LowPowerLab Forum
Hardware support => General topics => Topic started by: Lensdigital on February 03, 2015, 12:03:09 PM
-
I've build several versions of PCBs with RFM69C 915Mhz chip. Previous versions of PCBs work great, I can send and receive even without antenna if two boards are close. Latest one I build doesn't transmit or receive signal at all unless I physically touch antenna with my finger! Moteino and my board are inches apart, there's something seriously interfering on my board...The only difference between working board and this one is that I added footprint for RFM69W, modified VCC paths (I control power with 3 way switch between V-in and 3V battery which bypasses LDO) and moved around PCB traces.
Has anyone experience anything similar? I know RF design is an art form, but I build so many different prototypes and never seen this issue. So far I tried another RFM69C module, cut traces to second footprint and antenna hole and attached antenna directly to RF chip, still signal is completely blocked unless I use my body as antenna ???
-
You've probably done this, but you might want to check to see if vias on the RFM69 are overlapping contacts on your new board. I had exactly this problem with another RF module. I thought that the bottom of the module was insulated :(
Tom
-
You've probably done this, but you might want to check to see if vias on the RFM69 are overlapping contacts on your new board. I had exactly this problem with another RF module. I thought that the bottom of the module was insulated :(
Tom
Thanks Tom! Yeah it occurred to me as well, so I actually attached clear tape over unused contacts just in case. Didn't help :( I wish it was just that! :) I'm feeling like my PCB is cursed :)
-
Nah - the ghost is in the code. Driver compiled to expect RFM69(H)W?
-
Nah - the ghost is in the code. Driver compiled to expect RFM69(H)W?
Hmm arent C and W suppose to be the same other than physical dimention? But i will doublecheck the code, thanks for the idea!
-
_@Lensdigital,_
For the C variant, RegPaLevel (0x11) bits 5-7 must stay at the POR value. The remaining bits are the selected Tx power level - POR values are fine for testing since they are 0x1F which maps to max.
RegTestPa1/RegTestPa2 should not be changed from POR values.
The W variant needs PA0 OFF, PA1/PA2 ON. Same as above for the remaining bits of RegPaLevel.
RegTestPa1/RegTestPa2 get manipulated in the driver to respond to user power setting request.
-
_@Lensdigital,_
For the C variant, RegPaLevel (0x11) bits 5-7 must stay at the POR value. The remaining bits are the selected Tx power level - POR values are fine for testing since they are 0x1F which maps to max.
RegTestPa1/RegTestPa2 should not be changed from POR values.
The W variant needs PA0 OFF, PA1/PA2 ON. Same as above for the remaining bits of RegPaLevel.
RegTestPa1/RegTestPa2 get manipulated in the driver to respond to user power setting request.
I believe you are speaking about the HW variant, not the W variant...
Tom
-
CW is a small footprint version of RFM69W which is pin compatible to RFM12b. I discovered it accidentally when ordered it by mistake instead of RFM69W. I was very happy about it tho, because my Xronos clocks are still using RFM12b so only software needs to be changed.
In fact I successfully used RFM69CW before with same exact library and code as RFM69W...
-
@Lensdigital, yeah, I know. My comment was directed to @emjay. His statement about the PA1&PA2 was incorrect for the W version.
Tom
-
I think I found out what's going on... It seems that RFM69C modules were mislabeled as 915Mhz, but they are actually 433Mhz units! As soon as I changed frequency my boards are working fine! Does anyone knows how to check what frequency they are supposed to be? I remember on RFM12B there was some capacitor missing on specific frequency, but I'm not sure about RFM69Cs....
-
I think I found out what's going on... It seems that RFM69C modules were mislabeled as 915Mhz, but they are actually 433Mhz units! As soon as I changed frequency my boards are working fine! Does anyone knows how to check what frequency they are supposed to be? I remember on RFM12B there was some capacitor missing on specific frequency, but I'm not sure about RFM69Cs....
Wow, how did you discover that? In any case, Felix color codes his radios with a colored dot. We had a 'discussion' on that but Felix deleted the thread - I thought he should sticky it, but it did have some embarrassing content.
Look at your radios and see if they have the same colored dot. If they do, the 433MHz one was mis-colored. If they don't, you were sent to the wrong one. And, if you post the radio's color, Felix just might publish his color code again. Maybe in a separate thread or on his blog, but at least we'll have a permanent reference.
Tom
-
Wow, how did you discover that? In any case, Felix color codes his radios with a colored dot. We had a 'discussion' on that but Felix deleted the thread - I thought he should sticky it, but it did have some embarrassing content.
Look at your radios and see if they have the same colored dot. If they do, the 433MHz one was mis-colored. If they don't, you were sent to the wrong one. And, if you post the radio's color, Felix just might publish his color code again. Maybe in a separate thread or on his blog, but at least we'll have a permanent reference.
Tom
Well first of all, I didn't buy these from Felix :)
As how I discovered it, well here's short story.
I had 3 versions of my boards, let's call them V1,V2 and V3.
