advanced valve timing crazy experiment

[turboguzzi]

ok, here's the story

 

When I bought my 2nd legacy almost three years ago, car drove really great (still does), pulling better than my previous 1st gen, the only gremlin being a slightly lumpy idle (which changing plugs didnt cure). oh well i thought, can live with that with the car being a super clean 2nd gen leg with just 70K on the odo.

 

A month later, while taking the timing belt off to replace the leaking dreaded o-ring seal at the oil pump, I discover that one bank of cylinders had the timing pulley advanced by one tooth....(i.e. intake open earlier, exh. closing earlier)

 

I happily restore everything to stock, thinking about how incompetent some car mechanics are and how much better it will run now. well, after the work it indeed idled perfectly, but..... what do you know, pull from down low was actually better before fixing the bad valve timing... at least by seat of the pants it went back to pulling as my previous Leg did. And yes, i checked a few times, it is rightly timed now, it was 1 tooth off before...

 

50K on, the car is still a great runner but i thought back at the time that considering i dont give much importance to top HP , i'd be happy to have more torque down low. So why not advance valve timing for the complete engine? after all, even with just one side advanced i could tell it was pulling better at low-medium revs.

 

As i tune road race motorcycles, building my own motors, i am quite used to cam degreeing and usually retard cams for better top end power. so essentially i want to try the opposite. could even bring the bonus of better fuel economy.

 

I am near now to the car's second belt change at 120K, so really willing to give this a try on my 1995  european 2.0 5MT,

 

the camshaft pulley has 50 teeth, so a 1 tooth advancing equals 7.2 degs, not a whole lot actually.

 

ignition and injection events are governed by the crank sensor, my main doubt is if the cam sensor will "fell" something strange.

 

What do to the motor masters among us think? pros, cons, risks (engine is definitely non interference) or will it end in tears?

 

tnxs in advance for your feedback

Edited May 20, 2014 by turboguzzi

[1 Lucky Texan]

many newer cars have variable cam timing (and/or variable valve LIFT) so, what you noticed is likely a 'real' effect.

 

but, there could have been trade-offs longterm if left one tooth off; perhaps worse fuel mileage,  more likely to burn a valve, slightly worse emissions, less-smoth idle as you witnessed or ???

 

cars are a pile of compromises and I'd guess cam shaft timing (and other aspects like profile and lift) is on that list.

[turboguzzi]

tnxs texan

 

i relate the bit lumpy idling to the fact that it was only one side that was advanced.

 

I agree with the compromise thing, essentially saying that in my case, since i am often towing a trailer with a bike + car fully loaded, i'd rather compromise top end a bit for better low end.

[forester2002s]

This is an interesting topic. Unfortunately, I have no first-hand experience to offer.

 

But please do report back, if you do make this change. I, for one, would love to know what you find out.

[Fairtax4me]

I've tinkered with this idea as well. I may try soon. I need to put a new toothed idler on my 95 soon because its making noise. All the other bearings are nice and smooth.

 

The cam sprockets over here have 48 teeth so slightly more advance in timing at 7.5°. I just wonder if its too much advance since most of the low end setups I've done are v8s that usually get only 4°.

 

I've searched for adjustable cam sprockets but they only seem to be available for DOHC heads and they're prohibitively expensive. I've also not found proof that they will actually work on SOHC cams.

[1 Lucky Texan]

swapping in an engine with the AVCS system or w'ever it's called, then dyno tuning, is probably best. But I could see the attraction of experimenting with the idea.

 

might be a good idea to monitor exhaust temps but, just speculation on my part.

 

So easy to get a tooth off - may as well see it it's useful! hah!

[1 Lucky Texan]

OOoohhhh - found a coupla interesting links;

 

http://www.tuspeed.com/subaru-cam-timing-adjuster-p-1895.html

 

https://www.crawfordperformance.com/product/pulley-cam-timing-adjuster-for-timing-belt-on-subaru-engines/

 

and, be carefull - evidently too much timing change could lead to bent valves;

 

http://forums.nasioc.com/forums/showthread.php?t=1490257

[turboguzzi]

those adjusters are sure neat. could easily be of use to get less or more.

 

i've counted the teeth on the crank sprocket from a picture and it seemed like 25 = 50 at the cam, i might be wrong of course.

 

surely the beuaty of my proposal is that it's zero cost.

