Date: 19950220 From: Leo Kenner To: Bob Wier's (old) MC68HC11 mailing list Subject: CCD? I was wondering if anyone has attached a charged coupling device (CCD) to the 6811/6816 micro controlers to give them "vision"? I am interested in trying this out as a project, but I need a little push in the right direction. Any help would be great, including specs. on CCD's. Thanks. Leo Kenner Date: 19950221 From: Aras Vaichas To: Bob Wier's (old) MC68HC11 mailing list Subject: Re: CCD? I'm getting right onto this! But I'm waiting for my bloody MiniBoard to arrive. i here there are no 68hc11s around ... I'm intending to use the TexasInstruments TSL213 64 Pixel Linear CCD. This device is REALLY easy to use. It only needs a CLK and an ExposureSignal. The output is 0.2 to 2.2V analog (sampled and held!). I'm looking forward to getting this vision device powered up. I intend to use it for a stereo vision and tracking system. The IC costs about $7. It is a DIP8 and the image length is about 8mm - which makes fitting a lens easier (the angle of view is larger in relation to the focal length). WHen I get this thing going I'll send in code and schematics. Aras Vaichas Date: 19950221 From: Jaap van Ganswijk To: Bob Wier's (old) MC68HC11 mailing list Subject: Re: CCD? >I was wondering if anyone has attached a charged coupling device (CCD) to >the 6811/6816 micro controlers to give them "vision"? > >I am interested in trying this out as a project, but I need a little push >in the right direction. Any help would be great, including specs. on CCD's. Philips makes them. I have also wanted to do this sometime... Rumour has it, that if you scrape-off the top of the housing of a dynamic RAM, it acts as a CCD. Date: 19950221 From: Bob Wier To: Bob Wier's (old) MC68HC11 mailing list Subject: Re: CCD? Jaap van Ganswijk wrote: |Rumour has it, that if you scrape-off the top of the housing of a dynamic RAM, |it acts as a CCD. There have been discussions from time to time about this technique in the USENET sci.electronics newsgroup. It works, although it provides pretty low resolution (but cheap!). I saved some of those threads but that was several years ago - I'll see if I can turn that up from my tape archives (don't hold your breath though :-) Date: 19950222 From: Frank Henriquez To: Bob Wier's (old) MC68HC11 mailing list Subject: Re: CCD? You *COULD* remove the lid off of a ceramic DRAM (if you can find one) and jury rig a circuit to read out the DRAM. The result would be an awful, low resolution, poor (no) grayscale image. While you're at it, you could also try to make a 68HC11 out of discrete transistors and passive components, too...and it would make about as much sense. TI (and others) make fine, small CCD chips that cost under $50. You can buy complete CCD cameras on a 1" x 1" PC board for around $100. Most of the work, in any case, will be the 68HC11 software - a 192 x 192 pixel array will still require 36K of memory for a frame, if the plan is to read a complete frame into memory. The CCD will also require some clocking software, too. Frank -- Frank Henriquez UCLA Astronomy Department Date: 19950223 From: Jaap van Ganswijk To: Bob Wier's (old) MC68HC11 mailing list Subject: Re: CCD? >You *COULD* remove the lid off of a ceramic DRAM (if you can find one) and >jury rig a circuit to read out the DRAM. The result would be an awful, low >resolution, poor (no) grayscale image. As I understood it, the bits in the DRAM change slowly from 1 to 0 or the otherway round, depending on how much light they receive. If you look several times between intervals, you can determine the lightintensity per pixel, and thus make a grayscale image. The camera will be slow, but for still-pictures or for security-applications this is no problem. >While you're at it, you could also >try to make a 68HC11 out of discrete transistors and passive components, >too...and it would make about as much sense. Or build a DTMF-decoder with LM567's: It happens! >A 192 x 192 pixel array will still require 36K of memory for a frame. 36K bit == 5K byte. Date: 19950224 From: Frank Henriquez To: Bob Wier's (old) MC68HC11 mailing list Subject: Re: CCD? >As I understood it, the bits in the DRAM change slowly from 1 to 0 >or the otherway round, depending on how much light they receive. In any case, the dynamic range of the DRAM will be awful - certainly not much more than 4 bits. The DRAM is designed to *maintain* the charge in its cells - not real;ly what you need or want in a CCD. >Or build a DTMF-decoder with LM567's: It happens! Seen it done within the last 5 years. Amazing waste of time... >>A 192 x 192 pixel array will still require 36K of memory for a frame. > >36K bit == 5K byte. Actually, the CCD I was thinking of is the TI TC211, with 192H x 165V pixels; using an 8 bit A/D converter the image would occupy 192x165 bytes (31,680) Frank -- Frank Henriquez UCLA Astronomy Department Date: 19950626 From: Rizzardi Flavio To: Bob Wier's (old) MC68HC11 mailing