APRS to monitor your Airstream?

[alano]

I've already added the supercap to my schematic. I'm going with a smaller cap, about 5F, that will exhibit lower leakage current and be sufficient to guarantee sending a few messages out. I agree that the Lithium cell would certainly have issues at low temperature. I don't see the need to send more than one successful message indicating that power is completely interrupted. You could argue what happens in case the cell tower is out of range. I would counter that if the tower is out of range, then there would always be a larger super cap that wouldn't store enough energy until I was back on-line.

[AirstreamCSH]

Originally Posted by alano

The Arduino IDE is free from arduino.cc or you can download it for free from Microsoft's Store. Hologram describes how to configure the Arduino IDE for use with the Dash. They have a reasonably good forum where you can ask questions. You can always ask me as well

So when will you have a system for sale?

[alano]

Originally Posted by AirstreamCSH

So when will you have a system for sale?

No plans for the moment. Once I update the PCB, I'm happy if anyone wants to build a prototype and help beta test the monitor. Perhaps I'll ask for volunteers when I get everything together.

[rmkrum]

Please put me on the list of volunteers to build and test a prototype when it becomes available. Have all the tools and skills for building electronic hardware and troubleshooting software. Rich, KE4GNK.

[Scoutx]

You might upload the design, software and such to a DIY site and then post a link. Saves people wanting to do something similar from having the reinvent the wheel, as it were.

[alano]

Originally Posted by rmkrum

Please put me on the list of volunteers to build and test a prototype when it becomes available. Have all the tools and skills for building electronic hardware and troubleshooting software. Rich, KE4GNK.

Rich, thanks. Can you solder SMT components as small as 0402 resistors?

[alano]

Originally Posted by Scoutx

You might upload the design, software and such to a DIY site and then post a link. Saves people wanting to do something similar from having the reinvent the wheel, as it were.

I haven't really thought far in advance what my end goal is for this project. I'd certainly consider open-sourcing the project while exploring all alternatives.

[rmkrum]

Originally Posted by alano

Rich, thanks. Can you solder SMT components as small as 0402 resistors?

Yes, but I need to use a magnifying glass to do them. Bigger is a lot easier, IMHO, when hand soldering....the little stuff needs a steady hand, and a lack of caffeine to pull it off.

Sharp tweezers and a clean work area is key to pulling it off...and not sneezing[emoji3]

[alano]

It's getting more and more difficult to design stuff these days that folks can solder. The voltage regulator I specified isn't even available in a DIP or TO type package.

I broke down a couple of years ago and bought an inspection scope to work under, a pre-heater and hot air pencil. I find most of my soldering these days to be done with a syringe (for the solder paste) and hot air pencil!

[Al and Missy]

Originally Posted by alano

Rich, thanks. Can you solder SMT components as small as 0402 resistors?

I'd like to be involved as well. I repaired my 756 Pro III with a USB microscope and a Weller soldering station with the finest point they make. A nearby lightning strike (I guess) took out the speaker driver and remote control port components (0603s, SOT-223s, EMD-3s, 8SSOPs).

A hot air pencil looks like a much better solution.

Al

[alano]

Al, will do and thanks for reaching out.

[Scoutx]

For myself I still primarily use thru-the-hole components. In the odd, rare, case I have to use a SMT then I resort to an adaptor board to effectively convert it into a TTH component. But then I find that so much of what I use to have to do can be done quicker and easier with a simple microcontroller, premade modular boards, or a microcomputer.
The rest then becomes a breakout box and/or hardware controller board using TTH components. Code replaces so many components, and I don't have to redesign the whole board to make changes as I would have to do with hardwired logic.
For example, you're talking about a voltage regular, which I would assume is to handle the battery voltage to 5VDC needed.

$6 and I'm done. No muss, no fuss, no problem. Small, beefy, robust, and build in safeties.

https://www.getfpv.com/5v-step-down-voltage-regulator.html?gclid=Cj0KCQiA5t7UBRDaARIsAOreQtism hj-mneoITaOzM0xMpQxv1ZRsEGos0JbuwJHzVTW91K-9Ct66l0aAlnXEALw_wcB

soldier it to pins on my PCB, and move on. Even has a sleep mode if you really need to shut things down to save power.

[alano]

Scoutx, the Dash needs a 2 A supply although Hologram is somewhat vague on this point. I also need to set the +5V supply a bit higher than normal to allow the supercap to decay and still stay within tolerance. For example, I'm targeting 6.2 V which is OK since the Dash can accept 4.75 V to 7 V. If I used a fixed +5V regulator, I'd only be able to drop the supercap 0.25 V before exceeding the lower limit.

