chaosgk, the signal coming from the prop pins is pulse width modulated and produces a voltage on the 10nF capacitors. The wider the pulse, the higher the voltage on the capacitor. The capacitor filters out the high frequency pulses and leaves the smoothed audio frequencies. The 1uF capacitor blocks the DC component of the signal on the 10nF capacitors and provides an AC signal to the output.
I love companies that send out sample parts. I went on Texas Instruments site and found I could sample order most of the stuff I need, especially the TPics. There were a couple different versions with the same specs. The Tpic6595N looks to be the same as the Tpic 6B595N with the exception of a couple of grounds aren't used everything else is the same. I put the sample order in and forwarded it to the rest of the guys involved with the project and they are putting the same order in. Woo Hoo, I just saved myself $150 (that's like $500 canadian [noparse]:)[/noparse]· I also have some 5v regulators coming for the slave boards and some more drivers and receivers and some other stuff.· I put in for the 3.3v regulators but they were out of stock. I think I only need one of those anyway for the master so that's not a big deal I can source one of those from something else I'm sure.
I also ordered some press N' Peel to do the master with.· Also since I saved the $150, I might use that to get the slaves ordered from that PCB site we were lookin at earlier.·
Working on the Master again, I added in a couple of power regulators. 3.3v and 5v. Do I need caps and resistors on those somewhere and if so, what size? They will be fed by 12v. I read I should have a resistor on the + side coming in to reduce the strain on the regulators. There shouldn't be a very big draw on them either way. If needed, I can probably run 7v into them instead if it makes a difference.
Also, McTrivia, on the propmod, what connections do I need to be concerned with coming into it.· The Vss's are gound, the P0-P? are the I/O lines.· is Vin, the +, what·are the Vbat, RST, RX/TX Clock out, SCL and the Prop Plug connections for?
Does the Clock out need to go to anything on my board? Kwinn had assigned pins 6-10 to be the data pins for the slaves, does that need to change? Finally made some progress on finishing up the CNC, was having issues with other pressing matters (Ie my lawn neeeeeeedddded to be mowed. it's about 2.5 acres so takes around 4 hours with the riding mower) Looks like I'm not going to have the budget to get the slaves professionally made in time. I was expecting a big commission check this month to pay for all of this stuff from a couple of big projects we finished in April, but for some reason my manager decided NOT to bill them last month. So needless to say the entire budget that was supposed to be for the rest of my fireworks chemicals and to get the boards printed isn't going to be here in time so now I need to try and budget everything down to the penny for what we need. *&*(**(..... But I digress.
Anyway, It looks like the 5v current should be around 500mA.
$1/sq in is pretty spendy per board.· These are 8"x11" boards i am building on so $88 is pretty pricy.· Going through 4pcb.com would run me about $375 for 10 boards with 1/2 off for new customers.
Ok, if there are no other changes·I will finalize the board and get it printed.· I know we aren't using a lot of the pins on the prop because of the 74hc595 chips, but I still put connections on the board for them if they are needed some other time, I can't imagine what for, but they are there. ·
Well, after a few hours of frustration and broken bits, I finally have the CNC doing some mass drilling.· I had it run through one of my little sub distribution boards that the actual igniters connect to and about 30 drill points into it, I realized that I had not etched the board square.· The CNC was drilling square with the edges of the board but the layout was tilted.· So after about 45 min of letting it run through. I compared the drilled holes to a transparency of the board and everything was very very close. Close enough for me anyway.· So Just as a trial run for timing and to make sure the CNC could run non stop accurately for an entire board, I loaded up the master console gcode and started drilling that board.· Notice from the pics it isn't etched yet. This was just an experimental board. I may try to see if I can transfer the image after it has been drilled, but from what I read, it should be transferred, then drilled.
I haven't heard from Kwinn for a while, are we complete on the layout of the master board before I start it tomorrow night?· I am hoping to have the Master transferred and drilled tomorrow night and start soldering it on Tuesday.
