Everything nRF52840
-
@NeverDie soldering bottom/non visible parts needs at least some sort of reflow oven to make sure they melt and make connection. checking for bridges tho is impossible with microscope etc because they are under the chip. diagonal pin pitch is at around 0.55 and the ones on the same row are at 0.5 which is normal qfn package. if you can solder qfn package this is not a problem but the hardest thing is inability of inspection at hobbyist level :)
aligning is very simple, you just need to have correct courtyard printed on the silk layer of the pcb for the chip/module so when you place the chip/module on to the pcb, it needs to sit exactly on top of the courtyard drawing on the silk layer then you know your pads are exactly where they need to be.
sadly tho bigger parts like these modules cant benefit from the surface tension force is not enough to pull the weight but then again its very hard to misplace them especially the e73 module. raytac modules(blue one in the picture) are hard aswell they are like super tiny :) heres how it is under the chip but it baked nicely and it works as expected.
i hope to solder aQFN package aswell but im really not sure about that one :)) standard QFN with visible pads on the side is very nice i wish nrf52840 was like that.
@orhanyor Ah, I see now.
I wonder if you could use through-holes to gain access to the pads on the back of the module? I ended up doing that on one of my energy harvester boards to solder a ground pad that was on the back of a chip:
https://www.openhardware.io/view/733/Buck-Energy-Harvester
It worked, but it was a simpler case. In your case maybe you'd have to cut channels instead of holes. -
@orhanyor Ah, I see now.
I wonder if you could use through-holes to gain access to the pads on the back of the module? I ended up doing that on one of my energy harvester boards to solder a ground pad that was on the back of a chip:
https://www.openhardware.io/view/733/Buck-Energy-Harvester
It worked, but it was a simpler case. In your case maybe you'd have to cut channels instead of holes.@NeverDie yes that is definitely an option ive seen people do it with normal thru holes on those pads and it works. you might wanna pre tin the pads of the module and its good to go but good news for the E73(ebyte) module specifically none of the bottom pads are necessary for the module to work. they mapped every mandatory pin to the edge of the module which is very hobbyist friendly but of course you will have very few gpio pins to work with but still its very well thought by ebyte.
i dont know when i will order the pcb for the nrf52840 but i will definitely report the result and the antenna performance :) i also want to use this antenna for nrf24 but yea it is gonna have to wait not in a hurry atm. -
@NeverDie yes that is definitely an option ive seen people do it with normal thru holes on those pads and it works. you might wanna pre tin the pads of the module and its good to go but good news for the E73(ebyte) module specifically none of the bottom pads are necessary for the module to work. they mapped every mandatory pin to the edge of the module which is very hobbyist friendly but of course you will have very few gpio pins to work with but still its very well thought by ebyte.
i dont know when i will order the pcb for the nrf52840 but i will definitely report the result and the antenna performance :) i also want to use this antenna for nrf24 but yea it is gonna have to wait not in a hurry atm.@orhanyor I'm very interested in your results. I think you may be the first to upgrade a module antenna. It might save a lot of money if cheaper modules can be used. Fanstells are pricey for their good antenna, so if your approach is just as good or even better...
-
@orhanyor I'm very interested in your results. I think you may be the first to upgrade a module antenna. It might save a lot of money if cheaper modules can be used. Fanstells are pricey for their good antenna, so if your approach is just as good or even better...
