Missing laps with phidget 0/16/16 and 0.001 pulse width

Hi there

As written this morning, we are still struggeling with missed laps, even after installing the phidget card.

I have already mentioned this before, but the phidget card does "blink" when a car passes by, my logic question is following:

1. Our sensors are built into the slot = we use the guide as open/close mechanism. Does PCL have a minimum pulse length on the phidget boards or should that be detected no matter what? (just thinking i lines of a guide being only 2 cm and a car approx 20 cm long) be aware that the sensors work with 0 miscounts - have been using another program for months and never missed a lap, so it is not a sensor defect. Also see that the phidget board "blinks" to indicate a change in state.

2. could this error be caused by either program error (using the BETA version) - or perhaps an error en the new Phidget drivers??

Waiting for a quick reply - as we have a big race on the circuit tonight where we were planning to use PCL ...

Lars,
Your situation has me baffled.

Do your lap counting "misses" happen randomly with different lanes?
Do the misses only occur when multiple cars are passing the start/finish line in close proximity to each other?

Your slot-mounted, optical sensors should work fine with the Phidget and PCLC. I was going to ask the color of your guide flags, as I have heard of issues with blue guide flags and in-slot optical sensors. However, you mentioned that the Phidget board L.E.D.'s are blinking when a car passes, so that indicates the Phidget is receiving a proper signal.

As far as minimum pulse width, I don't think this is your issue.

I have never had a timing error to-date with a Beta or Released version of PCLC, so I seriously doubt that the PCLC program is the issue.

This leads us to the PC hardware/software setup.
What speed computer and Windows operating system are you using?
Are all other programs - besides PCLC - shut down in the background within your Windows operating system?
I am currently using a 1 Ghz desktop PC and a very clean installation of Windows XP Pro with all of the latest Service Packs.
As far as Phidget MSI drivers, I am still using v2.1.2 dated June 25, 2007 no issues.

As a last suggestion, and I realize it can be a pain, but try using another computer with the PCLC demo and Phidget MSI files installed.
Keep us posted.

Foiler in MI

Hi there

Thx for your reply

Will try to answer all the questions :-)

1. we use slot it standard guides (black) both blue and purple guides does not work with our sensors.

2. the sensors/phidget recognizes EVERY lapcount, it is only the PC/PCL that doesn't -

3. our timing PC is a P4 2.6 GHZ with 2GB RAM and a totally clean install of Win XP pro SP2 - only thing installed besides PCL is phidget and office. Alle updates have been made from Microsoft website

4. the strange thing is, that it is only with PCL this problem occurs - we have tried other programs and they all work without missed laps. But having bought Phidget, we want to use PCL

++

extra

We received advice from an electronic engineer today to adjust our input to the phidget card even further as he suspects the fault is within the on/off limit of the phidget card.

He glanced at the diagram for the interface card and noticed that in order to switch from one state to another the current must either be above 4v or below 1v - especially the last part with below 1 volt could be the reason as our pulse length is only 3/1000 sec - so willl try to add a resistor to make sure that one signal is 0v

Will keep you posted once we have tested - hope this works

If so will put exact schematics on the forum as others might expierence similar problems

:-)

Lars,
More thoughts about your missed laps...

1. Have you tried running the cars slower across your sensors to test your pulse-width concern?
If the cars are detected 100% of the time by running the cars slower across the sensors, then that would support the theory that the Phidget sampling rate is slower than what you need for your in-track optical sensors.
Try pushing the car through by hand for starters and speed up from there.

2. A fellow slot car racer/Electronics Engineer pointed out to me a few minutes ago that the Phidget's LED indicator is hardwired to the input circuit and is NOT an indication that the Phidget controller saw the input. I hadn't considered this.

Anyway, Lars, test per the above and let us know. You may be forced to go to an overhead gantry setup.

Foiler in MI

Thanks Foiler , Bad that LED indicator is not an indication that the Phidget controller saw the input.

Lars, as you know my Knowledge in Electronics is very basic and so i can't tell you what's wrong with your detector and the Phidget Interface.
Actualy you are the first who have missing lap with Phidget Interface so i think it's problem of compatibility of your detector and Phidget 0/16/16, hope Foiler and your electronics engineer will found the problem.
(I don't say that your detector is not good but only that your detector is maybe not compatible with 0/16/16 interface).

