View Single Post
  #159   Spotlight this post!  
Unread 19-02-2007, 13:03
Dave K.'s Avatar
Dave K. Dave K. is offline
Engineer/Mentor
FRC #0930
Team Role: Mentor
 
Join Date: Jan 2007
Rookie Year: 2005
Location: WI
Posts: 91
Dave K. is a splendid one to beholdDave K. is a splendid one to beholdDave K. is a splendid one to beholdDave K. is a splendid one to beholdDave K. is a splendid one to beholdDave K. is a splendid one to beholdDave K. is a splendid one to behold
Re: does anyone have OI and RC radio issues

Quote:
Originally Posted by Stud Man Dan View Post
Dan, We did the same thing as well today. Everyone in the room 1 by 1 would call their voicemail with their cell phones. We noticed an interesting pattern, while my service ( Sprint ) never seemed to cause the robot to lose radio contact, EVERY SINGLE TIME a cingular or AT&T serviced cell phone was used the robot would immediately lose radio signal. We also found that each time the phone would reconnect with the towers to re-establish signal the robot would break the frequency.

My service is Sprint, and the other sprint phones also never-( or atleast noticeably ever ) tripped the radio. Now the way I understand it Sprint uses a different cellphone technology than both Cingular and At&T. Its this technology that we believe is tripping the radios. We know for a fact that the cell phones would trip the radio, because we tested repeadtly for about 15-25 minutes, turning them off walking away hiding them then calling them or making calls with them.

I don't claim to have any clue whats going on, this is probably apparent due to the vagueness of my post however I do realise there is a problem. During our skrimmage yesterday, there were many times where people came very close to being hurt because of the radio frequency loss, then the robot suddenly coming to life near the sides of the fields. You don't have to be an electrical engineer to realise this is a problem. This type of problem not only can greatly corrupt the competition, it could also cause some significant safety issues, or even potential damage to our robots, I'd hate to see a robot lose radio contact as its driving up a ramp and flip over causings thousands of dollars in damage.

I really hope that IFI will do something soon. I completely agree with you Dan this is a HUGE issue. In the case of the bane bots, if problems occured, at the very worst we could always change companies, or come up with custom solutions. With IFI there is no other option. I know they realise this, and trust they will come up with a solution as they always do. Lets hope it comes before regionals, otherwise get ready to see some of the most interesting/unsightly regionals in years.
For those interested in some additional background...

Both the old (eWave) and new (IFI) radios use unlicensed (Part 15) frequencies in the 902-928MHz band. Both radio designs use relatively narrow channels, as opposed to a spread spectrum design. Both designs operate full duplex, with the OI and RC radios transmitting near either end of the spectrum.

"Cellular" carriers use spectrum allocations in the 824-849MHz and 869-894MHz, with the subscriber units transmitting in the lower channel block.

"PCS" carriers use spectrum allocations in the 1.8-2GHz range.

Cingular/AT&T uses GSM (Groupe Special Mobile) as the air access standard for their subscribers. GSM is a TDMA (Time Division Multiple Access) standard which assigns each subscriber a fairly precise time slot in which your device transmits a burst of data, representing compressed voice and control information. The packet frames are sent at a relatively low repitition rate of 217Hz or one of its submultiples.

That 217Hz packet rate results in a rapid on/off keying of the transmitter in the phone. The resulting modulation is AM (amplitude modulation) which can be problematic for many devices, whether they are radios or not. The succeptability some devices have is that components like diodes and transistors make good receivers because they will rectify the incoming signal, and by extension components like operational amplifiers (op-amps) that have transistor inputs are among the most succeptible for near field signals.

CDMA (Code division multiple access) is a spread spectrum technology that assigns each subscriber a PN (Pseudorandom Number) sequence with which to scramble their transmissions. The transmitter technically remains on the air while the frequency that it operates on is rapidly changed in a psueudorandmon manner. Phones that use this technology will typically not impact succeptible devices as much as a device using a TDMA/GSM type of modulation scheme.

Cingular and AT&T Wireless, pre-merger, both held "Cellular" and "PCS" licenses across the country, the local use of which was determined by how they came to own the licenses in a given market. Both companies heritage, in some markets, trace back to the wireline LEC (local exchange carrier) license grants in the early 80's where the LEC was given one 'cellular' channel block, and a wireless carrier was given the other 'celluarl' channel block.

Starting in the 90's, the newer 'PCS' spectrum was auctioned to new carriers, and companies such as Sprint got into the cell phone market, and established cellular carriers also bought spectrum in other markets.

Sprint uses PCS channels for their national system, and is CDMA.

Verizon Wireless came out of the merger's of Bell Atlantic Mobile Services (BAMS), GTE and Vodaphone, and the combined holdings and subequent acquistions and expansions have all utilized CDMA technology as their digital air access standard. In markets where Verizon's heritage is traceable back to the origional LEC, largely in the east, they primarily operate on celluar channels, though many areas both celluar and PCS are used for capacity. Local to me, Verizon uses PCS channels obtained through their acquistion of PrimeCo wireless which had constructed systems in the midwest, and if I remember correctly also had a prescense in Florida.


If celluar phones utilizing GSM/TDMA technology are found to drive the IFI radios nuts, then I suspect that Nextel subscriber units would also have the same potential for problems.

Nextel is neither a cellular nor PCS carrier, they are a SMR (Specialized Mobile Radio) licensee that has a patchwork of 800MHz spectrum holdings in which they have constructed a cellularized, full duplex, two-way radio network. Nextel uses a Motorola propriatary TDMA technology marketed as iDEN (integrated Digital Enhanced Network). iDEN's packet rate results in the subscriber devices transmitter typically being modulated at 22Hz.



In our teams situation, I've not attempted to torture the radio's with any cellphones, nor has anyone been on their phone's nor has any phone alerted (i.e. rang) during our tests. In many cases, I've been the only one present, and my phone is Verizon CDMA using PCS channels.

I'm not saying that there isn't a problem with susseptability from cellphones, just that in our case, I believe we are being affected by a different problem, as our packet loss events appear to always be tied to large demands on our electrical system.
__________________
--Dave