Opinion of circuit (motorcycle application)

[Miles]

Hello, I am designing a cirucit for an air shifter on a motorcycle. I
have included a link to the part of the circuit I am having trouble
with. I had one prototype tested, but it was triggering the 555 timer
falsly due to noise. I want to make sure that when the switch is
open, the output of the schmitt trigger will never go low, as it was
when I had inadquate protection from the swich. The wire from the
switch to the circuit is a few feet long.

My initial circuit had only a low pass RC filter on the input, then
directly to the 555. This triggered the 555 a few times a second for
no apparent reason, other than EMI I presume.

What do you think of the following signal conditioning circuit, given
the 74HC14 and 555 are powered off of a protected 5V regulator?

http://www.milesmckinnon.com/schematic.jpg

Thanks for the input!

 

[John Popelish]

Hello, I am designing a cirucit for an air shifter on a motorcycle. I
have included a link to the part of the circuit I am having trouble
with. I had one prototype tested, but it was triggering the 555 timer
falsly due to noise. I want to make sure that when the switch is
open, the output of the schmitt trigger will never go low, as it was
when I had inadquate protection from the swich. The wire from the
switch to the circuit is a few feet long.

My initial circuit had only a low pass RC filter on the input, then
directly to the 555. This triggered the 555 a few times a second for
no apparent reason, other than EMI I presume.

What do you think of the following signal conditioning circuit, given
the 74HC14 and 555 are powered off of a protected 5V regulator?

http://www.milesmckinnon.com/schematic.jpg

Thanks for the input!

I think you should include some series resistance (say 100k) between
R1 and C1, so that noise pulses entering through the switch wiring are
not applied directly to C1. You might also parallel C1 with an equal
resistor to reduce the voltage swing from 12 volts to about 6. Then
you can eliminate R7, so that the capacitors can do a better job of
protecting the inverter input from radiated spikes from the ignition.
Keep the C1 very close to the inverter. I doubt there is any good
point in having C2.

 

[Miles]

Thanks alot for your thoughts there. Would you say that, with your
changes, it should condition the signal well? Also a question....
Since the whole idea of this is to remove power to the coils for a few
mS (that part of the circuit works fine), do I have to worry much
about induced EMI from the wire to/from the power transistor which of
course is on the same small PCB?

Thanks again!

- Miles.

 

[Miles]

Does this one look better to you?

http://www.milesmckinnon.com/schematic2.jpg

Not sure if you noticed, but the inverting gate is also a schmitt
trigger. I figured some histerisis there would be a good thing.

Thanks.

 

[JeffM]

Your JPEGs would be much smaller
if you didn't have that stupid grid in the background.
Doesn't your software allow you to turn it off?

 

[Terry]

Miles threw some tea leaves on the floor

Hello, I am designing a cirucit for an air shifter on a motorcycle. I
have included a link to the part of the circuit I am having trouble
with. I had one prototype tested, but it was triggering the 555 timer
falsly due to noise. I want to make sure that when the switch is
open, the output of the schmitt trigger will never go low, as it was
when I had inadquate protection from the swich. The wire from the
switch to the circuit is a few feet long.

My initial circuit had only a low pass RC filter on the input, then
directly to the 555. This triggered the 555 a few times a second for
no apparent reason, other than EMI I presume.

What do you think of the following signal conditioning circuit, given
the 74HC14 and 555 are powered off of a protected 5V regulator?

http://www.milesmckinnon.com/schematic.jpg

Thanks for the input!

Firstly the Protel schema you have supplied is 1.2 a MB in size! It
needs to be about 10k max so modem users like me do not reject it as
taking too long to dl.

Secondly, in my humble opinion, you've gone well overboard trying to
protect the trigger input.

Why not just dump the HC14 in the input conditioning circuit, its
speed here may be part of your problem. You can loose all the support
zeners and resistors etc associated with it as well.

Lower R5 to say 470 Ohms and connect the switch directly across pin2
and 0 volts. This will provide a low impedance to the triger pin when
the switch is open, thus increasing noise immunity, and take advantage
of the wide voltage swing needed to triger the 555. If the "switch"
is a hall effect device with open collector the 470R should still be
ok.

A problem waiting to happen is that the LM2937 is sitting across the
vehicle supply, and if there is an alternator involved as opposed to
using a constant loss 12 volt battery (say on a drag bike) there can be
some pretty big spikes, and it may be only a matter of time before they kill
the regulator by exceeding its 60mS transient absolute max voltage of
60 volts.

Use a 1 ohm resistor and fuse in
series to the input of the regulator and crowbar it with a
decent C20 volt zener or similar.

Or any other better automotive dc supply protection circuit

Finally to test it, turn everything on and use a piezo gas lighter with
the spark end protruding and zap the frame, the wires everything you
can without damaging the 555, and see if you can induce some false
shifts.

It may be an idea to do that first actually, to see how the current
design handles it.

That's my 2 cents worth anyway. I've done a fair bit of automotive
work, including designing and building a programmable (PIC micro)
2nd stage clutch controller for a nitro injected Harley 8 sec drag bike.

 

[Active8]

Your JPEGs would be much smaller
if you didn't have that stupid grid in the background.
Doesn't your software allow you to turn it off?

his jpecs would be much smaller if they were gifs. gifs are good for
line (esp horizontal, IIRC) drawings. jpegs are for photos. i get
excellent results with pngs, also. IIRC they come out smaller than gifs.

brs,
mike

 

[John Popelish]

Does this one look better to you?

http://www.milesmckinnon.com/schematic2.jpg

Not sure if you noticed, but the inverting gate is also a schmitt
trigger. I figured some histerisis there would be a good thing.

Thanks.

Sorry for the delay. My ISP has been down for a few days.

The input section looks better. But I would like to see the whole
thing, especially around the coil driver and all the connections to
the 555.

I am worried that your schmitt trigger will drive the trigger input
well above the 5 volt bus when it swings low to high. While low, the
far side of the differentiator cap charges to +5, and then, when the
gate swings high, the cap tries to drive the trigger input to +10
volts. Another 5.1 volt zener across the trigger input would prevent
this.

 

[John Fields]

Sorry for the delay. My ISP has been down for a few days.

The input section looks better. But I would like to see the whole
thing, especially around the coil driver and all the connections to
the 555.

I am worried that your schmitt trigger will drive the trigger input
well above the 5 volt bus when it swings low to high. While low, the
far side of the differentiator cap charges to +5, and then, when the
gate swings high, the cap tries to drive the trigger input to +10
volts. Another 5.1 volt zener across the trigger input would prevent
this.