Oh, that wonderful feeling when the board boots up for the first time!
This is board 001 that I hand-built. I have some clearance issues, like the USB at the top. This is an oddball board, the part cost ended up being too high so the production board will move most of the I/O to daughter boards that can be added as needed.
There is also a 7" display board that connects via ribbon cable behind this. It isn’t done yet.
@Gus_Issa As it is currently shown there are:
2 - isolated serial ports that are software selectable between RS-232, RS-485, and CAN on one of them.
2 - 24Vdc Digital Inputs, 2- 24Vdc 100mA Sinking Digital Outputs, 2 - 0-10Vdc analog inputs
Isolated 4-20ma output with an additional isolated 500mA output.
1 Relay output
3 port ethernet switch, 4th port is SitCore, Passive PoE Input on Port 1, Software selectable passive PoE output on Ports 2 and 3 with current monitor.
1 Xbee socket
1 Nimbelink Socket for cellular
USB and uSD Card (below the SitCore)
FRAM Memory and EUI-48 EEPROM for MAC address
2 - I2C temperature monitors, one by the main power supply and 2nd in a “cool” area of the board.
40pin FFC connector to connect to the Display board. (Display board is separate to accommodate different sizes and makes it easier to accommodate supplier changes.)
FTDI Serial to USB on UART 5, (above SitCore to the left)
The 4 small plugs are “just in case, or just because”. They UART5, UART6, UART8 and I2C along with +5vdc.
SuperCap backup on the 3.3V supplying the SitCore and uSD and FRAM. (Currently not working, need to troubleshoot the circuit. I have it bypassed for now.)
Hopefully, I can get it all working smoothly. Tomorrow I hopefully will be testing the 7" cap touch screen.
@Dave_McLaughlin - I used an SMT solder past stencil and I have a desktop reflow oven. I use a vacuum pick up tool that I pieced together to place the parts. It has a switch to turn the vacuum on/off so I have better control. I did cobble together a fixture to help steady my hand to place the small pitch devices like the 200pin DIM. I have a USB microscope camera attached to help me line up the parts too. I did get some bridging on a few parts but I was able to clean it up with my soldering iron under my desk microscope.
I use 0603 for the passives, it is the smallest I can reasonably place. It took me 9hrs to do this one but I was still learning the board, I think the next one will go faster.
I bought a cheap microscope it has been invaluable! It has made a HUGE difference in what I could do in my own shop. I went with this guys recommendation (link to the microscope in his description): A Boy and His Microscope - A Love Story - YouTube
Two features that drew my attention:
Nimblelink cellular socket
Nimblelink cellular socket:
We have looked at using Nimblelink cellular LTE modems, but we have so much invested in the driver we developed for the Multitech MTSMC-LVW3. This driver/cell modem combo provides very high connectivity reliability.
Not questioning if Nimblelink is better or worse (I don’t know), just wondering what decision path was walked to choose Nimblelink over other vendor(s) wares?
We have been using a Sealed Rechargeable Lead Acid battery for our battery backup.
The battery goes from a high heat to nominally cold environment. With such, battery replacements are common.
We have shallowly looked at using SuperCap(s), but yet to go down that path.
How many amp-hours do you figure your SuperCap setup can provide?
Our BBU only needs to provide enough voltage & current to power the SOM, cell modem and other low-draw circuitry, for say 30s (connectivity good) to 2min (connectivity bad)…our devices are deployed in the hinterlands, so level of cell tower signal strength and connectivity ranges.
I choose Nimblelink primarily because of their wide selection of modems. I typically use AT&T/T-Mobile but some applications Verizon is the only option. I didn’t want to get locked into a particular service provider or technology. I feel like they do a really good job of keeping up with the current technologies and providers and that they will have a path forward as advancements come to the market.
The SuperCap backup will only provide enough power for the processor to close out and safely save any open files to FRAM or flash. The current board has an issue with that circuit so I don’t have a sense of the duration I will get. I am hoping for 1000ms or better. With an external capacitor bank, a few minutes is probably achievable.
If I was needing longer, like hours or days, I would probably go with Lead Acid as well. It has the best temperature range for both discharge and charging. LiPo doesn’t do well with cold temperatures, at least for charging. I haven’t really checked much but I don’t think you can get the power density from super caps as you can from lead-acid, at least not at a reasonable price point.