Data comms "over the air" (~5 miles)

Hi,

For reasons that need not concern us here, I would appreciate any ideas / comments /etc on the following "spec" (for want of a better term):

Background: I run in an event every year (OK, the Virginia City Hillclimb) which is timed by cutting the usual beacon at the bottom of the hill - This sends a signal to the (nearby) timing computer. When you reach the top, you cut another beam and that signal gets sent back to the computer via what I believe is a dedicated phone line - Good ol' POTS link. The trouble is this phone line is very often "dead" when we arrive for the event.....

Because of the nature of the event (it's just over 5 miles and climbs about 1200ft) "line of sight" doesn't exist. The corner workers (etc) all have voice communication enabled via a "repeater tower" that sits across the valley - Everyone can "see" the repeater and voice now works great.

What I'd like to try and do is eliminate the POTS line - I don't much care how, but it seems to me we should be able to send the top timing signal back via the repeater.

I have absolutely no experience with this stuff, so any pointers would be much appreciated - As Clarkson would say, "how hard can it be?"......

BTW: This question is purely for my own edification, I have not been asked by the FCA (or anyone else!) to investigate potential "solutions".

Cheers,
Ian

 

How about you get a laptop that someone has a CingularATT/Verizon data card and link it up that way. if you're on a hill you should be able to get some sort of data signal. You'll just need power to keep the laptop running.

Not sure what kind of accuracy you need, but you can sync up your computers using NTP so that they are within a milliseconds of each other and not reporting from different time clocks.

 

Very interesting idea - Thanks!

We do have power. "Absolute" accuracy is not as important as repeatability (IMHO) - As long as its reasonably constant any latency in the circuit should be "the same for everyone".

However, we are pretty "remote" in cell phone terms - I'm pretty sure my iPhone never sees 3G up there for example. Having said that, two "sync'd" computers deserves further thought....... Thanks again,

Cheers,
Ian

 

anthem's idea is nice - each computer creates a timestamp when the car crosses the line - the one at the end then sends his stamp back to the one at the start, so you don't care about eventual communication latencies. If your current system works by sending a direct signal over the phone line (and not the timestamp data), you'll face problems reproducing that with reliable and exact radio transmission, I think.

 

Definitely worthy of further investigation........ However;

I believe (but could be wrong), that's exactly the way it works - Cars are "queued" (into the 'puter) at the start - Trigger the beam - Start the timer -> at the top, trigger the beam -> stop the timer -> Send a "beep" to the bottom..... Note that there are many cars running at once, but they all arrive at the top in the same order (No overtaking! No walking between the beacons!.....)

?

We now have pretty reliable voice (only messed up by "operator error" IME] - The "repeater tower" does work. (*Huge* kudos to the guys who nursemaid it all W/E btw!) - "All" I want is to send the "beep" referenced above back to the bottom.

Putting transponders in all the cars was mentioned - Big $ - Who pays? - Best case scenario would require 20-30 xponders being rotated among cars - *Very* careful management of xponder movement mandatory......

Or, rent/buy a xponder if you want "accurate" timing?.....

I don't like either option! Any other ideas?

Seriously, thanks for the comments - Appreciated - Keep 'em coming!

Cheers,
Ian

 

It all depends on the way you send the data, the accuracy you want to achieve, and the reliablity of that accuracy.

a) You use internet, a local network, or generally any data connection to send your beep exactly in time. What if Windows decides that the network request has to be delayed by half a second? It will work in most cases, but one lag equals one worthless time equals one furious driver.

b) You use an analog connection, like you do with voice. The problem I see here is that a short bit of static interference or the random truck driver on the next highway using his CB might prevent the transmission from being "understandable". This isn't a problem when you talk with a person over radio as you can simply repeat it, but you have an exactly timed single event. Again, it will very likely work in most cases, but one failed transmission equals one worhtless time equals one furious driver...

