Really simple concept -- too complex for me.

[ratsplinter]

Hi

Desperately need help. I have NO electronics experience what so ever, so please bare with me.
The project is for a GSXR superbike.
I need a device that will momentarily cut spark going to the ignition coils (12V).
It needs to be triggered by a switch (normally open), with the ability to set the Kill Time (10ms to 150ms range), and the Kill Delay (10ms to 150ms range).
Sounds simple right ? For me .... not so much.
Any schematic / circuit drawing would help alot.
Also trying to keep the costs as low possible.

 

[timothy48342]

I will not have an answer for you, but maybe someone will. I don't know.

Some clarifying questions:
Make and model?
I found a website with wiring diagram for GSX-R750 and GSX-R1100, each with multiple years: http://gsxr711.com/content/frame.html
Is one of those your bike?

Do you happen to know how the ignition system works? Is it like a car? 12V to the ignition coil, and then high-voltage along spark plug wires to the plugs which are grounded to the frame??? Is it like that or something different?

Now you said, "momentarily cut spark going to the ignition coils" The spark comes from the coils. (in a car) You want to momentarily cut the 12V that goes to the coils, right?
(momentarily cutting the higher voltage spark would be much more difficult.)

While cutting power to the ignition coil would be easier... Will that work? I am thinking that cutting power to the coil for a very short time, (10 to 150ms) might not prevent it from functioning.

(Of course I'm assuming it works like a car.)
Also, what are you trying to accomplish by doing this? There might be more than one way to accomplish you end goal.

A little more info from you would be nice.

-tim

 

[GreenGiant]

I think that you could use some time on/time off delay relays, but those get really expensive really quickly

But as Tim said more info about the end goal would allow us to better help you

 

[timothy48342]

I clipped the attached image from the GSXR750w service manual. (That's not copyright infringement is it?)

I wrote the word "here" with an arrow where I think would be a good place to interupt the circuit. This isn't a recommendation. I don't know nearly enough about modern solid state ignition circuits. I'd hope some expert chimes in to say whether this could damage anything or possible confuse the CPU.
-tim

 

[ratsplinter]

thanx for the feedback so far.
It is for quickshifter as they call it.
A switch on the gearlever activates, which then cuts spark for a set time, which allows the gearbox to "slip" into the next gear.
That is the end goal.
Hope it make sense.
My current objective is to make it work for a 2004 GSXR 1000 Suzuki, and at a later stage for other models (for my mates).

 

[timothy48342]

Oh, ok! So your engine is at full power and you want a momentary loss of power to allow the shifting. (You want the engine to sort of hiccup a bit.)

Wow, cool!!

Here is a site for the service manuals: http://wotid.com/gsxr/content/view/22/35/
The thing is, there broken down as K1/K2, K3/K4, K5/K6, etc. (not by year)
(based on wikipedia) K1/K2 is pre-2003, so I took a guess that you might the K3/K4 manual might apply to the 2004 and only looked at that one.(This one: http://www.wotid.com/gsxr/manual/GSXR1000-03-04_OCR.pdf) The image they have on page 9-18 is similar to the one I posted above for the R750 model. Similar enough that I still think the spot I marked above is a good spot to interrupt the circuit. (I can't seem to steal an image from those PDF's of the manual. They are formatted different.) The spot is after the 12V splits between the "power source circuit" in the ECM and the coils themselves. Interrupt the lead to the right, the one going directly to the coils. (You wouldn't want to interrupt power to the CPU or it will deffinately have problems.)

Origianally I was worried that the CPU woud freak out at the loss of power, but the rpm's won't change much and this is so brief (150ms) that I don't think it will cause a problem. I'm looking at all the different inputs the CPU gets, like gear position switch and coolant temp and I don't see any reason the CPU would freak out. The cam position sensor would be affected by the change in RPM's, but this happens so fast I can't imagine the CPU would be able to tell that the RPM's have changed. Now, the CKP sensor... I don't even know what that is, so that's the reason I can't say for sure.

Anyway, you don't want to damage anything and you don't want the CPU to suddenly change the timing due to this "hiccup" your making, so look into all that and use your judgement.

(Also, I did a little guessing on identification. K3? K4? I'm not even sure I'm looking at the exact manual for your bike, so make sure were on the same page.)

The next matter is how to make the interrupt. A relay makes the most sense.

GreenGiant mentioned some relays and said, that they "get really expensive really quickly." Now you also said in your first post, "trying to keep the costs as low possible."

So... How low do you want to keep the costs? I checked pricing on these bikes and new they are more than $10K new for the 2011 model. I don't know what you paid, but.... Is a hundred buck low cost to you???

If there is some off-the-shelf component that in required or that makes this easier, how much would you pay?

To GreenGiant: How expensive are those relays with the "time on/time off delay" feature? (Also, are those physical relays or solid state circuits that work like physical ones? Just curiuos.)

Good luck to you ratsplinter,
--tim

(Will check back here on your progress, but I'm ready to sleep and working a double tommorow, so It'll be a while.)