The drilling is surprisingly quick from what I though it might be. It is averaging about 5 sec per hole, including transit time between holes. I can probably knock it down to about 4 seconds by setting my clearance height.· I have all of the motor speeds turned way down to keep the accuracy as high as possible to prevent losing steps and I set it to go .2in above the board between drill points, so I can drop that down to about .1 or .05 and save a second per hole.· 4 seconds a hole seems pretty quick· but that is still about 9 hours of straight drilling for all of the boards.· Hmm, that's not that bad now that I think about it.· There is no way a person could drill a hole accuratly in 4 seconds so I'm good with it.
Anyway, let me know if there is anything else you guys can think of to add to the master before I finalize it tomorrow.
I'm still alive and following this thread, but mctrivia had always answered your questions before I read them so I spent my time working on the software. Should have the changes tested and the code cleaned up by Tuesday morning.
Well, the CNC made it through about 1/2 of the master board before it broke a bit, which was good enough to test the resolution. It looks like it got off by a few steps by the end so I may have to slow it down a bit more, but it's coming along.
The bits i'm using are the Carbide high speed bits, if you look at the pictures, you can see a couple of broken ones of the type. I have them in all the way up to the size/depth marker and the cutting part of the bit is only about .75in long. It's running on a 30,000rpm rotary tool so speed should be high. I think it may have been an error with the steppers, or the bit just got dull. It had drilled about 600 holes by that point. I know they are carbide, but I"m not sure what kind of life I can expect out of them.
Any good suggestions on how to make simple rotary encoder for my drive screws? I think I am losing steps in there somewhere and I can't figure out a good way to get them back except slow it down even more.
Making your own rotary encoder to mount on the stepper motor may be possible if the steps are not too small. The encoder disk needs 1 line for every two steps of the stepper motor, and you need to print 2 disks. One mounts on the motor shaft and a piece of the second mounts in front of the opto sensor in a manner that has both clear spots line up on one step, and a line and clear spot line up on the second step. What you get out from the opto sensor is a square wave that goes from low to high or high to low for each step.
An opto sensor for the "Home" position as mctrivia suggested is also a good idea. If the stepper misses steps the system could go to the home position and restart from there.
BTW - With the encoder there to tell you when the motor has reached the next position you could move from point to point faster since you could use the information to accelerate and decelerate the motor speed.
That's what a "Home" position sensor is for. Most equipment I have seen (xy table, pipetters, cnc's, etc) will have a sensor (switch or optical) at one end of each axis of motion and will drive each axis to that position at low speed when powered up or reset. Once that is done the positioning can be done at higher speeds.
kwinn said... ·The encoder disk needs 1 line for every two steps of the stepper motor, and you need to print 2 disks. One mounts on the motor shaft and a piece of the second mounts in front of the opto sensor in a manner that has both clear spots line up on one step, and a line and clear spot line up on the second step.
Ummmm. What is the point of the disk being mounted on the opto sensor?· wouldn't the Alternating clear and black on the other disk produce the same effect, or is this just to aim the sensor. ·
Prop Galore? What's that?·
Hope we got everything correct on the master console, it's being drilled right now. Broke one bit so far, not sure why, I have all the speeds very low to keep up the accuracy.· I think I might have some holes that have multiple drill points on them so it drills it, raises and drills it again. I'll have to watch for that on the next boards. ·
Re: "What is the point of the disk being mounted on the opto sensor?"
The field of view of most photo interrupters is wide enough to include several lines on an encoder disk, so the intensity of light/IR hitting the phototransistor does not vary much (Figure 1). By putting a matching pattern over the sensor (Figure 2) the amount of light/IR hitting the phototransistor can be reduced to almost zero. This can produce an inexpensive and relatively high resolution encoder. All that is required is the ability to move and align the photo interrupter with the encoder disk.