My guess is that setting up your hardware to transmit the beep over radio, to filter the noise etc etc is about the same, rather more, work as telling each of the two synced computers to generate a timestamp the moment they receive their beep. As soon as you have the pair of timestamps, you don't really have to care if they have to be transmitted 1 or 10 or 100 times to arrive correctly. Even if your radio transmission (or rather network if we're sending around timestamps) completely crumbles, you can manually collect the (stored) timestamps from both computers to compare them.

 

Use NTP - widely available and generally what everyone in computer networks uses to sync timeclocks. Everyone over the internet is synchronized using NTP to atomic time clocks. You could easily use NTP on client computers like windows as well as to micro to milliseconds which is more than what you need here. All ntp does is sync the time between computers, and once computer timeclocks are sync'ed, you take a timestamp as they cross. Then you can send it over whatever network you need/want as it doesn't matter - its been stamped against a timeclock and is not dependent upon when the message arrives at the destination station. You can send it via email, via whatever.

I think the main thing is getting something that will transfer - whether that is an analog telephone line, a wireless connection of what not. Once you solve that - the rest is generally not that difficult.

 

Absolutely, that's what I proposed, or at least tried to do so But Ian's system now seems to rely on an analog signal that triggers the computer at the start to stop taking the time.

 

Fair comment - But, at this event at least a second or two is really neither here nor there - Someone will probably disagree but simply having times published on Sunday evening is wonderful! - Rumor is that a few years back the timing system *never* worked and all the times were "invented" by the organisers! - No big deal! - It's a "non-competitive" event after all! [Yeah, right!.....]

Having said that, I'm seriously thinking about the "two distinct PC's" approach - The whole thing revolves around everyone starting and finishing in the same order - If the "timestamp" records for starting and finishing were on separate 'puters I reckon they could be merged afterwards to generate times.

Cheers,
Ian
EDIT: PS, I am familiar with NTP btw - Thanks!

 

You do not necessarily need everyone to start and finish in the same order - I guess you'll have someone at the start and finish anyway, so they might add the number of the car (or whatever you use to identify them) to the timestamp. Create a database or an Excel table, add the # and the timestamp, and merge them when the event is over.

 

This is along the lines I've been thinking. There are "many" helpers at the bottom who "queue" the cars at the start. Many years back there were also people at the top getting 'em out of the queue - There are still people there, but these days they have nothing to do with "timing" - Thats handled automagically via the darn phone line......

This definitely needs more thought! - It's cast in stone that you "finish where you started" - If anyone stops on course (for example) the timing system is, I believe, to all intents and purposes, shutdown - If you were unlucky enough to be on the hill behind the broken car your time gets thrown out - Everything is "reset" and started again - I'm thinking the "timers" at the top could "batch input" the car #'s from voice instructions from the bottom - Wouldn't then need to "monitor" every car as it crosses the line.

I also hear you on the DB - I don't think it would be that hard (Famous last words!) - In fact, the S/W will "know" that anything below, say 3:00 is impossible and anything over, say, 5:30 should also be "ignored" - The "problems" come if anything gets "out of sync" of course. OTOH, it's not like missing an official time is the end of the world - My timing sheets for previous years have always had at least one or two runs that were simply labelled "---" - No big deal....

I guess I'm going to start thinking about the S/W.

Thanks again,
Cheers,
Ian

 

Only after reading the above BTW! - I'm not trying to claim I thought of it until you guys posted.

I guess it's this comment that surprises me a little - Good ol' "Morse code" is *way* more than I want to send - Plus, we've got a repeater available!.....

The other "advantage" is the S/W doesn't need to be changed........

Again, thx for the ideas,
Cheers,
Ian

 

Ian, our timing system does not depend on the phone system. We only use that for voice com.

The timing just uses a wire up the hill and we send a 48 volt signal. The lights open and close that circut. That system has been 100% reliable.

We do however need voice com from start to finish and the phone system has been the most reliable and least expensive. The problem comes from the annual hook up. To avoid paying phone service 12 months a year it is disconnected and reconnected every year and the brain dead idiots screw it up most times. That is where the delay is. In past years the management of the event would have someone go up a few days before to cover those sort of things but that was then and this is now.