list Subject: Home-made CCD Hello there. This is not strictly an hc11 message, but some months ago I read here the thread about the "home made" CCD devices, made with static RAMs. I remember that somebody talked about some chemicals to open the package without damaging it, to expose the wafer. Well, I have a lot of spare SRAMs here, and nothing to do for a while :-) Can anybody suggest a "painless" way to open them? Thanks a lot in advance, Flavio Rizzardi Date: 19950626 From: Tilman Sporkert To: Bob Wier's (old) MC68HC11 mailing list Subject: Re: Home-made CCD Are you sure about being able to do this with static RAMs? I thought this works with dynamic RAMs, where light affects the discharging rate of cells. Ho could this work with static RAMs? The cells are usually symetrical, and any effect light would have would affect both "sides" of each cell. > Hello there. > > This is not strictly an hc11 message, but some months > ago I read here the thread about the "home made" CCD > devices, made with static RAMs. I remember that somebody > talked about some chemicals to open the package without > damaging it, to expose the wafer. > Well, I have a lot of spare SRAMs here, and nothing to > do for a while :-) > Can anybody suggest a "painless" way to open them? > > Thanks a lot in advance, > Flavio Rizzardi > > Date: 19950628 From: Bob Wier To: Bob Wier's (old) MC68HC11 mailing list Subject: Re: Home-made CCD The following got directed off into the dead letter bin by mistake.... >the first use of this i ever saw was the xerox opti mouse, >and i think they used dynamic ram. light makes the stored >charge decay... > >Edmund. ===== Round Up the Usual Disclaimers! ====== Bob Wier, keeper of the Photo-3D, Icom radio (WB5KXH), Overland Trails, and Motorola HC11 mailing lists currently in Ouray, Co at 8,000 ft. elevation (this is a *high level* system :-) Date: 19950628 From: Steve Nordhauser To: Bob Wier's (old) MC68HC11 mailing list Subject: Re: Home-made CCD Decades ago, Byte mag ran an article for a cheap video camera called the Micron Eye. This used a DRAM from Micron that was fabricated with a clear window on top like an EPROM. To run it you refreshed all of the cells and waited. The exposure time was set by how long before reading out all of the cells. The light modified the time before the cells changed. I know Micron is still around but I doubt that the windowed DRAM is. Steve Date: 19950628 From: Frank Henriquez To: Bob Wier's (old) MC68HC11 mailing list Subject: Re: Home-made CCD The question is: why do you want to use a DRAM as an imaging device? The quality of the Micron (or any other DRAM based imager) was absolutely terrible - very poor dynamic range (a few levels) and due to the geometry of DRAMs, a very odd geometry. If the answer is "just to play around with an interesting idea" then fine - go have fun. If instead, the answer is "I want to add a simple imager to my project" then *don't waste your time with a DRAM imager* it would be a waste of time and effort. TI (and others) make small and easy to use *true* CCD chips - the TC211 comes to mind. I've also seen complete CCD cameras advertized in Nuts & Volts for around $100. Frank -- Frank Henriquez UCLA Astronomy Department Date: 19950628 From: Kenneth Atchinson To: Bob Wier's (old) MC68HC11 mailing list Subject: Re: Home-made CCD Frank Henriquez wrote: > [ stuff deleted ] > ..... TI (and others) make small and easy to use *true* > CCD chips - the TC211 comes to mind. I've also seen complete CCD cameras > advertized in Nuts & Volts for around $100. In the Circuit Cellar INK (July 1995), TIMELINE INC. have advertised Linear type CCD for as low as $15.00 (1728 elements) and a matrix type Sony CCD Imager for $49.00. This one has 384 (H) x 491 (V) elements. Has anyone played around with the any of the above items? Kenneth Atchinson Date: 19950920 From: PEDRO KULZER To: Bob Wier's (old) MC68HC11 mailing list Subject: IR CCD camera Hi! I'm looking for an IR CCD-like camera or something similar. All I want is to get a small CCD sensitive to IR, attached to the hc11 to make some data processing fora mobile robot. Any suggestions of where I can find such a device? Thanks Pedro Kulzer Date: 19950920 From: PEDRO KULZER To: Bob Wier's (old) MC68HC11 mailing list Subject: IR CCD camera Message sent by PEDRO KULZER to the mot-68hc11-apps Mailing List. Hi! I'm looking for an IR CCD-like camera or something similar. All I want is to get a small CCD sensitive to IR, attached to the hc11 to make some data processing fora mobile robot. Any suggestions of where I can find such a device? Thanks Pedro Kulzer Date: 19950222 From: L. Jeffrey To: Bob Wier's (old) MC68HC11 mailing list Subject: reading analog CCD output? Using a 68hc11 to read a CCD camera sounds great, but how do you actually read the signal? All of the CCD cameras I've seen have "standard" analog output. What does that mean? Reading a voltage with the 68hc11 is one thing, but how do you read a video