Post by thread:Re: [EE]: PC mouse optical detectors.

Hi. I just scraped an old mouse (a Logitech wheel mouse) to see what optical components they used. I noticed that the wheel had one LED (IR?) and one 3 pin "receiver" mounted. Now, the wheel just have a row of spokes, but someway, the unit must not only detect spoke/nospoke, but also the direction the wheel is turning. Anyone know how these receivers work ? Maybe I'll de-solder a pair and setup a testbed... No marking on the components, of course... Jan-Erik. -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

At 11:49 29/04/2003 +0200, Jan-erik Soderholm (QAC) wrote: >Hi. >I just scraped an old mouse (a Logitech wheel mouse) >to see what optical components they used. >I noticed that the wheel had one LED (IR?) and one >3 pin "receiver" mounted. Now, the wheel just have a row >of spokes, but someway, the unit must not only detect >spoke/nospoke, but also the direction the wheel is turning. >Anyone know how these receivers work ? There are actually two sensors on each of those receivers. These will receive the same signal but with a phase shift. e.g.: "Forward" movement: sensor 1 - 1111000011110000111100001111 sensor 2 - 0011110000111100001111000011 "Backwards" movement: sensor 1 - 0011110000111100001111000011 sensor 2 - 1111000011110000111100001111 By looking at the phase shift you can find out the direction the wheel is turning. G. -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

> -----Original Message----- > From: Jason Dini [SMTP:spam_OUTjason.diniTakeThisOuTORCHID-SA.CO.ZA] > Sent: Tuesday, April 29, 2003 11:03 AM > To: .....PICLISTKILLspam@spam@MITVMA.MIT.EDU > Subject: Re: [EE]: PC mouse optical detectors. > > I've often wondered about this myself. If you say that the receiver has 3 > pins, it could just be that they have 2 photo-diodes in one receiver > package > and simply see which gets ir light first. > > Jason > Correct. They are more likely to be a pair of photo transistors, but whatever they are there is two of them. Regards Mike ======================================================================= This e-mail is intended for the person it is addressed to only. The information contained in it may be confidential and/or protected by law. If you are not the intended recipient of this message, you must not make any use of this information, or copy or show it to any person. Please contact us immediately to tell us that you have received this e-mail, and return the original to us. Any use, forwarding, printing or copying of this message is strictly prohibited. No part of this message can be considered a request for goods or services. ======================================================================= Any questions about Bookham's E-Mail service should be directed to postmasterKILLspambookham.com. -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

The phase detection is explained here very clearly (page 3) http://computer.howstuffworks.com/mouse2.htm I understand that PICs have been used as the uC If there's anything that gets thrown away in big numbers it's mice. They're great for microswitches, leads, IR and other components -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

OK. I can understand that, and if they had 4 pins, I would have been even clearer. But with 3 pins (1 look as power and 1 as gnd), do they code/multiplex the two detectors in 1 signal on 1 pin ? Jan-Erik. Guillermo Rodriguez Garcia wrote : > There are actually two sensors on each of those receivers. > These will receive the same signal but with a phase shift. -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

At 15:56 29/04/2003 +0200, QAC wrote: >OK. I can understand that, and if they had 4 pins, I would >have been even clearer. But with 3 pins (1 look as power and 1 as >gnd), do they code/multiplex the two detectors in 1 signal on 1 pin ? You don't apply "power" to a sensor. An IR sensor is just a photodiode whose response will vary depending on how much light it receives within the IR band of the spectrum. G. -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

That sounds logical of course, *but*... There are three sensors with three pins each. Each has one pin to GND, one pin via a common inductor ("W1" ?) to power, and *ONE* pin each connected to the IC. So, in some way, they talk to IC via one pin for each sensor. And there os one sensor for the wheel and one for each "direction" of movement. So my bet is that this are some kind of sensors with some builtin logic. Well, maybe better to breadboard a sensor and take out the scope... Jan-Erik. Guillermo Rodriguez Garcia : >You don't apply "power" to a sensor. An IR sensor is just >a photodiode whose response will vary depending on how much >light it receives within the IR band of the spectrum. -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

Sorry about that ! If they can't get them together correctly again, you can always send them to me :-) (See my plan now ?) Jan-Erik. Richard Stevens wrote: > Jan-Erik, > In response to your posting, 2 mice in my office have been > spontaneously dismantled by their owners! If this is repeated across the > piclist, you could have a serious impact on productivity internationally! -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

