EEVblog
Вставка
- Опубліковано 1 лип 2014
- What's inside one of those omni-directional laser barcode scanners you use at the supermarket, and how does it work?
Motorola / Symbol LS9208
Datasheets:
Reflective Interrupter www.osram-os.com/Graphics/XPic...
www.motorolasolutions.com/web/...
www.keil.com/dd/docs/datashts/...
www1.futureelectronics.com/doc...
Forum: www.eevblog.com/forum/blog/eev...
EEVblog Main Web Site:
www.eevblog.com
EEVblog Amazon Store:
astore.amazon.com/eevblogstore-20
Donations:
www.eevblog.com/donations/
Projects:
www.eevblog.com/projects/
Electronics Info Wiki:
www.eevblog.com/wiki/ - Наука та технологія
When you set the scanner up, you have to power the scanner and scan the configuration barcode(s) to set it to the correct mode. It doesn't auto-detect from the cable.
And yes, you can be a nuisance and scan the reset barcode to mess up the scanner at the supermarket, as long as you have the barcode for the correct manufacturer/product family.
The spinning mirror not only sends out the scan lines, but it also collects the reflected laser light off the barcode (via the 5-segment mirror and the curved focusing mirror at the top). There is no timing information associated with the mirror (AFAIK). The bit clock is recovered from the pulsation of the reflected light as it sweeps across the barcode. That's what the "quiet zone" (the white area to the left and right of the barcode) and the start/stop symbols (the first and last few lines on the barcode) are for, as well as for determining the barcode type.
If you can, get your hands on a CCD barcode scanner, not a laser scanner. Not only can they read damaged barcodes better, but that kind of scanner CAN work with the illumination defeated.
The mirror wobble feature is actually fairly unusual for an omnidirectional barcode scanner.
Decoding the barcode is actually not that difficult. All barcodes break down to a simple bitstream. All the processor has to do, after decoding the light delta values into bits, is determine if those bits follow a certain pattern for each barcode type, decode the barcode into symbols, and apply the check digit test. Not that processor intensive at all.
I'm just guessing here, but the opto-feedback on the spinning mirror is probably used to set the vibration speed of the focusing mirror. I could be wrong, but I doubt it's being used for timing of the decode logic.
Source: My job. I'm a barcoding and RFiD specialist.
Thanks. What about the infinite number of angles and distances (and hence bit timing differences) that are possible as the user swipes the product. How does the processor handle all that? It could vary from one processed angle to the next as the product is swiped. I suspect the decoding algorithms aren't as trivial as one might think?
EEVblog There is some magic sauce in the firmware that looks at the data stream from the photodiode, and even if the scan was made at an angle, and the timing would be way off, it looks at all the received bits and calculculates the result. There is CRC error check, too.
EEVblog The barcode system works on changes, from dark to light or light to dark, not in thick and thin dark and light bands. If there is a change at a time point, that's a 1, if there is no change at that time point, that's a zero. The codes are tweaked so that there is no more than 3 'no changes' or zeros in a row, to allow it to detect timer drift. There are also three sets of two thin dark bands (actually, three patterns of 4 ones in a row), at the start, middle and end, to act as timing markers.
The scanner uses this to establish and adjust the timing as it scans the barcode.
EEVblog You're right, different angles and distances will cause the pulse train frequency to vary, but after the clock recovery happens, any bitrate differences due to parallax would be very minor - unless you have a ridiculously long barcode, like an EAN128 with many fields.
Now, I don't know exactly how the internal algorithm of this particular scanner works, but it is possible to adjust for any timing errors because there is both a start and a stop symbol. You could also do some clock adjustment based on transitions within the barcode.
My statement about processor usage was referring to the barcode decoding into characters, not about converting the analogue signal into the barcode bit sequence. I suspect the BedrockII chip is entirely responsible for this, as it is a very generic operation and probably best implemented in hardware. But as I said, I don't know exactly how the internals of this particular scanner work, so take that with a pinch of salt.
***** Unfortunately, barcodes do not contain a CRC, unless it is embedded into the barcode characters themselves. Barcodes do have a check-digit, but it is not based on CRC.
