Detailed description of the invention
Referring to the drawings, the mode being just used for implementing the present invention illustrates.
First, composition and the action of the acceleration switch in the most Japanese artistic conception document illustrates.
Fig. 5 is that the figure observed from upper surface side of the observable composition in the space within mass body with acceleration switch 100 opposite electrode, non-directional in Japan's artistic conception logs in No. 1310053 is described.101 is the periphery (housing) of acceleration switch 100, and 102 to 105 is the beam supporting weight 106.107 is opposite electrode.But, because up to four, beam is complex, therefore the Fig. 2 in the case of replacing Fig. 5 and using beam to be is described in detail.
Fig. 2 is the figure observed from upper surface side of the acceleration switch 200 in the case of beam is as described above.But there is the layer (first substrate) as lid the most above it and thereunder there is supporting course (the 3rd substrate).Fig. 3 is the sectional view cut off along the A-A ' face shown in Fig. 2, and the layer omitted in Fig. 2 is also included.It addition, Fig. 2 is equivalent to the figure cut off along the B-B ' face of Fig. 3.
As shown in Figure 2 to 3, acceleration switch 200 stacking from top uses the first substrate (cap rock) 205 of the insulant such as glass, uses second substrate 201(202,203,204 of monocrystal silicon etc. to be also included) and use the 3rd substrate (supporting course) 206 of the insulant such as glass to constitute.The monocrystal silicon of second substrate conducts to obtain and uses such as low-resistance silicon.It addition, through electrode 207 and 208 is formed by metals such as embedment gold, and become for by acceleration switch and the outside contact linked up.It addition, first substrate and the 3rd substrate are engaged with second substrate by methods such as anodic bonding.
Here then switch entirety and move in the direction of the arrow, owing to the mass body supported with beam does not moves, so the opposite electrode 204 being in mass body inner space contacts with mass body 203 when applying acceleration the most in the direction of the arrow.Additionally, omit beam and the periphery of mass body periphery in the diagram for ease of understanding.Accordingly, conduct and be connected with external connector by mass body 203, beam 202, substrate periphery portion 201, through electrode 207 from opposite electrode 204.Additionally opposite electrode 204 is connected with external connector by another through electrode 208.
Accordingly, if this acceleration switch is used as the switch of electronic device, the most only just make when detecting vibration the electronic devices such as acceleration transducer start, then make when not detecting vibration electronic device be off such that it is able to limit the unhelpful use etc. of battery.
By simulation, if silicon single crystal being used as the material of tectosome, calculating the displacement when the Z-direction of weight and X-Y plane direction are applied with the acceleration of 1G simultaneously, then can get following result.
The result of the acceleration being applied with 1G at Z axis and X-direction is as follows.
The thickness of beam is the displacement of weight when 20 microns.(unit micron)
The maximum displacement of X-direction: 23.67;
The maximum displacement of Y-direction: 3.42;
The thickness of beam is the displacement of weight when 40 microns.(unit micron)
The maximum displacement of X-direction: 12.17;
The maximum displacement of Y-direction: 1.87.
It addition, the result of the acceleration being similarly applied with 1G at Z axis and Y direction is as follows.
The thickness of beam is the displacement of weight when 20 microns.(unit micron)
The maximum displacement of Y-direction: 23.70;
The maximum displacement of X-direction: 0.65;
The thickness of beam is the displacement of weight when 40 microns.(unit micron)
The maximum displacement of Y-direction: 12.09;
The maximum displacement of X-direction: 0.34.
According to the above results,
1) maximum displacement of the X-direction when X-direction applies acceleration is about 0.12% with the difference of the maximum displacement of the Y direction when Y direction applies acceleration, demonstrates roughly equal displacement in the x, y direction.
2) when X-direction applies acceleration, as about the 14% and about 15% of the maximum displacement that amount is X-direction being tilted towards Y-direction displacement.
3) when Y direction applies acceleration, as about the 2.7% and about 2.8% of the maximum displacement that amount is Y-direction being tilted towards X-direction displacement.
According to above-mentioned 1) to 3) sensitivity that understands X-direction and Y-direction is roughly equal, and it is being applied with weight submissively displacement, the most less displacement on the direction of acceleration respectively.Thus, as being carried out the activity of perception such as people by vibration to carry out the connection (ON) of mobile electronic device, to disconnect (OFF) such usage, it may be said that have the most isotropic sensitivity, do not worry making due to biased sensitivity sensitivity to depend on the direction of acceleration and relatively low or do not have sensitivity etc..Can be enough to use accordingly, as acceleration switch.
