Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Base Station PCB - Request for comments
#1
I'm designing an appropriate base station PCB right now. The one I use until now is built on a breadboard and does not support dual voltage, so the ATMega can only run at 12 MHz. A higher speed is useful especially at the base station to speed up encryption / decryption of packets.

I created the schematics and described it roughly on the OHC website. The PCB is planned as a general purpose one where you have several assembly options depending on what you need.

If you have experience in developing such devices, your input about the following topics regarding the schematics is appreciated. This poll will be closed in one or two weeks (~5/10/2013).
  • Is the I/O connection between ATMega (5V) and RFM12 (3.3V) ok with the resistors or what could be improved?
  • Are the capacitor in place correctly for the voltage regulator? Do you have specific experience to use them differently / other types / other capacities?
  • Is it ok to connect the "MISO" pin of the ISP connector (which is the "output" from ISP connector point of view) directly to the SDO of the RFM (which is also an output)? Or what can be improved there? (Note: I did not have any problems with that in the past. It seems to work well.)
  • Is the selection of available I/O pins ok for a general purpose PCB, or which other pins would you like more than the ones currectly used?
Reply
#2
Thumbs Up 
The resistors look ok, but I am not an expert here. Have you measured the voltages?
Anyway: you may use RFM12 (not RFM12B) which works well at 5V. Available in Germany at Pollin for 868 or 433 MHz.
This is a nice, handy and small levelshifter used many times. Any small FET will work here: [Image: levelshifter.jpg]

Voltage regulator: I believe the LP2950Z is a 5V regulator, when used like this. The ones from recom have very low power consumption and don't need any cooling. Yes, they are a bit expensive... :-(
Or get the 8 pin version and use the sense pin.
You may reduce ripple with additional 100nF on both sides, but I don't think this is really neccessary. Anyway they cost a few cents.

MISO/SDO: unsure, everyone does it like this ;-)
Reply
#3
(04-29-2013, 08:56 PM)Inhumierer Wrote: The resistors look ok, but I am not an expert here. Have you measured the voltages?

No, I calculated them. Smile Should produce 3.24V.

Quote:Anyway: you may use RFM12 (not RFM12B) which works well at 5V.

My impression was that the RFM12 (5V) model was not available anymore. I didn't find them at the other companies I ordered the RFM12B before.
Are you sure they are available not only at Pollin? I don't want to risk that the base station uses the 5V model and we cannot buy it in the future anymore.

It they are available, I would remove all the level shifting stuff for the base station of course.

Quote:Voltage regulator: I believe the LP2950Z is a 5V regulator,
The one I used is a 3.3V one, available at Reichelt.de.
Reply
#4
I added a new "Generic Midi 2" schematics for a PCB without the dual voltage support.

If we use RFM12 5V modules, the voltage regulator and the resistors of the "Generic Maxi Speed 1" are not necessary and the design will be much simpler.

If you have further information how to improve the schematics (see the picture in the base station article), let me know.
Reply
#5
I found out that the RFM12 (without "B") seems to be an older model with some differences to the RFM12B. For compatibility reasons, I chose to use the RFM12B only, so the source code and the hardware schematics can be the same for all devices and to avoid the risk the circuit is not working (I have only tested the RFM12B model until now). (If someone is willing to test the RFM12 some day, we could also support it if we have a good reason.)

That means to use the previously mentioned "Generic Midi Speed" for the base station.

I created a 1-sided layout which can be etched manually. I plan to do so in the next couple of days. You can have a look at the layout in the base station article. If you have comments, please share them.
Reply
#6
(05-01-2013, 07:19 AM)breaker27 Wrote: I found out that the RFM12 (without "B") seems to be an older model with some differences to the RFM12B. For compatibility reasons, I chose to use the RFM12B only

I thought the only difference is the voltage and the price Big Grin Tell us about the differences. And I have both here, and I will make a prototype on a breadboard to play around. I believe a 40-pin ATMega 32 should work as well?
Reply
#7
(05-01-2013, 11:37 PM)Inhumierer Wrote: Tell us about the differences. And I have both here, and I will make a prototype on a breadboard to play around. I believe a 40-pin ATMega 32 should work as well?

The following is from the specs at hoperf:
  • Output power is 3dBm higher for the 12B (isn't it +50% signal strength?)
  • Reception is 2dB better at the 12B.

This alone makes me prefer to add 6 little resistors and use the 12B when I have almost double the reception signal strength!

From the datasheets of RFM12 and RFM12B:
  • The length of the synchronisation pattern can be set to 1 byte (instead of 2) at the 12B only. But I think we don't need this.
  • The RFM12B has an additional "PLL Setting Command". I think this is used when the module runs at a different clock. But I think we don't need this.
  • The RFM12B has a "Synchron pattern Command". You can set the synchonisation bytes (2 bytes) differently. This would improve that the modules don't react on other RFM12 modules' data if set differently for all devices (some specific two bytes for all your OHC devices, could also be set in EEPROM). This may be an option in the future!

As a summary, I definitely prefer to use the RFM12B. I improved the "Generic Maxi Speed" PCB layout a little bit and changed 3 resistors to zener diodes. Otherwise, the pullup R10 (1MOhm) for the RFM12B would not have worked.

If you (or another person) has bought several RFM12 (without B) modules, you can solder it also to the OHC PCBs. But you should be aware that you won't have the same signal strength then. It's your choice (maybe also depending on the distance the specific device has to support). But especially the base station PCB should definitely support the RFM12B from my point of view.
Reply
#8
The base station / Generic Maxi Speed PCB works, see this news item.

If you have any additional proposals for the schematics (other resistor values, placing of parts etc.), keep commenting...
Reply
#9
Your are sure that your PCB layout is good? Your traces have many curves 90 degrees, this can cause problems.
Reply
#10
(05-02-2013, 05:36 AM)breaker27 Wrote: [*]Output power is 3dBm higher for the 12B (isn't it +50% signal strength?)
AFAIR this is true for voltages, currents etc., but for power related stuff like antenna signals or noise it should be 100%. Correct me someone if I'm wrong. This stuff is ages ago. Indeed this is a very good reason for using the "B".
I didn't realize that. Never read this much from the data sheets ;-)
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)