Michael O'BrienWent back to the board layout as I noticed a couple things after sleeping on it for two days. 14V supply was running next to the switching plane for 3.3V so I'm expecting it to be a bit noisy, but I can filter that out if it's too much. Next up I went around my tight areas with a bunch of criss-crossing around the mDP female plugs. Adjusted some trace spacing and fixed a couple things I was unaware of initially. End result were the extension of a few ground planes without the need of vias.
Last night while looking around on a couple other sites for pricing, I noted that my board is <5 cm tall and <10 cm wide. A 5 cm x 10 cm board is apparently common for prototyping. Learn something new every day. Found another shop that can produce prototype boards a bit more quickly and with an e-test. I decided to up my board's size a bit to fill this area. I did so 'cause I only had 2 holes for mounting and figured 9 extra mm would space me for the other 2.
I changed the order of the main pressure sensor, EGT connection, and aux pressure sensor, centered the components inside the new dimensions and proceeded with similar cleanup to what I listed above. Dropped off 8 vias, added the mount holes and I'm now at 454 drills on the whole board, aka ~59 per sq. in.. It'll double manufacturing cost, but I'm sure I can half that number or decrease it even more than that if I bumped over to a 4-layer board.
Either way, I'll be ready to test and adjust the layout once I get the Osh Park boards and have at 'em in a few weeks.
Edit: Speaking of which, I ran the numbers for the LDO's temperature requirements and wow, I would burn up the 14V rail pretty quickly. Due to it's load and being driven by the 5V rail, I had to adjust a couple things for thermal dissipation. I've also switched it up to drive the 14V rail from the power for the whole board, not chaining it from the 5V output. Not sure if how I'll do this, but if I have a greater load on the 14 rail in the future, I will need a new LDO as it's not strong enough for more than ~250 mA.
Discussions
Become a Hackaday.io Member
Create an account to leave a comment. Already have an account? Log In.
http://www.digikey.com/Web%20Export/Supplier%20Content/Tyco_8004/PDF/Tyco_AutoNetwork.pdf?redirected=1
http://www.st.com/st-web-ui/static/active/en/resource/technical/document/application_note/CD00181783.pdf
Basically MOV/MLV/TSV's and PESVs
Recent cars have better power, but you can still see some nightmarish spikes/dropouts especially when someone else has installed something and mixed up chassis and signal grounds or added noisepaths etc.
Take a look at the DashDAQ too
cheers
Are you sure? yes | no
What are you using for power protection/ground differences/loops for in car? I've done a few things like this and as you've probably already come across, power in a car is a mess.
Are you sure? yes | no
I have am using 2 Vishay S07D-GS08 diodes for polarity protection on the power input. I forgot to add them on the first prototype board, but that is what what lack of sleep does to ya ;). They are rated for 750 mA @ 65C and the max I might possibly hit if all planned LEDs and the OLED screens were on full brightness and white, etc etc, would be ~550 mA.
If you have any other suggestions, by all means, I'm open to suggestions.
Are you sure? yes | no