Must-haves for EASY DIY controller?

DWZM, What is the plan for software changes and submissions from the users. I assume you are not going to open this up to the world so something like SVN patch would be in order. What are going to be the guidelines for submission of a patch? Have you established a review team for patch submissions?

hmmm.. that's a good question.
I was kinda under the assumption that the user would be responsible for his own coding (Well at least tweaks and mods) as I'm sure the idea is to have a base Cookie-cutter and from there individuals could tweak to their preference. I think that's what DWMZ and Thz where trying to get at a little earlier in regards to the code being general use.
 
From a philosophical point of view I'm curious as to why use the non-commercial aspect of the CC license. While attribution and share-alike make total sense, it seems like commercial use might be good for the project. Allowing people to commercially manufacture the board may increase the availability of the board for users, especially those who not very good at assembling the components.

My other question is if someone was going to assemble the boards and sell them, are they able to charge for the time it took to assemble them or would that violate the non-commercial aspect of the license?
 
Dustin does your lm317 get pretty hot mine seems to run very hot the controller is running off of it fine should I put a heat sink on it
 
DWZM, What is the plan for software changes and submissions from the users. I assume you are not going to open this up to the world so something like SVN patch would be in order. What are going to be the guidelines for submission of a patch? Have you established a review team for patch submissions?

The project is completely open source, hardware and software. The main idea that has been discussed is that we are going to have a basic feature set to include everything that is on the board. We will likely have examples for other common features that will be used. For any kind of customized features or anything outside the standards of a basic controller, it will be up to the end user to come up with that. The good thing is, we are going to have a wiki, so anyone that comes up with a cool feature will be able to submit their example/tutorial to the wiki for all users to benefit from.

From a philosophical point of view I'm curious as to why use the non-commercial aspect of the CC license. While attribution and share-alike make total sense, it seems like commercial use might be good for the project. Allowing people to commercially manufacture the board may increase the availability of the board for users, especially those who not very good at assembling the components.

My other question is if someone was going to assemble the boards and sell them, are they able to charge for the time it took to assemble them or would that violate the non-commercial aspect of the license?

I'm sure the main reasoning behind non-commercial is that this is an open source project developed by users of the community for the benefit of the community. Why should anyone be able to profit from the hard work of others? Besides, if someone can't solder the board, they are not going to have an easy time trying to get this thing to work. There's more to it than the board. Users should also be able to coordinate their own orders for batches of boards, after all, this project is listed as "DIY".

Dustin does your lm317 get pretty hot mine seems to run very hot the controller is running off of it fine should I put a heat sink on it

Yes, my LM317 does get quite warm at 7.2v with a 12v input. I'm using a radio shack heatsink I picked up locally. I'm also going to run a small 40mm fan in my enclosure.
 
hmmm.. that's a good question.
I was kinda under the assumption that the user would be responsible for his own coding (Well at least tweaks and mods) as I'm sure the idea is to have a base Cookie-cutter and from there individuals could tweak to their preference. I think that's what DWMZ and Thz where trying to get at a little earlier in regards to the code being general use.

To be honest this is something we haven't firmly decided on yet. Our vision for the code is basically that the project team will maintain a core firmware that will implement a core set of functionality (likely functionality for the hardware on the board, plus "obvious" extras like relay control). Users will be responsible for development beyond that if they have other functionality in mind. As far as users requesting or implementing changes to the core firmware, I'm sure we'll put something in place but haven't really decided yet. It certainly won't be open to the whole world.

From a philosophical point of view I'm curious as to why use the non-commercial aspect of the CC license.

Basically I don't want someone swooping in and deciding to sell Hydras purely for the money making potential. I want people who are involved with the project to be involved because they're interested in it (either from a contribution standpoint, i.e. helping with development, or from a use standpoint, i.e. building one for their own fish tank), not because they want to make a few bucks.

I understand that commercially manufacturing the board would increase the potential audience, but as far as I'm concerned, the goal of this project is to provide a framework for people who want to build their own controller. I see that as a void in the marketplace. If someone wants to buy an off the shelf controller (even and off the shelf Arduino-based controller) there are plenty of options - I don't think the Hydra really belongs in that segment of the market.

If a contributor to the project is serious about manufacturing/selling Hydras then we can have that discussion offline. Honestly, if the project was going to go down that route, it would probably make sense to redesign the hardware, as we've made a lot of concessions that would impede mass-production to keep it "beginner friendly."

My other question is if someone was going to assemble the boards and sell them, are they able to charge for the time it took to assemble them or would that violate the non-commercial aspect of the license?

In theory, that would violate the terms of the license. But again, I really have no plans of strictly enforcing the license or trying to scare people off of the project. If someone wants to build a few and sell them to their friends who happen to be interested in electronics but aren't confident in their soldering skills, that's totally fine - I am absolutely not going to stop them. (ReefCentral will stop you if you try to do that in this thread, but that's a different matter). But if someone starts soldering up a dozen every weekend and listing them on eBay at a price that's a clear profit over the cost of components, I'm going to be a little upset.

Again, I'm not bringing up the whole licensing issue as a means to scare people or halt their interest in the project. Just trying to help people understand what our intentions are. Our intentions are to get reef keepers interested in hobby electronics, and to provide a project for them to follow if they want to build their own controller. All we're really asking is that if you spread the word about the project (via blog posts, posts on other forums, etc.) that you link back to this thread so people know where the project came from, and that you don't steal the efforts of this team and go try to make a buck off them.

