Distance between a propellar processor and Ping

RS_Jim

Hi,
I am thinking about making a garage parking system for my wife's car. Basic concept is to use 2 pings along the side walls to determin the distance from the car to the side wall and one at the end to measure the maximum forward distance. The display would be tricolored led (ws5812) with different colors representing relative position. Two side by side both green, right down the chute. Left amber getting a little close to the left, red left big problem! Front green until stopping distance is close then amber, to red when stopping point reached.
The propellar board would be mounted on the ceiling near the opener to allow her to open the door from her Phone thru a wifi connection. Then on a hot day, she could open the door from a web page, then start the car using the utility she has for that.
Any thoughts?
Jim

Domanik

RS_Jim wrote: »

I am thinking about making a garage parking system for my wife's car. Basic concept is to use 2 pings along the side walls to determin the distance from the car to the side wall and one at the end to measure the maximum forward distance. The display would be tricolored led (ws5812) with different colors representing relative position. Two side by side both green, right down the chute. Left amber getting a little close to the left, red left big problem! Front green until stopping distance is close then amber, to red when stopping point reached.Jim

Or suspend a tennis ball with a string from the ceiling that touches the windshield right at the rear view mirror. As a temporary measure of course; you wouldn't want to infer anything about her driving ability.;)

kwinn

I don't think the distance to the pings would be a problem, but if it is a buffer for each ping could be added. Measure to see the length of wire needed if the board was mounted so all 3 cables are the same length and try it with a single ping if you have one.

Edit: I would try it with cat5/6 cable for connecting from the propeller to the ping and the leds.

RS_Jim

Yes, I was thinking of using cat 6.
Jim

RS_Jim

Domanik,
I am more concerned about the left to right spacing, the tennis ball is already in place.
Jim

Domanik

RS_Jim wrote: »

Yes, I was thinking of using cat 6.
Jim

I don't have experience with long lines used with pings; it seems to me:
During the transmit pulse the demand on current may be high. If VCC does sag you could add a large cap at the ping to compensate.
Since cat6 has so many wires you might want to use extras for ground to keep the prop and ping thresholds closer to the same ground reference. The signal will be quieter if you use a twisted pair: one for signal and the other line for ground. A 75 ohm at the prop and a second 75 ohm at the ping will improve signal quality going both ways.

kwinn

Extra grounds are a good idea but keep in mind that the current drawn by the ping peaks when it is transmitting and will be split between all the ground lines. The ping does not send an echo pulse back to the propeller until after it has stopped transmitting so the simplest way to avoid the noise is to wait for a period of time after the transmit pulse is sent before listening for the return pulse. Of course a capacitor on the ping end power lines will also help reduce the peak current.

RS_Jim

Ok, so I am hearing terminate both ends of the transmission line in 75 ohms and a surge reserve cap at the ping. Any suggested value for the cap? Also hearing doubling up on ground. How about power? How about pairing the data line with a ground,but only terminating the ground at the propeller?
Jim

msrobots

Hmm.

Not trying to be sarcastic here. But if your wife has trouble to center the car in a garage, wide enough to open the car doors, you may have a bigger problem here.

I would advice you to install the system on the car, not in the garage. So it would be able to assist her in other parking spaces also. Positive thing would be shorter wiring, negative thing more el. noise in the car.

One of the small TV screens used by back up cameras for RVs is cheap available, use one of the TV objects in the OBEX. Show distance front, back, left and right either as numbers or (better) as visual indicator with graphics.spin

I remember that Parallax lately put some water resistant/outdoor(?) Pings in their shop. (not sure here). Else some shielding against rain and car washes might be needed.

Enjoy!

Mike

kwinn

I would suggest using one pair for the power ground and V+ with a 100uF capacitor on the ping end. Pairing the power lines reduces EMI since the V+ and ground currents are in opposite directions. The capacitor will reduce sharp changes in current on the wires. Not sure where it would be best to put the terminating resistor and signal ground since a signal is generated at both ends. Try what you suggested but try to make it possible to terminate both ends if it proves necessary.

RS_Jim

@Mike, car has far better electronics on it than I could ever build, will park itself, but not in the garage.

@kwinn, thanks,I'll try that. Back in the day when I did audio systems we would do a single point ground regardless of where the signal originated.
Jim

Domanik

RS_Jim wrote: »

Ok, so I am hearing terminate both ends of the transmission line in 75 ohms and a surge reserve cap at the ping. Any suggested value for the cap? Also hearing doubling up on ground. How about power? How about pairing the data line with a ground,but only terminating the ground at the propeller?
Jim

The ground level is what’s important, not so much the power, because you want the three boards operating at near the same “ground reference” for the digital signals.

VCC / Power? Consider that 10 feet of 26g solid copper wire has a resistance of .43 ohms. The ping is spec’ed at 30 ma ICC. That give you a VCC drop of 13mv. Hardly worth fussing over if you consider your VCC has a spec of +/- 250mv.

The 75 ohm series terminations at both ends allow clean 2 way transmission. The trigger pulse width is spec’ed at around 5 micro seconds. If you look at the timing diagram for the ping it shows a blanking period after the fast burst, so waiting isn’t needed except, possibly, between measurements. If the ping was within a foot or so a simple 1K in series would work just fine. Note: The 75 ohms is in series, not a pull down to ground or to a termination network.

Twisting the signal wire with a ground will keep the signal clean by insuring the impedance will be at about 110 ohms (75 resistor + 35 source/both ends). Grounds should be connected on both ends for this wire.

The possible need for a large cap during the transmission burst was speculation on my part. Probably better to put it in than spend time researching. Plus it won’t hurt anything.

kwinn

RS_Jim wrote: »

....... Back in the day when I did audio systems we would do a single point ground regardless of where the signal originated.
Jim

Those cables typically had a common and signal wire (or + and - signal wires) as well as a shield that was grounded at one end only. The shield was to avoid noise pickup, and it was grounded at one end only to avoid 50/60Hz hum pickup due to ground loops. Current analog and digital signalling systems (audio, RS232/422/485, etc) are also grounded at one point in buildings. Amazing how much hum you can pick up by grounding both ends of a bus on a 100 foot cable.

Domanik

In the hookup I described there won’t be any hum, noise, excessive drops or ground loops. Signal fidelity between the prop and two pings will be very good. It'll do exactly what you want.

RS_Jim

I did some quick expermenting with 20 feet of cat5 doubling the power and ground, and used one pair of the cable as a signal path and floated the ground at the ping end. All seemed to work well. Now back to the drawing board for other parts that are not working so well.
Jim