I took fuse cord for what you meant. And we did try that for the first few years, but it can't be timed (reliably) and would be more expensive than the relay system we are working out.
2- TPIC6595
2- 0.1uF bypass capacitors
1- 100uF capacitor
1- 5V regulator
1- diode
16- relays (or 64 relays if you want one relay per output)
2- RJ 45 female connectors
1000- feet of cat5 cable (hopefully with connectors on)
64- 2 wire connectors to connect individual igniters
What type of master console do you plan on using to control the slaves? Are you going to want to manually control the firing or set up an automatic timed sequence?
I would suggest a propeller protoboard or propstick USB mainly because I am familiar with the prop. It is more than capable of handling this, but so would most of the other microcontroller modules.
Hey Kwinn, before someone stated that the tpics could be linked, correct? If so how? I have no idea what to use for the master board if the USB propeller Proto board will do it, the price is right at $39. ·The firing will be on a timed sequence. Do these Tpic6595 chips need to be programmed with a pic programmer and then inserted in the board, or do they get their signals directly from the serial or usb port?
I started drawing up the PCB for the relay matrix and holy Smile is that getting complex just with where all of the traces have to go, they go from the bottom to the top in 14 different places so far and that is only to get the relays connected to the tpics or Mosfets, or whatever they connect to on the coil.
If the Tpics can be linked, it may be less complicated in the long run to simply use 8 tpics per slave if they can be linked in some simple way and still able to turn on each output of the tpic to activate the relay. I'm really glad I have 5 more months until this has to do its job. I'm thinking I should have started this about July 5th of last year. If I need 100 tpic6595 chips, I will go that route, it's only another $100 and I think it will save me many many hours of work and probably be more reliable anyway. How does each output·get addressed in the programming when there is more then one PIC?
In your list you have 2 RJ45 females, what do you have planned for those exactly? Are they the link back to the master. The other question is if all of the programming gets stored on the chips, is there enough memory onboard to handle 400 different on and off commands for all of the relays (possibly more).· I may also look at getting the slave boards printed at a company if it isn't too much money per board.· Otherwise I may be getting my CNC router·built sooner then I thought to do all of the drilling 4000+ holes for me.
Each slave board will have a 12 port RJ45 panel on it that the relays connect to and the ignition wires will connect to the RJ45 jacks through another distribution box at the pyro racks.
500 relays for $100? not bad....would it be important to know if they have the protection diode built-in? I'd guess so....do you have a part number for them?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ <FONT>Steve
What's the best thing to do in a lightning storm? "take a one iron out the bag and hold it straight up above your head, even God cant hit a one iron!"
Lee Travino after the second time being hit by lightning!
Those 6595 chips are basically 74HC595 serial-parallel shift registers with high drive capability on the outputs, and use an SPI protocol for controlling them.
Leon
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM
Suzuki SV1000S motorcycle
So each of the 6595 chips can be independently controlled in the code right? Do they need to be programmed for a certain address before they are put in the boards?
No, the tpics do not need to be programmed with a pic programmer. They will need to get signals from the master console while the show is in progress. You will need a PC (windows) to program the master console (prop protoboard) using a USB connection unless you build a very fancy (expensive) master console. A laptop would be best so I hope you have one.
Yes, there is a lot of wiring, and it probably would be simpler to use 8 tpics per slave and drive each relay directly. It certainly is more flexible.
The 2 RJ45 female jacks are for the link between the master console and the slave. It may be necessary to have line drivers on the master and line receivers on the slaves depending on the length of the cat5 cables. Putting them on the circuit board with bypass jumpers is a good idea. Sending data will require 3 pairs of the 4 in the cable so you may have to provide power over a separate pair of wires. This may be a good idea anyway due to noise considerations. Standard lamp cord would be fine.
I have attached a block diagram of the setup based on the above and will follow that with details of connections for the slaves shortly.
The software for the prop to control the tpics is available, and stepping through a sequence of "time, pin to activate" entries in a table is easy enough to write in spin, but I do hope you can write the software (or know someone who can) to create the sequence portion of this. I do not think I will have enough spare time for that.