On PCLC side, Each time a car cross the S/F line the Phidget MSI trigger a event in PCLC and PCLC do the rest, so even 2 two cars cross the S/F line in the same time windows will put the untreated event in a queue so i don't think it's a problem in PCLC but maybe it's just a bad security lap time setting:
Did you have some message (in the main Screen of Pc Lap Counter) like "Lane x, lap rejected because..." or "Lap time not accepted lane:..." , if you have that kind of messages it's mean that your security lap time is too lower and you have to change some options:




Guy

Lars, did you have some news for us

Hi Guy

We are still struggeling with missed laps despite trying something new. Had our track rented out friday night but they used the old program and we had 0 missed laps.

Hopefully we should have a new piece of hardware tonight that gives us a 100% clean signal of either 0 volts (unblocked sensor) or 12 volts (blocked sensor)

Once this hardware is installed we are able to determine whether the phidget card is/isn't able to pick up the signal or if something else i wrong. No matter what, even if phidget isn't able to read the signal (3/1000 sec) we have a solution for that as well.

Our "expert" hopes to have the hardware ready for tonight but in any case latest during the week we will know for sure.

Will keep you updated as soon as we have results

:-)

Update on phidgets

WE ARE STILL MISSING LAPS

And now our hardware has a 100% detection - both on our amplifier and also the phidget card shows a "blink" meaning that the signal has been received - but still no lap is counted in PCL !?

We have done several tests and they all add up to the same result - either the computer or PCL is not quick enough when checking the USB port, because the phidget does see the input from our sensor.

I tested a lot also on the sensitivity but as far as I can see this makes absolutely no difference in counting laps - did everyting from 1 to 1000 with exactly the same result. For sake of good order we also did a lot of "slow" laps, and here every lap was counted - so the question is whether PCL or the PC is not fast enough to catch the input from the phidget card.

What is the minimum pulse length that PCL can register?
+what is the refresh rate of a USB 2.0 port?

Last but not least - after looking through the entire phidgets website i learned that the phidget card does 125 refreshed in 1 second = is impossible to count laps with 1/1000 sec ?

Hopefully you are able to answer some of these questions guy - further would like to know if the serial interface has the same problem as LPT (doing 4 things at once) if not we will try this for the detectors - until we have a new solution to the missed laps with phidget.

Really looking forward to hearing from you

B/Rgds
Lars

Hi Lars,

I check also the specificaton of the 0/16/16 interface kit and the refresh rate is 125 / sec , so it mean a precision of 0,008.
if your detector pulse width is only 0,001 sec you can have missing laps.
I don't know if the refresh rate is for one IO or for 16 IO so maybe the good pulse width must be 0,008 * 16 = 0,128.
So i think the solution is to increase the pulse width, i think it's possible to increase the pulse width by adding some little electronics component like condensator but as you know my knowledge in electronic is very basic.

Hi guy

Thx for your reply

As things look right now we are going to add condensators to our design - the unanswered question is off course whether the pulse will be long enough but that remains to be answered within the next week or so (have just ordered the parts needed)

FINALLY

We got our new piece of detection hardware built and tested - and this time with GREAT success - the missed laps that we expierenced does not occur anymore.

Very nice that we are finally able to check out all the features in PCL

++

For anyone wanting to use "slot detectors" we will shortly have a diagram showing how this is built - also if anyone is interested we could considder selling the complete package including starting lights and all.

Looking forward to a great race on monday - and many thanks to guy for his assistance

Hello Lars,

We also make use of slot-mounted, optical sensors and are struggeling with our hardware. We also encounter that blue and white guideflags can not be used for our sensors.
Our interface board is a Phidget 8/8/8 and now we are using the digital inputs in stead of the analog inputs (because of the better refresh rate on the digital inputs).
We now build a piece of detection hardware that creates the needed 0V when a car passes the sensor.
But we also encounter missing laps. Since we are not that experienced with building hardware: Is it possible that we can have a copy of your hardware diagram you are using now?
That would be very very great ! Thanx in advance.

Regards,
Danny

Hi Danny

I will ask my "expert" to draw the actual diagram for you + will take som pictures of our hardware

He is out of town until monday, but he will make the diagram beginning of next week

:-)

Hi Lars,

Wow that would be great.
Thanks !!

Hey everyone, I'm new at this forum.

