Adding more than 8 Antennas Using an Impinj Speedway R420 Reader

[Rivalry Events]

Hi Brian et al, 

Would expanding the number of antennas connected to an Impinj Speedway R420 4-port be compatible with Agee Race Timing software? I know that there are up to 8 green antenna lights on the timing software but was wondering what happens when we try to connect more than 8...

The schematic below shows how adapters and antenna hubs can be used to connect up to a whopping 32 antennas, which would be nuts!

(we only intend to use 6-10 antennas to widen our chip start with mat antennas)

[JK Race Timing Systems]

Hi
I would strongly urge you to read a post by dioxin where he explains perfectly the huge disadvantages of using more rather than less antennae.
These units are designed to allow plant fitters do more with less readers.
For example they all allow more read points on a conveyor with 1 reader rather than purchasing more readers. The time between switching between antennae is not as critical in those situations as product will be passing these read points at a way slower rate.
It's hard to explain but once the theory is understood you will try to use Less antennae to cover an area rather than more.
The best example is this.
If you were to line up 8 antenna on a start line and let the runners off and calculate the tag count.
Now take 4 of those antennae out and do the exact same test you may very well find that the test that was more successful was the one with less antennae.
The problem is simple. Only one antenna is on at any one time. It takes a finite time to stop transmitting/receiving on one before it moves to another. During this off time and up to the time that the next antenna in series is up to speed as it were absolutely no tags will have been read. Every time you increase an antenna you also increase the time that the reader is off.
If you still want to use an amount of antenna then there are 3 ways of making that work better. The first is staggering the antennae ultimately widening the area of catchment. That is why you see antenna running down the chute.
The second is run the readers in bistatic mode (out of the question on your reader) or the third is another clever way.
The default way the reader will read is antenna 1 to 4 in that order.
If you have the antenna alignment in the same order on the finish line it is not the best as the field runs across from left to right (picture the knight Rider light).
That means that while the antenna 4 is on then you can be sure that the area around 1 and 2 is dead and for a runner at full sprint he can be missed going through that point.
Now change you antennae configuration either in the setup or actually moving the antenna so that every second antenna will be active. Like the firing action on a car 1.3.4.2 . The reason behind this is that chances are if you have your line setup correct you will have a good crossover between 2 adjacent antennae so 1 covers the other as it were.
This I have tested many times in testing and the results are definite on that one.
Read Dioxins post. He gives the tech specs which I am too lazy to but it's all there
Best of luck

[Dioxin]

You won't only have the problems with timing just mentioned.

From the specs of the antenna hub:

https://support.impinj.com/hc/en-us/article_attachments/200774818/IPJ_SpeedwayAntennaHub_20120924.pdf

you'll also get a 1.3dB inserton loss (about 35% of your signal).

so you'll increase the area covered by having more antenna but each of those antennae will cover 16% less area than if directly driven by the reader.

[Rivalry Events]

Okay thank you! On the Impinj 4 port reader we intend to connect three of the ports to our 3 mat antennas at the finish (directly driven) and use the last port to connect an 8 port hub for 5 more panel antennas to stagger 50 feet along a narrower finish chute so when runners start walking they should be easily picked up if missed at the main finish line. Just trying to achieve 100% consistently.

[Dioxin]

Then beware of the timing. 

I'm not sure how the reader will share the time between the antennae as I don't have one of those but I'd guess it'll be around 100msec each so antenna 1, 2, 3 in the mat will each be seen for 100ms then there will be a 500ms gap while the 5 panel antennae are looked at.

It will be very easy for a runner to pass the mat undetected in that 500ms.

Test it thoroughly with quick finishers before you use it for real.

My preference would be to get a second reader for the chute and use that with 4 panel antennae.

It's 50 feet away from the finish so will not interfere with the mat. 

Also, 50 feet of low loss antenna cable for each antenna is expensive and, even very low loss cable, will cost you 40% of your signal strength.

Plus the cost in $

LMR400 cable at $1 / foot? for 5 cables of 50 feet? = $250

Antenna hub $265

GPIO adapter $150

That's $665

You can almost get a 4 port FX7500 reader for that and put it in the chute with short antenna cables.

The second reader also:

1) does not detract from the performance of the finish mat reader

2) gives you a significant amount of redundancy in your equipment

3) Is known to be supported by the ART software

[JK Race Timing Systems]

Yeah and have you tried using lmr400 . Completely unusable in the field. Impossible to work with

[Rivalry Timing]

We already have a stock of 50 ft cable and yes they are expensive. They are solid core metal and insulated so stiff and can carry the power, but you're right anything longer than that will cause significant gain and signal loss.

My question I guess is do you know any application where more than 8 antennas have been connected to ART software? 

I guess there is no feasible need to have more after reading what you've said, it's merely curiosity so I can understand better.

Also, the largest RFID reader I've seen is Zebra's 8 port reader. Are there bigger readers than that with more ports? How do big marathons do it? Just more readers?

[Dioxin]

"How do big marathons do it?"

You don't need more readers.

The Great North Run (world's biggest half marathon) uses 2 timing mats, each with 1 dedicated chip reader at each of the 5k check points and the finish line..

Although there are 40,000 runners, at the peak rate they pass at less than 30 per second.

Split that across 2 sides of the road and it's only 15 unique chip reads per reader per second.

[Dioxin]

Here's a picture of the start of the Great North Run:

https://huggity.com/wp-content/uploads/2015/06/PUMA-FanPic-at-the-Bupa-Great-North-Run1.png

You can see the timing mats on the start line. Two of them, each about 8m wide.

[Rivalry Timing]

So will ART connect to two 4 port readers at once?

Right now we have one 4 port reader, and we data is sent over a 20' CAT 5 Ethernet cord. 

Now we want to get an Ethernet "splitter" adaptor to have two 20' ethernet cords, one going to each of the 4 port readers. Is that possible? We want data from each of the two readers to get dumped into the main timing computer / database.

Thanks!

[Ohio Race Day]

Search this group for discussion on an 8 port network switch. I've posted a few times with supporting videos. It works. We time a triathlon using three reader on one cat5 coming into the main computer

[Ohio Race Day]

made by TP Link, model TL-SG1008P unmanaged switch

[Rivalry Timing]

Thank you I'll give that TP Link a try. And as long as ART will communicate to both as well we should be good to go! Excited to give it a try. I also read that each reader might need to be assigned a unique IP address in order for the software to communicate to both readers. I hope we don't have to do too much IT work.

[Ohio Race Day]

A couple of discussions about it the links below.  with videos.

https://groups.google.com/forum/?utm_medium=email&utm_source=footer#!msg/agee-race-timing/xcTRdcFCH_0/a2Y3N1HlBAAJ;context-place=msg/agee-race-timing/sSuKClkkG7g/jB7qh1X1EAAJ


https://groups.google.com/forum/?utm_medium=email&utm_source=footer#!searchin/agee-race-timing/tp$20link%7Csort:relevance/agee-race-timing/tJdw6tN9MBk/uSVHOcgLEAAJ