At 16:25 29/04/2003 +0200, QAC wrote: >That sounds logical of course, *but*... > >There are three sensors with three pins each. >Each has one pin to GND, one pin via a common >inductor ("W1" ?) to power, and *ONE* pin each >connected to the IC. So, in some way, they >talk to IC via one pin for each sensor. And there >os one sensor for the wheel and one for each >"direction" of movement. As I said you need two sensors per wheel to determine direction. These two sensors can be physically closer or far apart depending on which mouse you're looking at. From your description I assume that what you have is 2 leds + 2 sensors per wheel (can't see how it can be three sensors). The IR sensors will be biased appropriately and the current flow through them measured (ADC in the microcontroller) to determine the direction and rate of movement. >So my bet is that this are some kind of sensors >with some builtin logic. Don't think so :) G. {Quote hidden}>Well, maybe better to breadboard a sensor and take >out the scope... > >Jan-Erik. > > >Guillermo Rodriguez Garcia : > >You don't apply "power" to a sensor. An IR sensor is just > >a photodiode whose response will vary depending on how much > >light it receives within the IR band of the spectrum. > >-- >http://www.piclist.com hint: The PICList is archived three different >ways. See http://www.piclist.com/#archives for details. -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

At 16:57 29/04/2003 +0200, QAC wrote: >*ONE* (2 pin) LED and *ONE* (3 pin) sensor per wheel. >Jan-Erik. > >Guillermo Rodriguez Garcia wrote: > > From your description I assume that what you have > > is 2 leds + 2 sensors per wheel Then it is as I described at the beginning, one led and *TWO* sensors (mounted in a single receiver) per wheel. G. -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

(Well, I'm going to stop this thread soon...) Yes, but you described a 3 pin sensor with *two* signal pins. I have a 3 pin sensor with : One pin to GND One pin (via a inductor) to power One pin to the IC Jan-Erik. Guillermo Rodriguez Garcia wrote: > Then it is as I described at the beginning, one led and > *TWO* sensors (mounted in a single receiver) per wheel. -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

> OK. I can understand that, and if they had 4 pins, I would > have been even clearer. But with 3 pins (1 look as power and 1 as > gnd), do they code/multiplex the two detectors in 1 signal on 1 pin ? Very unlikely. One pin is either power or ground. One end of both photodectors are tied to this pin. The other two pins are the other ends of each photodetector. I seriously doubt you have both a power and ground rail in the three pins. ***************************************************************** Embed Inc, embedded system specialists in Littleton Massachusetts (978) 742-9014, http://www.embedinc.com -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

Yes, but then I'd expect *two* pins beeing connected to the IC , not ? As I said : one pin to GND (black lead in the cable) one pin to power (read lead in the cable, via an (common) inductor) one pin to the IC And the two first pins from each of the three sensors are tied together (directly, via the PCB). The third pin from each sensor is connected to three different pins on the IC. Jan-Erik. Olin Lathrop wrote: >One pin is either power or ground. One end of both >photodectors are tied to this pin. The other two pins are the other ends >of each photodetector. I seriously doubt you have both a power and ground >rail in the three pins. -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

At 17:04 29/04/2003 +0200, QAC wrote: >(Well, I'm going to stop this thread soon...) > >Yes, but you described a 3 pin sensor with *two* I know. >signal pins. I have a 3 pin sensor with : > >One pin to GND >One pin (via a inductor) to power >One pin to the IC I sincerely doubt this is the case. More likely you may have made a mistake somewhere and you'll actually have one pin to either GND or power, but not both, and 2 signal pins, no matter how they are connected. But just play around with it, you'll find out yourself. G. -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

> -----Original Message----- > From: Jan-erik Söderholm (QAC) [SMTP:.....Jan-erik.SoderholmKILLspam.....PAC.ERICSSON.SE] > Sent: Tuesday, April 29, 2003 4:12 PM > To: EraseMEPICLISTspam_OUTTakeThisOuTMITVMA.MIT.EDU > Subject: Re: [EE]: PC mouse optical detectors. > > Yes, but then I'd expect *two* pins beeing > connected to the IC , not ? > > As I said : > one pin to GND (black lead in the cable) > one pin to power (read lead in the cable, via an (common) inductor) > one pin to the IC > > And the two first pins from each of the three sensors are > tied together (directly, via the PCB). The third pin from > each sensor is connected to three different pins on the IC. > > Jan-Erik. > If this is correct then there must be some decoding logic built into the detectors as you suggested. Do you have a scope to check the third pin with? My guess would be some kind of pulse width encoding: i.e. short pulse = one step forward, long pulse = one step back. Regards Mike ======================================================================= This e-mail is intended for the person it is addressed to only. The information contained in it may be confidential and/or protected by law. If you are not the intended recipient of this message, you must not make any use of this information, or copy or show it to any person. Please contact us immediately to tell us that you have received this e-mail, and return the original to us. Any use, forwarding, printing or copying of this message is strictly prohibited. No part of this message can be considered a request for goods or services. ======================================================================= Any questions about Bookham's E-Mail service should be directed to postmasterspam_OUTbookham.com. -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

