. All 25 samples must agree.
However, large values of Cx on the SNS wires can load
down the pins to the point where the 1M pull-up resistors
cannot pull high fast enough, and the pins are read
erroneously as a result. Cx should be below 50pF to prevent
errors; this value can be read with a conventional
capacitance meter with the QT240 removed.
The option setting resistors are mandatory and cannot be
deleted. They must be strapped to either Vdd or Vss.
lQ
5 QT240R R1.11/1006
Speed option (Strap S1): This jumper selects whether the
device acts in a slower, low power mode with a response
time of approximately 100ms, or in a Fast mode with a
response time of 40ms typical. Fast mode consumes
substantially more power than the Slow mode, but also
enables the use of spread-spectrum detection. Only Slow
mode supports the use of external Sync (Section 2.3).
Response time can also be modified by changing the
oscillator frequency (Section 3.3).
Recalibration / toggle select (S2, S3): See Table 2.2.
There are three recalibration timing options (‘Max
On-Duration’, see Section 2.1.4) and one toggle mode
option. The recalibration options control how long it takes for
a continuous detection to trigger a recalibration on a key.
When such an event occurs, only the ‘stuck’ key is
recalibrated. S2 / S3 should be connected as shown in
Table 2.2 to achieve the desired Max On-Duration of either
10s, 60s, or infinite.
Toggle option: One option is toggle mode, which allows all
four keys to behave with flip-flop action. In this mode, each
key’s corresponding OUT pin will toggle High / Low with
successive touches on the key. The underlying Max
On-Duration is 10s in this mode. If a timeout occurs in
Toggle mode, the toggle state is not affected. Toggle state
flips only when the corresponding Out pin goes High.
This is useful for controlling power loads, for example in
kitchen appliances, power tools, light switches, etc . or
wherever a ‘touch-on / touch-off’ effect is required.
2.3 Synchronization
Sync capability is only present in Slow mode (Section 2.2). If
SYNC is not desired, SYNC/SS should be connected to Vss.
Adjacent capacitive sensors that operate independently can
cross-interfere with each other in ways that will create
sensitivity shifts and spurious detections. Since Quantum’s
QT devices operate in burst mode, the opportunity exists to
solve this problem by time-sequencing sensing channels so
that physically adjacent keys do not sense at the same time.
Within the QT240 the four channels operate synchronously,
so it is not possible for these channels to cross interfere.
However, two or more adjacent chips will cross-interfere if
they are not synchronized to each other. The same is true of
the effects of unsynchronized external noise sources.
External noise sources can also be heavily suppressed by
synchronizing the QT240 to the noise source period. External
noise creates an ‘aliasing’ or ‘beat’ frequency effect between
the sampling rate of the QT part and the external noise
frequency. This shows up as a random noise component on
the internal signals, which in turn can lead to false activation.
Mains frequency is one common source of interference. A
simple AC zero-crossing detector feeding the SYNC pin is
enough to suppress this kind of periodic noise. Multiple
devices tied to SYNC can be synchronized to the mains
frequency in this fashion.
If two physically adjacent devices are to be synchronized to
each other, they should be connected via the SYNC pin to a
clock source that is slower than the burst rate of either
device. For example, a 50Hz clock can synchronize two
QT240’s running with burst spacings of up to 10ms each.
The two QT240’s should be synchronized on opposite
phases of the clock source, i.e. the clock source should feed
one part and its inverted phase, the other part.
A sync pulse on SYNC/SS in Slow mode acts to break the
QT240 out of its sleep state between bursts, and to do
another burst. The device will then go back to sleep again
and await a new SYNC pulse. If a Sync pulse does not arrive
within about 90ms, it will wake again and run normally.
External sync pulses can be used to accelerate response
time (at the expense of power) in Slow mode. Sync pulses
running at 25Hz for example will improve response time by a
factor of 2. Sync cannot be used to slow down the device.
Sync Mode Effects on Timings: In the absence of a Sync
signal, the Max On-Duration timings and drift compensation
rates in Slow mode are nominally correct. It should be
understood that the Max On-Duration timings and drift
compensation rates are slaved to the burst interval in Slow
mode, and that changing the burst interval will have direct
effects on these parameters.
Since the most common use of Sync is to synchronize the
device to Mains frequency (50 or 60Hz) the device makes an
assumption that the presence of a Sync signal is at 55Hz,
and the timings are made to be correct at this frequency.
Should the Sync pulses vary from this frequency, the Max
On-Duration timings and drift compensation rates will vary
proportionately. Thus, if the Sync pulses are 25Hz, the
10-second Max On-Duration timing will become 10*55/25 =
22 seconds nominal. Only at Sync=55Hz will the 10s timeout
be 10s (the same as if there were no Sync signal, or the
device was in Fast mode).
3 Circuit Guidelines
3.1 Cs Sample Capacitor
Charge sampler capacitors Cs can be virtually any plastic
film or low to medium-K ceramic capacitor. The ‘normal’ Cs
range is 4.7nF to 47nF depending on the sensitivity required;
larger values of Cs require higher stability to ensure reliable
sensing. Acceptable capacitor types for most uses include
plastic film (especially PPS film and polypropylene film) and
X7R ceramic. Lower grades than X7R are not advised;
higher-K ceramics have nonlinear dielectrics which induce
instabilities.
lQ
6 QT240R R1.11/1006
VddSlow / Sync
VssFast / Spread Spectrum
Table 2.1 S1 Speed / Sync Options - SNS1K Pin 4
Timings assume 100 kHz operation
infiniteVssVss
DC Out
10sVddVdd
Toggle
60sVssVdd
DC Out
10sVddVss
DC Out
Max On-Duration
S3
SNS4K pin 18
S2
SNS3K pin 20
Table 2.2 OPT Options

QT240-ISSG

Mfr. #:
Manufacturer:
Microchip Technology / Atmel
Description:
Interface - Specialized 3.9 - 5.5V 4 Chan Touch Sensor
Lifecycle:
New from this manufacturer.
Delivery:
DHL FedEx Ups TNT EMS
Payment:
T/T Paypal Visa MoneyGram Western Union

Products related to this Datasheet