I was reading about PC Lap counter for some time now... It seems really promising.

I come from Argentina.. have basic skills in electronics and advanced skills at informatics. The tracks here are like "third world tracks", so I am planning to start using phidget on my own track.. and I want to install it myself. That's why I wanted to tell you lars, that i would REALLY appreciate if your expert drew that diagram!

Thank you very much for everything in advance.

Hi Gaspacho

My "expert" arrives back in Denmark tonight, so will ask him to prepare the schematics soonest possible.

However if you want to use the phidget for lapcounting only, I suggest that you try out the new parrallel interface that Guy has made for PCLapcounter.

The reason for this is, that though we now have 100% reliable lap counts, the system is no longer accurate down to 0,001 sec - simply because the phidget card uses 0,008 second between each "refresh" - this was seen very clearly yesterday evening where many driver succeded in making identical lap times - something that should only happen maybe once or twice a year. So for the perfect lap count use the new improved parrallel interface.

I you want to use some of the other features that Phidget also gives (e.g. starting lights,track call) then it's a good choice because of the 16 input/and 16 output channels. But for lapcounter must be aware of the above fact.

As promised will post the schematics soonest possible

Good luck with the track

Thank you very much lars, I will be using phidget cause I'll be needing starting lights and fuel control. I am not really worried about the 0,008 refresh rate against the 0,001, here in Argentina everybody is using 0,01 refresh rates anyway... told ya we were a third world country... ;-)

Again, I would be really grateful If your expert drew that plans, I am really lonely here, so I am going to need as much help as possible.

I will be testing everything using a 4 lines track, using an 8/8/8 Phidget. Are the inputs/outputs enough to do starting lights, fuel control and lap counting? I think I will be needing a 16 Input/Output pidget, won't I?

Really, thank you very much lars.

Hi Gaspacho,

Take in mind that the electrical diagrams for a Phidget 0/16/16 must be adapted for usage with Phidget 8/8/8.
The 0/16/16 needs a positive pulse as the 8/8/8 needs a negative pulse to trigger the digital input. The analog input on the 8/8/8 is even slower then the digital one and not wise to use it for counting laps.
We are also using the 8/8/8 and struggeling with our hardware. Now (yesterday) we designed a new version and today we will build and test this. When this (hopefully) should work I will post our diagrams as well.
Our hardware now is triggered via slot-mounted IR-sensors (TCST2000), and the pulse which is send to the Phidget 8/8/8 is now extended.

Regards,
Danny
http://www.slotmanege.nl (sorry it is only in Dutch)

Hi Lars, I don´t know if you're still "playing" with this stuffs however I'd like you to help me.
Did you fix missing laps problems? I'd like to know if you have a schematical electronic diagram of any hardware that could make the pulse longer. I have been using track mounted sensor on parallel port and PLC doesn't detect pulses under 0.5 sec. Could you help me with your experience? Thanks!

lars_jacobsen wrote:

Hi there

As written this morning, we are still struggeling with missed laps, even after installing the phidget card.

I have already mentioned this before, but the phidget card does "blink" when a car passes by, my logic question is following:

1. Our sensors are built into the slot = we use the guide as open/close mechanism. Does PCL have a minimum pulse length on the phidget boards or should that be detected no matter what? (just thinking i lines of a guide being only 2 cm and a car approx 20 cm long) be aware that the sensors work with 0 miscounts - have been using another program for months and never missed a lap, so it is not a sensor defect. Also see that the phidget board "blinks" to indicate a change in state.

2. could this error be caused by either program error (using the BETA version) - or perhaps an error en the new Phidget drivers??

Waiting for a quick reply - as we have a big race on the circuit tonight where we were planning to use PCL ...

Hi Silvio,

Please take a look at this thread.

Regards,
Danny

http://www.slotmanege.nl

Hi Danny,

many thanks!
It seems good!

I'll try it as soon as we can!

Regards

Silvio

slotmanege wrote:

Hi Silvio,

Please take a look at this thread.

Regards,
Danny

http://www.slotmanege.nl

Great! It works!! Thank you!!

Silvio Donato wrote:

Hi Danny,

many thanks!
It seems good!

I'll try it as soon as we can!

Regards

Silvio

slotmanege wrote:

Hi Silvio,

Please take a look at this thread.

Regards,
Danny

http://www.slotmanege.nl