10 replies to this topic

[SlotCarsten]

Recently, I have been made aware that certain specific motors, high-timed Group 12 in particular, failed to register counts on deadstrip lap counter systems.

 

So I have done a little research, and I think I can explain what is going on.

 

We have all learned in school that when you turn an electric (DC) motor it produces a DC voltage. Certainly if you put a multimeter on it seems that way.

 

However it turns out the truth is more complicated than that. If you use a scope to measure the waveform generated it looks more like this:

 

 

The blue curve is the voltage from the motor, the red is the signal after the LapMaster input filter.

In this instance (Driving a Parma S16D with a Dremel), the motor produces ~3.5 volt, but with small fallout to 0 volt.
This voltage is added to the LapMaster deadstrip voltage of 5 volt, so that the sum is between 5.0 and 8.5 volt.
That is always more than the 3.5 volt which is the minimum required voltage to make LapMaster count.

However when I twist the endbell on this modified motor to simulate higher timing, look what happens:

 

 

The negative spikes on the blue curve goes much deeper, down to -4.8 volt. The motor is outputting reverse voltage pulses.
Now if you add this to the +5 volt, you only get +0.2 volt, not enough to make a count.

 

In this meassurement it is only for very brief moments of time, but it is affecting the filtered voltage response as well, dragging it down.

 

In this case, not enough to be a real problem. But I have not had a change to meassure on the actual troublesome motor.

This is just a simulation. However I think we can conclude that this is the direction the problem is going at.

So what to do about it?
You can boost the deadstrip voltage for a quick fix, but the better solution will be to put a diode in series with the deadstrip wires.

 

When this is done even a high-timed motor will look like this:

 

 

(High-timed motor with diode in series and 3 volt deadstrip voltage.)

 

Note: The deadstrip voltage should have been 5 volt, but was reduced to keep the meassurement within the scale of the instrument.

Now I have described this in a LapMaster context, but it is really not a problem unique to LapMaster, but something in common for all deadstrip driven counter systems.

 

Practical fix

My recommendation would be to put a diode (e.g: 1N4001, 1 amp 50 volt - many others would also work), in the deadstrip wire coming from the right side of the dead strip. The marked leg (scribe/circle) should point away from the deadstrip, and towards the counter electronics. You would need one diode for each lane.

[Tex]

We have all learned in school....

 
Well, not really; I was not a good student.
 
But, I did read your explanation and looked at the charts. I've heard of lap counting problems and it seems you may have some definitive info regarding such problems. I hope a track or two, who had reported such problems, will try out your diode fix and report their findings.

 

Thanks, Carsten!

[Ramcatlarry]

[NSwanberg]

Nice write-up and scope captures!

[The Number of]

Thank you, Carsten, for doing a thorough and professional job of researching this problem. It will help all tracks.  

 

Bill 

[Mike K]

Brilliant analysis!!!

[Tom Thumb Hobbies]

Just another reason LapMaster is THE best system available.

[Dan Ebert]

We recently had some counting issues at our track.  The cars in question that did not count had one common denominator.  The racers were gluing their braid to the guide to keep it flat and prevent twisting in wrecks.  We tested and found cars that didn't have the braid glued counted with no issues.  The racers believe that the cars are bouncing across the dead strip and not counting.  My question is,  could it be possible that gluing the braid may create a low contact or low voltage feedback because of partial insulation.

[jimht]

Braid will normally wear from the rear of the guide forward and as it wears is naturally self-cleaning.

 

Braid with adhesive in it will lose contact and will be quite flat, allowing the front of the braid to get coated with track crud, also insulating it = bad contact. Laps will be missed.

 

Better than glue is a small strip of threaded carpet tape (really gooey stuff) 1/4" x 1/2" inserted between the braid and guide, leaving the braid at the back of the guide unstuck, keeps the braid in place quite well, especially on rental cars.

[SlotCarsten]

Mechanical failure to make contact with the deadstrip is not what this story is about.

 

Of course any failure to make reliable electrical contact with the deadstrip will not trigger a count.

This being because of bad braids, mechanical uneven construction of the deadstrip, car jumping the deadstrip, or even an actual case where my car coasted over the deadstrip without counting then coming to a halt. That caused a little headscratching until we discovered that the motor wire broke a few feet in front of the counter! If the car is not in the slot when it parses the counter it will not count either.

 

These kinds of problems are easy to understand. Why certain (rare) motors don't count at all is in a different league.

[jimht]

Your explanation of the problem with high-timed motors does not explain clearly why the same problem occurs with zero-timed FK motors.

 

Will your solution eliminate the issues with the FKs?

 

Will upping the dead strip voltage solve both problems?