The trap you guys are falling into is called "jumping to implementation before design". The suggestion to throw away the existing board elicited immediate reactive responses to pick a new one. Be proactive on the design instead. Do not propose implementation solutions before deciding what the device is even going to do, for simple example, what LED's it is using to show heating/cooling cycles. Jumping immediately to "OK let's pick a board! Let's pick a chip!" (and "Let's use arduino!") is a classic type mistake. Compare designs of peers. Look at the strengths and goals of the designs. I already mentioned that there is not a "one size fits all" with embedded boards so attempting to find one is not likely to reach a conclusion. [Note Simon I would suggest if you're giving away boards to random people, maybe try handing out Teensy 3.x's or Teensy LC's instead of the nano's.]
There will almost always be a custom board in an embedded project. Including a thermocycler. It does not have to be made in CAD or etched by a fab. It does not have to be complex. The power relays are mounted to something and wired to something, that something is a board. The temp probe is going to be wired to something. The user LED's are wired to something. Don't make a rat's nest of jumper wires. These should be neat and connected on their own board. It could be a pre-drilled copper bus board. It could be a custom etched board with traces drawn with a ruler and a sharpie permanent marker. These are possibilities. In discussing design (which is what I have been attempting to start here) compare the competing designs and list the likely possibilities. If you go with a kit board vs components, then immediately what comes to mind is: does this board have a prototyping area to add components to it, or what components is that board missing that will need to be put on a second board anyway. No, the arduino boards do not have prototyping areas. What I didn't blatantly say, and I guess I should have, and will now, is: start out with the idea of creating the custom board so the form factor will fit in the space you want with the mounting options you want, and ideally everything will be on one PCB. [Note: The small Arduino boards do not have mounting holes. The only way to mount them is to use it's header pins.]
I don't know where this "electronic novices can't do these complex things" perspective comes from. Is it any different than crufty liberal arts school teachers saying, "Ohh, watch out for calculus, you don't want to try that, it's tooooo hard for you!!" I was a novice once. I took a piece of 1/8" thick waste acrylic out to a workshop and drilled 0.1" holes in it for an IC and components to fit into because I couldn't afford to purchase an FR4 copper clad board, then I soldered the IC and components together point-to-point [note, my holes were incredibly off-center but I bent the pins to fit and it worked fine]. That was probably my first "board". I didn't have a schematic, I drew one on a spare piece of graph paper based on the circuit I saw in the datasheet and a project idea I read in a magazine. I didn't have an electronics soldering iron, I used an automobile-grade type, I didn't have access to the chemicals for etching. I think I was 12 at the time. I had no guidance. I didn't have instruction from school or a mentor. I didn't have youtube videos to watch as a guide. I didn't have new components, I ripped them out of an old radio and tape player. I didn't have datasheets for anything in the world available via a mouse click. I didn't have a digital logic probe or O-scope. Look at how much easier CAD and fab technology and component procurement has become since then and all the examples and videos online! It is much easier now if anything. It is not difficult even for a "novice". Do the homework. RTFM. End the "must spoon feed them" mentality. It doesn't take "excessively long" and you don't "need to get it done right now". Any biology protocol will take longer to incubate than to design and build the electronics. It's all good experience and not wasted. BTW that circuit worked, I still have it, I found it in a move a while back, quite a trip to find. Today, use the latest and greatest tools available for the newbie, and why not start with the professional tools compared to starting with the lego kit, or at least, quickly switch over to the professional tools and ditch the toys, take off the training wheels as soon as possible.
A complete beginner without ever having touched any electronic component can easily download kicad and draw out a schematic from other examples found elsewhere [online, datasheets, magazines, whatever], can absolutely do a layout, then very easily print out the copper traces and etch a board and drill the holes and even bake on the surface mount parts to complete a new board. I am not saying this is necessary, I am saying, drop the idea that it is sooooo hard and sooooo complex. It is no more complex today than brewing beer in the kitchen or making a batch of yogurt from scratch or canning jam. I hope everyone does those things too. They will be helpful skills to have for when the zombies come.