No worries man. Now it makes sense Kwinn on the encoder why you should have the matching pattern. Would a solid pattern on the open slot do the same thing or wouldn't it give as high of reading when it is open.
So if I understand the encoder correctly, on startup, the machine will go home to figure out where it is so the encoders have a reference point, once they have been set to home, every time the motor puts out a step, the software should be checking for the signal from the encoder to verify the step has been taken. I'm assuming that if the motor has put out 3 steps and the encoder only saw two of those, the system would tell it to take another step to catch up to where it should be. how simple of circuit do I need for that? The CNC software calls for two encoder inputs per motor. A port A and a port B. I wonder if I can get by with just one (i'm running out of pins on my parallel port quick). I would need another 6 pins if that's the case. one for each of the X axis drivers and one for Y. Z seems to be fine. It' only travelling a total of .25 inches each time and seems to be holding it's position fine. I also have the option of counts/unit so I'm guessing I don't actually need to have 100 marks on the encoder because they are 200 steps/rev. Maybe 50 would be fine, that would be 7.2 deg or 1/500th of an inch in theory. Hell, I'd be happy with 1/100th.
Comments
p4 goes through a 10k to a 10nF to ground. Between the 10k and 10nF you also have a 1uF going to the audio jack. Yes the extra pin goes to ground.
Your mixer ay or may not except stereo inputs. if not use 2 mono audio jacks.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
propmod_us are now in stock. propmod_1x1 arrive on 26th. Only $30
Need to upload large images or movies for use in the forum. you can do so at uploader.propmodule.com for free.
I also ordered some press N' Peel to do the master with.· Also since I saved the $150, I might use that to get the slaves ordered from that PCB site we were lookin at earlier.·
Post Edited (chaosgk) : 5/27/2009 6:08:43 PM GMT
Also, McTrivia, on the propmod, what connections do I need to be concerned with coming into it.· The Vss's are gound, the P0-P? are the I/O lines.· is Vin, the +, what·are the Vbat, RST, RX/TX Clock out, SCL and the Prop Plug connections for?
Post Edited (chaosgk) : 5/29/2009 2:47:01 AM GMT
R=3/max current
P=10*max current
Caps are not as important with a battery for input.
10uF min
100uF or more is better
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
propmod_us are now in stock. propmod_1x1 arrive on 26th. Only $30
Need to upload large images or movies for use in the forum. you can do so at uploader.propmodule.com for free.
Post Edited (mctrivia) : 5/29/2009 3:21:08 AM GMT
Clock out is P29
P? Is io lines
Gnd is vss
Vin bottom right not used
Top right is vdd
Prop plug is where you plug programmer in look at assembled pics on my site
On phone so I can't see pic did I miss anything
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
propmod_us are now in stock. propmod_1x1 arrive on 26th. Only $30
Need to upload large images or movies for use in the forum. you can do so at uploader.propmodule.com for free.
Anyway, It looks like the 5v current should be around 500mA.
His schematic was made considering this board
I can get your boards made for $1/sq inch if you can output in gerber
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
propmod_us are now in stock. propmod_1x1 arrive on 26th. Only $30
Need to upload large images or movies for use in the forum. you can do so at uploader.propmodule.com for free.
Ok, if there are no other changes·I will finalize the board and get it printed.· I know we aren't using a lot of the pins on the prop because of the 74hc595 chips, but I still put connections on the board for them if they are needed some other time, I can't imagine what for, but they are there.
·
I haven't heard from Kwinn for a while, are we complete on the layout of the master board before I start it tomorrow night?· I am hoping to have the Master transferred and drilled tomorrow night and start soldering it on Tuesday.
The drilling is surprisingly quick from what I though it might be. It is averaging about 5 sec per hole, including transit time between holes. I can probably knock it down to about 4 seconds by setting my clearance height.· I have all of the motor speeds turned way down to keep the accuracy as high as possible to prevent losing steps and I set it to go .2in above the board between drill points, so I can drop that down to about .1 or .05 and save a second per hole.· 4 seconds a hole seems pretty quick· but that is still about 9 hours of straight drilling for all of the boards.· Hmm, that's not that bad now that I think about it.· There is no way a person could drill a hole accuratly in 4 seconds so I'm good with it.