@NeverDie price wise i think e73 modules are unbeatable they are 3.. dollar or something with the current discount in aliexpress. just for the nrf52840 Soc price in mouser is $5.5 which is still ok if theres going to be a noticeable difference in signal strength. considering much bigger ground plane and bigger antenna i expect better performance but need to test it :) aside from the soc theres a low pass filter which is .30 and a crystal which is probably around .20-.30
that blue raytac module was https://www.raytac.com/product/ins.php?index_id=24 8 or $9 which gave almost the same performance as the e73 so at that price point ive not tested it but BT840F($7) looks better and much easier to solder as the bottom pin pitch is whooping 1.5mm or so.. you can park a truck between those pins :)) -
To free up this thread for its primary purpose, I just now started a new thread for anyone wanting to continue the wristwatch topic: https://forum.mysensors.org/topic/10917/wristwatches-that-invite-development
-
Today i thoroughly tested nrf52840 modules that i have(2 with ceramic antenna and 2 with ipex connector external antenna) all modules were set to max (+8 dBm) TX power and my observation was kinda disappointing. Setting up services, characteristics etc is fun and all but signal strength is disappointing. I tried ceramic to ceramic, ipex to ipex, ceramic to ipex and ipex to ceramic none of them stood out they gave the same performance. As soon as i put the module in a plastic container or theres more than 1 wall in between packets become unreliable and they just keep dropping, connection slows down etc it just doesnt have enough power whereas nrf24l01 with pa module is rocking even if theres like 10 brick walls in between its not even a competition. I dont think its an antenna problem because it doest get any better than ipex connector so the only option is to add some front end module. There are many options for FEM with different capabilities and functions but I will try to add RFX2401C because im familiar with it and its fairly simple. It is controlled with RX, TX pins but im not really sure how to implement it to a code.
seems like softdevice has a support FEMs but its a question if its gonna work in arduino environment :) Gotta try and see..https://devzone.nordicsemi.com/nordic/nordic-blog/b/blog/posts/pa-lna-support-in-s132
also in nordics front page you can see they released a FEM for nrf52 and nrf53's its called nRF21540
quoting from the product page
https://www.nordicsemi.com/News/2019/12/Nordic-samples-nRF21540-RF-Front-End-Module
"The nRF21540 RF Front End Module is a range extender optimized to boost the link budget of Nordic’s nRF52 and nRF53 Series advanced multiprotocol wireless SoCs. When combined with an nRF52 Series SoC, the nRF21540 RF FEM’s +21 dBm TX output power and 13 dB RX gain ensure a superior link budget for up to 16x range extension"
so that kinda proves my point connection is somewhat weak and it needs a FEM any kind to at least bump the power to around 20dBm and it will all be good and usable.
And keep in mind nrf52840 has the most power of all with max +8dBm power, 52832 has 4 and their new workhorse dual processor nrf5340 comes with +3dBm power so probably thats why they are pushing nRF21540 forward to ensure best possible connection. -
Today i thoroughly tested nrf52840 modules that i have(2 with ceramic antenna and 2 with ipex connector external antenna) all modules were set to max (+8 dBm) TX power and my observation was kinda disappointing. Setting up services, characteristics etc is fun and all but signal strength is disappointing. I tried ceramic to ceramic, ipex to ipex, ceramic to ipex and ipex to ceramic none of them stood out they gave the same performance. As soon as i put the module in a plastic container or theres more than 1 wall in between packets become unreliable and they just keep dropping, connection slows down etc it just doesnt have enough power whereas nrf24l01 with pa module is rocking even if theres like 10 brick walls in between its not even a competition. I dont think its an antenna problem because it doest get any better than ipex connector so the only option is to add some front end module. There are many options for FEM with different capabilities and functions but I will try to add RFX2401C because im familiar with it and its fairly simple. It is controlled with RX, TX pins but im not really sure how to implement it to a code.
seems like softdevice has a support FEMs but its a question if its gonna work in arduino environment :) Gotta try and see..https://devzone.nordicsemi.com/nordic/nordic-blog/b/blog/posts/pa-lna-support-in-s132
also in nordics front page you can see they released a FEM for nrf52 and nrf53's its called nRF21540
quoting from the product page
https://www.nordicsemi.com/News/2019/12/Nordic-samples-nRF21540-RF-Front-End-Module
"The nRF21540 RF Front End Module is a range extender optimized to boost the link budget of Nordic’s nRF52 and nRF53 Series advanced multiprotocol wireless SoCs. When combined with an nRF52 Series SoC, the nRF21540 RF FEM’s +21 dBm TX output power and 13 dB RX gain ensure a superior link budget for up to 16x range extension"
so that kinda proves my point connection is somewhat weak and it needs a FEM any kind to at least bump the power to around 20dBm and it will all be good and usable.
