Single-Board Universal Module

moss wrote: scraping off the solder mask under microscope with an X-acto blade is not that hard, easy no, but not that bad. Once the mask is well scraped off, add some solder to the copper.
Then you just need to cut a gap in the scraped trace, and reflow on the cap over the gap. Adding "pads"? Not necessary.

Look at eBay for "10X-20X Boom Arm Widefield Stereo Microscope" and you can find a decent solder scope for under $200.

Heh...as if it were actually so easy. Let me tell you how it goes at my house....scrape resist off...looks good...add solder to exposed lines...oh it won't stick...add more heat, crap I jut bumped it and got solder on the neighboring cap....carefuly try to remove it, now the cap came off too...and brought the pad with it...damn, now try to put the cap back, looks ok, but there isn't a connection..add more solder...go back up a couple steps and repeat...house is now on fire and everyone is running for their lives...the board is a smoking ruin of course. Well, maybe not quite, but you should have seen the massacre that resulted from my trying to clean up a few solder bridges on the STM32 processor.

In all seriousness, I have decent equipment (including a temperature controlled iron, hot-air rework, and binocular scope). What I don't have is a steady hand or much experience working with these tiny components. I am amazed I was able to assemble one of these boards on my own and actually get it to work. I am not going to push my luck and damage one further unless absolutely necessary

We've been looking for such a chip for years, and never found one with sufficient overlap in capability. There appear to be a couple of possibilities ofr the CYRF to communicate with a CC2500, but not for the common protocols. The nrf and a7105 also have some overlapping modes, but not at the same data rate.

Sorry I haven't made much progress lately. there were some issues at home that distracted me this week. I have finished some software for testing the RF performance though. I expect I'll get to it this weekend.

I was able to add the 100pF cap to the board as well. It really wasn't too difficult.

I have checked in my latest code which adds a test-mode for the RF modules.

There is good and bad news.
The good news is that I seem to be able to send a signal out of the module:



The down-side is that the signal strength is very weak. it is about 13dBm down from the BUYCHINA cyrf module I have at the same settings. Note that you should ignore the absolute number, I don't have the calibration right for the attenuators I have installed.

I haven't done any debug yet other than to verify the same code works on the Discovery board. It is possible I have some settings wrong. I also haven't gone through and tested all 4 modules yet. But getting any signal at all out is a good start.

Well, I cycled through the 4 transceivers.
CYRF: -41dBm @ 54mA
A7105: -31.5dBm @ 54mA
CC2500: -48dBm @ 73mA
NRF24L01+: no signal @ 33mA

At full power, I would expect the current to be ~400mA, so I think we're nowhere near max power. I need to check that the stm32 I/O is working as expected. Something definitely looks fishy here.

Note that you'll need the following:
* U.FL to SMA connector
* DC-Block SMA connector (I use this as it also has a 6dB attenuator and power limiter: 'Power Limiter for RF Explorer', but you can find similar items on Ebay)
* 30dB attenuator (I found one on Ebay searching for 'DC-3 Ghz 30dB SMA attenuator')
* sma cables and adapters to connect the RFExplorer to the board.

You can also use antenna on both sides, but this can be tricky if there is any wifi or bluetooth in the area since you'll get a lot of interference.

After further experimentation, I don't think the switch is working.
The original code wasn't ever changing the switch controls. However even after fixing the code and verifying the pins are switching, I got the exact same result.
I will next check everything at the switch and make sure I have proper connections there.

It is also possible that the PA isn't working, and what I'm reading is some sort of RF coupling from the board. But I've verified that the PA is enabled as expected.

Also, the CC2500 signal IS there. It is very weak (just above the noise threshold) but I can see it is definitely there.