Making a system to monitor fuel/oil ratio.

[Fl_Richard]

Re: Making a system to monitor fuel/oil ratio.

And easier. Pulses increase as volume increases. Simple comparator circuit could yield simple red,yellow, green guage. Add more sophisticated LCD driver and processor and poof VERY accurate guage.

Remember the people on this board are the future buyers; they seem to be leaning toward flow based measurements.

 

[Lyle29464]

Re: Making a system to monitor fuel/oil ratio.

First I think It is neat that you like to do projects like this. I wish you luck.

Second It seems like a real uphill battle. If you have a system that is 100% fool proof and free. Why replace it with a system that is so sensitive and subject to the many thing that will change the shade of the mix. It seems to me that the age of the mix also has a lot to do with the color.

 

[madgadget]

Re: Making a system to monitor fuel/oil ratio.

Nothing ventured, nothing gained

This is for on board blending systems, not for tank premixes you have done yourself (in case I was not clear). So for anyone who believes in nothing but diy pre-mixing then this will not interest them at all.

I have just done a few more experiments with a three sensor setup one for oil fuel and mix. It looks as though there will be no need to calibrate this at all (other than initial sensor calibration). Now I just need to freshen up on my maths make some predictions and then some measurements and see if they give me the expected results. If so then it looks as though for my own personal application at least. I am onto a winner...

 

[turbinedoctor]

Re: Making a system to monitor fuel/oil ratio.

Well Madgadget I for one like the way you think and apploud you for posting here. Yes there are skeptics but don't let them talk you out of anything. As for me I am a premix guy because I don't have an auto blend engine but if I did I would want a device such as yours to ensure the auto blend is working. The way I read your posts is that you are looking to monitor and not control the blending. After some trial time with a successful product I could see where it would or could be use as a blending controller. I will PM you a message.

 

[kenmyfam]

Re: Making a system to monitor fuel/oil ratio.

Nothing ventured, nothing gained

This is for on board blending systems, not for tank premixes you have done yourself (in case I was not clear). So for anyone who believes in nothing but diy pre-mixing then this will not interest them at all.

I have just done a few more experiments with a three sensor setup one for oil fuel and mix. It looks as though there will be no need to calibrate this at all (other than initial sensor calibration). Now I just need to freshen up on my maths make some predictions and then some measurements and see if they give me the expected results. If so then it looks as though for my own personal application at least. I am onto a winner...

Glad to see you are moving forward. I have one question though.
What if one of the sensors fails ???
Will it cut the motor ??

 

[ondarvr]

Re: Making a system to monitor fuel/oil ratio.

The problem isn't that he's trying something and being creative...that's great....the problem is IMHO he's chasing the issue in the wrong direction, he's relying on things he can't control (oil and fuel color) and that will vary, to monitor the exact, or very close to the exact ratio. You need to find something that will be a constant.

 

[tboydva]

Re: Making a system to monitor fuel/oil ratio.

Just happened along this thread. I love to tinker and I think one of the original posts had the right idea - 2 sensors, one before and one after the mix. As has been pointed out, the sensors would have to be appropriate, but if you always use the same oil, it seems that the "background" gasoline signal could be canceled. I got the sense the design involved "electrical" canceling. Have you considered using a microcontroller? There are lots of options there (I've used PICs for many projects). Most of the low end series (16F, for instance) have multiple A/D converters. You could read both sensors and use the PIC to subtract the signal. Do your calibration curve with several grades of gas and see what you get as a "final" signal. I'd hazard to guess if your optical windows and geometries are consistent, you'll get fairly accurate measurements. You could also use one of the PICs digital output pins to drive a solenoid (kill switch). If you proportion got below a certain level - stop the engine. Of course, I'd make sure my kicker ran on pre-mix!!!

Neat project!

 

[madgadget]

Re: Making a system to monitor fuel/oil ratio.

The problem isn't that he's trying something and being creative...that's great....the problem is IMHO he's chasing the issue in the wrong direction, he's relying on things he can't control (oil and fuel color) and that will vary to monitor the exact, or very close to the exact ratio. You need to find something that will be a constant.

I cant control - no. But can I compensate for (accurately) - Yes I think so.

 

[madgadget]

Re: Making a system to monitor fuel/oil ratio.