Dustin does your lm317 get pretty hot mine seems to run very hot the controller is running off of it fine should I put a heat sink on it

The heat dissipated in a linear reg is basically equal to the voltage it has to drop times the current it's passing. So if you're sucking 500mA through it and it's dropping 5v, you're asking it to dissipate 2.5w. You can look up the max power dissipation in the datasheet for the particular LM317 you're using as it'll vary from manufacturer to manufacturer and package to package. As long as you're under that limit you are theoretically OK. If you are at or near that limit you definitely want a heatsink if you can fit one.

Another option is to reduce the input voltage, if that's possible. If you ask it to drop less voltage, it'll be dissipating less power.
 
I understand that commercially manufacturing the board would increase the potential audience, but as far as I'm concerned, the goal of this project is to provide a framework for people who want to build their own controller. I see that as a void in the marketplace. If someone wants to buy an off the shelf controller (even and off the shelf Arduino-based controller) there are plenty of options - I don't think the Hydra really belongs in that segment of the market.

Thanks for the clarification. I imagine if someone were to mass produce them the price would increase pretty significantly and would probably get to the point where the users might as well buy reefkeeper or neptune.

Thanks for patiently explaining it!
 
Yeah I basically copied the latest schematic you did and did some arranging. Dont have the silk screen complete yet. I incorporated the tie in and seperated out the address and GP pins. Everything is routed but I dont like the location of the DC input jack because it faces a different direction than the serial jack. I dont want to make the board larger because this satisfies the 5cm by 5cm boards size for patchpcb and makes it a little cheaper.

Schematic:

RelayBoardv14schematic.jpg


Board:

RelayBoardv4.jpg



DustinB I can send you the schematic files if you want to take a closer look at them.
 
Last edited:
One option would be to move X1 and the DC jack off the PCB - put headers for each on the PCB and use panel-mounted connectors in the wall of your project box. That frees room and solves alignment problems.

Otherwise if you move the standoff hole next to X1 I bet you could get enough room to rotate the DC jack 90 degrees and get everything to fit along one edge. Speaking of the DC jack, what's up with the big SMT-looking pads under it connected with those thick traces?

I'd also do a ground plane on both sides of the board. It makes routing ground connections painless. Once the ground plane was in place, I'd route the power traces in a "logical" format, then work the signal traces around that.

And, I'd probably up the width of all power traces significantly.

Finally, I see the .1uf ceramics are on there now, but I'd try to position them next to whatever they're supposed to be filtering - i.e. leave one about where it is, near the MCP. Put the other two near each of the voltage regs.

For the silk, the convention we used on the mainboard was to include component names, but not values. Also we included pinout labels for headers. You may find that you have to relocate some of the headers to make enough room for the labels. Regarding the jumper headers for the address pins though, I don't see any reason to not just push those together into one big 3x3 block.

I see an errant "+5v" near the upper right corner of the MCP in the schematic.

Also it would be nice to break out the INT pin on the MCP in case someone dreams up a reason to use it in the future.

Finally, carefully plan the location of the standoffs and make sure there is enough clearance for the hardware you'd use with that hole size. In a couple spots they look really close to neighboring components.
 
Good suggestions.

I believe the pads are for surface mounting, at least that's what they were in my version.
That was one part I was confused about. Most of the pcb mounted jacks require a couple slots instead of standard holes. From looking at itead/seeed it seems they do not allow any slots like this, only drills. Have you ordered boards before with these types of slots?

One other suggestion, maybe rotating the MCP 180 degrees might be better?
 
For DC jacks I always just specify a hole big enough to fit the tab in. Look at the spec sheets for a few common DC jacks and make the holes big enough for the one with the largest tabs. You end up filling a huge void with solder but that makes for an even more secure connection. Here's the back of a partially assembled Typhon to show what I mean:

2010-11-16_14-25-14_344.jpg
 
Ill try to work up another version tonight. I like the idea of moving the some of the components off board and mounting them to the wall of the box. The DC jack came from the sparkfun library so I made an assumption for spacing. Ill rework this version and make the changes.
 
Oh, another space saving thought. Leave the top voltage reg where it is. Flip the other one and push it to the edge of the board. Now, you can put a heatsink on that extends out past the edge of the board, instead of one that takes up board space (imagine if the yellow "u" shaped heatsink shown for the top reg was flipped around 180 degrees).
 
I like to get started but I dont know where to start, like parts,etc. anyone have this?

Also anyone working on a saltinity and Orp Sensor?

Thanks
 
drillsar, read back through the thread and find a link to the google code site. There is a BOM (bill of materials, aka parts list) on that site. You can buy everything you need from 3 or 4 different vendors. You'll also need a PCB. There are Eagle files on the site that describe the PCB. You will need to find a board manufacturer and use those files to get a board from them. batchpcb.com and dorkbotpdx are the best bets unless you want 10 or more units.

It's worth noting again that this project is still very much at a development stage and we have not yet really put much effort in to documentation or making it "easy to swallow" as we are still refining the design. if building one of these based on the info in this thread is not obvious then you might be more comfortable waiting a few months until we have the project in a more presentable state for the general public.
 
drillsar,
We don't have any plans to have salinity and/or orp on the main board. That doesn't mean that if you have a working circuit we can't design a "shield" and add the functionality.
 
ORP would be really simple since it's just a modification of the same amp circuit we're currently using for pH. Salinity I have zero knowledge or experience with, but I'd imagine it would essentially be similar. For the most part, these probes all put out a really weak signal over a useless voltage range, so the struggle is to amplify and map it to something meaningful.
 
Back
Top