Well, I'll take a sensor away and setup a testbed. I just made a last quick check with my ohm-meter, and, yes, pin 1 from each sensor are tied together (and to GND), the same for pin 2 (and to power). Pin 3 goes to the IC. Anyway, I'll be back after some evening testing ! Jan-Erik. Guillermo Rodriguez Garcia wrote: >> I have a 3 pin sensor with : >> >>One pin to GND >>One pin (via a inductor) to power >>One pin to the IC > >I sincerely doubt this is the case. More likely you may have >made a mistake somewhere and you'll actually have one pin to >either GND or power, but not both, and 2 signal pins, -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

>One pin to GND >One pin (via a inductor) to power >One pin to the IC I suspect your sensor may not be two photo diodes, but rather two photoresistors in series across the supply, and the input on the chip is actually an analogue one measuring the voltage of the divider. This way it would be easy for the micro to see the "black stripe" of the spoke in the wheel move past and trigger an interrupt at midpoint, and then figure out which way the bar is moving by determining if the voltage moves up or down. It may not be CDS type sensors, but that was what I had in mind, but I don't think they are fast enough response for this. Anyone remember the dance arm sensors on reel to reel tape drives ???? -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

> -----Original Message----- > From: Alan B. Pearce [SMTP:@spam@A.B.PearceKILLspamRL.AC.UK] > Sent: Tuesday, April 29, 2003 4:34 PM > To: KILLspamPICLISTKILLspamMITVMA.MIT.EDU > Subject: Re: [EE]: PC mouse optical detectors. > > >One pin to GND > >One pin (via a inductor) to power > >One pin to the IC > > I suspect your sensor may not be two photo diodes, but rather two > photoresistors in series across the supply, and the input on the chip is > actually an analogue one measuring the voltage of the divider. This way it > would be easy for the micro to see the "black stripe" of the spoke in the > wheel move past and trigger an interrupt at midpoint, and then figure out > which way the bar is moving by determining if the voltage moves up or > down. > Not convinced about this. The output would have the same mark/space irrespective of direction. With no reference to compare this signal to, how would the direction be decoded? In fact, you'd get the same output with just one sensor AFAICT. Mike ======================================================================= This e-mail is intended for the person it is addressed to only. The information contained in it may be confidential and/or protected by law. If you are not the intended recipient of this message, you must not make any use of this information, or copy or show it to any person. Please contact us immediately to tell us that you have received this e-mail, and return the original to us. Any use, forwarding, printing or copying of this message is strictly prohibited. No part of this message can be considered a request for goods or services. ======================================================================= Any questions about Bookham's E-Mail service should be directed to RemoveMEpostmasterTakeThisOuTbookham.com. -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

>Not convinced about this. The output would have the same >mark/space irrespective of direction. With no reference >to compare this signal to, how would the direction be >decoded? In fact, you'd get the same output with just >one sensor AFAICT. Forget mark/space, think analogue voltage. As the spoke moves clockwise (say) the spoke will shield a portion of the sensor so the voltage on the output pin will go (say) high for a midpoint voltage. This could be enough to put the trigger point outside the window of the chip, thereby generating an interrupt, or whatever is used to sense the position. As the spoke keeps moving the voltage will go towards midrail again as the spoke shields both parts equally, and then as the other half of the sensor gets shielded the voltage goes low, and again outside the window, generating another interrupt. I believe doing this with a 16F630/676/627/628 or any others with comparators would allow you to sense the mouse mickeys quite successfully. -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

> One pin to GND > One pin (via a inductor) to power > One pin to the IC One pin to rule them all, and in the darkness bind them. ***************************************************************** Embed Inc, embedded system specialists in Littleton Massachusetts (978) 742-9014, http://www.embedinc.com -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

There are two sensors in a slight angular shift. They are designed such that there is overlap between there pulses. When you rotate the wheel in one direction you get A B 0 0 0 1 1 1 1 0 0 0 And when in the other direction you get this sequence reversed. A simple finite state machine is used to extract from these sequences Increment and Decrement pulses. The finite state machine is designed such that even if the wheels changes direction arbitrarily, you don't accomulate any error. Tal > {Original Message removed}

Yes, I understand that, questiton is, how do you represent that using a single pin ? Jan-Erik. PS. And I think you should have read the rest of the thread before replying... DS. Tal wrote: There are two sensors in a slight angular shift. They are designed such that there is overlap between there pulses. When you rotate the wheel in one direction you get A B 0 0 0 1 1 1 1 0 0 0 And when in the other direction you get this sequence reversed. A simple finite state machine is used to extract from these sequences Increment and Decrement pulses. The finite state machine is designed such that even if the wheels changes direction arbitrarily, you don't accomulate any error. Tal -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