Robert Backhaus From the top of my head, each barcode "symbol" contains a fixed number of black and white bands. The total number of "bits" in each symbol also remains constant. By "bits" I mean, a 1-thickness band is 1 bit, a 2-thickness band is 2 bits, etc... The bands in each symbol can be expanded into a fixed number of bits, and vice versa. The only exception to this rule are start and stop symbols, and possibly control symbols of some of the more exotic barcodes.
You're correct; barcodes are like NRZ signals. It's the transition you're looking for, not the color of the bar. Some barcode formats even alternate between white and black for the first bar in each symbol, simply because each symbol contains an odd number of bars. Really neat stuff.
However, most (nearly all?) barcode formats, and in particular the EAN13 format he was using, do not contain "sync" bars or symbols in the middle of the barcode.
EDIT: Actually, I just double-checked, and you're right. EAN13 barcodes DO have the "center guard" bits. My apologies, I stand corrected.
I think the opto-interrupter is keeping the mirror speed locked. in a feedback loop. Probably if they do this, then they can approximately determine the time its going to take for that beam to scan. it could be easier to recover the bit clock.
I bricked a barcode reader that I was evaluating. It was always on and had no trigger. I absentmindedly scanned the setup barcode that took it out of 'always on' mode into 'trigger mode' as might be required for models with a trigger.
When I 'phoned the engineer that I had been given as a contact at the supplier he listened patiently, said that what I had done was impossible and tried it himself.
Result, two bricked barcode readers.
This is the greatest thing I've read all day.
Are you in Arduino forums also ?
+Malhar Deshmukh Yes
I get a kick out of the giant barcodes on some Aldi food packaging. They've clearly optimized for fast checkout.
This thing essentially takes your one-dimensional handheld scanner and positions it in a whole bunch of different ways very quickly. I would hesitate to call this "image processing" because the processor is simply waiting for a serial stream of meaningful data. Once you realize that this is still a one-dimensional scanner, your "angles" only mean different scan speeds. Most don't work. too high, too low, left, right... It's waiting for one meaningful scan.
I was a point-of-sale programmer for Sears (now Sears Holdings Corporation) for ~14 years. I've had lots of time with these Symbol devices. A joy forever! Thanks for the vid.
Also it's interesting that the parabolic mirror is vibrating. I didn't think it was supposed to do that. I remember doing thousands upon thousands of scans in our retail systems labs, and the lines were discrete without moving up and down. We also had another weird Symbol scanner that had I think 2 rotating mirrors, and the whole design spun around in circles on the paper. Also, you aren't seeing 1's and 0's as you said in the video. There are different types of bar codes, and they actually can use different widths of bars and blanks in different sequences to represent different characters. In other words, it is NOT binary.
JLTSoft
They probably have different models. Simpler ones without wobble, and this more advanced one with wobble (trademark!)
JLTSoft
Maybe it's changing focus real quick to make sure there's always one line sharp enough, regardless of distance to object?
@EEVblog, With laser barcode scanners the laser sensor is simply detecting the on off pulses from the reflection from the dark and light stripes of the code. The sensor is 1 dimensional and only receives a serial stream of data. The sensor doesn't need to know anything about the angles the beam is scanning.
Here is the way I keep it straight. They also make pen type barcode scanners which use your hand to swipe the laser across the barcode. The rotating mirrors simply automate that process. At any given moment the laser is simply a point of light so it doesn't need a 2D sensor.
Edit: Actually, I didn't mean 1D sensor, I meant 0D like a photodiode. In my mental picture the 1D was the time axis. Sorry about the wording above if it confused anyone.
Also, the laser reflection return path follows the same path the laser took but in the reverse direction. The photodiode is either in the LLED package or there is a beam splitter to divert the return reflection into a detector located on the laser assembly.
Dave, the "Pebbles" chip is the image decoder for the image coming from the camera, and the second STI chip is used for the interfacing and detecting the RJ-45 cable used, USB or PS/2.
The image decoder chip is just used for decoding the barcode sent from the "Pebbles" chip and then sending it to the Atmel that is sending it to the second STI chip, to the Hitachi and the to the POS through the cable
Great teardown! That's definitely a 10-pin modular connector, not RJ-45 (RJ-48 maybe?). Apparently additional pins are needed for all the protocols this thing supports.