Perfect.· Now I know what i need to order and will get to it in the next week or so, I have some projects that NEED to get done before I really jump into this with all of my spare time. I have guys that know programming so I will have them help with the software for this if I can't figure it out.
Here is my parts list, let me know if I am missing something important.
To build one slave you will need the following:
100 - TPIC6595 (actually cheaper to buy 100 then it is 75)
16- 0.1uF bypass capacitors
8- 100uF capacitor
8- 5V regulator
8- diode
500- relays
10 - 12 port RJ45 panels
8 - single RJ45 jacks
1 - USB propeller proto board
8 - Cat 5 cables
Let me know if I missed anything important.· I may throw LEDS in there for contenuity testing, but I already have the design for those coming off the relays in the NC circiut
With those proto boards, is there a way to put a pause in there that would be manually controlled·from a switch?· For example if we have a grass fire during the show and I want to pause the program. Can i flip a switch and have the program pause then continue when I turn it on or off again?· And yes, we do have minor grass fires every year but we are alway prepared for them and the grass is short anyway.· Last year I just let the show go on while I ran and put them out.
Are the Tpics linked simply by the ser-in and ser-out and all of the other connections are linked to the main power and grounds on the slave boards?
Here is the basic schematic for the slave unit. It is based on using 8 TPIC's and 64 relays per slave as mentioned in the last post.
I have not specified the line driver yet since I have not had time to look for one. Being a network engineer you may have access to them. Each slave will need a quad differential line receiver, and the master will need a quad differential line driver for each slave.
I have put a switch on the IG+ line for safety. The state of the outputs from the TPIC's may be on or off on power up so the IG+ switch should always be off before powering up. For safety you may want to make it a self latching relay so it will automatically turn off if power is lost and require a button press on the slave to activate it.
I would also recommend that you put a 1K2 resistor and LED across each relay coil so you can see if a relay is energized before wiring up the igniters and turning the IG+ on. This will also help with debugging the hardware since the LED's will indicate what relays are on. Less work and expense than hooking up fireworks. The software to send the data to the slaves is already available from the OBEX on the Parallax web site. The program to present the data for transmission will need to be written, and most likely in spin. The prop has 32K of Hub ram so there should be no problem accomodating the program and sequence data. What I picture is having a table with the time from the start of the program, the slave to send the data to, and 64 bits of data for the relays. The sequence would be:
On power up or Reset button is pressed
output all 0's to all slaves
wait for start button
Start button is pressed:
time is set to 0
read sequence
if sequence is stop exit program
wait until time is equal or greater than sequence time
output sequence data
goto read sequence
Each command would occupy 3 32 bit longs in the prop ram and would be generated by a program on the laptop.
PS - You do know that id does not have to be 64 relays per slave. It could be more or less. Multiples of 8 make the best use of the chips and the software easier. It could be done with 9x48, 6x72 or 10x40 if that is more convenient.
The schematic for the prop line drivers didn't load, can you post it again.
Kwinn, where are you from? I feel like I need to take you out for a beer or something.
I am aware there doesn't need to be 64 channels on each one, but from a production standpoint, 64/board is a good number and it's easier to do 8 boards exactly the same than 8 that are all different and that will give me over 500 cue points total.
Thanks to everyone again for all of your help on this you have no idea how much I appreciate it.
Maybe I just forgot to attach it. I live just outside Toronto Canada. You?
Don't forget to put a 0.1uF bypass capacitor on each chip. Are you going to make pc boards for this?
I am near Minot, North Dakota. are quad drivers and receivers universal? I found some cheap ones on ebay and If they will work I'll order 8 sets of the drivers and receivers.
I'll be making all of the PC boards myself unless I can find some place to print them for me cheap (doubt it).
We have a big LAN party this weekend I am hosting so I will be busy for the next few days getting my garage cleaned out again for it and get the heaters cranking so I won't be doing any work on this till next week now.
I've also got a Model RC hobby that is in need of some attention around here too.
When I get some PCB's drawn up, I'll post them here and let all of you check my work to make sure when I turn the system on, I don't get 500 devices triggering at the same time. Boom, Big Badda Boom.