So, now I'v had my scope connected to the "third" pin of the sensor. In the Idle state (wheel standing still) it outputs two "1" pulses, 2.5 us and 3.5 us long and with a 15.5 us delay between them. This is repeated each 190 us (or with 5.3 kHz). It doesn't matter if there is a spoke or not in between. When the spokes moves in one direction, the second (3.5 us) pulse becomes longer (aprox 13-14 us) and delay between the pulses shorter with the same amount so the total time for both pulses stays the same (21.5 us). When the spokes moves in the other direction, I had problem with the triggering of the scope, but it looked like the first puls gets longer instead, so that most of the 170 us delay becomes a long pulse (part of the "first" pulse). Anyway, one would realy need a data sheet on these, I suspect. While soldering tiny "probes" on the PCB, I happend to burn the sensor for the "wheel". I tried to replace it with one of the sensors for the "movement" detector, but obviously they are a little different. They detect movement (spokes moving by) in two different directions (of by 90 deg). There are a picture of this actual mouse on this page : www.saitosite.com/pc/0007/pachimouse.html. Let the page load and search for "SAT48a". There are two pictures (the page is in Japanes, I think...). Now, how do one find any datasheet on these ? Thanks for your time ! Jan-Erik. -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

At 07:55 PM 4/29/03 +0200, QAC wrote: >In the Idle state (wheel standing still) it outputs >two "1" pulses, 2.5 us and 3.5 us long and with a >15.5 us delay between them. This is repeated each >190 us (or with 5.3 kHz). >It doesn't matter if there is a spoke or not in between. > >When the spokes moves in one direction, the second (3.5 us) >pulse becomes longer (aprox 13-14 us) and delay between >the pulses shorter with the same amount so the total >time for both pulses stays the same (21.5 us). > >When the spokes moves in the other direction, I had problem >with the triggering of the scope, but it looked like the >first puls gets longer instead, so that most of the >170 us delay becomes a long pulse (part of the "first" pulse). Is there a capacitor connected from that signal pin to ground (or Vdd?) If you disconnect the signal lead from the processor, are there pulses on the sensor line or the processor line? I'm thinking that the processor is outputting pulses and the sensor is modulating them in some way. An RCtime type measurement comes to mind. dwayne -- Dwayne Reid <spamBeGonedwaynerspamBeGoneplanet.eon.net> Trinity Electronics Systems Ltd Edmonton, AB, CANADA (780) 489-3199 voice (780) 487-6397 fax Celebrating 19 years of Engineering Innovation (1984 - 2003) .-. .-. .-. .-. .-. .-. .-. .-. .-. .- `-' `-' `-' `-' `-' `-' `-' `-' `-' Do NOT send unsolicited commercial email to this email address. This message neither grants consent to receive unsolicited commercial email nor is intended to solicit commercial email. -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

Nothing between the "output" pin of the sensor and the IC. Could be a pull up inside the IC, don't know. Nice sharp pulses. Jan-Erik. Dwayne Reid wrote: >Is there a capacitor connected from that signal pin to ground (or Vdd?) If >you disconnect the signal lead from the processor, are there pulses on the >sensor line or the processor line? > >I'm thinking that the processor is outputting pulses and the sensor is >modulating them in some way. An RCtime type measurement comes to mind. dwayne -- http://www.piclist.com hint: The PICList is archived three different ways. See http://www.piclist.com/#archives for details.

Hi, >So, now I'v had my scope connected to the "third" pin of >the sensor. > >In the Idle state (wheel standing still) it outputs >two "1" pulses, 2.5 us and 3.5 us long and with a >15.5 us delay between them. This is repeated each >190 us (or with 5.3 kHz). >It doesn't matter if there is a spoke or not in between. Maybe the receiver has variable-R sensors or something? As someone already pointed out, it would be interesting to know whether the receiver is just modulating pulses generated by the IC. It seems strange that the receiver itself will generate the pulses on its own... [...] >Anyway, one would realy need a data sheet on these, I suspect. Yes :) G. -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email TakeThisOuTlistservEraseMEspam_OUTmitvma.mit.edu with SET PICList DIGEST in the body

Well, after a second thought (an a second test on the mouse electronics) I'v found out that the pulses are there even when the sensor is de-soldered. So they must come from the IC then. Still, I'd like to know how the sensor modulates the puls train and how it detects which way the spoke is moving. Well, this was just a late-night-thing, and I'm not going to take it any longer than this... Nice component, anyway... Regards and a happy celebration of Walpurgis night ("Valborgsmassoafton" in Swedish...). Jan-Erik. Guillermo Rodriguez Garcia wrote: >Maybe the receiver has variable-R sensors or something? >As someone already pointed out, it would be interesting >to know whether the receiver is just modulating pulses >generated by the IC. It seems strange that the receiver >itself will generate the pulses on its own... -- http://www.piclist.com#nomail Going offline? Don't AutoReply us! email RemoveMElistservTakeThisOuTmitvma.mit.edu with SET PICList DIGEST in the body