Author Topic: Poor range out of the box [SOLVED]  (Read 38137 times)

mitchese

  • NewMember
  • *
  • Posts: 2
Poor range out of the box [SOLVED]
« on: December 01, 2013, 01:17:41 PM »
Hi Everyone,

I am having a range issue out of the box with a Motino R4 that I received.  I purchased three with RFM69HW @ 433mhz, and soldered the piece of wire into the ANT slot.

After programming with the struct_send and struct_receive example codes from github, they do work and communicate, however the range is around 30cm.  Anything beyond 30cm and it loses connectivity.

According to what I ordered, it was supposed to be R3, but I received R4 (no big deal though).  In case something got mixed up with the frequencies, I also tried setting the frequency to 868 and 915mhz, just to see if that helped improve the range.  All frequencies work equally poorly.  Is there any way to confirm the receiver I have is the one I ordered, 433mhz?

I have tested to ensure I don't have a short between the ANT and ground - multimeter shows this as an open circuit.

On the serial console of the receiver, when the antennae are within 5cm of one another I can see:
Listening at 433 Mhz...
SPI Flash Init OK!
[99]  [RX_RSSI:-74] nodeId=99 uptime=16200 temp=91.23 - ACK sent. Pinging node 99 - ACK...ok!
[99]  [RX_RSSI:-74] nodeId=99 uptime=16500 temp=91.23 - ACK sent.
[99]  [RX_RSSI:-75] nodeId=99 uptime=16800 temp=91.23 - ACK sent.
[99]  [RX_RSSI:-75] nodeId=99 uptime=17100 temp=91.23 - ACK sent. Pinging node 99 - ACK...ok!
[99]  [RX_RSSI:-76] nodeId=99 uptime=17400 temp=91.23 - ACK sent.
[99]  [RX_RSSI:-75] nodeId=99 uptime=17700 temp=91.23 - ACK sent.


If I move it to 30cm, the RX_RSSI increases to high 90s then drops:
[99]  [RX_RSSI:-90] nodeId=99 uptime=103200 temp=91.23 - ACK sent.
[99]  [RX_RSSI:-88] nodeId=99 uptime=103500 temp=91.23 - ACK sent. Pinging node 99 - ACK...ok!
[99]  [RX_RSSI:-90] nodeId=99 uptime=103800 temp=91.23 - ACK sent.
[99]  [RX_RSSI:-87] nodeId=99 uptime=104100 temp=91.23 - ACK sent.
[99]  [RX_RSSI:-92] nodeId=99 uptime=104400 temp=91.23 - ACK sent. Pinging node 99 - ACK...nothing

What can I do to increase the range?  My range requirement is not great, the project is a pair of suits for a costume that need to communicate with one another.  The people wont be more than a few meters away from one another, with some potential objects between them.
« Last Edit: July 28, 2014, 03:06:30 PM by Felix »

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 6866
  • Country: us
    • LowPowerLab
Re: Help diagnosing poor range out of the box
« Reply #1 on: December 01, 2013, 01:30:39 PM »
White dot on the radio is 433 mhz, grey dot is 868/915mhz. Match the settings in the sketch with your hardware.
Make sure you are using radio.setHighPower() if you got the HW version.
If you have everything right, the HW will give you several hundred meters in open air.

jbeale

  • Jr. Member
  • **
  • Posts: 62
Re: Help diagnosing poor range out of the box
« Reply #2 on: December 02, 2013, 12:45:56 PM »
I got several feet of range with a 915 MHz module programmed for 433 MHz.  Sounds to me like wrong frequency or no antenna connection.  Might be worth posting a photo of your module just to confirm the antenna placement, and the actual Tx and Rx code you are using.

mitchese

  • NewMember
  • *
  • Posts: 2
Re: Help diagnosing poor range out of the box
« Reply #3 on: December 03, 2013, 02:59:49 PM »
Hi Guys,

Thanks for the replies, it turns out I was not enabling the highpower mode (missed that line in the setup - reading has a big advantage).  After adding range has increased significantly (more than what I need it for)

KanyonKris

  • Full Member
  • ***
  • Posts: 113
Re: [SOLVED] Poor range out of the box
« Reply #4 on: December 10, 2013, 02:29:07 AM »
This is a gotcha. When I started with Moteino, looking at code examples, I assumed if I had a RFM69HW I could choose to use the high power mode or not. Luckily, I turned it on just because I thought I might as well use high power while I setup my Moteinos. Otherwise I would have run into this same problem.

At the very least the comments in the code should make it clear that you must use high power mode if you have the RFM69HW radio module.

A better solution would be if the HW69 library auto-probed the radio module to find out if it's a W or HW, and if HW set the PA1 and PA2 on. I looked at the HW69 code but I'm not sure if I can make it work. I'll keep tinkering and see if I can come up with something.

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 6866
  • Country: us
    • LowPowerLab
Re: [SOLVED] Poor range out of the box
« Reply #5 on: December 10, 2013, 08:12:17 AM »
The amplifier is implemented in hardware, and is not a firmware switch, so you have to know what you have. Sometimes lunch is not exactly free ...

KanyonKris

  • Full Member
  • ***
  • Posts: 113
Re: [SOLVED] Poor range out of the box
« Reply #6 on: December 10, 2013, 05:13:31 PM »
Bummer. So if you turn PA2 on (0x11 bit5=1) and read it back, both the RFM69W and RFM69HW will report that it's on even though the RFM69W doesn't have a PA2? Same with RegTestPa2 (0x5C)? Ugh.

Could RegVersion (0x10) be used? Does it report a different version for RFM69W and RFM69HW? But then what if they revise one of the parts and it changes? Again, ugh.

Felix, reading into your reply are you saying that these two modules are register identical, thus there's no way to tell them apart from software? Triple ugh.

