Timy and USB

Subsequently to the long email of March 2, 2011 about dropping the USB connection between a Timy and computer:
My best hypothesis follows from the fact that we were providing power to the Timy via a Generator set that had a load governor. I think it did not produce reliable power, and the Timy was not getting sufficiently charged. A result of this could be that it was not putting out the standard 5 volts at the USB port.
I could only recreate the problem by allowing the batteries to get low.
Contrary to those profs who taught us to not change more than one variable at a time, and consistent with having a race next week, I have upgraded all of the Alge software, am using an amplified USB hub, and have bought new batteries. Currently I have no problem.
Original Post
The USB host controller may drop the device if it 'detects' EMI on the interface. I noticed it when using a dodgey long USB cable to a Timy.

In linux the kernel reports something like:

hub 3-0:1.0: port 2 disabled by hub (EMI?), re-enabling...
There are several threads on this issue, but this has the latest post, so I'm adding my comment here.

This is still an issue. I have a very hard day in the timing room today. Timers took times perfectly. USB connectivity between the Timy and the computer was dropping many times.

As other people have pointed out, there are too many variables.

I did not have this problem even once last year. What's changed? 1) A brand new building with other evidence to suggest that the power is not super reliable. 2) A brand new timing computer (cheapest Toshiba they make) running a new operating system (Windows 7 as opposed to XP). 3) New printers and monitors plugged in on the same circuits. 4) Just updated to the new Timy firmware this past week (but issue existed earlier this season). 5) Latest version of Split Second.

At this point, I'm just glad to know that it's a known issue with no clear cause. And, I'm pretty much just posting this to keep the discussion alive.

I'm frustrated, but my results are up on FISSKI.com, and everybody got a system A time (eventually). I'll keep working on it...

-DI
With a little bit more time to reflect, here are some further observations. There are still too many variables, but it suggests some paths forward.
1) I decided to abandon System A and run all on System B, so I switched the announcing monitor (old-school CRT display) to the System B and then I started to have problems on System B, but no problems on System A. However, I cannot say with certainty that the switch in dropped connections exactly corresponded to switching the external monitor.
2) Given the uncertainty of the times we were announcing, I told the announcer to stop. So, the amplifier was no longer being used.
3) I don't recall if the monitor was disconnected. I don't recall if the amplifier was turned off.

Shortly after things got really bad, the problem "resolved" and both systems worked without issue for the remaining 40 minutes of the race. But, they have run for 40 minute blocks without issue before. My gut says that it has something to do with the external monitor and the amplifier.

-DI
The simple solution is to connect your TIMY by serial connection. This is what I have almost always done and it never drops. I have done the USB thing also, with failures, in the past and this is why I go serial. There is no brain damage in figuring out that this works.
Agreed John. When I do occasionally mess with USB, I have my serial cable connected and am ready for a quick switch to dump direct USB connection. We still keep a small number of "timing" computers with Com 1 on board serial ports to go direct from timer to computer as a fail safe.
Well: John highlighted the **ONLY** safe solution.

We tried the Timy's USB connection extensively, with many PCs, different versions of Windows, different hardware connected (i.e. score boards,
modems, mountain wires, etc.), and after all, we were simply left without hope.

Basically, any glitch, any ripple generating EMI/RFI happening anywhere around the Timy, will hang all communications.
Whether the problem resides in the USB driver or in the Timy hardware, or more probably both is still unknown, but if you want to work smoothly, without being afraid of everything happening around you, you must definitely avoid the USB connection with the Timy.

Yes, the classical anti-EMI/RFI strategies may help somehow, but with some side effects. We tried using optocoupling on inputs and outputs, HI-Q shielded USB cables (max 2 ft) and insulating transformers for power. It was
partially working, but with an inacceptable number of extra devices around the Timy. Partially means that even the proximity of the big CC motors of the cable cars gave high noise on the wires, and every day we discovered something new that could easily block the communications.