I no longer run timing but did for a while and overhauled the system so am asked to intervene when things go bang. Last year that happened because a brand new finish line light just decided to quit working. Not much we can do to prevent that.

When management of the event was better there were no delays. After the timing system and com system were both overhauled a few years ago both were up and running before the drivers meeting was finished and ran like a train all weekend. Management of the event has gotten so bad the people who were responsible for that level of operation have left and the event has resumed Gong Show like operation.

Until the club fixes that nothing will improve.

 

Firstly, thanks to Brian for the comments. [I'm glad I mailed you as I suspect "tech toys" is not one of your usual haunts....] I obviously now have even more Q's:

Once again, I stand corrected!

Another wire outside of the phone line pair? Did "we" install it? Or, is it using one of the spare lines in the phone circuit? [ie, the phone's got to be "live" for it to work?]

That I certainly didn't know! - I figured "race control" to the top of the hill comms were handled via the radios. Wrong again! [Having said that, my Q becomes "why do we need voice between the two - At least if things are working as advertised?]

I'm guessing now (always dangerous!), but it sounds as if at least some of last years times were "stopped" by someone talking into the phone at the top?

Cheers,
Ian

 

It is a spare phone wire but it is in no way connected to their system. We are connected at both ends.

For the timing we need to be able to talk. For one we have to turn the lights on and off for non timed vehicles to pass through. If a car starts at the bottom but a resident want out at the top and we let him go the system will stop the timer for the car on course. That scenario alone happens about 40 times a weekend. To expand the radio system to 2 frequencies would be expensive, we have explored that option.

Since we have been using paid course workers there have been no assurances that they are capable using their brains or the equipment properly so yes that has happened.

 

Hmmm..... This is getting serious - May be time to take it out of the "public domain"

I guess I'm "volunteering" to try and improve things - Which raises even more Q's:

OK - So, even without voice comms [which may or may not work depending on Nevada Bell and/or mgmt?] we can hook into a "spare wire"? - *If* that could be "guaranteed" (connectivity between the beacons and the 'puter) then the "two separate, but sync'd PC's" idea discussed above becomes moot.

Understood - As always, it's the "exceptions" that cause problems. [Although, "Timestamping" cars into a DB for later merge would, I think, handle this issue.]

I thought our radios already had multiple channels? With apologies in advance if I'm missing something, but why can't we have a radio in the "timing van" and another at "post finish" operating on a different channel to the default (corner worker) channel?

Lack of "training" & awareness of the way the system works maybe? Raises another good (and maybe offline) question - IIRC, the "paid corner workers" were put in place so as to avoid idiots like me screwing up the radio transmissions between race control and the workers. [I seem to remember a claim that it would also lessen the time "wasted" moving volunteers (and their radio "etiquette") onto and off the hill. That seems to have failed IMHO.

As always, my 02c,
Cheers,
Ian

 

I thought our radios already had multiple channels? With apologies in advance if I'm missing something, but why can't we have a radio in the "timing van" and another at "post finish" operating on a different channel to the default (corner worker) channel?



Lack of "training" & awareness of the way the system works maybe? Raises another good (and maybe offline) question - IIRC, the "paid corner workers" were put in place so as to avoid idiots like me screwing up the radio transmissions between race control and the workers. [I seem to remember a claim that it would also lessen the time "wasted" moving volunteers (and their radio "etiquette") onto and off the hill. That seems to have failed IMHO.

As always, my 02c,
Cheers,
Ian[/QUOTE]

The radios have 2 channels but the repeater system can only carry one at a time. That system would need very expensive upgrading. Also we really need those channels just for race control. As you can imagine safety is of paramount importance and between RF interference or just someone else on one of the channels the ability of race control to switch channels at any moment is required. Again from experience I know this is a common problem and is done several times a day. It in not the exception...it is the norm.