Well, North Dakota is a bit far to go for a beer, but if I am ever heading that way I'll let you know. As long as it is a quad RS422/485 line driver it should be fine. Anyway just keep in mind that that is not a detailed schematic. Nicetie like 0.1uF bypass capacitors and such have been left off the drawing but should be placed near the chips. Also the LED in series with a 1K2 resistor across each relay coil is a good idea. I would put them on the board and house it in a case with a transparent cover so you can see the status of all the relays before connecting anything that goes boom.
Just so I'm clear on this. When the boards get powered up, do they run their programs as soon as they get power, or do they all wait for the commands from the computer?· If they all wait for the commands from the computer (which is what I hope for) I was planning on using key lock switches on the +24v side coming into the board.· Each relay will have 2 LEDSs assigned to it. One green one on the NC side to check contenuity of the ignitor and one red one on the NO side to make sure the relay isn't in the·fire position at startup time.· I'll also have to build some test boards to simulate the ignitors for timing.
I recently built a project where I used 4 - 74165 shift registers and it worked great.· I was reading this post and thought I would also attach my file so you can use it as a reference also.· In the past I have actually used up to 8 registers without much problem so your system should work great.
If I may offer a suggestion...· I would·add a remote monitoring circuit that would display the status the of firing units from the main controller.· This way you can see if all the ignitors, power, and communications are working before pressing the start button.· This shouldn't be too hard as you will have extra pairs of cables in the cat5 and you can use steering diodes off the relays NC connection to make sure each unit is ready to fire...
I am from down by Sioux Falls, SD so it might be time for a field-trip... Always wanted to build something like this...
The slave boards do not run anything when they power up. They are completely controlled by the master console. The problem is that the flip-flops in the 8 chips can come up either high or low at random so they have to be reset to low by the master console. That is why I put the switch in the circuit. A key lock switch is even better.
The LED's on the relay are a good idea. How are they powered? Can you send me a schematic of the connections?
The test board could be one TPIC with 8 relays. In fact that could be the building block for all the slaves. Put 8 boards in a slave console.
Just to be sure we are on the same page, you are aware that the control signals from the master console needs all 4 pairs (all 8 wires) of the cat5 cable?
To give me a better idea of how everything connects together and the distances involved can you fill in the info in the diagram below and correct any misconceptions I may have.
The more I thought about my last post and the 8 relays to a board idea the more sense it makes. Easier to lay out, easier to build, easier to troubleshoot when you have problems. Attached is my quick thought on layout.
P1 and P2 are outputs from the relays to RJ45 connectors.
P3 is the input from the master console to the line receiver. Only one board per slave would have the line reciever installed. the rest would be daisy chained using P4/P5.
P4 and P5 are for daisy chaining the control signals to all the slave boards.
Ok.· Based on the highly advanced graphic I attached to this document,· (I used MS paint cause I forgot to get visio from the office again today).· Attached to each slave board is a 12 port RJ45 panel.· Each of the relays from the 6595's get connected to 7 of the wires in the 8 wire cat5 cable.· #8 gets used as a common ground for that line.· The RJ45 cables get connected to the ignition distribution panels that have varying numbers of RJ45 jacks.· The Distribution panels split the 7 hot wires and the ground back into a distribution terminal where the actual leads from the igniters connect to.
I know I am losing one ignitor per tpic connecting them this way, but for ease of design, each RJ45 will have one ground and 7 ignition points. Because there are 8 expected tpics per slave, I will route the extra relay from each tpic to another rj45 jack and use this as a spare or extra set of lines if needed.· I have 12 port RJ45 jacks specified because I got a set of 10 for $20 on ebay.· I figured the price was right and that gave me more than enough.· I will probably buy single PCB mount RJ45 jacks for the distribution boards or use some of the 100+ RJ45 jacks I have left over here.
I was thinking the slaves only needed three of the pairs, but using 4 makes no difference anyway.· The 24v power lines coming from the master console are seperate feeds anyway like we talked about for noise reasons, and I want higher gauge wire anyway for those since there is some current on those.