Yes, I didn't count them at the time.
Brilliant (no pun intended) tear down! I've always wanted to take a deeper look a these things and am reminded of that every time I go to the grocery store.
One of the coolest teardowns in a while! Also a vote for a 2nd video with the photodiode waveforms.
That weird package is most likely Osram SFH9201 "Reflective Interrupter in SMT Package" or one of the competitors.
Yep, it is, exactly as I thought. Link is in the description.
That scanner is just seeing a serial stream of light and dark from the photo-diode. However many scan lines and directions it has. Standard bar codes normally start and end with 2 thin lines which lets you find the start of the bar code and the approximate scan speed(for manual scanning). Resolution is the reason laser is used.
Great video Dave! I've always wondered how these things tick! Keep it up, I love your work. -Peace from Slightly Further Down Under (NZ)
Covering the laser and being surprised it doesn't work at all? Uhh? it works by reflecting the laser!!
Gotta check...
Oliver O'Neill Sorry to correct it, but no it would not. This is a laser scanner, not an image scanner. Completly different sensor. This scanner can only detect two states: Laser is being reflected by surface, laser is not reflected by surface. It can not detect an image. Even if it had an imagesensor (which it does not have, it just has an light sensor), it could never work with standard light, becaus it can only see 650nm (typical laser light). Thats why it can still read the barcode when he shines a flashlight on it: The scanner just doesnt see normal light.
Image Scanners are much more complex and expensive. Simplified: They are digital cameras with a software to detect Barcodes.
CCD Scanners (the usb model he also shows) are image scanners with just one (1) line of detectors. Those could probably work with any lightsource in the range of 650nm.
Right, the sensor is just one photo diode, or whatever, not an array. The scanning of the laser converts the bar code from the spatial domain to the time domain, so to speak, and the time domain signal is decoded by the electronics.
I'd be surprised if there even is a synchronization between the logic and the spinning mirror. The bar code probably has all the right start and stop bits that it can just get picked out of the stream.
Photodiodes can work at normal lights, there are different types centered at differents wavelenghts, and I bet this dude is much smarter than you and the only reason he is checking if it works without the laser is for the people who dont know that kind of stuff...
EtikettenWissen It has many similarities with a flying spot camera and it could in principle use the pattern to make images, it would be cumbersome due to the scanning pattern, a simple 2D scanning grid is generally used for flying spot cameras. Of course, since this is a scanner for 1D barcodes, it has no real need to reconstruct an image, it probably just filter and gain control the input and constructs a stream of bits that represents the high and low lights along the line currently scanned.
The DS9208, which can detect 2D barcodes has square scanning field, I believe, so it might be using a flying spot approach, though, since it also can detect codes from display devices like mobile phones, I suspect they may have 2D image sensing. Though you can read some display technologies with a flying spot scanner, they would have to be reflective types, not the emissive/backlit type displays generally used on phones these days.
Thanks Dave , always wanted to see the guts of one of those, and see what makes them tick !
awesome teardown dave
Wow! Excellent video, as always! :)
I can still remember, back in the early 80's where the laser and prism system was built into the item belt system. I remember being like 4 years old and mesmerized by it.
Cool Dave! Funny thing is I recently did a similar tear-down of a datalogic scanner of very similar type. I salvaged alot of the parts inside (motor, mirrors etc).
Always wondered how those buggers work. Thanks Dave for teardown...
This video has been very informative to me. Many thanks for making it :)
wonderful video. I learned a lot. please do more of these tear-down videos
Good one! I've taken a couple of those types of scanners to bits... Symbol has very nice mechanical design, at least the ones that i had pre-moto. Great vid
As i worked for a PA rental we had some lasers as well, they used piezos for moving the tiny mirrors in the latest optic banks before we closed down....So for 14:27, i presume it´s a piezo as well...let´s see if i´m right ;)
No it´s working like a inverted "loudspeaker"....very cool anyway...:)
Thanx for this great one,Dave!
