Society of Robots - Robot Forum

Electronics => Electronics => Topic started by: ed1380 on December 20, 2009, 07:44:22 PM

Title: High voltage sensing
Post by: ed1380 on December 20, 2009, 07:44:22 PM
I'd like to replace a tachometer on a motorcycle with a led bar type tach. I'd like to use the atmega8 i have from the $50 robot. the only problem I have is sensing the spark. I could make an iduction coil around the plug wire (cant figure how to wire it up) or just use the same signal wire used by the old tach. only problem is that theres 20kv in that wire.
How would i tune that down to something that wont fry my mcu?

Thanks
Title: Re: High voltage sensing
Post by: SmAsH on December 20, 2009, 08:28:50 PM
Hundreds of Google links out there.
An i idea i saw to solve the problem, have a photo transistor in front of the propeller and have the mega count how many times it spins past. I was thinking, wouldn't something that doesn't even contact the high voltages be better?
But, here are some examples that i found through Google:
http://www.stefanv.com/electronics/tachometer.html (http://www.stefanv.com/electronics/tachometer.html)
http://www.t-r-j.com/Auto/Tach/tachometer.htm (http://www.t-r-j.com/Auto/Tach/tachometer.htm)
Title: Re: High voltage sensing
Post by: ed1380 on December 20, 2009, 08:37:54 PM
lol :D ive seen both links. first one wouldnt work for me, and I already sent out an email to the author of the second one.
any idea on how to do inductive pulse sensing? I know you wrap a wire around the spark plug wire, but what does ground connect to? Do you even use ground?
Title: Re: High voltage sensing
Post by: SmAsH on December 20, 2009, 08:41:34 PM
Ah, i see... Wouldn't ground connect to the ground of you circuit?
I was just thinking, does a such thing as a reverse relay type thing exist,
Using high voltages to control low ones?
Title: Re: High voltage sensing
Post by: ed1380 on December 20, 2009, 09:06:13 PM
maybe put alot of diodes in series. couple volts drop on each
Title: Re: High voltage sensing
Post by: SmAsH on December 20, 2009, 09:12:16 PM
Do diodes even exist for 20kV?
Title: Re: High voltage sensing
Post by: ed1380 on December 20, 2009, 10:04:18 PM
well more reading shows 60-100v+ so it shouldnt be that far fetched
Title: Re: High voltage sensing
Post by: SmAsH on December 20, 2009, 10:34:32 PM
Ahh, fair enough then. Will you just have it hooked up to a digital pin that counts how many per minute?
Title: Re: High voltage sensing
Post by: hopslink on December 21, 2009, 03:08:19 AM
What type of ignition system does your bike have. Is it not possible to get a signal from somewhere on the Low Tension side of your system?
Title: Re: High voltage sensing
Post by: Soeren on December 21, 2009, 03:50:07 AM
Hi,

well more reading shows 60-100v+
Measured with what?
A DMM wouldn't see the short 20..30kV peaks.

An inductive sensor (a wire loop) connects to Gnd and the input.
But, you usually connect a tacho to the ground (switched) side of the ignition coil of a Kettering ignition and even if you have an electronic ignition, it will often be wired with the ground switching, so look into that before going inductive.
Title: Re: High voltage sensing
Post by: ed1380 on December 21, 2009, 08:50:51 AM
it has CDI ignition. The CDI box outputs a pulse to the ignition coil. That same pulse goes to the frequency to voltage converter on the tach.
i remember reading that this pulse is around 20kv and the ignition coil boosts it up to 200kv. more reading last night about cdi said the pulse is only 60-100v

The thing is that I know what wire the stock tach uses, but i have no idea how much power is going through it and dont want to burn my chip
Title: Re: High voltage sensing
Post by: Soeren on December 21, 2009, 11:47:05 AM
Hi,

it has CDI ignition. The CDI box outputs a pulse to the ignition coil.
Yes, CDI stands for Capacitive Discharge Ignition and it works by charging a cap and then discharging it into the ignition coil when ignition should occur.

