TTL interface

[Johnm48]

I have a project to interface a dental camera system to a printer. The control circuitry on each machine uses TTL logic. I built an optical interface but the signals are different. The basic method of operation is to send an external signal to the printer (Mitsubishi CP910u) through pin 2 from hi to low to capture an image then on pin 6 take it from hi to low (momentary) to print. The catch is that the camera sends other logic signals to pin 2 when different inputs are given at the camera and the printer goes off on other selections that are not desired and it is difficult to return to the basic operation.
The question at hand is there a way to take the pulse signal from the camera system and convert it to a momentary signal (one shot) only for the printer?

[haklesup]

By optical interface, do you mean a digital interface for the picture (optical) data or did you actually opto-isolate the signals between the camera and printer? By TTL Logic do you mean that the interface is strictly 5V?<p>Sounds like a solution using only cables (wire to wire handshaking)is unlikly. <p>A solution using only simple combinational or sequential logic may be possible. Look at using the take picture signal as the clock instead of as an input or you could free-run a clock and use all the pins as inputs. <p>If the signals are too complex to handle with logic only then a small microcontroler may be appropriate. You could then handle each signal independently acording to a program you write. A PIC or Basic Stamp or other controler would probably work fine and be economical enough to do.

[Johnm48]

The interface I am working with is strictly the control signal, nothing to do with the video signal.
The interface is opto-isolated between the camera and printer, both systems are running at 5 volts.
The printer will take the command to capture the image when pin 2 is brought low by physically shorting the pin to ground once, on the next crude command it will toggle back to the live image. This is the way I would like to see it respond. The complication comes when the printer gets a multi pulse signal from other source connected to the output pin on the camera system, IE: there are number of buttons on the fiber optic wand that is used for inspection. One of the three is for "capture", one for "multi frame capture", one for "print". It seems that the "capture"and "multi capture" both are outputted on the same pin. How can the unwanted signal be blocked?

[Chris Smith]

A 555 one shot

[Dimbulb]

you must connect a dentist camera directly into printer #CP910U. You get a capture function when you give pin 2 a one shot. ok so far.<p> Now theres more
then along comes these pulses from the something
a part of a dental system called "other source" which has a hand operated optic devise called a wand with three buttons. capture, continious low res and print.<p>You are trying to hack the wand pulses because the capture and multicapture are both on the same pin ? <p>You have a project.
You have a dental photo system.
You want the camera and printer to bypass the system (software) and do a single print.<p>Let me guess the software is awkward and requires fiddle file flim flam and somewhere there otta-be some way to get this silly thing to print. You called the company. It would be nice if<p>[ May 20, 2003: Message edited by: dimbulb ]</p>

[Chris Foley]

db is on the right track. Try getting in touch with the manufacturer of the dental camera. The company that manufactures the camera might have applications people who could give you a hand. By the way, you have read the manual, right? Go through it one sentence at a time, and see if there's any information there you can use.

[Johnm48]

Let me see if I can make it a little clearer. <p>I have a device that is used to produce a picture. There are 2 parts to it. The main body of the device called the docking station, and the optic head of the camera called the wand. The wand is attached to the docking station by way of a cord. This wand has along with the optics on it 3 buttons that the user uses to give commands to the docking station. The docking station has out put pins feeding an RJ45 jack that is mounted on the rear with Cat 5e cable to a printer arrangement that is no longer used by this camera system.<p> The next piece of equipment is a color printer, Mitsubishi CP910U. The printer has an 8 pin DIN connecter to input commands from a remote source. The source will be from the wand via the docking station. <p>I am not sure that the 2 pieces of equipment are directly compatible so rather than damage a $13,000 dollar camera or printer I am trying to interface them with out electrically coupling them. So that is were the opto-isolation comes in to play. This will be between the camera docking station and the printer. In the past the original system had a programable interface on it but it came configured for the system and they seem to have fallen off the edge of the earth.<p>Working with the manual for the printer there is a circuit diagram that shows what pin needs to go low to capture a picture. When the line is brought low again a second time, the picture will be live. There is the ability to go between live and capture until the desired image is found.<p>Again working with the manual for printer there is a circuit for the print input, when a pulse is given the line is pulled low the printer will start to print the from the last captured image.<p>That is the explanation for 2 of the 3 buttons on the wand. The bed button on the camera wand, when pressed will give a set of pulses on the same pins as the capture/live thus causing the printer to go into a different mode. It's passable to get it back to the basic mode but it's hit & miss until the correct sequence is given. Not acceptable.
The camera company has been bought out and no longer supports the equipment so I only have the pins identified by use of a VOM. <p>It sounds like the 555 might be a way to go.

