Jump to content

Starlet Ep91 4efe turbo conversion.


Recommended Posts

On 4/12/2020 at 3:29 PM, Sam44 said:

 Over the past two weeks I've been busy putting some ideas into action?

This involves improving the base to mid range power levels on the setup I'm running. Smoothening out the power graph. Also giving great throttle control. 

I've made a adaptor plate to fit this the Audi 80 sport twin plate throttle body to the 5efe inlet manifold I'm currently running. This has a 45deg angle introduced into the adaptor plate so as to ramp the intake air charge across the manifolds back chamber wall so as to feed cylinders 1 & 2 better improving cylinder power balance. 

I have had to switch back to the ae101 gen1 4e ECU because the Audi throttle tps works with this ecu. At a later date I will convert the Audi tps to the ep 91 UK tps and switch back the ECU this has a better fueling map. 

 

 

e8b4108282c4.jpg

Really interesting build

Re: the inlet manifold feeding cylinders unevenly, I read a discussion on a skyline forum about Greddy front inlet manifolds vs the centre entry manifolds found on rb25det. It seems that some cylinders were leaner on the front entry due to plenum shape or distance from the inlet.

Would it be possible to mount the standard starlet throttle body on the side face of the starlet manifold (facing the front of the car) to feed the manifold centrally? Might end up feeding the centre cylinders more than the outer two but could be better balanced? Could keep the good TPS then. Just a thought.

Also really interested in more details / pics of the closed back water pump impeller and Manifold mounted MAP sensor you mentioned in another thread.

 

Good Work 🙂

Edited by Claymore
detail
Link to post
Share on other sites
  • Replies 267
  • Created
  • Last Reply

Top Posters In This Topic

Top Posters In This Topic

Posted Images

TOYOTA-MR2-1990-93-3S-GTE-INTAKE-MANIFOLD-Celica.jpg.34e68c5450ea589d8a0bfde1f5b65249.jpgI'll get some pictures up very soon. I just need to tidy afew weilds and mounts the map sensor with screw to the manifold instead of zip ties lol. I was trying different positions in the plenium and measuring map sensor response to throttle position for fueling response. With the best position closest to the back of the throttle for reaction time but this can create a sensitive throttle pedal if you run a single large throttle body. The twin plate throttle I run has stopped this and is giving great throttle\boost control on the foot as well as great power build. 

The manifolds you mentioned with the center feed like the acis on the 3sgte are alot better for power deliver. Plus if you look at the 3sgte unit it has 2xrunners feeding 1 cylinder this allows the manifold to head acis butterflies to control low rpm air velocity this system beat Honda's VTEC in torque levels with the 3sgte engine having a long stroke producing crazy torque on the high reving K20 and b series engines. 

The 3sgte engine is 1 of the best high powered units of the last 30 years for its size. 

The time delay on the standard map sensor against throttle position is crazy due to the vacume pipe length and manifold position and pipe diameter and the inline filter. 

Be ready for the sharpness of the throttle if you change this. Its really very impressive. Use a rubber buffer placed in-between the manifold and map sensor if you are mounting it directly to the manifold because engine vibration can be seen on the signal its that sensitive. The rubber buffer does stop the vibrations signal error.  

Moving onto the water pump I'm removing it soon replacing it with a fully electrical unit that has a controller monitoring engine temps I'll run this with a restrictor plate instead of a thermostate to produce required engine block pressures. The reason is I can't count on the standard water pump I'm getting alot of fan belt slippage coming on boost. I can also find best engine temps for best power normally found around 50 to 70degc. I can also up this in winter to give better in cab heating and defrosting. It also removes the risk of a thermostate failure.

I'll take some pictures of what was done to improve the standard unit which has a very poor open back impella design. 

At present I'm working threw the standard factory productions errors in the system. There's alot of them. 

Picture above: the great 3sgte intake design.

 

 

Edited by Sam44
Link to post
Share on other sites

I've got a C63 or c64 out of the lotus 2zz.

I've started to strip the box today and have found some differences that are going to course me some problems. the main difference is the final drive differential drive shaft diameter is larger that just at present I'm stuck on, and these will not take the starlet drive shaft cups. If any one knows of any starlets running this box, or a cup to fit the box and the starlet shaft or CVJ otherwise I will have to get a final drive diff gear made at work to house the starlet lsd unit. The lotus and Toyota C63 & c64 boxes have much better lsd units and stronger cup stubs. 

The good news is the 6th gear is thicker (in width) and the final drive gear is alot thicker against the c160 Levin box I have here. 

I've managed to rip the rear gearbox mount off the c160 box last week. Practicing my launch in the dry.

I'm on 165 wide tyres so I thought I'd be safe. I hit just over 5sec to 60mph. @1bar on a tdo4l 13t hyb around 210hp (yet to be dynoed)

Edited by Sam44
Link to post
Share on other sites

Fitting the C4 high output ignition coil. 