Anyway, let me know if there is anything else you guys can think of to add to the master before I finalize it tomorrow.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
propmod_us are now in stock. propmod_1x1 arrive on 26th. Only $30
Need to upload large images or movies for use in the forum. you can do so at uploader.propmodule.com for free.
1) Keep as little of the drill bit out of chuck as possible.
2) Keep your chuck speed high. I used 35,000 RPM
I was using a drill press though and not a CNC but hopefully that is helpful.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
propmod_us and propmod_1x1 are now in stock. Only $30. PCB available for $5
Need to upload large images or movies for use in the forum. you can do so at uploader.propmodule.com for free.
Any good suggestions on how to make simple rotary encoder for my drive screws? I think I am losing steps in there somewhere and I can't figure out a good way to get them back except slow it down even more.
Post Edited (chaosgk) : 6/1/2009 3:47:16 PM GMT
For my lego scanner I used a switch it touched after
Each pass.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
propmod_us and propmod_1x1 are now in stock. Only $30. PCB available for $5
Need to upload large images or movies for use in the forum. you can do so at uploader.propmodule.com for free.
An opto sensor for the "Home" position as mctrivia suggested is also a good idea. If the stepper misses steps the system could go to the home position and restart from there.
Post Edited (kwinn) : 6/1/2009 5:27:43 PM GMT
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
propmod_us and propmod_1x1 are now in stock. Only $30. PCB available for $5
Need to upload large images or movies for use in the forum. you can do so at uploader.propmodule.com for free.
·
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
propmod_us and propmod_1x1 are now in stock. Only $30. PCB available for $5
Need to upload large images or movies for use in the forum. you can do so at uploader.propmodule.com for free.
Hope we got everything correct on the master console, it's being drilled right now. Broke one bit so far, not sure why, I have all the speeds very low to keep up the accuracy.· I think I might have some holes that have multiple drill points on them so it drills it, raises and drills it again. I'll have to watch for that on the next boards.
·
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
propmod_us and propmod_1x1 are now in stock. Only $30. PCB available for $5
Need to upload large images or movies for use in the forum. you can do so at uploader.propmodule.com for free.
The field of view of most photo interrupters is wide enough to include several lines on an encoder disk, so the intensity of light/IR hitting the phototransistor does not vary much (Figure 1). By putting a matching pattern over the sensor (Figure 2) the amount of light/IR hitting the phototransistor can be reduced to almost zero. This can produce an inexpensive and relatively high resolution encoder. All that is required is the ability to move and align the photo interrupter with the encoder disk.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
propmod_us and propmod_1x1 are now in stock. Only $30. PCB available for $5
Need to upload large images or movies for use in the forum. you can do so at uploader.propmodule.com for free.
So if I understand the encoder correctly, on startup, the machine will go home to figure out where it is so the encoders have a reference point, once they have been set to home, every time the motor puts out a step, the software should be checking for the signal from the encoder to verify the step has been taken. I'm assuming that if the motor has put out 3 steps and the encoder only saw two of those, the system would tell it to take another step to catch up to where it should be. how simple of circuit do I need for that? The CNC software calls for two encoder inputs per motor. A port A and a port B. I wonder if I can get by with just one (i'm running out of pins on my parallel port quick). I would need another 6 pins if that's the case. one for each of the X axis drivers and one for Y. Z seems to be fine. It' only travelling a total of .25 inches each time and seems to be holding it's position fine. I also have the option of counts/unit so I'm guessing I don't actually need to have 100 marks on the encoder because they are 200 steps/rev. Maybe 50 would be fine, that would be 7.2 deg or 1/500th of an inch in theory. Hell, I'd be happy with 1/100th.