signal? -Jeff Date: 19950222 From: Aras Vaichas To: Bob Wier's (old) MC68HC11 mailing list Subject: RE: reading analog CCD output? Question is: Using a 68hc11 to read a CCD camera sounds great, but how do you actually read the signal? If you look in the TI databook on OPTOELECTRONICS AND IMAGE SENSORS ... it will show you that the output from *some* CCDS is a wicked analog which needs all sorts of conditioning. BUT, the signal is controlled by a CLK so everything can be synced. One method of conditioning is to convert the signal to a standard video signal - for this you will have to check out a data sheet, etc. This signal can be directed into the "VIDEO IN" of a VCR .. or whatever. To recover the orignal analog signal for AD conversion you need to sync the camera and the AD convertor. THis is doen with so-so complex analog circuits (ERGH!). THere is an CCD IC WHICH PROVIDes (bloody CAPS LOCK@!@!) dig. and anal. outputs. It is: They are produced by: visioTEC GmbH Postfach 421151 55069 Mainz Germany Phone +49 6131 95883-72 Fax +49 6131 95883-77 The first one is a chip (44pin LCC chip) featuring a complete micro- controller compatible CCD-camera. It contains an 160 x 160 pixel image sensor array and lots of on-chip-logic. Some of the features (information from visioTEC): - 160 x 160 pixel array - four image formats (120x120, 120x160, 160x120, 160x160) - pixel size 10.5 um x 10.5 um - exposure control (range 2,000 : 1) - gain control - variable frame rate (0.5 - 24 frames/sec) - on-chip 8-bit ADC - automatic ADC offset calibration (dark current compensation) - on-chip optical shielding - serial or paralle digital output (!!) - analog output - power consumption < 100 mW This chip does obviously not need many external components to build a complete camera. The only problem could be the low resolution (only 160x160). The chip will cost about DM 90,00 (should be around $50 - $60). ____________ taken from comp.robotics________________ Whatever, I've got work to do. ADIOS! Aras Vaichas ____________________________ They're watching you ... ____________________________ Date: 19950222 From: Leo Kenner To: Bob Wier's (old) MC68HC11 mailing list Subject: Re: reading analog CCD output? L. Jeffrey wrote: > Using a 68hc11 to read a CCD camera sounds great, but how do > you actually read the signal? > > All of the CCD cameras I've seen have "standard" analog > output. What does that mean? Reading a voltage with > the 68hc11 is one thing, but how do you read a video > signal? > > -Jeff > Well I supose you could set up some cct. to convert the analog signal to some sort of usable digital signal, but what I was talking about is connecting a CCD 'chip' to the hc11. This would be a lot simpler since the chip outputs a digital signal. Most of the circutry in a video camera is to convert that digital signal to an analog signal. By attaching the CCD to the hc11 you are eliminating two signal conversions. Leo Date: 19950222 From: Aras Vaichas To: Bob Wier's (old) MC68HC11 mailing list Subject: Re: reading analog CCD output? The output of a CCD is an ANALOG signal. A CCD itself is a chain of charge 'wells' that fill up under light radiation, these are the Charged Devices. These wells are then connect by a regular network of electrodes. They each pass their charges into 'charge paths' - this is the Coupled bit. In between these charge packets are clk signals and reset signals and the signals that are sued to 'move' the charges out of their electrically isolated wells. This is normally a very messy analog signal. But TI instruments have made one chip, that I know of, that performs all sorts of good things to this analog signal which can then easily be converted to digital USING ANOTHER circuit. Refer to TI's TSL213 64 pixel linear CCD. It is very *nice*. hope this helps. The TI data books also tell you how CCDs work. Date: 19950223 From: Frank Henriquez To: Bob Wier's (old) MC68HC11 mailing list Subject: Re: reading analog CCD output? >Leo Kenner wrote: > >Well I supose you could set up some cct. to convert the analog signal to >some sort of usable digital signal, but what I was talking about is >connecting a CCD 'chip' to the hc11. This would be a lot simpler since >the chip outputs a digital signal. I don't know what kind of CCD chip you're talking about, but most modern imaging CCD chips output an *analog* signal corresponding to the amount of light that has fallen on the pixels. There are a few CCDs with built-in A/D chips, that do put out a digital value, but the CCD itself is an analog beast. Most of the circuitry in a CCD camera is there to drive the CCD clocks (which usually fall in some bizzare voltage range, like -12 to +3) with a fairly small section devoted to amplify the *analog* video data and massage it into an RS-170 video signal. I can think of a couple of circuits to digitize a video signal with a 68HC11. One circuit would have an LM1881 (or equivalent) sync separator, an 8 bit flash A/D and a PAL or