85 GLH Turbo experimental testbed build

[rgl10]

Another little job I did was make my own swing valve. I had several 2.25 Garrett sv's and a few Mitsubishi's but no 2.5 Garretts let alone a 3 inch aftermarket one. Looking at the Mitsu part I noticed how it was the same as a 2.5 Garrett but smaller where it bolted to the turbo and didn't have a valve. So I though I would try to combine the two and save a few bucks. The Mitsubishi unit is much bigger right where the Garrett is smaller. By combining the two you can have a nice high flow unit for the cost of a few nickel welding rods. I watched a few youtube videos on how to weld cast iron as I wasn't really sure. I bought a bag of fine play sand to bury it in after it was welded up. I tacked it at the corners and kept heating it with a torch while I welded it. I would run a bead about 3/4 to 1 inch at each corner till I connected all the welds and to keep the heat even. Then I covered it with sand that I had warmed up and let it set all night. Some grinding and some paint and best of all less than 20 dollars invested!

[IMG][/IMG]

[IMG][/IMG]

[IMG][/IMG]

[IMG][/IMG]

[IMG][/IMG]

[IMG][/IMG]

[rgl10]

Still trying to list most of the stuff I did to the car in 2013. I do drive this car in nice weather everywhere. I took it up into PA last fall to get a dyno run in. It was a 300 mile trip and the car averaged 28 mpg. For such trips I needed cruise control. However with modded computers you usually loose speed control since those inputs and outputs are moved to more performance oriented things like staging rev limiters and dual boost control. I remembered putting on Mopar cruise controls back in the 80's so I went about trying to find one in the local junk yard to fit my car. When Chrysler went to fuel injection they added the speed control into the computer on all cars from 86 up. Minivans with the 2.6 and Diplomats still had carbs so they came with their own stand alone computer, nearly the same as the one Mopar sold for dealership install. Only two wires are switched in the module connector if I remember correctly. Also the module programming is a little different. Minivans were heavy with weak motors so those modules are more sensitive to speed changes, the throttle moves more when it drops say 3 mph. The Mopar kit for turbo cars and Diplomats had the same or nearly the same calibration for the module. Anyhow I was able to get the parts gathered up, the module, the small wiring harness and the vacuum solenoid pack etc. My car already had the factory switch on the column and the cruise control brake switch since I had installed them years ago with the factory computer. It was a simple matter of just pluging eveything up and now I have factory speed control again.

[IMG][/IMG]

[IMG][/IMG]

[IMG][/IMG]

[IMG][/IMG]

[IMG][/IMG]

[Turbo224]

So much ingenuity in this thread!

[Reaper1]

Very nice on the meth injection system. I've been thinking on how to make a system that uses individual injectors and a dedicated pump, regulator, and tank for many many years. I've just never built it. The biggest issue is the injector control. The way I want to do it is exactly like the fuel system already on the car. Recently with all of the returnless fuel systems out there, even running the regulator in the tank would really simplify things. An aftermarket injector controller or ECM could control the injector pulse width. The reason for that would be similar to the reason the current meth injection systems out there vary the pump output. For this system the pump pressure is always the same and the amount of meth is regulated by the injectors. Yeah, it's more complicated than what's out there now and what you've come up with, but for people that want to run a LOT of it, I think it may be a good way to go. Better spray pattern for atomization, no changing of jets, full control of how much is being injected, common parts, and it could be super stealthy...even more so than current stuff because it would look like you are just running twin fuel rails.

[blk86trbo]

So much ingenuity in this thread!

Agreed, lots of pics too which is always great!

[rgl10]

Very nice on the meth injection system. I've been thinking on how to make a system that uses individual injectors and a dedicated pump, regulator, and tank for many many years. I've just never built it. The biggest issue is the injector control. The way I want to do it is exactly like the fuel system already on the car. Recently with all of the returnless fuel systems out there, even running the regulator in the tank would really simplify things. An aftermarket injector controller or ECM could control the injector pulse width. The reason for that would be similar to the reason the current meth injection systems out there vary the pump output. For this system the pump pressure is always the same and the amount of meth is regulated by the injectors. Yeah, it's more complicated than what's out there now and what you've come up with, but for people that want to run a LOT of it, I think it may be a good way to go. Better spray pattern for atomization, no changing of jets, full control of how much is being injected, common parts, and it could be super stealthy...even more so than current stuff because it would look like you are just running twin fuel rails.