I will draw up a schematic of how the Leds are wired at work tomorrow, but basically,· (+24v
resistor
green led
line to RJ45 jack)· the LED·circuit·will bypass the relay.· It will be the same on the opposite side of the relay, the one that gets switched on.· but when that one activates, it will also send power to a·red·LED.·The red led will either bypass the out loop through the RJ45 or it will run through the RJ45 to test the jack.· Probably just bypass it.·
The reason for building the green led in line is to check the contenuity of the igniters when everything is hooked up and ready.· The reason for the Red led is for programming and testing firing sequences.·
This way when we run the program, we will see LEDs light up as the relays trigger, letting us know what sequence they are firing in and that the relay is actually activating.
I understand now about the 6595's possibly being in an on state when they get power and that they recieve all of their commands from the main board. I'm still not quite clear on the main board program.
Does the propeller chip run it's program when it gets power, or is the PC constantly sending it data and it is sending that data to the correct PICS?· Anyone know if I can have some type of input from a switch that will pause the program if a button is pushed.· Pause is better then killing the system.· I would like the pause as a function to be able to hold the program and continue it when the button is let up.· I will of course have a master switch at all times that kills all 24v power to the relays.
by the way, if anyone knows a source of cheap 4", 5", or 6" tubes, please let me know. We have plenty of 3" tubes (400 1/2" thick cardboard tubes newspapers are printed from) Whatever the suggestions are they have to be extremely strong to be able to hold up to the forces of the shells lifting off. we currently only have about 6 4" tubes and last year, we used 6" PVC buried in the ground all the way to launch our larger shells. The large shells lifted off fine, but one of my guys suggested we put an extra ounce of lift powder in to make sure we had enough height and everyone of the tubes were destroyed. We already knew that PVC was a no no, but didn't have anything else at the time. All of the larger stuff is at least 300 yards from anyone and this yearwill be pushed to around 400 yards. The 4" tubes are currently steel high pressure pipe and those are buried in the ground as well. It looks like we will have a couple of 8" tubes this year too [noparse]:)[/noparse] for the finale.
Don't forget, we are still amatures with limited budgets but what we lack in money we make up for in performance.
I think we are the only ones in our area that will be putting on a show to music (hopefully). Everyone is more then welcome to come up (or down if you're in Canada).
I guess that makes mine a highly advanced graphic as well since I use paint for block diagrams as well. In any case it does clear up most of my questions. Here is another highly advanced graphic that will help to explain how the relays get activated from the prop and my idea of how the software would work. To start with, the prop is a 32 bit chip, and a long is a 32 bit memory location.
Time in the system will be measured in seconds (unless you need finer resolution) from the time the start button is pressed (T0).
Pressing the pause button will stop the timer at that point. Pressing pause again will start it again.
A command or instruction for the master console will be 3 longs.
The first long will contain the time in seconds (28 bits so up to 268,435,456 seconds) from T0 to execute the command, and the slave number to send the command to (4 bits so 1 of 16 slaves)
The second and third longs will contain the bit pattern to send out. One bit for each relay.
There are 4 lines from the master console that control each slave, SRCLR clears the data in the shift register, RCK clocks the data from the shift register to the storage register, SER IN is the serial data from the prop, SRCK is the clock signal that shifts the data bit in.
The sequence of events to shift a command in would be, output the bit, pulse the clock high and repeat 64 times. At the end of 64 shift pulses the first bit sent out will be in the TPIC on board 1 and the RCK would be pulsed high to transfer the bits to the output register. If a bit is 1 the relay activates and something goes boom.
Well that answers a bunch of questions. I will be ordering a propeller board next week to do some experimenting with and some of the other parts are on the way.· I got a auction of 2000 green LEDs on ebay for $20 so·I will have a lifetime supply of those.··Do you know if these drivers would work http://datasheet.digchip.com/477/477-07034-0-MC3486P.pdf·instead of the ones you suggested?· I'm looking for a cheap source of those as well.
Comments
2- TPIC6595
2- 0.1uF bypass capacitors
1- 100uF capacitor
1- 5V regulator
1- diode
16- relays (or 64 relays if you want one relay per output)
2- RJ 45 female connectors
1000- feet of cat5 cable (hopefully with connectors on)
64- 2 wire connectors to connect individual igniters
What type of master console do you plan on using to control the slaves? Are you going to want to manually control the firing or set up an automatic timed sequence?
I would suggest a propeller protoboard or propstick USB mainly because I am familiar with the prop. It is more than capable of handling this, but so would most of the other microcontroller modules.