 

[ratsplinter]

Hi Tim

Thnx for the interested reply. Much appreciated.
You are spot one with the "hiccup". This allows for full throttle gear changes.
The model to be exact is for a 2004 GSXR 1000 (K4).
I'll be using the K4 as a prototype. If I can get it working for only this one, I'll be very happy.
Regarding the comments :
- the CPU would not freak out - the motor is still running.
- define really expensive ? I dont have a problem buying a already made one, but not all my friends are so fortunate. This really is to help a couple of buddies out, using my bike as a test unit.
- The existing unit on the market comes with the following wires/instructions :
Black - Ground
Red - +12V, Feeding Coils
Blue - Shifter Switch (the one on the footpeg - normally open
Yellow - Switched 12V output to coils

I hope this info helps, let me know if you can shed some more light on this, or need other info.

 

[GreenGiant]

programmable relays run from around 100 all the way up to around 300 depending on voltages, and capabilities

 

[ratsplinter]

Thats running quite expensive...
Is it possible to do it with some form of microprocessor ? I have a little bit experience with the Parralax Basic Stamp 2 interface - but really only limited.
Correct me if I'm way off here please.
From what I understand, the BS2 can take voltage between 5 an 15v.
It then bring it down to constant 5v, on which the processor operates.
I have written a simple piece of code which basicly does what I want, but the output voltage is only 5v. How to I get this up to 12v ? Am I on the right track, or heading in the wrong direction ?
Tx again for all the input so far.

 

[timothy48342]

I tried to look up the programmable relays that GreenGiant mentioned, but I just don't know enough about them to filter for what you need. (My search was getting muddled by timing relays that can be programmed with large times, measured in seconds or minutes.)

Sound like your getting there, though. The BASIC Stamp just might work.

In this manual:
http://www.parallax.com/Portals/0/Downloads/docs/prod/stamps/web-BSM-v2.2.pdf
I'm looking at page 14, Table 1.2 and for the output pins labled p1 -15 it says:

General-purpose I/O pins: each can sink 25 mA and source 20
mA. However, the total of all pins should not exceed 50 mA (sink)
and 40 mA (source) if using the internal 5-volt regulator. The total
per 8-pin groups (P0 – P7 or P8 – 15) should not exceed 50 mA
(sink) and 40 mA (source) if using an external 5-volt regulator.

So, I think you just need a relay that can switch 12V, but is opperated at 5V and draws less than 20mA. (Opps, I just found out that is really low for a relay. If you can find one like that or find the solid state equivelent of a relay, but with the low current draw, then great...)

(...otherwise) Here is a link to an article where someone is using the output of an Arduino to control a relay:http://www.glacialwanderer.com/hobbyrobotics/?p=9
I didn't read the article, but when I saw the schematic, I thought this is what you need.
Schematic:http://www.glacialwanderer.com/_blog/blog2008/04_April/hb_relay5.jpg
(The article might be a good read, becuse he probably explains his part selection.)

I made a schematic!!! (Really, I just copied his layout and re-labled it for your bike. It's attached.)

Unknowns:

1. Where he has a 1k resistor coming from the Arduino you might have a different value. It would be based on the current draw of the coil of the relay you choose, and characteristings of the transister used. (And an easy calculation for the masters on here. I might be able to do it if someone else here confirms my calculation, but we need the other info first.)

2. The diode can probably be the same, but doesn't have to be. It protects the transistor. When the transistor shuts off, there is energy in the coil of the relay that has to be disipated, that all. Lots of different diodes might work.

3. He does this with 5V and you'll use 12V. If you had a 5 V supply, you could duplicate his circuit alsmot exactly, but 12 is fine. You just might have slightly different parts.

4. The transistor. He uses a 2n2222. I don't have transistor characteristics memorized. (Not even for this very common one) Here is what it needs, I think:
The max collector current needs to be more than the current draw of the relay coil. And there is a characteristic that represents the ratio between the collector current and the base current. (I don't remember what it is called.) For instance if the collector current is 5 times the base current and we have a maximum available of 20mA that we can take from the BS2, then the collector can pass 100mA, which still might not be enough for the relay. (need to know more about the relay) If the collector current was 20 times the base current, we would have 400mA available for the relay, which should be plenty.

5. The relay. It has to be fast and that's about it. You want to control the width of the gap in power to the coils between 10ms and 150ms, so you don't want a relay that takes 200ms to re-engage power when it's done. I guess if you can, you want to pick one the has a low current draw at the coil, so that transistor selection is easier. If it happened to have a built-in protection diode, that would be a nice plus.



To the experts out there... will this work? Or could this work? Or Is there maybe a much much easier way to do this? Ha, ha!!

Are there relays that switch at speeds of 10 to 20 ms or so or will he need something solid state?

Do relays usually draw in the 50mA to 200mA range?

What is that transister characteristic called that's used to calculate back and forth between the base current and the collector current? I think for audio amplification it is "beta", but I can't remember if that same value would be used here for power switching or if it is called something else.

-tim
Edit:
Its the "gain" of the transistor I was thinkning of. On the data sheet it would be the "DC beta.) (bDC)

That value should be AT LEAST about 20 or so, but the higher that value is, the less restrictive your relay selection will be.

As an example, I looked up the 2n2222 and it would be fine.
A gain of 75 and max collector current of 800mA.

Also I did a google search about relay switching speed and the first worthwhile hit came from THIS SITE. YE-AH!
Here https://www.electronicspoint.com/relay-switching-speed-t119276.html they say 50hz to 200hz for a typical relay which works out to 5ms to 20ms, so it shouldn't be difficult to find one that works. They also talk about alternatives for faster switching.
:end of edit