My priority with this project hasn't been to make it easy for a DIYer to build, I'm more focused on performance and cost.

If you'd like to review the hardware design I'm more than happy to share it with you. Just PM a note.

[Scoutx]

Are you using a voltage regulator or a DC/DC converter?
I ask because the converter is far more energy efficient.

PM sent. Once I have a better idea of your design parameters I can see if I can aid you.
Certainly you want to keep the supercap just below the max operating voltage, but have you considered putting the supercap directly across the battery leads with an extremely low forwards voltage diode to prevent discharge if battery disconnected?
the run your DC/DC converter from that and will have a much higher initial energy charge to start with, and can prolong it longer by running everything a bit above minimum voltage requirements.
Of course cost is a factor to keep in mind, so reasonable limits must exist. Also you must consider limits. Maybe cheaper to run a supercap just below 7V, than buy one that can handle up to 15V.

[alano]

I need to use a DC/DC converter since the input requirements are to accept 24 V. Since a 24 V supercap would be expensive, I decided to put the supercap at the output of the converter. The supercap needs to be derated as well, so I selected a 7 V supercap, and derated it to handle the max temperature.

[Scoutx]

24V?

I would have thought either 15V or 30V depending on battery configuration.

Either way looks like you've got a handle on it. What's your expected power draw on sleep, standby, active and transmitting?
Have you figured out how long you will need to stay in the other modes for each transmission cycle?

[alano]

Correct, the monitor can accept greater than 24 V, I just threw out that number to indicate that it can measure up to two batteries in series.

A goal is to minimize the overall energy consumption of the monitor so that it's consumption isn't a significant factor for folks monitoring their trailers without shore power or an external charger.

While the device is asleep, the current drain is about 1 mA @ 5 V, so hopefully, I should see less than 500 uA @ 12 V.

For the moment I wake up every 15 minutes (synchronized to the top of the hour) and stay awake for 3 minutes. There's a minimum period that's needed to connect and retrieve messages. I haven't played around with reducing this period yet.

If there are no pending incoming SMS messages I go back to sleep. If there are incoming messages, I stay awake for 5 minutes and reset this 5-minute timer on every received SMS message. My thought here is that if there's an incoming message, it's likely that another one may follow.

The current drain while active hovers around 25 to 75 mA @ 5 V and then occasionally spikes up to 100 to 200 mA when transmitting. This number is quite variable and depends on the distance to the tower.

The times are easy to change and it may make sense to parameterize these for the user. When connected to shore power the device will be active 100% of the time.

There's a tradeoff here that perhaps wasn't obvious: a direct SMS-based system that occasionally sleeps will have an inherent latency until the pending message can be acted upon. I could eliminate this latency by implementing a server-based model, but this model has another set of challenges with respect to cost, maintenance, data volume and app development. Knowing that the system will respond on the 1/4 hour mark, the latency seems a small price to pay for overall simplicity.

[Scoutx]

Yes. That all makes sense. Further I suspect in almost all cases an incoming message is more to force a report than anything else. In rare cases you might want to tweak a setting, but I would think that would be fairly rare.
I think the timers should be variables as people may want to adjust then to suit their own needs/wants.
Power demands seem reasonable, and you may want to consider a converter with a sleep mode and run off the supercap during the sleep cycle. Should drop your power demands during sleep to single digit microamp leakage

[Scoutx]

Off hand the only thing I see to comment on is that you might want to be able to switch off the resister network for the battery sense except when you're actively taking a measurement.
The leakage is very low about 25 microamps from each battery. Not sure it's worth it, but it is a power bleed if only a minor one.

All in all though looks like a nice low loss design.

[alano]

I revised the PCB and built a prototype of the v2 monitor. This includes the ability to monitor two separate battery systems, each up to 24 V, a set of general purpose inputs and outputs, and the ability to power the system and be continuously active when connected to shore power.

However, I recently learned that the Dash module that I'm using is going to be canceled next week by the manufacturer, Hologram. They are no longer taking orders on their website. They are planning to move support over to a new Arduino board, the MKR GSM 1400 distributed by Arduino.cc. It's comparable in many ways to the Dash except it doesn't support over-the-air software updates. Hologram support isn't yet available for the MKR GSM 1400, but I'm told it's coming within the next few weeks.

I've already ordered an MKR GSM module, but I'll need to redo the PCB and wait until Hologram releases support before I can make any progress.

The cancellation of the Dash module will certainly delay the project. I'll report back when I have a new prototype that can be duplicated.