The paid cornerworkers were hired for 2 reasons, neither of which you cite. One is to reduce the number of times worker changes take place so everyone gets to have longer run periods. It has helped that a great deal. The other is to keep a safe minimum number of people on the hill required everyone to report to their job as required every time or there would be a shortage and we would have to stop the event and look for volunteers. We have had personal emergencies, people whose car broke so they just went home etc. It was a rare year when there wasn't trouble in that area of some sort. The paid corner workers have fixed that. The other problems are a byproduct of that fix. We contract a company to do it and they send bodys, some good, some bad. There have been cases of the company refunding some of the fee and apologizing for the quality of the help.

Improving the system is always a good idea. I have no problem with that but getting our signal up and down the hill isn't one of them. The lights at the bottom and top have been trouble intermittantly which is one of the reasons you and I spoke of the transponder option.

Finding a cheap and reliable way to expand our voice traffic capabilities is another that would be good. One thing to keep in mind though, for a couple of reasons I will not go into here the club is not excited about any large unreimbursed expenditures for equipment that will be primarily for the hillclimb.

Also another thing which you might not be aware. After all the timing and scoring equipment got squared away and I stepped out of it, in the off season our equipment trailer was broken into and some of the equipment was stolen. I offered to assist in recreating what was there because I knew the system better than anyone. I was not taken up on that offer. The event took place with a bunch of new untested equipment assembled by someone who only had a passing familiarty with it. Since it didn't work and I was there I was asked to make the big pile of parts work again. It was a real firedrill and is a symptom of a larger problem that is the real reason I am no longer involved.

If you want to take over the event I will be happy to help. Until then you can just find me at the Gold Hill in the pub. I think you know the way. I'll save you a seat.

 

Whoa! I never meant to imply I wanted to "take over the event", just that I was willing to help......

I *thought* (mistakenly as it turns out!) that the timing system was one of the "weak spots" in the setup and figured I'd ask a (somewhat rhetorical) question to see if anyone had any ideas for "improvement" - I'm still a little surprised nobody jumped in with an RF based solution (albeit possibly way out of our budget range), but as we can see from the above, there are (or could be) alternatives. However, given my new found knowledge it seems this is moot.

Again, I didn't know that. But, I do remember our discussion that any budget would be, basically, zero. [Thereby ruling out the transponder approach of course.]

+1 [I remember well the pre-corner worker days and struggling to find people to volunteer] - Again, I wasn't trying to imply they're not a good idea, and apologize if I gave that impression.

OK - I'll change tack in order to try and do something "worthwhile"!.... [BTW, before anyone suggests "use cellphones you dummies!" reception up there is "spotty" at best]. In this vein, I'm still not 100% clear on what copper is used for the timing signal - Is it using one of the 4 wires in the phone line? [The other two obviously are used for the land line] or is it a completely separate feed? - Reason for asking is we're obviously using one pair for the landline - If the other pair is unused we could hook up a two-line phone system between race control and "post finish" - Would that help, or are "you" looking for comms with others? Who and where? [It would probably be possible to multiplex the timing signal over a voice line, but may be a PITA [Read "needs $"].

As always, thanks for the comments - If nothing else we've got something to BS about in the GHH when I buy your G&T.....

Cheers,
Ian

 

We have 4 phone lines going up that hill. We have 2 of them hooked up to the phone co with 2 numbers. One for start, one for finish. Start calls finish and now they have voice com. The other 2 are not connected to the phone company, they are just wires. We use them to run a 48 volt signal from a battery pack up and back down the hill. The finish lights make or break that circut and that is converted into a mouse click which stops the clock. There is very little load on it so the small wire is ok and the battery pack could probably run it for a week straight. The timing lights themselves run on 12 volt car battery jumper packs and they just make or break a circut that is literally a horn relay and makes and breaks the 48 volt circut. It was all dreamed up years ago. It is simple, self contained and since we cleaned it all up has been reliable until a bunch of it was stolen.

The entire timing system is battery powered because we have no wall sockets. Even the computer is on battery. We run a generator at start but it serves several purposes (and runs out of gas etc) so we can run for some time independant of any AC power. The limiting factor is the laptop battery.