I like the idea of a separate multipoint system. If you wanted to make huge power by running straight methanol that would be the way to do it. The system would only be needed under boost to add octane and fuel. You could tune BoostButtons cal to take into account the extra fuel under boost. This way you could run pump gas and conserve your methanol to be used just when needed. Distribution would be perfect too. I think I could whip up a system like that pretty quick. If you wanted to do it on the cheap I would think you could use a socketed 88 or 89 logic module and steal some inputs from your main computer just to get enough infomation for the secondary computer to run. Then mod one of Robs cals and turn off all unnecessary stuff and adjust your injector control to what you want. Now you've got me to thinking. If you want to do it first you might want to hurry because I might just put that on my list of stuff to do during 2014 lol!

[Reaper1]

You are WAY ahead of me...you've got a running car! LOL If you can make it work, please go for it!!! I'd LOVE to see it come to fruition.

[rgl10]

I think that pretty much covers the major changes I did to the car last year. There were other things like beefing up the trans, trying different gear ratios, swapping governor valves around to change shift rpm, rebuilding the window regulators, refurbishing the instrument cluster ect. Even though taking a lot of time, I would not consider those kind of things experimental so no need to go into detail on them here. The project that I'm on now though is by far the most experimental thing I've attemped on this car. Its an addon variable cam timing system. I wanted something that could be just bolted on and easily taken back off. Like a lot of projects, I've thought about doing this for a long time (over 25 years). I started gathering parts and mocking up the items to see how or if they would fit (always important on an Omni) and getting an idea of how much I could change the cam timing without major surgery. My first idea was to use a drawer slider to move the idler pulleys. It was cheap and easy - two very important things in my book. However I knew the slider, even though being a heavy duty unit, would fail. The more I thought about it the more it bothered me. So I came up with plan B.

Here are the parts I intended to use the first time around.

[IMG][/IMG]


The idlers are from a subaru wrx sti timing belt. They were the smallest I could find that I know were designed for just this kind of use. After some research I did find out that they are usually the first idler on wrx's that fail though. Probably can't be helped since they are so small the bearing speed is several times faster than the bigger idlers. Its said they normally last over 60 thousand miles so it probably won't be a problem on my engine anytime soon. I bought them off ebay for about 27 dollars each. I hated to pay that much but I spent way too much time trying to find a better deal and it just didn't happen. I shortened both of them to make the overall width of the unit smaller. Here one is already cut down.

[IMG][/IMG]



My original mock up looked like this.

[IMG][/IMG]


Here is the principle of how it would work. By just removing the timing belt idler pulley you get 10 degrees of slack. This slack can be moved from the front to the back. I am going to make my system run 5 degrees advanced at idle and low rpm. Somewhere above 3500 or 4000 rpm I will move the idlers toward the front of the car. This will put the slack on the back side retarding the cam timing 5 degrees for good top end power. I'm not shooting for big gains here. Realistically I'm hoping for 10 ft lbs of torque and a smooth idle on the bottom end and hopefully 20 hp on the top end. A more radical engine than mine might benefit more.




Everything was going to be hid under the a/c hoses. I didn't like that very much since no one would be able so see how it worked. And as I mentioned before I really didn't want to use the slider after thinking about it. So plan B was put into action. I wanted to build it into the mount bracket since that could be changed easily. I decided to try a pivot system instead of a slider type. Here are pictures of that undertaking.

[IMG][/IMG]


[IMG][/IMG]


[IMG][/IMG]

Anyone here old enough to have seen the 1979 tv series "Salvage One"? I was about 16 when that show aired. I loved it! Here is the opening intro-

Narrator: Once upon a time, a junkman had a dream.

Harry Broderick (played by Andy Griffith) : I want to build a spaceship, go to the moon, salvage all the junk that's up there, bring it back, sell it.

Narrator: So he put together a team. An ex-astronaut. A fuel expert. They built a rocketship, and they went to the moon. Who knows what they'll do next.