What's cool is that nowadays image-recognition technology is so good that you don't need any moving parts, or even a laser at all, to decode bars; just use a camera such as the one built into a smartphone, and the app that drives the recognition (which is how QR has already been done for a good while now, but the camera may still not have quite the angular omnidirectional flexibility as this does yet.
I don't know much about electronics but this was really fun to watch! I always wondered why the bar code scanners had moving parts.
Transducers are common in laser scanners to produce different patterns, it's basically a speaker (minus the cone) that responds to different input frequencies and levels.
The supermarket I work at has omni-directional scanners, but they are ancient things that take up a lot of the counter space and have huge windows for the lasers.
Really interesting to see the mechanical isolation for the optical assembly, That's not something I usually think about.
This thing just took be back to my factory automation days back around 2000.
The industrial 232/485 devices were a pain in the ass to keep running.
What a beautiful design
Very interesting video. I'm a cashier at a grocery store and I have always wondered how a scanner works
Great video, love it!
There is no image processing. It just scans (hence the name), lines and measures a reflection using single sensing element. Laser makes reflection very bright and easy to distinguish. Only remaining processing is to synchronize the code, as the barcode is flat, but at unknown distance or angel, which shortens or lengthens the barcode in time. Arm seems a bit like over kill, but maybe they really need a speed , as this thing needs to try to detect about 80 codes per second. I still feel they should be able to process that amount of scans with just the h8s they have on board.
My guess is they needed some customization features (i.e. upgradability in the field, including ability to add support for new code types), and h8s was just too old to be messed around anymore.
How the 2d codes with this kind of laser sensors works I have no idea. I think they do have two perpendicular mirrors.
These vids are so cool. This guy is a genius. He's like, oh look that thing is shaped funny I have no idea what it is. Two seconds later he'll deduce what it is based on the tiniest piece of information and give you this amazing reasoning for it.
@22:15 I believe that feature is there entirely for very fine weight balancing of the prism which is probably done with some sort of stroboscope hence the shiny square corner.
The “line” scanners can be useful. In fact we use them at work for scanning serial numbers and that sort of thing - UPC, CodeWhatever and Code128, DataMatrix, and QR. the reason is that there are frequently many barcodes in a small area: SN, MAC, PN, Rev#, etc.
I was going to try and get you one of the old bench-scanners (from an old-school checkout) to you - but was having problems with the cost - guess somebody got you one before I did! Oh well, saved me a small fortune ;)
I use a handheld Symbol scanner at work that only shoots a single laser line, far more effective than the omnidirectional one that doesn't work good on the barcodes I scan. The handheld is in a holder that has a magnet that keeps the scanner active via a magnetic reed switch.
04:00 Yes, you can, but either as service personnel, or someone close to them. There's calibration pattern sheets similar like this, for every known bar code scanner. Different patterns exists even for same models with either USB or RS-232 models, so you need to know EXACTLY what are you dealing with.
Fun? Best part of this fun is probably when they catch you on security cameras, no doubt. Good lock then!
And laser diode is so critical part that motor in bar code scanner will not spin at all (returning only beep error code) if diode is not functioning.
Now I know why these devices are that expensive. Delicate and well made with high quality custom components.
Excellent.
Very good video.
So awesome, Its the everyday simple items that I don't know about. You could ask me how some obscure item works but when you ask me how an everyday item like this works I couldn't tell you. Well before now.
you ware right about that photo-transistor ... so this is what they call it... SFH3201-2/3 NPN Photo-transistor SMD 460 to 1080nm OSRAM
thanks Dave, that was very interesting :)
I don't think it actually does image processing. The laser just illuminates the barcode and the reflected light level changes while it travels across. The sensor just picks up pulses in the brightness level.
You could try putting the barcode in front of the sensor with the built in laser covered, and then flick a laser pointer across! Or use the laser from the other scanner to illuminate it.
I think the mystical non-magnetic thing on top of the rotating part ( 22:35 ) is to balance it so it does not vibrate at full speed. They may manufacture this part and do a spinning vibration test. If required they can grind of a tiny little amount of the material to improve rotating stability.
