Maker Pro
Maker Pro

Drag race Xmas tree project continued.

P

pimpom

Jan 1, 1970
0
Continuing the saga of my drag racing Christmas lights project, I
have another question.

Since the race is to be held in an abandoned air strip an hour's
drive out of town, mains power is probably not reliable there,
and I don't want to depend on a portable generator either. So I
thought I'd design everything to use batteries for this event - a
12V truck battery for the lights.

LEDs would be efficient but I won't be able to get readymade
LED-array lamps in time - another restriction due to my location.
And I want to avoid the addititonal workload of manually
assembling 14 lamps of several dozen LEDs each and having to
improvise reliable mechanical holders for them. So I thought I'd
use incandescent automobile light bulbs for the lights.

I've finished a tentative design of the entire timing, sensing
and control unit, using MOSFETs to drive the lamps (Relays would
introduce additional delays when times will be measured in small
fractions of a second. Unless a better idea comes up, I'll just
have to accept the turn-on delay of the incandescent lamps as
unavoidable).

The actual wire lengths from the control unit to the lamps will
be 2-3 dozen feet. At first, I thought I'd avoid having to run
more than a dozen thick gauge wires over that distance by placing
the MOSFETs near the lamps and use a cable of thinner wires to
send the gate drives. But I've had second thoughts and this is
what my question is about: Will cable capacitance, inductance and
cross-coupling pose problems with that scheme? Thanks in advance.
 
P

pimpom

Jan 1, 1970
0
D said:
Ghetto style:
A piece of plywood,
4 nails,
A strip of spring steel,
and the mechanics from a microwave dish spinner.
Use inverter between battery and motor.
This is a professional contract between the race organisers and
me. An amateurish lash up has to be avoided as far as possible.
Isn't all that matters is premature start and distance time.??
The tree lights can be doing random timing. When the Go light
lights...it means go and that can happen at any time.
Atomic timing not required.

Unless it's standard that the 'Go' light is predicted by
rhythm.
Tap ...tap ...tap ... press gas!
I suppose without rhythm and the race is part machine
performance and
part eye/foot response time.

That's just it. It's entirely based on rhythm. Pros time their
reaction and skill at anticipating the moment of 'Go' to the
millisecond - literally. This is partly why I intend to change
the lights to LEDs after this race. I'm using a crystal
oscillator for the countdown timer and false starts will be
detected to within a millisecond.

Anyway, my concern is not about possible loss of timing
precision. It's about cross-coupling causing unwanted switching
and cable reactance causing instability and slow turn-on, with
consequent over-dissipation of the MOSFETs.
 
P

pimpom

Jan 1, 1970
0
D said:
??? Huh
I dunno but this doesn't seem like the type of project to have
that
type of problem.
Wouldn't the only cable going to the tree be a pair of wires to
a
single 'start sequence' button?
Battery and all control electronics are on the tree.

No. It's not just a visual countdown system. The lights are
linked with four optical sensors to detect and indicate correct
positioning of the racers before the start and also unambiguously
indicate a false start.

In addition to driving the lights, the control unit has to drive
four pulsed IR LEDs and receive inputs from four IR receivers.
The IR emitters have to be located directly in between the
racers' starting position, the receivers on each side of the
tracks at least 15 ft from the middle. The lights have to be some
distance in front of the racers. All of this means a lot of wires
going in and out of the main box. The start and reset buttons can
be anywhere within a reasonable distance.
 
P

pimpom

Jan 1, 1970
0
Wingsy said:
I would use a low value R at gate-source on the FETs, something
close
to the characteristic impedance of the cable, if you know that.
Something around 200-500 ohms would seem appropriate. Delays
caused by
cable capacitance & inductance should not be an issue for your
timing
resolution. I did a project for a Christmas light show years
ago and
had success driving gates from 300 feet. I think you're going
to be
far removed from any source of external interference, so
anything
that gets you will probably be of your own making. I'd also
suggest a
TVS on the gates.
Thanks for the reply. It's reassuring to know about your
successful experience. The distance I'll be working with will be
about a tenth of yours. Switching speed is the one factor that's
not crucial here - even a 100 µsec rise time shouldn't be a
problem as far as timing resolution is concerned. As there won't
be time for a major redesign, I was being extra careful in case
someone could anticipate a problem I hadn't thought of.
On your problem of timing... could you not measure the on-time
of the
headlights and build in a delay so the timing start trigger
occurs at
a point where the lights become visible? I know you said you've
already finished the design so maybe this suggestion is not
feasible
now.

That's a thought, and the design is not set in stone. But with
the limited time remaining before the race, I'd like to avoid
introducing anything but minor changes unless something makes it
imperative. I'm working on the pcb design right now. For this
particular event, I have to work with materials that I can
quickly collect. For later races, I'm thinking of using triacs to
switch readymade mains-voltage LED lamps, coupled to the control
circuitry via optoisolators.
 
A

Antoon

Jan 1, 1970
0
Wingsy said:
I would use a low value R at gate-source on the FETs, something close
to the characteristic impedance of the cable, if you know that.
Something around 200-500 ohms would seem appropriate. Delays caused by
cable capacitance & inductance should not be an issue for your timing
resolution. I did a project for a Christmas light show years ago and
had success driving gates from 300 feet. I think you're going to be far
removed from any source of external interference,

I'd be carefull with that assumption. He's got high performance race cars
idling just a couple of feet away from his electronics. Their design
criteria
might not include EMC guidelines!


Antoon
 
J

JosephKK

Jan 1, 1970
0
I would use a low value R at gate-source on the FETs, something close
to the characteristic impedance of the cable, if you know that.
Something around 200-500 ohms would seem appropriate. Delays caused by
cable capacitance & inductance should not be an issue for your timing
resolution. I did a project for a Christmas light show years ago and
had success driving gates from 300 feet. I think you're going to be far
removed from any source of external interference, so anything that gets
you will probably be of your own making. I'd also suggest a TVS on the
gates.

On your problem of timing... could you not measure the on-time of the
headlights and build in a delay so the timing start trigger occurs at a
point where the lights become visible? I know you said you've already
finished the design so maybe this suggestion is not feasible now.

Not worth the candle, by using the same type bulb it will be very
nearly equal over all the bulbs and not disturb the rhythm.
 
J

JosephKK

Jan 1, 1970
0
Thanks for the reply. It's reassuring to know about your
successful experience. The distance I'll be working with will be
about a tenth of yours. Switching speed is the one factor that's
not crucial here - even a 100 µsec rise time shouldn't be a
problem as far as timing resolution is concerned. As there won't
be time for a major redesign, I was being extra careful in case
someone could anticipate a problem I hadn't thought of.


That's a thought, and the design is not set in stone. But with
the limited time remaining before the race, I'd like to avoid
introducing anything but minor changes unless something makes it
imperative. I'm working on the pcb design right now. For this
particular event, I have to work with materials that I can
quickly collect. For later races, I'm thinking of using triacs to
switch readymade mains-voltage LED lamps, coupled to the control
circuitry via optoisolators.

That might turn into a step backwards. Look into just how mains
supply LED lamps work, there is a little SMPS in there.
 
Top