And keep in mind nrf52840 has the most power of all with max +8dBm power, 52832 has 4 and their new workhorse dual processor nrf5340 comes with +3dBm power so probably thats why they are pushing nRF21540 forward to ensure best possible connection.@orhanyor Yes, that's been more or less my experience as well for modules without FEM and/or good antennas. At the very least you want a gateway with a good PA and LNA and antenna. At least for me the best module I tested was the Fanstel BT840X module, which unfortunately has a high price that hasn't come down, and it's not particularly small either. If you have a BT840X/BT840XE on both ends of the link, then you probably won't experience any meaningful connection issues in, say, a 4000sf house built using typical sticks and bricks construction, which is what I tested in.
That's why for anything needing guaranteed good range where size doesn't matter, like a leak detector, I use inexpensive LoRa modules instead. In contrast, the area where nRF2840 really shines the most is in remote controls, because it can be configured to very efficiently listen for inbound packets using just battery power for a very long time and with no perceivable latency. Thats due to its tight integration with the mcu and shared memory and the nifty PLC, not to mention the 2mbps datarate. It'd be a lot harder to do that with LoRa, which has none of that.
-
@orhanyor Yes, that's been more or less my experience as well for modules without FEM and/or good antennas. At the very least you want a gateway with a good PA and LNA and antenna. At least for me the best module I tested was the Fanstel BT840X module, which unfortunately has a high price that hasn't come down, and it's not particularly small either. If you have a BT840X/BT840XE on both ends of the link, then you probably won't experience any meaningful connection issues in, say, a 4000sf house built using typical sticks and bricks construction, which is what I tested in.
That's why for anything needing guaranteed good range where size doesn't matter, like a leak detector, I use inexpensive LoRa modules instead. In contrast, the area where nRF2840 really shines the most is in remote controls, because it can be configured to very efficiently listen for inbound packets using just battery power for a very long time and with no perceivable latency. Thats due to its tight integration with the mcu and shared memory and the nifty PLC, not to mention the 2mbps datarate. It'd be a lot harder to do that with LoRa, which has none of that.
@NeverDie Well BT840X has built in FEM SKY66112 if im not mistaken, thats why you are getting good results but how did you make it work in the code. Because for TX and RX you need to tell the NRF when to listen and when to transmit so the FEM can flip switches inside and work its magic.
Tomorrow i will try a remote application where client transmits analog read data every 50ms and reports it to the peripheral device, i will put it in a plastic box and wrap my hand around it and see how many packets it will drop tho im really skeptical after today. -
@NeverDie Well BT840X has built in FEM SKY66112 if im not mistaken, thats why you are getting good results but how did you make it work in the code. Because for TX and RX you need to tell the NRF when to listen and when to transmit so the FEM can flip switches inside and work its magic.
Tomorrow i will try a remote application where client transmits analog read data every 50ms and reports it to the peripheral device, i will put it in a plastic box and wrap my hand around it and see how many packets it will drop tho im really skeptical after today.@orhanyor said in Everything nRF52840:
how did you make it work in the code
IIRC, it's in the datasheet. It is your code which has to flip the switches before and after a Tx or Rx. The nRF52 itself doesn't know the FEM even exists. Think of it as an aftermarket bolt-on, which is what it really is.
-
@orhanyor Your breakout board from the E73 board looks really nice. Is this something you have for sale somewhere? Would you considering selling any extra bare pcbs?
@Jon-Raymond selling them was not my intention but i have around 8 bare pcbs and 2 built ones. i can fix you up with one of those if you want to or i can cook totally brand new one up to you.
-
@Jon-Raymond selling them was not my intention but i have around 8 bare pcbs and 2 built ones. i can fix you up with one of those if you want to or i can cook totally brand new one up to you.
