Hi !
Thanks for addressing my query fast and simultaneously updating the product web page.

Yeah, testing the RFD900 on high temperature is actually very good idea and mandatory requirement for most of the FPV fliers as the modem gets heated up in the sun, while working at the time when plane is continuoysly heading towards one long waypoint and sun rays are falling on the modem non stop.

Would love to hear the test results soon from you.

Would RFD900 operate from the +5V supplied from APM ?
Would it also operate with 5V FTDI as well ?

I am unable to make the payment as I can not find Paypal option anywhere, Pl. guide.

Regards

Rana

Originally Posted by Seppos

Thanks for the heads up! - yes, thats a pretty key bit of information - I have updated the product pages and website.

The air data rates supported are from 4k up to 250kbit/sec, these are settable via AT commands. The serial datarates are from2400 to 115200 baud. At the moment, the list of rates and commands are listed on the 3DR wiki from the RFD900 page. I will bring this information into the RFD900 pages for easy reference.

Two antenna ports are used for diversity. The Si1000 has built in diversity support such that it will select the best antenna for every packet. Both antennas get used for transmit and receive, depending on which has the best signal.
If you have a specific requirement for one antenna to transmit, the other to receive, it is possible to modify the open source firmware to do this.

Generally I have used APM 1 and 2, with the default settings, 57600 baud, 64k air data rates. If higher rates are needed, 115200baud, 250k air rates are available. It is also possible to turn off ECC and opportunistic resend to increase throughput. Typical throughput rates for mavlink data are in the order of 80% or so due to overhead from the TDMA framing overhead.

Due to popular demand I will undertake some mean HIGH TEMP tests on the modems. One modem will get a good roasting while at full power, maximum data transmission. . Any guesses on where it will get to before it kicks the bucket ?

I'll say it should give up the data ghost at 160F (72C) - maybe less.

Originally Posted by CenTexFlyer

I'll say it should give up the data ghost at 160F (72C) - maybe less.

Ahaha, not even close - we just had it in the oven at *110C*! It just kept chugging along and the data just kept coming through, albeit with reduced duty cycle (due to thermal shutdown) and receive sensitivity (because the LNA doesn't work quite as well when it's that hot.)

Seppo is a bit reticent to push the temperature higher, because even though he doesn't mind sacrificing a module (FOR SCIENCE!) the testing cables are another story...

Prior to releasing RFD900 modems, I had done high temperature testing up to 90 deg C with nil adverse effects. This was sufficient to satisfy specification requirements, but was not an absolute test.

Test setup: Duty cycle limiting disabled, Vcc=5.0V, Radio set to maximum output power, monitoring current consumption, power output, and AT&T=RSSI on the radio. Sending "0123456789" repeatedly. With this we can read the radio temperature on the CPU die, using its internal temp sensor. (uncalibrated at the moment, but a reasonable indication)

After some careful roasting in the environmental chamber, the limit at which RFD900s drop link is at 125deg C . Up to this point, they work fine, with a gradual decay in output power, and approx 8dB or so reduction in RSSI level.
The output power gradually dropped from 29.5 at 25 degC, down to 27.3 at 120degC.

Below are some pics of the modem in the chamber at 120degC.

The cool thing is, after cooling back down, the modem started working normally !
I am sending it back down to -70 degC to see what happens.
This kind of *really* extreme testing is completely over the top, but kind of fun !.

End result: I couldn't kill the RFD900 ! and melted strain relief boots on my nice test cables.

PS:
More information on how duty cycle offset (DCO) works... During the transmit cycle, the radio monitors the radio temperature and if it exceeds a preset threshold (70degC for currently released modems), it will gradually reduce the duty cycle to maintain 70 deg C on the radio. You can see what the duty cycle offset it by typing in AT&T=RSSI in command mode. The last parameter (dco= ) will tell you what the offset is in (%). The maximum offset is 80%, which will still allow data to go through, without risking damage.
From these results, it can be seen that I have been very conservative, and this limiting temperature can be comfortably raised to 100 degC.
Typical difference between the CPU temperature and heatsink temperature are around 15 deg C. The maximum junction temperature on the PA is around 150degC so there is good margin with most normal applications.

PPS2:
Operating an RFD900 at these extremes is completely out of specification and not recommended. Standard disclaimer: *Don't try this at home*

-60degC wasn't enough, -73degC is pretty close to the minimum that our chamber can do. It would be really interesting to send a couple of radios to the edge of space..... Any high altitude balloon people around

I am very glad to see the results of the RFD900 at high temperature extreme conditions. Its very great thing about this modem that even if the temperature in the testing exceeded 125degC, the modem shut down and again started when it cooled down below that temperature.

In fact such high temperature conditions will never come, most people consider +85degC the highest point for normal operation of any RF Modem.

Let me tell you this modem is going to hit in the FPV and UAV market very high and this product will be the largest revenue generating product of your company.

Suggestions for future enhancement;

1. For commercial UAV application, the the telemetry data needs to be encrypted so that hackers can not hijack it. If encription option is already there, then its great.

2. Point to point and point to multipoint network topologies should be supported so as to find place of this modem in the industrial use as well. If it is already there, then its great.

3. Benchmark product for point 1 & 2 could be DNT900.

Many many thanks for shipping my pair of RFD900 modems within 4 hours of making the the payment. Even big MNC's can,t meet such fast shipping.

Are your prices in USD or AUD?

Its AUD

Folks, I have received mine today, quality of manufacturing is really majestic.
It worked out of the box with APM 2560 board, loaded with Arducopter 2.71 firmware.
It is even working with the pair of 3DR Radios I bought from DIY Drones.

Hi narpat007,
Thanks for your feedback!. Looking forward to hearing how you find the performance in your application.
We have a distributor in Thailand, Jani Hirvinen from Jdrones ( jdrones.com ) who has modems listed in $USD. Modems from Australia are sold at $AUD. These days, the exchange rate is pretty close to 1:1, but for those who prefer USD, that option is available.
We are also gearing up for the next build of modems to make sure they are always available from stock.
To date, 99% of orders placed before 1pm Australia time (UTC+10), get shipped the same day, we aim to keep it that way !

Hi Seppos !
My GPS getting locked without any issue even when the RFD900 modem is operating at distance of hardly one feet.
Long Range performance is yet to be checked, for checking this I will use 13dBi Yagi antenna at the ground station and 5dbi omni whip antenna in the air. But since there is rainy season in northern part of our country, till October end, so really difficult to find chance.

Hi Seppos !
I see, there is latest RFD900 firmware update available, what is new / amendment in the new firmware ?

The latest firmware has updated temperature limiting thresholds set to 100 deg, from 70 deg. This will let the board transmit at full power, maximum duty cycle to the rated temperature of the RFD900 of +85degC. Above this temperature, the modem will start to limit the transmitter duty cycle. Based on earlier testing, this is still a safe limit.

Would it be feasible to create a verision that runs at 868Mhz to allow for use in Europe?

Hi Air,
Certainly it is feasible and being looked at. The certification costs for Europe are significantly higher so it is planned to get a certified RFD900 version first for AUS, USA, CAN and NZ.