Great video, with some mechanical/optical in addition to the usual electronics. At 11:57 you try to scan without the laser diode. I think that the laser doesn't just illuminate, it effectively *points* the two-wire sensor at each particular part of the barcode while scanning, while the sensor just picks up all light from the area (i.e. it doesn't "look" into the mirror). So even with a bright light, the sensor won't be scanning the barcode, just looking at all of it at once as a gray.
the sensor just looks at the total brightness of total area.
Due to the fact there is just one laser beam "at a time" passing of the bar, you just differenciate the signal, no matter what the ambient light is, as long as it is continuous.
So it will NEVER work with ambient light, as long as you don't swipe with (another laser pointer) over the code, with decent speed.
I just repaired a waterlogged LS9100 that I found on the street. It's extremely similar to yours. Mine doesn't have the extra vibrator coil but the rest of the design is very similar (though with mid-90s components). The prism on mine doesn't have any reflective segment. I suspect the
timing of the rotation is completely unrelated to the optics - I think it'll decode as long as the detector sees *any* valid train of pulses, regardless of their timing with respect to the mirror. You could probably swipe a barcode past the detector "really fast" and it might still work. I suspect the photoreflector on your prism is needed to align the scanning with the action of the vibrator coil, to prevent one from interfering with the other. The extra plastic bit in the corner might be for balance.
The top of the prism with the black paint and a shinny surface is to optically measure rpm of the prism...
Same a a optical sensor on a throttlejockey, and using a black and shinny surface on the cooling fan on the nitro motor to control the rpm on a nitro r/c heli for hard 3d flying when the motor loads up the throttle will be increased to keep motor at a set rpm
Yes that 6 pin package is a sensor. On the 2 pin side you have a infrared diode and on the 4 leg side is a sensor. We have them at work and are used to detect the rotation of a shaft on which we have a silver and black pattern.
There is a good reason behind using the laser as illumination and not ambient light. If ambient light was used the detector would need large lenses on it to focus on a tiny spot on the barcode and scan across. Much harder to do and would have a focal point a certain distance away from the reader where its capable of scanning codes. While a tight laser beam can easily be swept across in a line using mirrors and has no focal point while the detector looks for light bouncing back from anywhere in its operating angle (Hence the bare detector with no lenses on it apart from filtering to reduce ambient light interference).
That little package is probably the sensor that causes the laser to turn off when you stop the prism from rotating. And perhaps for timing.
I doubt it was to bind ground to ground cause the metal case would connect those grounds anyway, more likely it was for shielding afterthought. The optical position sensor has IR GaAs-emitter in combination with a silicon NPN phototransistor.
The part where he stalls the motor and the Laser switches off, is actually a safety feature. Even though laser is class 1, when focused on a single point (in this case) can heat up the point of illumination quite a bit. Please correct me if I'm wrong.
Thanks
That gold plating on the mirrors makes it look like a piece of space equipment.
Nice and interesting!
I've done a little image processing, and the computational bottleneck in processing the scan is probably in the rotation. So I guess that's what the "flintstones" chip is doing, maybe some sort of eigenvalue computation DSP?
Nice teardown.
Here's my take on the full path of the light.
Just as David says, the laser reflects on the center circle in the wobbly mirror, down to the rotating prism, sweeps across the 5 sided reflector, and then sweeps the barcode.
But how does it get back to the photo diode?
I think the most visual way to show that would be to replace the photo diode with a LED (say, a blue one), and see what it lights up. The area that is lit blue will represent the area that the photo diode "looks at".
The opening on the photo diode casing is directed at the wobbly mirror, and that's most likely why the mirror is much larger than needed for the laser, and it's also concave to concentrate the light. So the blue light bounce off the whole wobbly mirror, down to the rotating prism, over the 5 sided reflector, and on to the barcode as a fuzzy blob around the sharp red dot from the laser.
That would actually be a cool experiment to view with a high speed camera. You'd see how the area that the photo diode looks at actually follows the laser dot sweeping around.
I don't know how big this area is, but I'm sure it's too large to see any details in the code. The details comes from the small laser dot, the tracking "viewing area" is just to reduce noise.