Just happened along this thread. I love to tinker and I think one of the original posts had the right idea - 2 sensors, one before and one after the mix. As has been pointed out, the sensors would have to be appropriate, but if you always use the same oil, it seems that the "background" gasoline signal could be canceled. I got the sense the design involved "electrical" canceling. Have you considered using a microcontroller? There are lots of options there (I've used PICs for many projects). Most of the low end series (16F, for instance) have multiple A/D converters. You could read both sensors and use the PIC to subtract the signal. Do your calibration curve with several grades of gas and see what you get as a "final" signal. I'd hazard to guess if your optical windows and geometries are consistent, you'll get fairly accurate measurements. You could also use one of the PICs digital output pins to drive a solenoid (kill switch). If you proportion got below a certain level - stop the engine. Of course, I'd make sure my kicker ran on pre-mix!!!

Neat project!

I've moved on from this. I have three inputs now. Oil, Fuel and mix. I have been working on a mathematical formulae which will relate them all. I'm no maths expert, in fact its been years since I have done anything like this so that's taking a little time. As for Microchips, yes for my personal use I will be using one. I cam program ASM so that's something

In fact my own version will have a lot more to it but the initial one I will just keep simple.

 

[tboydva]

Re: Making a system to monitor fuel/oil ratio.

Sounds like you've got it under control... If you can program ASM, you're doing well. I use C - which for your application could be very easy as you would perform an A/D conversion, then directly subtract (i.e. newvariable = after_read - before_read. No need to convert to analog voltage value (unless you wanted to output to an LCD or something). You could calibrate this and you'd be in good shape. Microchip's C compiler is free (for small programs). When you get it done, post your specs (I'd like more info on the sensors). Good luck!

 

[madgadget]

Re: Making a system to monitor fuel/oil ratio.

I'm still optimising the sensors and playing with different ones. The method I am using at the moment does work, but the sensors output impedance is VERY high i.e. in the region of 1.7M ohm going to infinity as the light pass through falls. I am not happy with that at the moment, although it does actually seem to be very stable now I am using unity gain buffers. (even my test equip was loading the results!)

I actually have a dsPIC Pro 3 development kit with a dsPIC30 mcu on it and full C compiler, but ill be honest even though I have had it for nearly 2 years I have done nothing with it, and I know ASM a whole lot better than C so to begin with it will be done that way. Only thing is I am not sure a 10 bit A/D is going to be enough, 1024 sample points is not a great deal, I need to make sure I optimise that as well as I can, or use one of the 18F range like a 18F4523..

But again I am jumping ahead of myself now.

I keep telling myself to keep this simple, but my mind goes crazy when I get consumed by a project like this.

 

[tmcalavy]

Re: Making a system to monitor fuel/oil ratio.

Sounds confusing enough to ruin an outboard if one of the sensors fails. I prefer the old timey method...my $3 pre-marked, premix bottle that I got from Wallyworld...no electronics and I always get the mix right for the motor.

 

[madgadget]

Re: Making a system to monitor fuel/oil ratio.

The outboard doesn't/wouldn't rely on it. There are no moving/mechanical parts to fail.

This idea/system is intended for INDICATION PURPOSES ONLY.

Correct sensor operation is easy to test and very predictable.

Everyone likes doing things their own way, and like I said earlier this or anything like it would never convince those who swear by premixing. Its not for them, it never will be.

 

[madgadget]

Re: Making a system to monitor fuel/oil ratio.

WOW!!! What can I say.

IT NOW WORKS!!!!!!!!!!!!!!!!

:> :> :>

AWESOME!!!!

 

[madgadget]

Re: Making a system to monitor fuel/oil ratio.

Just thought I would add, using non real time (I.e. inital fuel/oil figures) taken first then the sample mix I have a 19% error with no correction at 50:1 ratio, this was done with 200ml of fuel and 4ml of oil. With "realtime" base corrections and also a fixed sensor positions I suspect this error will drop A LOT.
If I can get it down to 10% or less at 50:1, which I think will be EASY, that's an acceptable error to work with I feel

So you could be confident of knowing your mix is say 45-55:1 which would be fine for any engine

 

[Faztbullet]

Re: Making a system to monitor fuel/oil ratio.

Here is a tidbit you might could use, you can test/monitor the rotational sensor on your V-6 by using a meter set to Hz. Attach red lead to blue/white(tapped in,not disconnected) and blk to engine grd. Frequency should be: 8Hz at 1000 rpm, 17Hz at 2000 rpm, and 33 Hz at 4000 rpm. The oil gear is stripped or failing if the multi-meter shows no frequency or erratic frequency.

 

[madgadget]

Re: Making a system to monitor fuel/oil ratio.

Thanks for that, that was something I was thinking to look into for another project later. I want this to be totally non specific so that I could if I wanted fit it to any engine.