I started drawing up the PCB for the relay matrix and holy Smile is that getting complex just with where all of the traces have to go, they go from the bottom to the top in 14 different places so far and that is only to get the relays connected to the tpics or Mosfets, or whatever they connect to on the coil.
If the Tpics can be linked, it may be less complicated in the long run to simply use 8 tpics per slave if they can be linked in some simple way and still able to turn on each output of the tpic to activate the relay. I'm really glad I have 5 more months until this has to do its job. I'm thinking I should have started this about July 5th of last year. If I need 100 tpic6595 chips, I will go that route, it's only another $100 and I think it will save me many many hours of work and probably be more reliable anyway. How does each output·get addressed in the programming when there is more then one PIC?
In your list you have 2 RJ45 females, what do you have planned for those exactly? Are they the link back to the master. The other question is if all of the programming gets stored on the chips, is there enough memory onboard to handle 400 different on and off commands for all of the relays (possibly more).· I may also look at getting the slave boards printed at a company if it isn't too much money per board.· Otherwise I may be getting my CNC router·built sooner then I thought to do all of the drilling 4000+ holes for me.
Each slave board will have a 12 port RJ45 panel on it that the relays connect to and the ignition wires will connect to the RJ45 jacks through another distribution box at the pyro racks.
Post Edited (chaosgk) : 2/3/2009 5:00:58 AM GMT
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
<FONT>Steve
What's the best thing to do in a lightning storm? "take a one iron out the bag and hold it straight up above your head, even God cant hit a one iron!"
Lee Travino after the second time being hit by lightning!
Leon
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM
Suzuki SV1000S motorcycle
Manufacturer: Tyco
MFG Item #: TSC-106D3H
Contacts Rating: 1 Amp 120VAC/24VDC
Coil Rating: 6VDC
So each of the 6595 chips can be independently controlled in the code right? Do they need to be programmed for a certain address before they are put in the boards?
Leon
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM
Suzuki SV1000S motorcycle
Yes, the tpics can be linked.
No, the tpics do not need to be programmed with a pic programmer. They will need to get signals from the master console while the show is in progress. You will need a PC (windows) to program the master console (prop protoboard) using a USB connection unless you build a very fancy (expensive) master console. A laptop would be best so I hope you have one.
Yes, there is a lot of wiring, and it probably would be simpler to use 8 tpics per slave and drive each relay directly. It certainly is more flexible.
The 2 RJ45 female jacks are for the link between the master console and the slave. It may be necessary to have line drivers on the master and line receivers on the slaves depending on the length of the cat5 cables. Putting them on the circuit board with bypass jumpers is a good idea. Sending data will require 3 pairs of the 4 in the cable so you may have to provide power over a separate pair of wires. This may be a good idea anyway due to noise considerations. Standard lamp cord would be fine.
I have attached a block diagram of the setup based on the above and will follow that with details of connections for the slaves shortly.
The software for the prop to control the tpics is available, and stepping through a sequence of "time, pin to activate" entries in a table is easy enough to write in spin, but I do hope you can write the software (or know someone who can) to create the sequence portion of this. I do not think I will have enough spare time for that.
Here is my parts list, let me know if I am missing something important.
To build one slave you will need the following:
100 - TPIC6595 (actually cheaper to buy 100 then it is 75)
16- 0.1uF bypass capacitors
8- 100uF capacitor
8- 5V regulator
8- diode
500- relays
10 - 12 port RJ45 panels
8 - single RJ45 jacks
1 - USB propeller proto board
8 - Cat 5 cables
Let me know if I missed anything important.· I may throw LEDS in there for contenuity testing, but I already have the design for those coming off the relays in the NC circiut
With those proto boards, is there a way to put a pause in there that would be manually controlled·from a switch?· For example if we have a grass fire during the show and I want to pause the program. Can i flip a switch and have the program pause then continue when I turn it on or off again?· And yes, we do have minor grass fires every year but we are alway prepared for them and the grass is short anyway.· Last year I just let the show go on while I ran and put them out.
Are the Tpics linked simply by the ser-in and ser-out and all of the other connections are linked to the main power and grounds on the slave boards?