Well, I guess the HopeRF folks thought most customers would use one or the other module. BTW, is there any downside to dropping the RFM69W and just use the RFM69HW? Seems like there's always the option to turn down the RFM69HW power, and thus make it the same as the RFM69W? Well, hats off to you for making both available so customers can choose, I just hope the support questions don't drive you crazy.

P.S. Those HopeRF datasheets are, um, NOT the best I've seen.
« Last Edit: December 10, 2013, 11:24:20 PM by KanyonKris »

jbeale

  • Jr. Member
  • **
  • Posts: 62
Re: [SOLVED] Poor range out of the box
« Reply #7 on: December 10, 2013, 06:10:16 PM »
Maybe the code could do auto-sensing simply by trying both settings of the power switch, and see which one works?  But if the power amp is simply not stuffed in the -W version, is there any harm in turning on the high-power setting anyway?

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 6866
  • Country: us
    • LowPowerLab
Re: [SOLVED] Poor range out of the box
« Reply #8 on: December 10, 2013, 07:28:40 PM »
Unfortunately if you use the W with high power turned ON or the HW with high power OFF, they will simply not work at all.
I think physically the difference is obvious. So I guess blind people would have somewhat of a better case on this...

The RF chip is supposedly a Semtech SX1232, which has the ability to drive high power amps to get to +20dBm output. HopeRF implemented that on the RFM69HW.
I take that as a benefit rather than a nuisance. I'd rather deal with the same chip, same format, same pinout, same registers, same lib, 1 line difference, than having 2 different libraries with 2 different sets of code and 2 different everything else.

See page 23 in this doc where it explains some about the different possible RF amp circuits that are implementable: http://www.semtech.com/images/datasheet/sx1232.pdf
HopeRF's RFM69 datasheets like pretty much all other datasheets are just copycats of the original chip makers, branded with their own logos and deleting the extra amp stuff for radios like the 69W.

I think the HW is nice to have when needing the extra power and can certainly substitute an upgrade or replace the W altogether with a minor difference in price.
However I'm finding the W is sufficient for all around my property - my mailbox notifier being the farthest and still kicking at around 85dBm below grade in my basement where the receiver is. So I could not be happier. The RFM12B was struggling to get that far and half the time would not make it.

jbeale

  • Jr. Member
  • **
  • Posts: 62
Re: [SOLVED] Poor range out of the box
« Reply #9 on: December 10, 2013, 08:15:11 PM »
If you want some extra signal margin you might also consider putting the receiver somewhere besides the basement :-)  But it's impressive that it works even there.

john k2ox

  • Full Member
  • ***
  • Posts: 111
Re: [SOLVED] Poor range out of the box
« Reply #10 on: December 10, 2013, 11:19:19 PM »
I have defined the first 31 locations in Moteino eeprom to store individual radio info.  The info includes the radio ID, freq adjustment coef., temp cal factor, ... and options.

The option byte has a bit for external ram an another for the high pwr version.

After I init the radio I read in the "state" from eeprom. This way my sketches work with any combination of hardware without modification.

If anyone is interested I can share the structure and the editor.

john

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 6866
  • Country: us
    • LowPowerLab
Re: [SOLVED] Poor range out of the box
« Reply #11 on: December 10, 2013, 11:42:12 PM »
I keep all the brains in my lab in the basement so it's easy to monitor everything when I'm working here. Also keeps the house nice and tidy without wires and power adaptors somewhere in a closet.
The mailbox notifier is out there in bitter cold transmitting every half minute, I get every packet and the signal is strong at -80-90dBm, so I'm happy. It's a RFM69W by the way, not HW.
The original 9V battery died mid November so I changed it to a fresh 9V and upgraded to RFM69W. The battery voltage drops significantly in the cold winter and rises when it's hot. A 6x9V batt pack is $13 at HD, so $2 gets you 10 months of mailbox notifier service .. not bad I think.

KanyonKris

  • Full Member
  • ***
  • Posts: 113
Re: [SOLVED] Poor range out of the box
« Reply #12 on: December 10, 2013, 11:57:31 PM »
Unfortunately if you use the W with high power turned ON or the HW with high power OFF, they will simply not work at all.
I think physically the difference is obvious. So I guess blind people would have somewhat of a better case on this...

But being new to Moteino, I didn't know what differences to look for, or even that I MUST know which RF module I have and use the correct mode in software or it wouldn't work.

Here's the define from MotionMote:

//#define IS_RFM69HW    //uncomment only for RFM69HW! Leave out if you have RFM69W!

I suggest:

//#define IS_RFM69HW    // Must be set correctly or radio will not work! Uncomment for RFM69HW, comment out for RFM69W.

Thank you for the link to the Semtech datasheet. It is clearer than the HopeRF cut and paste job.
« Last Edit: December 11, 2013, 04:48:27 PM by KanyonKris »

KanyonKris

  • Full Member
  • ***
  • Posts: 113
Re: [SOLVED] Poor range out of the box
« Reply #13 on: December 11, 2013, 12:05:58 AM »
John, that's a very good way to address this issue. Yes, please share.

I have defined the first 31 locations in Moteino eeprom to store individual radio info.  The info includes the radio ID, freq adjustment coef., temp cal factor, ... and options.

The option byte has a bit for external ram an another for the high pwr version.

After I init the radio I read in the "state" from eeprom. This way my sketches work with any combination of hardware without modification.

If anyone is interested I can share the structure and the editor.

john

ssmall

  • Full Member
  • ***
  • Posts: 158
  • Country: us
Re: [SOLVED] Poor range out of the box
« Reply #14 on: December 12, 2013, 01:44:38 PM »
Does the RFM69HW use significantly more power than the RFM69W?