EVIC home page

EVIC Update #4

January 7, 2001

In my December report I told you about my switch to the 14 pin PIC16C505 micro to get more I/O lines. At that time I had wire up 2 inputs to allow me to make step changes in the intake and exhaust valve closing times. I now have a total of 6 switches wired. One is used to enable the Hit & Miss mode and the other 5 have been used for various timing tests with interesting results.

Some of the feedback I received as a result of the previous E-mail updates set the direction for the experiments I am about to describe. One suggestion was to use a PC to vary the engine controls. This idea led to the test switches. Another tip was to look at a web site where experiments with Atkinson cycle engines were described.

http://perso.wanadoo.fr/mc-engine/inside.htm 

Other feedback has prompted me to expand and update the software description in the EVIC book. I appreciate all your feedback and suggestions. Please keep the E-mails coming. 

Although the engine ran at close to 7000 RPM back in July I had not been able to repeat this result. When I got a good top speed it didn't idle very well. At high speeds it had always spit fuel out the carburetor intake, a sign I interpreted as the intake valve closing too late. I November I found a French site on the web where they were suggesting late intake valve closing as a means of "throttling" an engine with reduced pumping losses. They discusses the Atkinson cycle where the power stroke is longer than the intake stroke and suggested a pseudo Atkinson cycle by late intake valve closing. Their approach was mechanical but it seemed to me that I could do it more easily by just keeping the intake valve open longer. I set up a test where I could delay the intake valve closing by 1, 2 or 3 milli-seconds. With the engine running at about 5000 RPM I flipped the first switch introducing a 1 ms delay. The engine sped up. The same thing happened with 2 and 3 ms delays. At the higher speeds it was spitting fuel and operation was somewhat erratic as it had been back in July. This time I concluded that the intake valve was floating or sometimes closing more slowly. I checked that the valve wasn't binding in any way and then decided a stronger intake valve spring was in order. I had tried this previously without success. This time I measured the spring forces when compressed to a know length, 1/2 an inch and made a spring that gives a force of 11 oz, half way between the one I was using and the one I had tried several months ago. This proved to be a good choice. The solenoid can handle the extra force and the test results indicate that the valve operation at high speed is more consistent.

Over the past few weeks I have programmed the switches to try changes in valve opening and closing times and in the spark advance. This enabled me to tune the valve timing to improve the top speed and still keep a good idle. After idling at under 2000 RPM it accelerated to just over 7000 RPM. I was measuring speed using an oscilloscope. The time per revolution was consistently 8.5 ms per rev, 7058 RPM. This seems to be a limit for now. There is an exhaust valve response time hardware/software problem that creeps in here cause the exhaust valve to open later than is desirable as the speed approaches 7000 RPM.

Instead of trying to find a way to push the top end I returned to the concept of a pseudo Atkinson cycle. I had improved the high speed performance with later intake valve closing. What happens if I go even later? I found that I could in fact throttle the engine by late intake valve closing. The down side is again I have fuel coming out the carburetor intake. This time it isn't spitting out but a finger held in front of the carburetor intake soon is wet with fuel. This isn't an unexpected result since some of the mixture is flowing back out of the cylinder as the piston rises. I see promise here for a combination of a normal mechanical throttle and late intake valve closing to reduce pumping losses and give better air/fuel mixing at low speeds. This will require some more intake manifold volume than I have now. The technique would be to close the carburetor throttle part way. This results in reduced pressure in the intake manifold and cylinder on the intake stroke. The intake valve remains open as the piston starts to rise and until the cylinder and manifold pressure reach atmospheric. It must close at this time to prevent mixture being expelled from the carburetor intake. Because the valve closes when the piston is already part way up there is reduced air/fuel charge in the cylinder. In a normal engine the intake valve is closing when the pressure in the intake manifold and the cylinder are still well below atmospheric pressure. This also results in a reduced charge in the cylinder. In both cases the result is less power output and hence reduced engine speed. In the normal engine the increased negative force on the piston during the intake stroke increases the pumping losses. There is an interesting research project here!

This experimenting has kept me away from doubling acting solenoids and all the other interesting things I could do with the EVIC. But there is always tomorrow.

Dave Bowes

Contact Info : dbowes12@cogeco.ca