Additionally, this acceleration switch is characterized by single acceleration sensitivity.Thus, by goods different for sensitivity are pressed series of products, user is obtained in that the acceleration switch of required sensitivity.
For example, it is contemplated that this acceleration switch to be used for the situation of pedometer.When the action that people stands, the acceleration of about 1.5G acts on vertical direction, so the acceleration switch of acceleration sensitivity 1.5G is installed on pedometer.Accordingly, in the moment that people stands, this acceleration switch is connected, it is possible to detect standing state.As long as detecting, by certain means, the state that this acceleration switch is connected, and acceleration transducer is made to start according to this signal, it becomes possible to start the counting of step number behind.Accordingly, acceleration transducer is made to stop when people does not carries out walking action, it is possible to prevent battery consumption to be capable of long-life pedometer.
It addition, the most commercially there is the wireless mouse of personal computer, but battery consumption needs frequently to change soon.In order to prevent this point, built-in acceleration sensor occurs, has made the power on wireless mouse with beginning action of mouse when mobile mouse.But, acceleration transducer price is high, and flows through consumption electric current because of action all the time, so present situation is the failure to fully meet the requirement of user.If instead of acceleration transducer, this acceleration switch is used for mouse, then disclosure satisfy that the requirement of user.It practice, in order to detect that people moves the action of mouse, as long as the acceleration switch of about 0.5G is installed on mouse.As long as by using acceleration during mouse to make this acceleration switch connect, and only started the action of wireless loop by this detection signal, it becomes possible to cut down consumption electric current during non-usage.
In response to above such user's requirement, as long as such as getting out, with 0.2G step-length, the product that sensitivity is different from 0.1G to 2G, just can be readily available the element of the threshold value with required acceleration sensitivity, and need not be made oneself specially acceleration transducer is adjusted to the circuit of required acceleration sensitivity.In order to achieve this it is needed the dimensional parameters in change switch element.Thus, investigate how dimensional parameters has influence on acceleration sensitivity.
As this acceleration sensitivity being brought the component parameters of impact, there are electrode gap, the size of beam and the weight of weight.Below, electrode gap, the size of beam and the weight of weight are illustrated respectively with the relation of sensitivity.
First, the relation of electrode gap and sensitivity is described.The sensitivity of switch is driven by the displacement of weight and the distance of electrode gap.In this switch, the distance inside central electrode and weight becomes electrode gap.The acceleration of such as 1G puts on switch, during weight displacement 10 μm, if electrode gap is 10 μm, two contact electrode, because of switch be turned on thus can perceptual signal.Now it may be said that switch has the sensitivity of 1G.If now electrode gap is 5 μm, with the acceleration 0.5G of the displacement of half, i.e. half and two contact electrode, it is possible to say the sensitivity with 0.5G.So electrode gap and sensitivity is in proportional relationship.
The etching of electrode gap needs to carry out size etching little, high-precision.Generally speaking the MEMS at silicon uses a kind of rich of dry ecthing to execute process (Bosch in processing
Process), but restricted.This is because the phenomenon such as side etching and scallop phenomenon (scalloping) and be affected in dimensional accuracy., in electrode gap, there is deviation in its result.The deviation of electrode gap becomes the deviation of sensitivity as described above.It is believed that also want about 3 μm in order to avoid this impact is minimum, if considering repeatability and the precision manufactured, it is the electrode gap of reality about 5~10 μm.
Here, it is known that if wanting to change the sensitivity of this switch, as long as changing this electrode gap.If setting electrode gap 10 μm, processing dimension produces the deviation of 0.5 μm because of phenomenons such as side etching and scallop phenomenons at each product, then 1G sensitivity is produced to the deviation of 5%.When setting this electrode gap as 5 μm, the deviation of 0.5 μm processing dimension becomes the sensitivity variations of 10% for 0.5G.The method of the sensitivity i.e. changing switch by change electrode gap is feasible.It is, however, possible to become the sensitivity for desired value and the different product of departure.
The size of beam illustrates with sensitivity with that.
In this structure switched, sensitivity is inversely proportional to the spring constant of beam part.Spring constant k formula (1) represents.
[formula 1]
Here, E: Young's modulus;
The thickness of w: beam;
The width of t: beam;
The length of L: beam;
That is, can conclude that the following fact according to formula (1).
1.
The thickness of beam is as a proportion with sensitivity.If beam is thickening, spring constant also becomes big, and sensitivity is also deteriorated.
2.