[Dimbulb]

This description identifies the difficulty as a glitchy signal when attempting to change from live to single frame. It works only once in a while and after repeated attempts.<p>In isolating the source of the failure.
Is there a way to determine if the problem is in the wand, the docking station or the printer.<p>Your guess is that it may be that the signal to the printer at pin 2 is not correct and this would be in the docking station or a button sticking.<p>Your inclination is to improve the signal to the printer by using an opto-isolator.<p>My inclination is to observe this signal on a scope hoping to see what the correct signal looks like. In looking at the scope paying attention to
the height and width of the correct signal. Is it weak or strong what does pushing the button show ?<p>The signal may be three different (widths)multiples of the clock which are three control signals, If the signals are seen by the printer chip as weak unpronounced or noisy they may not be meeting the requirements of that printer model.
And much speculation could be made. Information on what is the correct signal should be available from the printer manufacturer.<p>[ May 21, 2003: Message edited by: dimbulb ]</p>

[Johnm48]

I believe that the signal identification is indeed one of the problems. At the docking station I have
identified the voltage to be at 5 volts then being pulled low momentarily . Up to this point I was
unsure how to look at it other than with a VOM. The scope will solve that.

On the docking station I see the signal from the first button on the wand being clean. It will toggle
from live to memory with no problem, outputting on pin X, to pin 2 on the printer. It’s been in
service for 5 days now with no problem. If and when the second button is pressed on the wand it
will output on the same pin X. I can see the pull down from 5 volts. The sequence is not easily
returned to the live / memory without pressing the first & second buttons hit & miss. This is
where the scope will help. At this point I have not looked into trying to disable the second button
within he docking station, but I’ll give that a shot.

In the printer manual it states that the timing of the input signal needs to be 38.4 ms/per1 word
and giving examples of the print code utilized. So it is clear what the input signal should be
there.

[haklesup]

Sounds like a logic "state machine" would work here. You should be able to make one using a hand full of flip flops and logic gates. the design method can be learned from any good logic design book but let me try to set the stage:<p>Imagine a black box with inputs outputs and a clock input.<p>Your inputs are the signals produced by pushing the buttons on the wand. you may use AND, OR or NAND logic to purify these signals if needed. <p>The clock is any signal that changes when any button on the wand is pushed. If this dosen't already exist, you may have to OR some signals to make one. It may only clock when a button is pushed or it may run all the time<p>The outputs are the pulses that the printer needs to go into its various modes. If it is just a matter of having the right logic on your output pins at the right moment when the printer looks at them then thats not too hard but if you need to send a serial sequence of pulses to satisfy the printer, you have to rethink the clock.<p>It looks to me that the wand/docking station produce serial data and that the 555 is an attempt to clean that up into a single pulse suitable for combinational logic. The printer however, is expecting serial input (multiple pulses synched with a clock - it seems) so if you give it a single wide pulse, it might or might not understand what you want it to do.<p>Before you can design any kind of logic, you need to figure out the timing diagram for the wand and the printer seperately. You need to determine if a sequence of pulses is needed or if a static pattern of 1,0 is required for inputs and outputs. This determines the fundimental design requirements and method.<p>Once you have the timing diagrams, you can make the state-next state logic diagrams (which define what you want it to do) from which you derive the boolean equasions for the logic that goes between each flip flop stage in the state machine design.<p>Sounds complicated but the hard part is reverse engineering the signals on the two devices. There are undoubtedly other functions that you don't care about but you will have to avoid activating.<p>Chris

[Johnm48]

Contributors, Thanks for your input. Your ideas gave me some good leads to follow. I checked out a site on the many uses of the 555IC and looks like the way to go utilizing the one shot application. It may be a few weeks until I get back to the project. Time to get the garden in. Thanks again.
John