There will almost always be a custom board in an embedded project. Including a thermocycler. It does not have to be made in CAD or etched by a fab. It does not have to be complex. The power relays are mounted to something and wired to something, that something is a board. The temp probe is going to be wired to something. The user LED's are wired to something. Don't make a rat's nest of jumper wires. These should be neat and connected on their own board. It could be a pre-drilled copper bus board. It could be a custom etched board with traces drawn with a ruler and a sharpie permanent marker. These are possibilities. In discussing design (which is what I have been attempting to start here) compare the competing designs and list the likely possibilities. If you go with a kit board vs components, then immediately what comes to mind is: does this board have a prototyping area to add components to it, or what components is that board missing that will need to be put on a second board anyway. No, the arduino boards do not have prototyping areas. What I didn't blatantly say, and I guess I should have, and will now, is: start out with the idea of creating the custom board so the form factor will fit in the space you want with the mounting options you want, and ideally everything will be on one PCB. [Note: The small Arduino boards do not have mounting holes. The only way to mount them is to use it's header pins.]
I don't know where this "electronic novices can't do these complex things" perspective comes from. Is it any different than crufty liberal arts school teachers saying, "Ohh, watch out for calculus, you don't want to try that, it's tooooo hard for you!!" I was a novice once. I took a piece of 1/8" thick waste acrylic out to a workshop and drilled 0.1" holes in it for an IC and components to fit into because I couldn't afford to purchase an FR4 copper clad board, then I soldered the IC and components together point-to-point [note, my holes were incredibly off-center but I bent the pins to fit and it worked fine]. That was probably my first "board". I didn't have a schematic, I drew one on a spare piece of graph paper based on the circuit I saw in the datasheet and a project idea I read in a magazine. I didn't have an electronics soldering iron, I used an automobile-grade type, I didn't have access to the chemicals for etching. I think I was 12 at the time. I had no guidance. I didn't have instruction from school or a mentor. I didn't have youtube videos to watch as a guide. I didn't have new components, I ripped them out of an old radio and tape player. I didn't have datasheets for anything in the world available via a mouse click. I didn't have a digital logic probe or O-scope. Look at how much easier CAD and fab technology and component procurement has become since then and all the examples and videos online! It is much easier now if anything. It is not difficult even for a "novice". Do the homework. RTFM. End the "must spoon feed them" mentality. It doesn't take "excessively long" and you don't "need to get it done right now". Any biology protocol will take longer to incubate than to design and build the electronics. It's all good experience and not wasted. BTW that circuit worked, I still have it, I found it in a move a while back, quite a trip to find. Today, use the latest and greatest tools available for the newbie, and why not start with the professional tools compared to starting with the lego kit, or at least, quickly switch over to the professional tools and ditch the toys, take off the training wheels as soon as possible.
A complete beginner without ever having touched any electronic component can easily download kicad and draw out a schematic from other examples found elsewhere [online, datasheets, magazines, whatever], can absolutely do a layout, then very easily print out the copper traces and etch a board and drill the holes and even bake on the surface mount parts to complete a new board. I am not saying this is necessary, I am saying, drop the idea that it is sooooo hard and sooooo complex. It is no more complex today than brewing beer in the kitchen or making a batch of yogurt from scratch or canning jam. I hope everyone does those things too. They will be helpful skills to have for when the zombies come.
## Jonathan Cline ## jcline@ieee.org ## Mobile: +1-805-617-0223 ########################On 3/25/15 5:13 PM, Simon Quellen Field wrote:
I respect you, and your engineering talents.
But are you really are recommending that the electronics novice design his own PC board using a CAD system, ship it off to a PC board manufacturer, and solder in all the parts of the power supply and microprocessor and clock system, and program the chip on a separate programmer board?
My recommendation would be to use a ....
0 comments:
Post a Comment