Those reflective opto-sensors are common as anything, nothing mysterious about them. They were in millions of VCR's, to sense reel rotation - the reels having mirrored stripes on the underside that reflects the IR or UV light back to the sensor in the same package
About bricking the supermarket's barcode readers and toying with its configuration: DEFCON 16: Toying with Barcodes
IT USES INFRARED!!! That is why it has gold plated mirrors and a red laser.
rather than screwing with the lighting conditions expose the sensor to ared laser (or a array of them with a few wavelengths for good measure and then see if it can reed it
Cool enjoyed watching this
No, you cannot get it to work under ambient light conditions. The laser is absolutely required to sweep through the barcode. Only a CCD based scanner can read from ambient light.
What happens if you use the hand held line scanner on the bar code and try to read it with the symbol device sensor? I would assume it would pick it up still.
I've always wondered whats in these things. Incidentally was watching in the morning around the time our parcel postie arrives. Each time Dave scanned a barcode and the beep occurred our dog went ballistic at the front door cause he thought the postie was out the front...smart little guy...or not...
From memory there is a set-up barcode that locks the settings from changes, Supermarkets 'should' be set-up like this
Should, yes, but in practice? you gotta wonder..
Oh no supermarket laser scanner hacking, the ques and long enough all ready.
I service these and other super-market POS equipment for a living.
Some models of scanners support a kind on "Lock" code, but many do not. All the models that I know of that do have a "Lock" code, also have ways that it can be bypassed. (Otherwise, it would be impossible to reconfigure after the first time it is "Locked")
And in my experience, there is no need to wonder weather it is used in practice. None of the customers I service use this feature. In fact, a good ~10% of the service calls I run are to reconfigure scanners that have been "accidentaly" bricked either by some one messing around with the programmign barcodes, or simple random noise patterns beign detected as a barcode by the scanner. Barcodes and scanners are designed to make the latter less likely, but it does happen.
the rotating prism probably provides some cooling function for the laser
I think the light sensor isn't focused enough to read between the stripes of the barcode and needs the flicker of the laser scanning across the barcode.
That fascinating package @20:30 is most likely a reflex sensor, IR or visible emitter and detector separated so there is no leakage between them. If you look what is in front of it when assembled, you see that spinning mirror, which has an reflective area on it.
As to the flatflex ground plane, my best guess is that the photodiode or transisor is fed into an OP in current mode, and I know from school that current mode inputs (i.e. an inverting amp fed directly at the - input) are sensitive creatures, needs guard traces, shielding, the works.
Agreed on the "special package. Looked like an IR emitter/detector pair.
My best stab at the datasheet:
docs-europe.electrocomponents.com/webdocs/08b0/0900766b808b01c9.pdf
9:30 to 10:30 Aren't the number of parallel lines more than 4? The vibrating mirror (Solenoid + permanent Magnet) is producing those extra lines (which are, of course, parallel to the four parallel lines when he force- stalls the vibratory motion). Please correct me if I'm wrong.
Thanks
That connector looked like an actual USB symbol, lol.
Warning: just because the manual says it's a class I laser product doesn't mean the laser diode isn't a higher safety class. The manual is most likely assuming you aren't going to open it up and play with the diode directly. Please be careful with unknown laser diodes near your eyes and/or cameras.
13:18 I don't think so. The sensor might be just a simple phototransistor, and laser ensures that the sensor gets a _stream_ of data (gets maximums and minimums and interpret the signal). If there's no light that moves across the barcode, it won't sense anything.
It would be interesting to use a smoke machine to see all of the beams coming out of that.
or dave could light up a cigarette for science! but they also sell smoke in a can :)
jusb1066 Awesome!!! I want some :)
i have tried the idea with the fog machine - thats not really fun -
you dont see beams - the laser is a way to week -
you have to remember: the solid laser beam is distributed all over the different lines - so if you can see the beam out of the laser-diode after the rotating mirror you only see a really small amount of light.. -
If it works just as keyboard, is it possible to just print a barcode containing "alt-f4" or something else (or is barcode only numbers)?
Newer ones are based on cameras and image processing and are really fast compared to the cheap laser models.
You said one of the chips was made in the 53rd week, do you have more weeks in a year then the 52 we have here in the U.S.A.??