The most important pro with CDI is that the very short rise time gives a high voltage and due to the short rise time, the air (and grime) around the ignition cables is less prone to ionize (which is the prelude to discharges in wrong places).
The major con, on the other hand, is that the ignition pulses are very short, so they demand a better air/fuel mixture and better timing (than regular inductive ignition) to ignite the gas fully in this short time.


That same pulse goes to the frequency to voltage converter on the tach.
i remember reading that this pulse is around 20kv and the ignition coil boosts it up to 200kv. more reading last night about cdi said the pulse is only 60-100v
Your ignition system can't handle 200kV, but I think it should get higher than 20kV, as this can be accomplished on a simple Kettering (contact) ignition with a so called "sports coil".
I have been up to 30..35kV with an ignition transformer (which is way efficient than a coil, due to the closed field).
But I haven't really done that much work with factory made CDIs, so I wouldn't know if they're coiled down, relative to other types of ignition.


The thing is that I know what wire the stock tach uses, but i have no idea how much power is going through it and dont want to burn my chip
You could make a voltage divider for measuring, that's what I do when I want to measure an ignition system.
Code: [Select]
GND o--[10k]--+--[10M]--[10M]--[10M]--[10M]--[10M]--[10M]--[10M]--[10M]--[10M]--[10M]--o HT
              |
            DMM +V
Put the resistors in a plastic tube and fill it with epoxy when the terminals have been made, then you have an HT probe that'll last quite a while.

For the project at hand, you need a voltage divider which isn't quite that extreme. Whether it's resistive or capacitive won't matter. Find the voltage level at your system and just make a simple voltage divider with resistors and then feed it through a limiting circuit (resistance and diodes towards B+ and Gnd).
Title: Re: High voltage sensing
Post by: ed1380 on December 21, 2009, 03:08:20 PM
well I just lost my post cause i cant upload bmp. time to retype  >:(

I understand the part about CDI and the HV probe.

Ive looked at voltage dividers online and the calculators are confusing. they dont say what data to input. BTW it runs on a 12v system. are my calculations correct? (pic)

also all the limiting circuits online are overly complex. I'm a mechanic not an electrician ;D

maybe i should just use an impedence coil? but is the ground near the coil or the spark plug? will it produce over 5v? why dont i just get a replacement tach for $20?  i want a led one  :D

Thanks
Title: Re: High voltage sensing
Post by: Soeren on December 21, 2009, 05:32:01 PM
Hi,

Ive looked at voltage dividers online and the calculators are confusing. they dont say what data to input. BTW it runs on a 12v system. are my calculations correct? (pic)
You need huge resistors to pull that off... 100V in 100 Ohm is 1A, hence 100W would need to be dissipated in that resistor. You would kill the ignition totally that way, however.

Try this (http://That.Homepage.dk/PDF/Ignition_tap.pdf) instead
It has got protection for the I/O-line as well.


why dont i just get a replacement tach for $20?  i want a led one  :D
Have you considered using one or two LM3904 chips for the display?
Title: Re: High voltage sensing
Post by: ed1380 on December 21, 2009, 09:32:30 PM
that diagram looks like overkill  :D thanks

Id like to keep everything software, cause i understand that slightly more than electronics. and to me its easier to tweak :)
Title: Re: High voltage sensing
Post by: ed1380 on December 22, 2009, 10:17:03 AM
i looked at the stock tach. and the ignition pulse goes through a 18k resistor and straight into the v-f converter. number T2259A cant find anything on it. maybe the pulse is weak enough to do that or maybe the chip is made to handle that voltage?
Title: Re: High voltage sensing
Post by: ed1380 on December 22, 2009, 01:39:07 PM
this is pretty awesome

http://www.walktronics.de/html/led-drehzahlmesser.html (http://www.walktronics.de/html/led-drehzahlmesser.html)