·
Leon
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM
Suzuki SV1000S motorcycle
Leon
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM
Suzuki SV1000S motorcycle
Here is the basic schematic for the slave unit. It is based on using 8 TPIC's and 64 relays per slave as mentioned in the last post.
I have not specified the line driver yet since I have not had time to look for one. Being a network engineer you may have access to them. Each slave will need a quad differential line receiver, and the master will need a quad differential line driver for each slave.
I have put a switch on the IG+ line for safety. The state of the outputs from the TPIC's may be on or off on power up so the IG+ switch should always be off before powering up. For safety you may want to make it a self latching relay so it will automatically turn off if power is lost and require a button press on the slave to activate it.
I would also recommend that you put a 1K2 resistor and LED across each relay coil so you can see if a relay is energized before wiring up the igniters and turning the IG+ on. This will also help with debugging the hardware since the LED's will indicate what relays are on. Less work and expense than hooking up fireworks. The software to send the data to the slaves is already available from the OBEX on the Parallax web site. The program to present the data for transmission will need to be written, and most likely in spin. The prop has 32K of Hub ram so there should be no problem accomodating the program and sequence data. What I picture is having a table with the time from the start of the program, the slave to send the data to, and 64 bits of data for the relays. The sequence would be:
On power up or Reset button is pressed
output all 0's to all slaves
wait for start button
Start button is pressed:
time is set to 0
read sequence
if sequence is stop exit program
wait until time is equal or greater than sequence time
output sequence data
goto read sequence
Each command would occupy 3 32 bit longs in the prop ram and would be generated by a program on the laptop.
Kwinn, where are you from? I feel like I need to take you out for a beer or something.
I am aware there doesn't need to be 64 channels on each one, but from a production standpoint, 64/board is a good number and it's easier to do 8 boards exactly the same than 8 that are all different and that will give me over 500 cue points total.
Thanks to everyone again for all of your help on this you have no idea how much I appreciate it.
Don't forget to put a 0.1uF bypass capacitor on each chip. Are you going to make pc boards for this?
I'll be making all of the PC boards myself unless I can find some place to print them for me cheap (doubt it).
We have a big LAN party this weekend I am hosting so I will be busy for the next few days getting my garage cleaned out again for it and get the heaters cranking so I won't be doing any work on this till next week now.
I've also got a Model RC hobby that is in need of some attention around here too.
When I get some PCB's drawn up, I'll post them here and let all of you check my work to make sure when I turn the system on, I don't get 500 devices triggering at the same time. Boom, Big Badda Boom.
If I may offer a suggestion...· I would·add a remote monitoring circuit that would display the status the of firing units from the main controller.· This way you can see if all the ignitors, power, and communications are working before pressing the start button.· This shouldn't be too hard as you will have extra pairs of cables in the cat5 and you can use steering diodes off the relays NC connection to make sure each unit is ready to fire...
I am from down by Sioux Falls, SD so it might be time for a field-trip... Always wanted to build something like this...
The LED's on the relay are a good idea. How are they powered? Can you send me a schematic of the connections?
The test board could be one TPIC with 8 relays. In fact that could be the building block for all the slaves. Put 8 boards in a slave console.
Just to be sure we are on the same page, you are aware that the control signals from the master console needs all 4 pairs (all 8 wires) of the cat5 cable?
To give me a better idea of how everything connects together and the distances involved can you fill in the info in the diagram below and correct any misconceptions I may have.
Master Console (1)
(8 cat5 @ xxx feet long)
Slave consoles (8)
(8x16 cat5 @ xxx feet long)
(***??????***) what's here?
(2 wires to igniter)
Do individual pairs go from the slaves to the igniters or is there an intermediate connector board/box with 8 terminals?
P1 and P2 are outputs from the relays to RJ45 connectors.
P3 is the input from the master console to the line receiver. Only one board per slave would have the line reciever installed. the rest would be daisy chained using P4/P5.
P4 and P5 are for daisy chaining the control signals to all the slave boards.
(8 cat5 @·100-200·feet long)
Slave consoles (8)
(8x16 cat5 @·30-300 feet long)
(***Ignition Distribution boards***) what's here?