I has been frustrating. The final decision was to rely on serial.
Cheap, rock-solid, highly reliable..... could you ask for something more than that?
Just keeping this forum moving. Thanks for all of the replies...

I just ordered a USB-Isolator for $99. It won't get here for awhile. Later experience last season showed me that I could definitely generate false starts and finishes on my Timy by connecting or disconnecting external monitors or printers from my laptop. I can make this happen even when the Timy is not hooked up to my main timing cable that runs up the mountain. Therefore, the "blinky-box" or the Tag "line conditioner" do not seem to be things that will work in my case.

Also, I just went and changed the USB power management on my two computers.

I used this link for my XP laptop:
http://www.ehow.com/how_521027...b-power-options.html

I used this link for my 7 laptop:
http://www.sevenforums.com/har...sb-power-output.html

These two fixes make more intuitive sense to me than other things I have suggested.

The USB Suspend Select thing definitely sounds like one of my issues: dropped connections between Timy and PC which is restored if I unplug and replug the cable.

The USB-Isolator website definitely describes what I'm experiencing. I'm less concerned with isolating my Timy from the mountain and more concerned with isolating it from the other electronics in the building.

Apologies to Dirk, but I'm a tinkerer. If not me, who? If not now, when?
I just got the USB-Isolator in the mail. As advertised it is "transparent." In other words, I hooked a USB cable from my computer to the isolator and a second cable from the isolator to my Timy and as far as the computer can tell it's just connected right into the Timy. So, no setup required, just plug and play. It's working in my office. We'll have to see how it works under race situations with lifts running, snow-making, announcing, printers, etc. For now I'm optimistic.

I purchased one of these:
http://www.acromag.com/catalog/746

In talking with Fred he mentioned this:
http://www.blackbox.com/Store/...Isolator-2-kV/SP386A

-DI
Has anyone had success in overcoming this issue on a laptop using PCMCIA to Serial converter? Our experience has been less than fully successful. We get a few pulses from the TIMY to the SST software and then nothing. Backing out of Setup and Test and restarting seems to allow a few more pulses. Then nothing. Any thoughts?
We had the same problem and our fix was pretty simple. First, we always use an octocoupler for our start impulses. Second, the USB cables are sensitive to interference from other cables so we began to isolate the impulse signals from the USB by routing them under our timing desk crating both distance and material as isolation. No problems. Our lesson was a neat timing wire setup was the cure.
I have used PCMCIA serial card adapters and they worked reliably for me. This is a physical port unlike USB which is a software generated virtual port. As such USB is subject to possible interference as the posts in this thread attest. Serial-to-USB adapters can be problematic for the same reason.

My opinion is that the best practice is to have dedicated timing computers with physical serial port/s so that USB connection from timer can be eliminated. A good machine is a Dell D600 laptop. They date back to 2002-04 and have serial and parallel ports, USB ports, wifi and CD burner. They were made in the millions and leased in bulk to large corporations, so you can pick up good used models for under $100 on eBay. They come from the XP era, but with 1GB RAM installed, they will run Win7. Machines of that age will have dubious battery life, but Li-ion replacements can be had for under $30. It's not a big stretch for a ski club to purchase four of these so that both timing and race office have equipment that's not clogged up with family email and pictures and is guaranteed to have the right drivers and software.
Wow, I had no idea there was so much history on this topic ...

My club just purchased a TIMY2 for backup timing. I decided to save some money, and just use the USB cable that comes with the timer instead of purchasing the special serial cable. Installed all the drivers (apparently different for TIMY and TIMY2), did the ComToFile thing, etc., and had everything working in my basement.

Last Sunday was our 1st big race, and I've got no connection. On a whim, I try a different USB cable, and that seemed to fix it. Three days later at our club's Family Fun Race, I can't get a USB connection with either USB cable. Since we weren't using back-up timing, I was able to just switch to serial.

Needless to say, I'm ordering another serial cable. I can't handle the stress of USB ...