Did you try blocking the sensor that was reading the top of the spinning prism?
min 23: 14 , that weird device look like an IC , is an actual detector that captures the light from the reflecting square on the motor, in order to align it, However, the way of working is simple, the slit in between is usually dark , once the light reflected from the up surface of the motor (the square reflected area) with a specific angel, it will go into that slit and illuminate it.Then a detector will do its job. (its my theory) . so the slit has a function (i call that geometrical engineering)
no capture of the photodiode waveform? :(
Right! Detecting the light-waveform would be interesting. He could prepare a barcode from a I2C signal and set the scope to detect and decode I2C? :D Ah, no, it would require two signals. Maybe RS232 or something else, then? :D
Michał Gawron
that is a nice idea :-)
i think somewhere i have a old laser barcode scanner... will have to try this..
Stefan Krüger i have saved a barcode pen from the junk. Those things are quite interesting, i might take it apart
I can't wait to buy something heavy again!
Lovely, lovely..
Please help! I have a symbol hand held scanner model LS2208 - SR20001R-NA. The laser is very dim, you could barely see the laser. Just wondering what's the name of the part called that needs to be replaced? I'm trying to fix it my self. Please let me know soon. Thanks
this guy is way to excited about his scanner he would do great at a supermarket
I'm confused about why Dave checked to make sure the thing didn't work without the laser under normal light conditions. My understanding is that the photodiode detects the changes in brightness of a particular wavelength of light (the same as the laser), and converts that into an electrical signal which is then decoded by the rest of the circuitry. So as the laser sweeps across the bar code at a particular speed, the scanner can decode it as if it were some kind of Morse code.
Assuming I'm not too far off, why would Dave think it might work with ambient light?
The discussion about how the atmel chip is doing "image processing" and "different angles" is also confusing. Unless I'm way off, it doesn't see any angles. It just sees a never ending string of bits and watches it for certain patterns. (probably a header, followed by a number of digits, and another header)
Am I missing something here?
EDIT: and if I'm right, you wouldn't have to program the scanner at all. It would read any code and output it. The point of sale computer on the other end of the cable would compare the barcode to a database to apply the price/itemname/whatever, and then talk to the inventory system and do the rest of the stuff...
EDIT: This was meant to be to your prior post. Sorry. Commented in the wrong place.
Ambient light: All wavelengths. Including those of the laser diode. Intensity? Possibly lacking.
Peel back the filters on the receptor, and perhaps it'd be happy to work with ambient.
Possibly the circuits (Atmel and maybe others) need more contrast than ambient can provide - especially with the filters in place.
Electronics are not good at "Seeing". It takes work and engineering.
Gavin Callard A photodiode only detects changes in overall brightness. If you're not sweeping a beam of focused light across the barcode, why would you expect it to be able to detect the pattern of the code? In that case, turning off the laser is just stupid.
Also, given that the photodiode can't see an image, but rather sees changes in overall light, why would anyone think it could see angles?
Can you please open and show latest barcode scanner and explain it
23:10 *Infra red. Probably just counting the rpms.
24:00 "Vibrating shaft" lol
If you hook up the oscilloscope to the sensor, would you see the "barcode" as pulses on the screen as the sensor scans each bar on the strip?
If the object with the barcode is totaly stationary with the sensor, kan it still read the code or do you have to move the barcode across the laser beam for the sensor to read it?
I used a hand held omnidirectional scanner at work. Obviously all the mechanics doesn't fit in there since it's so small. How does it work then?
Oliver Guy maybe optical (like a camera inside)?
Not sure the CPU would need to process any angles, the mirrors do that it hardware so the CPU just sees a bitstream essentially and has to watch it for barcodes.
I've found this video trying to repair a reader like this. Great work!
My reader restart itself sometimes while is reading a barcode.. anyone has a clue? I can't figure out :)
The split opto coupler looks like a reflective opto sensor
Yep, it is, link in the description.
53 rd week '05 ?? I must have a funny calender.
I noticed that too, but if it is 53rd week of 05 that means it was produced on 31 december of 2005 (52 weeks cover only 364 days so 53rd week cover only one or two in leap year remaining days).
Whenever I hear spinning mirrors and lasers, I can't help but think "instant blindness!" I know a barcode scanner is probably not that harmful to look at.
20:30 optical reflection sensor that looks at the mirror RPM.