I’m new at .NET Micro and tinyclr forum. I want to use .NET Micro for my next project, but I’m not sure what gadgeteers to choose. I will be happy if someone helps me out.
The project is a handheld device that contains Graphic LCD, SD Card, Thermal Printer and GPRS. The user enters subscriber number and current status number, the device calculates subscriber’s usage fee from current status number and old status number, prints it, stores that information on SD Card, and sends it to the server if GPRS network is available.
My question: Are the following equipments fit to it?
I’ve read some topics that say the Panda II doesn’t support PPP. Is that nessecary for voice communication or SMS? I’m planning sending data over TCP/IP.
Depends on what sort of GUI you’re planning to build. The Panda-II + Touch is a very slow combination. If you want something more powerful, I’d recommend a Spider + LCD.
What about GPRS? Does Panda make it easy or same as any other platform? Can I use .NET classes, or do I have to do it with Serial Communication and AT commands?
And does SD Card require more RAM? For searching through entries in files, I think I can fetch line by line, and remove after checking. But I saw similar device that uses 2Mb RAM, and when I asked them what this RAM size for, they answered “to use SD Card” (their platform uses 16bit 22MHz MCU, and think they might have coded FAT file system on their own).
We’re planning 100+ device for start. And 500+ or 1K+ later. To reduce software development costs I’m trying to prefer .NET platform while keeping hardware costs acceptable.
I think you’d be best off calling GHI sometime and talking this over with them. They’re the people who can really help you target the right platform that will help your product be a success.
You are on the right track for sure. Like Brett suggested, it is easier to chat a bit about your project over the phone then give you pointers on what will fit your needs best.
Web site says “Please visit our TinyCLR Forum for comprehensive support. Our engineers along with the .NET Micro Framework community actively monitor the forums to provide excellent support.”
@ CE Let me see if I can’t help address a few things for you.
Using the SD card shouldn’t add any noticeable amount of overhead to your project, so go ahead and use it all over the place. SD is a REALLY good option for the Panda II’s small amount of RAM.
GPRS shouldn’t have an issue on the Panda II, you might not get the highest baud but it should do.
Same goes for the thermal printer.
How are you connecting the printer? Will it be another serial connection, I2C?
As mentioned touch will be SLOW and not extremely precise. Please see:
FEZ Touch UI Controls (Free open source, available in code share)
and
Skewworks Sprial (commerical, closed source)
Both are pretty slow when it comes to graphics, but depending on your needs it may be enough.
I would agree that contacting GHI first thing Monday morning might be a very good idea. They’re always very willing to talk options and work with you.
I will have to wire Touch manually, and modifying some code on managed driver may be required. But I think it’s possible because it uses parrallel interface. Timing may differ.
MSDN says on Cpu.SlowClock Property: “The Slow Clock is a slower and more precise clock than the system clock. It is used for the timing of the GPIO pins.” Cpu.SlowClock Property | Microsoft Learn
Someone from GHI already said “that’s a string we don’t use”. But if actual values are close, my plan doesn’t work.
Does that mean, if I continuously switch high and low the GPIO pin without delay, I got SlowClock frequency?
I will be happy to hear what are the probabilities.
Thanks in advance.
With a Cerberus the FEZ touch is not plug and play yet. There is quite a bit of work to get this going.
[quote]Some POS Printer with Serial Interface
[/quote]
Looks like a hand held metering (power, heat, gas etc.) device. Did you look at the kind of enclosure you need and if everything will fit in ? What kind of printer ? or will you use a mechanism + interface board + cutter + paper well ?
What kind of battery ? How will it be charged ?
Do you really need a touch ? Will a simple display + keyboard work ? Will the unit be used in daylight in the open ?
I may be asking vague stuff but trust me, I have been there and done this for metering and transit ticketing. You need to engineer this product to make it robust and not fail out in the field. Will you need the ability to to transfer data from device to device directly ?
[quote]Does that mean, if I continuously switch high and low the GPIO pin without delay, I got SlowClock frequency?
[/quote]
What is the requirement here ? You want to toggle a GPIO ? What does that do ?
Do you actually have a FEZ device with you ? You can not expect to have a working unit on the first set of boards and modules you purchase. Prototype, check the performance against your acceptance criteria, make changes, test again (iterative development).
@ Rajesh
Thank you for your comment. It leeds me somewhere.
Device is exactly what you say. Meter Reading Handheld Device (without any sensor for now). It will be used in the field, but most of time in buildings. Let me explain more detailed.
Every home has counter for usage metering. Electric, water and natural gas. The government distribute the data collecting job to the companies. And these companies read all the user’s data from that counters in their area (entering the subscriber number and counter value). There are different types of devices. Some works offline. Worker collects all the data on SD Card, and move it computer when he/she comes back to the office. And some works online. The device store data on SD Card, but when the network is available, it sends data to the server. There is also a server software that communicates to the database over API that is provided by government agency.
I didn’t highlight thermal printer and battery because it’s not .NET or gadgeteer related. I will do it in one way or another.
I asked that because I thought connecting Touch to Cerberus (parallel of course). And if speed of GPIO depends on that clock, I may get same behavior as Panda.
It’s very good point. I always forget that. Is Touch bad at this?
Yes, it was our first plan. But one of my friend works as an accountant in a company that is already doing this job. And he says that they try to extend features of devices. If we start with nice hardware (not so expensive), this may be easy for us in the future. We didn’t eliminate keypad. It may be nessecary if touch doesn’t work after a while. We think small keypad with silde funtionality.
1.GUI with touch
2.Ability to add modules at a later date
I am also very sure that you want the device to be a little secure and that someone will not duplicate your firmware (easily)
A very safe bet for your Rev1. device will be to use a
[ul]FEZ SPider
T35 Display module
SD module (may be the micro SD module)
USB DP
…anything else you may need.[/ul]
For future iterations you can layout a simple 2 layer PCB with the EMX module and display for a compact solution. Very soon the G120 SoM will be available, you can use then on a future iteration and benefit the lower cost for the same features.
I do understand the the cost of the Spider and T35 will be a little high to start with but you will get returns with a shorter development cycle, performance and user experience.
[ul]With the Spider and T35 you can use native NETMF graphics (GHI Glide will make it a breeze)
Firmware on Spider can be pretty secure
Features like IFW will make maintenance easy, when you want to push new firmware for rate changes, GUI changes or a new functionality[/ul]
The BOM cost will be cheaper with the USBizi (Panda) or Cerberus (STM32F4) but you will incur increased development cost.
@ Rajesh
Thanks for your comment. It’s nice to hear from someone who did pass on similar way.
I think spider is a great device, and I may prefer that later. With a device with that capabilities, I can calculate good path for workers who work in same area, and display that path on map (probably on google maps) to leed them. I think it increases productivity. And maybe other workers on map. That time, I may need a bigger display like T35.
I’m getting closer with your help.
I’ll call GHI and ask some combinations in this week.