The width of beam becomes three ratios with sensitivity.If the width of beam becomes big, spring constant also becomes big, and sensitivity is also deteriorated.
More than according to, if the width changing beam just can maximally change sensitivity.Thus, by each switch being changed the width of this beam, so that it may so that the product that the acceleration sensitivity of switch is different.But, if this is on the contrary, the departure of beam width can make sensitivity variations the biggest.If as it was previously stated, the width of beam because of phenomenons such as side etching and scallop phenomenons difference, machining accuracy produce deviation.Its result, sensitivity also produces deviation with cube.
If it addition, such as use SOI wafer, can remain untouched the thickness as beam by this active layer, so the thickness of beam does not has big restriction dimensionally, additionally can also keep the precision of size.Accordingly, if using the wafer of active layer variable thickness sample, and each switch is changed the thickness of beam, so that it may so that the product that the acceleration sensitivity of switch is different.But, whenever the goods making sensitivity different are accomplished by obtaining the silicon wafer of the thickness of the active layer 209 changing SOI wafer.
Then, the weight with regard to weight illustrates with sensitivity.
Determine weight weight be the volume of mass body 203, here it is the thickness of silicon wafer is with the surface area in the XY face of the mass body 203 shown in Fig. 4 long-pending.
Firstly, since the thickness of silicon wafer is intactly as the thickness of weight, so displacement is more and become high sensitivity if silicon self is the thickest.But there is the restriction of dry etch process here.A kind of rich process of executing applying dry ecthing in the etching of silicon can etch thicker silicon at high speed.But even with rich process of executing in the case of the thickest, the process time of silicon etching is the most long, cost increase, additionally the size of weight, the control of shape can become difficulty.Thus be preferably used about 350 μm.
Then, the surface area in the XY face of mass body 203 identified below.Switch periphery must guarantee as the region that anodic bonding etc. engages.This region is the broadest, and anodic bonding is the most stably reached.If by engage location for paste be set to square for 1mm about 20%, chip peripheral part will be ensured, then about the outer circumference diameter size of mass body 203 the most about 800 μm in the region about totally 200 μm of unilateral 100 μm, chip both sides.
Here, the weight of weight is inversely proportional with acceleration sensitivity.That is, if weight is heavier, acceleration sensitivity is preferable.If that is, the surface area changed in the XY face of mass body 203 or the thickness of silicon wafer, it becomes possible to change acceleration sensitivity.But, whenever the goods making sensitivity different are accomplished by obtaining the wafer of the thickness changing silicon.
On the other hand, the method changing the surface area in XY face is effective.Be because while that the processing of decision table area uses the rich dry ecthing executing process, even if but surface area change in its XY face and be also not changed on etching condition.Thus, even if surface area change but the deviation that produces because of phenomenons such as side etching and scallop phenomenons is also constant.And, even if as it was previously stated, each product produces the deviation of 0.5 μm processing dimension, the impact for the peripheral radius of mass body 203 is the most little.If the such as mass body of radius 400 μm is then the sensitivity variations of 0.25%.If additionally the mass body of radius 200 μm is only then the sensitivity variations of 0.5%, even if making radius be reduced to half, sensitivity decrease is to 1/4, within its deviation also converges on 0.25%.The i.e. goods of sensitivity 2G and the goods of 0.5G are restrained with the deviation within 0.25%.
And, it also is able to obtain same above effect by the inner side cutting tip D of etching quality body 203 as shown in Figure 1.It is i.e. inversely proportional with the weight of mass body 203 due to acceleration sensitivity, it is possible to etching inboard portion D changes the weight of mass body 203 to change sensitivity.Specifically, if cutting tip D inside Jia great ing, the weight of mass body 203, so acceleration sensitivity is deteriorated.This etching can be carried out when etching forms mass body 203 simultaneously, does not has operation increase and productivity ratio to deteriorate this advantage so there is also.It addition, the method is compared with the method for the surface area of above-mentioned change mass body 203, the peripheral radius of mass body 203 does not changes, and can suppress this advantage of sensitivity variations further so there is also.Although it addition, be illustrated with the inboard portion of mass body 203 as cutting tip, but either Outboard Sections etches the mid portion in outside and inner side, its effect is not changed in.
Though additionally, describe to illustrate in the present invention as preparing multiple with the step-length of 0.2G from 0.1G to 2G, this numerical range be not limited fixed.
More than according to, can easily be provided the acceleration switch with the little multiple types of the threshold value of single acceleration sensitivity, low consumption electric current and deviation by the present invention.