-- Mark
One limitation of using timers via serial is it ties up the single serial port most laptops have, leaving nothing to drive the display board. You can use a PCMCIA slot serial adapter if the laptop has one. If it doesn't, you can either admit defeat and drive the display from the B laptop, or hook up another timer and laptop via the second output on the A side of the blinky boxes.
Hi, I'm new on this site but I've heard about this USB-connected Timy EMI issue in hills in our region (Quebec).

We have NOT experienced USB connection failures despite the presence of *many* electronic devices, 3 x computers, printers, wireless internet connectivity, amplifier, extra monitor, even an outdoor HD Monitor to display live-timing.

One difference perhaps: We have our clocks and timing computer powered via a big UPS. Could this be the source of protection? Comments welcome.
I use a UPS too, mainly for surge protection of delicate and expensive electronics. I do think that Alge have done some significant work on the Timy driver to deal with these problems. The number of posts in this thread has dropped off a lot, so either everyone switched to serial connection or the failures aren't happening nearly as much. Nonetheless I still need to run the Com To File utility for my laptops to see Timy with the new driver installed.

I've been running an extra Timy via USB this winter just to see if it will fail; you could call it the B' timer. Timing signals come via blinky box and the computer is one of the 'race use only' machines so there is no background junkware clogging up the works. So far through two weekends of SL it's 100.00% successful, same as the other two timer/laptop pairs (with serial connections). But then we had no problems in the past to speak of either. I do think it was software issues that were taking down USB connection in the past, mostly due to 'dirty' computers.
Hi all,
During the last winter, I was using Timy 2, for Alpine skiing races and training, connected at my laptop with the USB port without any bad issue. The original problems seem to be solved, but I agree that a serial connection would be always safer anyway.
Jacky
skboyd, I had the same problem.

ALGE offers some cables, that you can have through your local dealer (P/N 205-02).

Yet there are some other possibilities. I went to a local computer repair shop, and they hacked me a wonderful and elegant solution: they took a thin 25P Cannon gender changer, took off the internal connections and put a 25p M on one side, 25p F on the other. All pins were soldered pass-through, but pins 10, 11, 12 have been extracted through a lateral hole, to a 10ft cable with a standard 9p F connector. All was for roughly 10 bucks.

Then I used a serial to USB converter. MOXA and BlackBox are my preferred. Although not the cheapest available, they are rock-solid and drivers are absolutely reliable. Some models are also isolated, this is mandatory if you work in a harsh electrical environment.

My personal experience is not to go 'el cheapo' on this connection: yes, obviously an USB A-B cable comes for just 3 bucks, but then.... good luck.
As you may see from the previous threads, while some users seem to be quite happy with the USB connection, the vast majority had to go back to serial to get rid of the many issues related to the USB connection. To be honest, I must tell you that I did not try the recent updates of the USB driver, it might be better than it used to be.

But... serial is simple, works seamlessly and you can go many feet away from the Timy.
Just what everybody needs. Nothing more, nothing less.

I had connectivity issues this past week with Timy via USB for the first time in years. The new FIS rules have abolished sync between runs on a two-run event like SL or GS, so we leave timers running during reset and second inspection. When we test pulses before second run, they are received at the timer but not into SST, despite the Setup and Test tab still showing Timy status as connected. Looking further, the Timy connection dialog box has lost its 'heartbeat'. Unplugging/replugging USB and restarting SST fixes the problem, but it's a glitch nonetheless.

Not sure what changed to bring this about, but it's something to be aware of. Make sure your pulse tests are good all the way to the software, not just to the timers. 

Interesting. I have a different issue, but possibly related.

We use a Timy3 through a serial to USB adapter, which of course shows up as COM3 or COM4 to Split Second.  A couple of races ago, SS hung because of port issues.  Unplugging the USB adapter and plugging it back in cleared it.  I didn't think much of it, but then it happened again at the next race.  This "hanging" occurs after power-down for sync.

I mention it only because it's never happened before.  Perhaps it's the result of a SS update?

-- Mark

Add Reply

Likes (0)
×
×
×
×