(2 wires to igniter)
Ok.· Based on the highly advanced graphic I attached to this document,· (I used MS paint cause I forgot to get visio from the office again today).· Attached to each slave board is a 12 port RJ45 panel.· Each of the relays from the 6595's get connected to 7 of the wires in the 8 wire cat5 cable.· #8 gets used as a common ground for that line.· The RJ45 cables get connected to the ignition distribution panels that have varying numbers of RJ45 jacks.· The Distribution panels split the 7 hot wires and the ground back into a distribution terminal where the actual leads from the igniters connect to.
I know I am losing one ignitor per tpic connecting them this way, but for ease of design, each RJ45 will have one ground and 7 ignition points. Because there are 8 expected tpics per slave, I will route the extra relay from each tpic to another rj45 jack and use this as a spare or extra set of lines if needed.· I have 12 port RJ45 jacks specified because I got a set of 10 for $20 on ebay.· I figured the price was right and that gave me more than enough.· I will probably buy single PCB mount RJ45 jacks for the distribution boards or use some of the 100+ RJ45 jacks I have left over here.
I was thinking the slaves only needed three of the pairs, but using 4 makes no difference anyway.· The 24v power lines coming from the master console are seperate feeds anyway like we talked about for noise reasons, and I want higher gauge wire anyway for those since there is some current on those.
I will draw up a schematic of how the Leds are wired at work tomorrow, but basically,· (+24v
resistor
green led
line to RJ45 jack)· the LED·circuit·will bypass the relay.· It will be the same on the opposite side of the relay, the one that gets switched on.· but when that one activates, it will also send power to a·red·LED.·The red led will either bypass the out loop through the RJ45 or it will run through the RJ45 to test the jack.· Probably just bypass it.·
The reason for building the green led in line is to check the contenuity of the igniters when everything is hooked up and ready.· The reason for the Red led is for programming and testing firing sequences.·
This way when we run the program, we will see LEDs light up as the relays trigger, letting us know what sequence they are firing in and that the relay is actually activating.
I understand now about the 6595's possibly being in an on state when they get power and that they recieve all of their commands from the main board. I'm still not quite clear on the main board program.
Does the propeller chip run it's program when it gets power, or is the PC constantly sending it data and it is sending that data to the correct PICS?· Anyone know if I can have some type of input from a switch that will pause the program if a button is pushed.· Pause is better then killing the system.· I would like the pause as a function to be able to hold the program and continue it when the button is let up.· I will of course have a master switch at all times that kills all 24v power to the relays.
Don't forget, we are still amatures with limited budgets but what we lack in money we make up for in performance.
I think we are the only ones in our area that will be putting on a show to music (hopefully). Everyone is more then welcome to come up (or down if you're in Canada).
Post Edited (chaosgk) : 2/5/2009 4:24:28 PM GMT
I guess that makes mine a highly advanced graphic as well since I use paint for block diagrams as well. In any case it does clear up most of my questions. Here is another highly advanced graphic that will help to explain how the relays get activated from the prop and my idea of how the software would work. To start with, the prop is a 32 bit chip, and a long is a 32 bit memory location.
Time in the system will be measured in seconds (unless you need finer resolution) from the time the start button is pressed (T0).
Pressing the pause button will stop the timer at that point. Pressing pause again will start it again.
A command or instruction for the master console will be 3 longs.
The first long will contain the time in seconds (28 bits so up to 268,435,456 seconds) from T0 to execute the command, and the slave number to send the command to (4 bits so 1 of 16 slaves)
The second and third longs will contain the bit pattern to send out. One bit for each relay.
There are 4 lines from the master console that control each slave, SRCLR clears the data in the shift register, RCK clocks the data from the shift register to the storage register, SER IN is the serial data from the prop, SRCK is the clock signal that shifts the data bit in.
The sequence of events to shift a command in would be, output the bit, pulse the clock high and repeat 64 times. At the end of 64 shift pulses the first bit sent out will be in the TPIC on board 1 and the RCK would be pulsed high to transfer the bits to the output register. If a bit is 1 the relay activates and something goes boom.
·
Just Kiddn.· I found them on Ebay.· Here is the link to his store.
http://stores.ebay.com/Electronic-Components-Wholesale
·