@orhanyor I'd be interested in a couple bare pcbs as I have all the components. It would save me some time rolling my own and waiting for the Chinese pcb supply chain to sort itself out.
-
i just gave this pcb to a production, it has nrf52840 and pa/lna module, 3v3 LDO on it, practically a stand alone board with nrf52. i did not include a battery connector or charger because its for prototyping. i copied the front end part from my nrf24 modules so if i can solder it should work smoothly. tho this aQFN footprint looks very intimidating :) some of the tracks and gaps are at the manufacturers design capability tolerance limits so lets see if they gonna accept it.
-
i just gave this pcb to a production, it has nrf52840 and pa/lna module, 3v3 LDO on it, practically a stand alone board with nrf52. i did not include a battery connector or charger because its for prototyping. i copied the front end part from my nrf24 modules so if i can solder it should work smoothly. tho this aQFN footprint looks very intimidating :) some of the tracks and gaps are at the manufacturers design capability tolerance limits so lets see if they gonna accept it.
@orhanyor Yes that aQFN footprint does look daunting. Do you have a special paste layer for the center pad that limits the paste? I've seen issues where too much paste on the ground pad will actually float the chip off the pin connections.
-
i just gave this pcb to a production, it has nrf52840 and pa/lna module, 3v3 LDO on it, practically a stand alone board with nrf52. i did not include a battery connector or charger because its for prototyping. i copied the front end part from my nrf24 modules so if i can solder it should work smoothly. tho this aQFN footprint looks very intimidating :) some of the tracks and gaps are at the manufacturers design capability tolerance limits so lets see if they gonna accept it.
@orhanyor AFAIU you're going to solder it manually, right? Will you use stencil or are you going to presoak pads? For the stencil it might be also reasonable to order a custom PCB pallet which may help to precisely align the board to the stencil. Similarly, another pallet may help to place the chip right into the center. It will be very interesting to hear from you how it went. Wish you best of luck!
I've ordered some boards with nRF52840/nRF52833, but also had to ask for assembly for exactly this reason. The PCBWay did machine soldering of the aQFN chip, and the rest were soldered manually.
By the way, the PCB price were raised two times: first, for 5 mil tolerance (the board uses NFC so I was unable to route antenna differently), and second, for ENIG finish because of the aQFN-73 package (the factory insisted on it in order to avoid soldering issues). I still have no room for the RESET, but perhaps I need to go your way and drop some unnecessary pins - that's nice idea.
-
@orhanyor Yes that aQFN footprint does look daunting. Do you have a special paste layer for the center pad that limits the paste? I've seen issues where too much paste on the ground pad will actually float the chip off the pin connections.
@Jon-Raymond exactly too much in the middle will raise it and bam you have a failed board :)
@Mishka its a wise choice to ask for assembly because of the reasons you mentioned. ENIG is kinda mentioned everywhere for this aQFN footprint otherwise surface might not be even for every single pin and it will result in failed soldering BUT i still went with HASL just to prove them wrong :))
capped via service is expensive so i just went through every single pin to examine and decide if i really need it or is it necessary for it to work and spaced them out as much as i can. i still got plenty, 3 analog pins, 9 high speed digital pins and rest is low frequency pins which are not recommended for high speed operations. it can interfere with the bluetooth signal. nordic specialist approved the design so if everything goes well it should work.
my only worry was the at the bottom side XC1 XC2 pins you can see they are really close to each other i think it was 5 mil but still it was in the capabilities so it went through. i ll see how it goes its gonna be interesting.
and yes i use stencil and a reflow oven.
how do you like the antenna performance of the nrf52840? sadly it doesnt meet my requirements so im trying this pa lna method to see if i can make it work. -
@Jon-Raymond exactly too much in the middle will raise it and bam you have a failed board :)
@Mishka its a wise choice to ask for assembly because of the reasons you mentioned. ENIG is kinda mentioned everywhere for this aQFN footprint otherwise surface might not be even for every single pin and it will result in failed soldering BUT i still went with HASL just to prove them wrong :))
capped via service is expensive so i just went through every single pin to examine and decide if i really need it or is it necessary for it to work and spaced them out as much as i can. i still got plenty, 3 analog pins, 9 high speed digital pins and rest is low frequency pins which are not recommended for high speed operations. it can interfere with the bluetooth signal. nordic specialist approved the design so if everything goes well it should work.