That show agreed with my idea (and my wallet) of building stuff.

Everytime I look in a junkyard, salvage pile or at the stuff stacked behind the shop, I see possibilities.


[IMG][/IMG]



I made it so I could change the bearing in the pivot if I needed to with out removing the whole mount.

[IMG][/IMG]


[IMG][/IMG]

[IMG][/IMG]

[rgl10]

After I got the bottom part of the pivot made I bolted the mount bracket to the motor. I cut cardboard to fit then use it as a pattern to cut the real part. Most of the parts to make this are just things I found laying around the shop. There is an old alternator brace in there, a pushrod from a diesel motor, a solenoid bracket from an old pickup etc. The list just goes on and on.

On a side note- My dad showed me how to weld when I was young. He said you know you got it right when it sounds like you're frying eggs. My welds look like scrambled eggs, close enough I guess.

[IMG][/IMG]

[IMG][/IMG]

[IMG][/IMG]

[IMG][/IMG]

[IMG][/IMG]

[IMG][/IMG]

[IMG][/IMG]

[IMG][/IMG]

[IMG][/IMG][IMG][/IMG]



The cylinder can use air or hydraulic fluid to move. I think I will try air first since that will be the simplest and least dangerous way to go. The unit will be spring loaded advanced 5 degrees as shown in the picture below. All the slack is pulled to the front. Then I will use boost pressure or a small compressor and air tank to pull it forward which will retard the cam 5 degrees by moving all the slack to the back side of the belt.

[IMG][/IMG]

[rgl10]

Now even though I move the two pulleys together the belt gets a little slack in it when the idlers are near the middle of their travel. Part of it has to do with the fact that the slack side on the front of the belt is shorter than the slack side on the back. The other part is because my idlers pivot in an arch and do not move straight forward and backward. If I knew mathmatics better I could have changed the angles and made it pull the right amount of slack on the front and back while keeping the belt tight all the time. No tensioner would be needed if done that way. Sometimes though when I build something I have to build something else to make the first something work right lol. So I built a tensioner like one that is used on a serpentine belt. Nearly all serpentine belt tensioners are too strong for a timing belt. I found one that had the pulley inline with the spring which is what I had to have to make it work. However it was tensioned the opposite direction from what I needed and was way too strong. So I cut it apart and removed the stops. Then I cut another notch in the housing and flipped the spring over. This way I could adjust the amount of tension on the spring and it would push against my timing belt instead of moving away from it. I removed the plastic grooved pulley and installed a flat steel idler.

[IMG][/IMG]



I bolted it to the power steering pump bracket. It lines up perfectly with the timing belt and keeps it tensioned just right.

[IMG][/IMG]



After I got everything lined up I removed it all and added some reinforcements and braces. I ground the welds down and painted it all. These are the parts of my regulated 10 degree variable cam timing unit.

[IMG][/IMG]

[Vigo]

I cant remember where i posted this but i was talking about this same basic concept in another thread here a while back and mentioned some of the same issues: that a serpentine belt tensioner would be ideal but would have to be weakened since they are too strong as stock. Part of my idea, though, was to push on only one side of the belt and allow the tensioner to do all the work on the other side. I feel this would be a mechanically simpler setup once it was 'all figured out', but the amount of travel and tension from the tensioner would have to be 'just right'.

I do like the system you've come up and you definitely seem like the kind of guy who would build a better one once you have results pointing towards any actual functional failings that may or may not arise.

[Keito]

Funny, I have a box of Mitsu/Garrett swingvalves cut up on the shelf I was going to weld and make 2.5" swingvalves
Been on the shelf for about 10 years, found a cheap Garrett one and didn't bother
I was actually going to weld up a fixture and sell a few

[rgl10]

I cant remember where i posted this but i was talking about this same basic concept in another thread here a while back and mentioned some of the same issues: that a serpentine belt tensioner would be ideal but would have to be weakened since they are too strong as stock. Part of my idea, though, was to push on only one side of the belt and allow the tensioner to do all the work on the other side. I feel this would be a mechanically simpler setup once it was 'all figured out', but the amount of travel and tension from the tensioner would have to be 'just right'.