two to act as a memory counter. The HC11 would just be used to generate the CCD clocks and synchronize the A/D converter. Later, it could transfer the completed frame to a host computer (or whatever). A second design would use the LM1881 and A/D converter, but would replace the PALS with the HC11. The CPU would generate the proper clock sequence, digitize the resulting pixel and store it in memory. It may be a tough programming project to keep the CCD exposures short, though. Frank -- Frank Henriquez UCLA Astronomy Department Date: 19950223 From: Aras Vaichas To: Bob Wier's (old) MC68HC11 mailing list Subject: Re: Reading the CCD chip with a 68hc11 CCDs are a linear memory storage device. They can ONLY be clocked to obtain a sequential output of the charges. I am DYING to mess with a CCD chip but I'm waiting to get my 68hc11 IC ... Think of a CCD as a shiftregister in base infinity (or some larg number of discrete electrons!) Each charge 'well' reacts to integrates the light falling on it and forms a voltage. These voltages then are are released into the shift register paths (using the CLK pulses) to the output. THe output can be between 10V and -2V. The image data is contained within the 0 to -2V RANGE. Chips like TIs TSL213 output 0.2 to 2.2V. this is because they have inverters and differential amplifiers on board. LAST WORD: READ THE DATA BOOK! Texas Instruments: OPTOELECTRONICS AND IMAGE SENSORS :] Aras Vaichas Date: 19981209 From: Robert L. Smith To: Bob Wier's (old) MC68HC11 mailing list Subject: Re: ccd camera You wrote: >I am using a ccd line scanner that produces an output signal of 1.2V >max. >The black level however is between 400 and 780 mV for the different >camera's of the same type.. >Can anyone tell me if this is normal, and how can I make a simple >circuit that fixes >this problem? The black level output (actually 'dark current') is a thermal phenomena caused by electron motion in the indivadual CCD bins and is related to the temperature of the chip. It is a normal effect. This chip-to-chip variation is normal and must be accounted for if you wish to perform accurate light level measurements (photon counting)with your camera. I have never heard of fixing it with a 'circuit', the normal method used in medical imaging systems is to close the CCD sensor to light sources and measure the dark current level. This is done as part of the camera setup and calibration process. This 'reference' current is then subtracted from the measured value to compensate for the residual dark current. It should be noted that for precise measurements in scientific grade CCD cameras it is customary to run a calibration 'curve' of sensor output level vs. light input level. As a rule the output of the CCD sensor is a linear function of the light level. This curve deviates from linear, of course, as the input light level approaches the dark current value under very low light conditions. The customary compensation means involved discovering (via the calibration process) the coefficients of a best fit linear transform that related camera output, on a bin- by-bin basis) to the input photon count. This transform was then applied to every measurement (again, on a bin-by-bin basis) to correct the observed reading to the desired light level. For large sensors this would involve many thousands of coefficients and MAC (multiply - accumulate) operations per image exposure. A 1024 x 1024 sensor had a million bins!!! 2048 x 2048 or even 4096 x 4096 sensors are not uncommon (but VERY expensive). Needless to say this was a job for a DSP engine. For your line sensor the total bin count, of course, is much lower and (depending on your frame rate) may be able to be handled by one of the fast microcontrollers of the day. Again, in precision applications, it is also not unusual to cool CCD chips to rather low temperatures to extend the useful light sensitivity range down to very low light levels (and thus, extend the useful dynamic range of the instrument). My memory is a bit dim here, but I think that dark current was halved for every 7 degree C. of cooling, so dark current could be reduced to very low values by cooling into the temperature range of -20 to -30 degree Centigrade. The range was within the range of thermoelectric (Peltier junction) coolers so cooled camers without the hassle of cryogenic systems could be built and operated effectively. The usual method was to place the cooling subsystem in an external cabinet and cool the chip with a heat exchanging fluid (typically air or an inert gas). I have also seen mid-range performance cooled cameras with the CCD sensor mounted directly to the cold face of the thermoelectric cooler and the hot face mounted directly to a finned heatsink on the back of the camera. Happy imaging, Bob Smith ---- Avoid computer viruses -- Practice safe hex -------------- * * Specializing in small, cost effective embedded control systems * * Robert L. (Bob) Smith Smith Machine Works, Inc.