my only worry was the at the bottom side XC1 XC2 pins you can see they are really close to each other i think it was 5 mil but still it was in the capabilities so it went through. i ll see how it goes its gonna be interesting.
and yes i use stencil and a reflow oven.
how do you like the antenna performance of the nrf52840? sadly it doesnt meet my requirements so im trying this pa lna method to see if i can make it work.@orhanyor said in Everything nRF52840:
BUT i still went with HASL just to prove them wrong :))
I'll be not too wrong if say that literally everybody out there is waiting for the result :-)
capped via service is expensive so i just went through every single pin to examine and decide if i really need it or is it necessary for it to work and spaced them out as much as i can...
nordic specialist approved the design so if everything goes well it should work.Yeah, HDI is too expensive for small batches. I'm just wondering how Adafruit managed to make those boards at that price. As to me the only reasonable explanation is that they produce thousands of them.
my only worry was the at the bottom side XC1 XC2 pins you can see they are really close to each other i think it was 5 mil but still it was in the capabilities so it went through. i ll see how it goes its gonna be interesting.
You might want drop the P0.26 GPIO I think.
how do you like the antenna performance of the nrf52840? sadly it doesnt meet my requirements so im trying this pa lna method to see if i can make it work.
IMHO the SoC itself is fantastic. Unfortunately, so far the only boards I have on hands are a couple of those E73 by CDEBYTE - just received them few days ago, but still had no time for them. Also, a number of nrf52833 based Raybeacons are stuck in fabrication. They're not about long range though - the board has low-performance zero clearance antenna.
The built-in 8dBm TX amplifier is a great add-on. It comes at cost of higher current consumption, of course, so it's subject of proper battery selection for the application. An external 20dBm amplifier may require up to 150mA which will drain battery like a hog. IMHO for the cases when you need it really long a directional antenna is must.
On the other hand, if you have to have good coverage in the center of a building it might be reasonable to stick with an amplifier. But while you're waiting for the boards, I'd suggest to give a second try to your IPEX modules with this simple DIY biquad antenna: https://martybugs.net/wireless/biquad/. It's very easy to build and promises to add more than 10dBi to your link budget. I remember I used it to feed a pair of 1m dishes (i.e. about +25dBi more) for long range wifi - worked like a charm.
Another interesting option might be a phased array antenna. The active array will add up about 10-15dBi which is perfectly aligned to peer direction. Not sure will it be any affordable though. A passive array might be much easier to implement in PCB, but you have to align it manually.
-
@orhanyor said in Everything nRF52840:
BUT i still went with HASL just to prove them wrong :))
I'll be not too wrong if say that literally everybody out there is waiting for the result :-)
capped via service is expensive so i just went through every single pin to examine and decide if i really need it or is it necessary for it to work and spaced them out as much as i can...
nordic specialist approved the design so if everything goes well it should work.Yeah, HDI is too expensive for small batches. I'm just wondering how Adafruit managed to make those boards at that price. As to me the only reasonable explanation is that they produce thousands of them.
my only worry was the at the bottom side XC1 XC2 pins you can see they are really close to each other i think it was 5 mil but still it was in the capabilities so it went through. i ll see how it goes its gonna be interesting.
You might want drop the P0.26 GPIO I think.
how do you like the antenna performance of the nrf52840? sadly it doesnt meet my requirements so im trying this pa lna method to see if i can make it work.
IMHO the SoC itself is fantastic. Unfortunately, so far the only boards I have on hands are a couple of those E73 by CDEBYTE - just received them few days ago, but still had no time for them. Also, a number of nrf52833 based Raybeacons are stuck in fabrication. They're not about long range though - the board has low-performance zero clearance antenna.