V1 I only tested with 915Mhz RFM69W (large module) and it worked flawlessly as far as radio. I made whole bunch of them, they all work great.
For V2 I decided to switch to 433Mhz and soldered RFM69CW (small module). Again this one worked flawlessly I made two of these, great range, no issues with signal.
When working with 433Mhz modules I realized that antenna is way too long for my projects and decided to switch back to 915Mhz. But I also discovered interesting fact, that 433Mhz modules can also work at 915Mhz frequencies, at least at close range....
By the time 915 RFM69CW modules arrived (I got 10), I got V3 boards from PCB manufacturer and soldered one.
That's when I discovered that it's not working at all. I tried all I could think of, soldered another board, then desoldered RFM module cut all traces between small and large footprint, tried different antenna lenghs, etc.
I was kind of bothering that I never tested V2 board with 915Mhz module, so I soldered one last night. And guess, what? It had same issue as V3 board, and even touching antenna with fingers, had intermittent success.
I started to have strong suspicion that it was something with RFM modules. Than this morning I thought, what if they mislabeled... So I uploaded "transmit" 433Mhz sketch to V2 board that had "915Mhz" chip, and "receive" sketch to V2 board that had 433Mhz chip. And it worked! Not a single dropped packet, and I even still had shorter antenna on first module.
Then to verify that I'm not crazy, I uploaded 433Mhz "Transmit" sketch to my V3 board, and guess what, it worked.
So most likely possibility is that these chips are mislabeled.
BUT, it is still possible that my PCB design is flawed, and it just blocks 915Mhz band somehow...
I setup my photo rig with macro lens and took pictures. RFM chips are different, one (labeled as 915Mhz) is missing capacitor, but I don't know if it's due to frequency or setting or version of the chip, as actual IC is also slightly different.
Here's photo of 433Mhz RFM69CW (http://lensdigital.com/mote2/SQM_433Mhz.jpg).
And here's one labeled as "915Mhz". (http://lensdigital.com/mote2/SQM_915Mhz.jpg)
BTW labeling RFMs with color dot is a great idea. Especially when it's on the front of the chip! :)
-
Well first of all, I didn't buy these from Felix :)
LOL! Now you're paying the price for your sins!!!
<...snip>
So most likely possibility is that these chips are mislabeled.
BUT, it is still possible that my PCB design is flawed, and it just blocks 915Mhz band somehow...
I setup my photo rig with macro lens and took pictures. RFM chips are different, one (labeled as 915Mhz) is missing capacitor, but I don't know if it's due to frequency or setting or version of the chip, as actual IC is also slightly different.
Here's photo of 433Mhz RFM69CW (http://lensdigital.com/mote2/SQM_433Mhz.jpg).
And here's one labeled as "915Mhz". (http://lensdigital.com/mote2/SQM_915Mhz.jpg)
BTW labeling RFMs with color dot is a great idea. Especially when it's on the front of the chip! :)
I strongly suspect that the capacitor is part of the antenna matching network, especially since it's only present on the lower freq radio. IMO, I'd think its not unreasonable to switch a radio between 915 and 8xx, but going between the higher and lower frequencies would only be done in a test environment as an expedient. I've seen too many SWR charts to think one dish serves all!
In any case, you can probably marked this one [SOLVED]... Good job ;-)
Tom
-
In any case, you can probably marked this one [SOLVED]... Good job ;-)
Haha, not quite yet :) That stupid capacitor is bothering me...
But I had another idea. Going to try RFM12B 915Mhz module. If that doesn't work, than it's my board ...
EDIT: Yup, RFM12b works at @915Mhz on latest PCB... So I'm pretty sure it's something with RFM69CW modules... Gonna contact seller...
-
The BOMs for 433 and 915 are quite different even if the rest of the parts look similar. The values are different.
Especially on the RFM69HW this is a lot more obvious.
For the 868/915mhz there is only 1 passive value difference in BOM, other than the 0ohm reisistors used to route the output signal.
-
Hey Felix, where did you find BOM for these modules?
Anyway I got replacement modules today in the mail (this time it's RFM69W) and soldered to same PCB where I had issues before, and everything works great at 915Mhz! Even without any antenna attached :)
-
Hey Felix, where did you find BOM for these modules?
Anyway I got replacement modules today in the mail (this time it's RFM69W) and soldered to same PCB where I had issues before, and everything works great at 915Mhz! Even without any antenna attached :)
The SX1231H datasheet is a good place to start. Then I reverse engineered the Hope version, they add a bunch of passives for their output routing path. This is because they use the same PCB for many different versions of the RFM69/65.
I would recomment soldering the antenna otherwise you might risk reflecting that signal back into the transceiver and potentially damage it if you run it like that for a longer time.
-
Hi Lensdigital,
I had the exactly same issue, module labeled as 915Mhz but it only work as 915Mhz if I turn on the "HiPower", but the range is just a few meters. Switching to 433Mhz, they work perfectly! My modules don't have that extra capacitor as per your picture (http://lensdigital.com/mote2/SQM_915Mhz.jpg).
Well, at least now I have another person with same issue... contacting seller again!
Cheers,