I do like the system you've come up and you definitely seem like the kind of guy who would build a better one once you have results pointing towards any actual functional failings that may or may not arise.

I think the one idler system would work fine once you got the spring loaded one on the back adjusted right. The two idler idea was something I thought about a long time ago so I thought I would try it out first. I will get it dynoed this summer if all goes well. If it does not make at least some more torque on bottom end and some more hp on top end then I will just go back to stock and that idea won't bother me anymore. If it does work I will probably build a second one making improvements from what I have learned on this one and from ideas I get from you all on these forums.

[fishcleaner]

Looking good. Those machine screws are weak on the adjustable cam gear, weld those nuts on so they don't back out, you can snap them off if you need to.

[rgl10]

Funny, I have a box of Mitsu/Garrett swingvalves cut up on the shelf I was going to weld and make 2.5" swingvalves
Been on the shelf for about 10 years, found a cheap Garrett one and didn't bother
I was actually going to weld up a fixture and sell a few

Combining the two does make for a nice cheap alternative. Mine turned out 3/4 of an inch deeper since I just added the Garrett piece right onto the Mitsu housing. Makes a nice smooth radius inside with less restriction.

[glhs875]

I think the one idler system would work fine once you got the spring loaded one on the back adjusted right. The two idler idea was something I thought about a long time ago so I thought I would try it out first. I will get it dynoed this summer if all goes well. If it does not make at least some more torque on bottom end and some more hp on top end then I will just go back to stock and that idea won't bother me anymore. If it does work I will probably build a second one making improvements from what I have learned on this one and from ideas I get from you all on these forums.

Oh I am pretty sure there will be a power increase like you mentioned. I have been retarding the cam on my engines up to 6 deg. for years. On my most current combo the engine was making power to 7K using a stock 89 T1 cam in a retarded setting. Still had good bottom end as well!

[rgl10]

Looking good. Those machine screws are weak on the adjustable cam gear, weld those nuts on so they don't back out, you can snap them off if you need to.

Thats good to know. I will need to adjust my cam gear after I get everything set up to fine tune it in. I will look into getting some healthier bolts to replace the stockers.

[rgl10]

Before I got into this vct project I tried to find some info on how a single cam motor benefits from changing the cam timing. There are literally hundreds of dyno charts that show the benefits of vct on dual cam motors but not so many for single cam engines. Here is one chart for a Volvo engine that they added a vct system to and how the power improvement looks compared to a non-vct motor.

[IMG][/IMG]




Also I programmed every spec on my motor I could find into engine analyzer pro. The torque and hp that it generated from this info is really close to how my engine performed this past fall on a mustang dyno. So I changed the cam timing in the program to 5 degrees retarded and ran a synthetic benchmark. After that I changed the timing on the cam to 5 degrees advanced and overlaid that dyno run onto the earlier one. This is what it projected the engine would make at those two cam settings with all other engine specs left the same. Also these tests were run at 18 psi boost, just like I normally run my engine. It appears that the motor will make 5-8 more foot pounds of torque below 3000 rpm and 10 to 20 horsepower from 4000 rpm to the redline. I also know from adjusting my cam that it idles smoother and has much better low end throttle response with the cam advanced but the engine struggles after 5 grand. The engine pulls good to 6000 though if I retard the timing a few degrees but idles like it has a big cam. So I'm hoping for the best of both worlds all the time with this vct setup.


[IMG][/IMG]

From looking at the chart above it appears that I can change the cam timing from advanced to retarded anywhere between 3 and 4 thousand rpms and not make much difference in torque or hp.

[Vigo]

Well, in that case i would set it to cycle in the high 3s to avoid unnecessary cycling from cruising around at low throttle since you may often get over 3k but rarely get to 4k when driving casually. Or set it low and make it cycle as much as possible for durability testing.

[rgl10]

I never thought about it cycling. Thanks for bringing that to my attention. Maybe I will make it so that it has to be under boost or over a certain throttle opening before it can retard the timing. With my 2.60 final drive I never cruise over 4 grand so I think it could just stay advanced all the time during normal driving. Only when I need the most power would it retard the cam. I appreciate the input Vigo. That gives me something to think about.