The built-in 8dBm TX amplifier is a great add-on. It comes at cost of higher current consumption, of course, so it's subject of proper battery selection for the application. An external 20dBm amplifier may require up to 150mA which will drain battery like a hog. IMHO for the cases when you need it really long a directional antenna is must.
On the other hand, if you have to have good coverage in the center of a building it might be reasonable to stick with an amplifier. But while you're waiting for the boards, I'd suggest to give a second try to your IPEX modules with this simple DIY biquad antenna: https://martybugs.net/wireless/biquad/. It's very easy to build and promises to add more than 10dBi to your link budget. I remember I used it to feed a pair of 1m dishes (i.e. about +25dBi more) for long range wifi - worked like a charm.
Another interesting option might be a phased array antenna. The active array will add up about 10-15dBi which is perfectly aligned to peer direction. Not sure will it be any affordable though. A passive array might be much easier to implement in PCB, but you have to align it manually.
@Mishka i was looking at that long range support of nrf52840 by lowering the data rate to 500kbps or 125 kbps you gain a significant amount of range like 4 fold but it requires a coded PHY which is not available in arduino environment at least not that i know of. adafruit marked it in their to do list but know knows when they gonna implement it probably thats why nrf24's make a significant difference in signal strength when you lower the data rate to 250kbps.
Yes that amplifier draws quite abit of current i think it was up to 350ma when tx is on but ofc it is not on all the time so but still its not for everyone.
i will look into that antenna you mentioned while im waiting for the pcbs thank you for the suggestion!
antenna that i used in the pcb is this one: i read its better than chip antennas but ofc at a much bigger footprint
http://www.ti.com/lit/an/swru120d/swru120d.pdf -
@orhanyor I can confirm this pcb antenna works well. I used it with 52832 a while ago. You'll certainly need to tune it to get the exact target freq though, usually to do on each new pcb design&shape (so it's not shifted too much compared to your others nodes), but it's a good antenna, easy to use, resistant to detuning. This is the recommended ant by TI for 2.4ghz.
Here I just have one bt840xe on custom gw (dual band gw, using BLE for 2.4ghz) with good external ant, it has very good range. I didn't design the 52840+PA because I preferred to have ce/fcc for PA, and for one device only, it wasn't worth the shot. For others nodes, I don't need PA, or prefer mysensors subghz for very long range.
For a devboard, this looks nice, but depending on the target application and sensitive sensors etc, maybe adding an rf shield could help in some cases, especially as you're planning to use a power amplifier. Spectrum analyzer (or a vna) can help to check this (bad rf emissions too, passive filtering can help) and optimize range.good luck for your soldering :)
-
@orhanyor I can confirm this pcb antenna works well. I used it with 52832 a while ago. You'll certainly need to tune it to get the exact target freq though, usually to do on each new pcb design&shape (so it's not shifted too much compared to your others nodes), but it's a good antenna, easy to use, resistant to detuning. This is the recommended ant by TI for 2.4ghz.
Here I just have one bt840xe on custom gw (dual band gw, using BLE for 2.4ghz) with good external ant, it has very good range. I didn't design the 52840+PA because I preferred to have ce/fcc for PA, and for one device only, it wasn't worth the shot. For others nodes, I don't need PA, or prefer mysensors subghz for very long range.
For a devboard, this looks nice, but depending on the target application and sensitive sensors etc, maybe adding an rf shield could help in some cases, especially as you're planning to use a power amplifier. Spectrum analyzer (or a vna) can help to check this (bad rf emissions too, passive filtering can help) and optimize range.good luck for your soldering :)
@scalz thanks for the input, ive seen they tune it by cutting a length from the longer end but in order to do that i need equipment which i dont have so its going to be a hit or miss :) if it doesnt work well i can use SMA antenna but i think thats the least of my worries with that pcb. I like to experiment and see for myself so even if its wasted its ok :)