 

Maybe this wasnt so clear in my post: The car HAS been already running for something like 15K with one side advanced, (belt was replaced by PO at 55K and i bought the car with 70K)

 

no damage to the engine whatsoever, after setting the  timing right, i've done already another 50K. Car doesnt consume oil, does 26-27 mpg. If the advance would have done any damage it would have shown by now.

pretty tempted to try but keep your thoughts coming

Edited May 20, 2014 by turboguzzi

[1 Lucky Texan]

try some more searching the Net, I only did a brief search. don't forget to search youtube too. It's sometimes surprising what shows up there.

Edited May 20, 2014 by 1 Lucky Texan

[Fairtax4me]

That is genius! I think I can make one of those! Well... Not as pretty as that, but something along the same lines.

 

Funny that both web sites have the exact same description...

[turboguzzi]

hei fairtax

 

when you think about it, al there is to it is calculating how much bigger the tensioning roller's diameter need to be in order to pull the belt the ammount needed to create the advance required and simply turning on a lathe a bigger sleeve.

 

Y've done something similar on a Rotax 500 single motor for flat track to compensate for lower deck height, it also has toothed belt cam drive.

[Gloyale]

The passenger bank can be moved without affecting spark timing.

 

The drivers bank will affect Valve and Spark timing......

 

If it was the drivers bank it could be the spark advance you were feeling.

[turboguzzi]

ok, finally the hot topic,,,,, the cam sensor

 

though i for one think  that it is the crankshaft sensor that  sends the key ign/inj timing event and that the cam sensor was more for a warning in the event of a belt failure.

 

this also makes more sense to me as if the cam sensor was controlling spark and injection, any elongation/wear in the belt would send them off spec....

 

by your logic, those belt tensioners linked by lucky texan would throw everything off tune.

 

so here's a more focused question to the experts: what is the task of the cam sensor? is it just to control that the cam is generally rotating or does it also send a timing signal? and if off by 7-8 degrees, then would it confuse the ECU ?

[1 Lucky Texan]

I dunno about AVCS, but I though on 'simpler' systems, the cam sensor was just there as a 'check' to shutdown the engine if it's signals varied 'too much' from the crank sensors signal - hopefully to save starting the car with a broken timing belt and bending valves?

[Fairtax4me]

Engine doesn't have to actually start to bend valves. Just needs someone to turn the key. The ECU won't prevent cranking.

I'm not sure how far out the timing can be before the ECU determines there is a problem and sets a code for crank/cam sensor.

I do know that misfires will occur, and the ECU will set codes for those, if only the drivers side cam is off by one tooth. I've had that happen. I don't recall which way it was off.

[turboguzzi]

not worried about bending valves, as i said before, one bank was advanced one tooth to begin with for quite a while and no damage occurred.

 

bending valves in tuned motors (like in one of texans' links) is related to retarded exh closing with hot cams, i know this from my bike race motors that are really on the limit in that sense. that shouldn't be a problem with ADVANCED timing.

being a single cam motor, in/ex overlap is not changing either, so no danger for valve clash either.

 

Only showstopper is how will the cam sensor / ECU react. I dont mind living with a red light in the dashbord for an error code, but not with missfiring....

 

so fairtax, essentially you are saying that the one tooth advance at the cam wheel will cause  a misfire? interesting as over the life of a belt there's surely a bit of de-sync occurring as the belt stretches and the tensioner takes up the slack... I believe that the cam sensor would have quite a high tolerance for allowing that.

 

so here's the one thing  i cant find the answer for: What's the REAL task of the cam sensor?

the spark event is surely timed off the crank sensor, I bet my you know what on that. There's no way spark would be allowed to drift according to belt wear. 

So  to me it looks like:

Option 1 - it's a safety thing that just tells the ECU cam is waaayy off or not turning.

Option 2 - The above + it also times the injection event.

 

Actually in the second case, it would be great as it will maintain the right relationship between injection timing and inlet valve open/close.

 

But surely if it will cause a misfire it's a no-no....

[Fairtax4me]

so fairtax, essentially you are saying that the one tooth advance at the cam wheel will cause a misfire? interesting as over the life of a belt there's surely a bit of de-sync occurring as the belt stretches and the tensioner takes up the slack... I believe that the cam sensor would have quite a high tolerance for allowing that.

As I said I'm not sure which way it was out. I didn't pay attention at the time.

I was getting misfire codes for all 4 cylinders though, so I'm inclined to think that it was affecting spark timing for all 4.

 

It has been suggested in the past that it could be determined by unplugging each sensor separately and checking for which process is affected. I've never tested this theory.

[turboguzzi]

ok, tnxs!