After testing the 4efe coil out kv 21kv, over 3krpm this drops off to 18kv. At 1bar of boost pressure this low kv output is starting to affect the spark duration.  And dwell angle ultimately delaying ignition timing impacting egt and the flame propergation in the cylinders. (With air being an insulating element the forced induction\pressure increase affects the spark plugs ability to arce. This with the increased cylinder temps also impacting component resistance) on the ep91 4efe it really starts to become a problem around .8bar. 

I've also installed the E3 plugs these deal with heat alot better than most heat range 6 at 1bar.

The C4 coil delivers 64kv , 85:1 ratio internal windings. A ballast resistor is required to run this coil on the 4efe ignition system because of its low primary winding resistance. I've chosen the higher 1.6ohm resistor for a longer life span of the C4 coil. 

IMG_20200503_084002.jpg

IMG_20200503_083125.jpg

IMG_20200503_084109.jpg

IMG_20200503_083049.jpg

IMG_20200509_103657.jpg

How the system looks as parts: modified 4efe distributor and added king lead, ballast resistor and 64kv C4 coil. Also in the picture is standard 18kv Toyota coil . And a standard distributor.

 

IMG_20200712_180751.jpg

Edited by Sam44
Link to post
Share on other sites

I'm also building this today, so as well as running my nock gauge I'll be listening to the engine threw the car stereo using a 2wire vauxhall nock sensor and the super stereo ear kit. Costing around £18 delivered. 

It took about 1hr to build really easy if you have experience with electronics. If you dont take your time and follow the instructions it's not hard. 

IMG_20200503_110115.jpg

IMG_20200503_154146.jpg

IMG_20200503_154752.jpg

the key to using this to have all and every engine item/engine bay, and drive train item super healthy to start with, use good noise suppression wire connected to the sensor going to the stereo ear  as well as placing the stereo ear device into a insulated box all to reduce phantom knock and have the frequency's your listening to clean. 

use the device alot at first, even recording the sound over a set rout for around 20 to 30 min. you can use this to hear any changes to health of your engine, gearbox/clutch, drive train and detonation over the length of ownership, recording and listening to how these components age wear. i also use the data logged from my engine sensor along side the sound recording to try and pin point the source eliminating false positives. 

i even installed a clutch switch so i knew the clutch position and more importantly the time its being operated as well as gear selector/syncro noise. even helping pinpoint some gearbox noises in the process. 

its very clear even gear and clutch usage can easily be heard. if a problem starts to develop you will become aware of it. at first and only if everything is secure and healthy it does pick up detonation quite easily. the sound of which does change in the upper rpm band.

 i had to filter out road noise and grease my drive shaft cv joints making the sound easy to listen to and pick up det. 

we have also used op-amp ic on an additional beard board to eliminate the need/use of battery's 

running the recorded sound threw the pc software allows you to compare recordings and pin point areas for any further investigation. .

im overall very impressed with this little cheap device, but do ask that it not be used onits own to pinpoint detonation, rather use it as an aid to help reduce the risk of detonation, as well as an early warning device of an up and coming component failure.       

 

Edited by Sam44
Link to post
Share on other sites
On 5/3/2020 at 7:41 PM, glanzagee said:

also whats the frequency that you should be listening at??? (im not asking for a answer ) 

Typically for this engine around 5khz to 7khz. But you can hear it. This works in the human hearing range it's an amplifier. 

It will not display specific frequency's and phantom knock is always a problem. But it's a handy little peace of kit. I came across this on other sites and on u tube. 

From what I can tell it operates in the range of 20hz to 20khz. This could also be recorded and there's some handy audio software out there that can analyse and narrow in on specific frequency's. 

I've wired the knock sensor to the stereo ear using correct 2xpin vauxhall electrical plug and coaxial cable correct diameter to help reduce interference and signal contamination. 

This is what it sounds like as expected human hearing range. The sound on mine is much clearer reduced interference. 

 

https://www.google.com/url?sa=t&source=web&rct=j&url=http://www.ti.com/lit/pdf/spra039&ved=2ahUKEwi7u7imnpjpAhXQTMAKHZslAfYQFjAKegQIARAB&usg=AOvVaw2Egyuh8uWWHFLJVWFJ0nBn

Some useful info. 

I have been researching audio software and this is totally free and looks really good. 

https://www.google.com/url?sa=t&source=web&rct=j&url=https://www.audacityteam.org/&ved=2ahUKEwj8-JauzpnpAhXsRxUIHYM8BqMQFjAAegQIDBAD&usg=AOvVaw3-_gEDCN7gzc2bfAseMMKD

Other useful videos 

I'm also running with catalytic converter off just at present using this gas analyser to monitor exhaust gases. Around 1bar of boost is the tipping point on the tune I'm running. 

 

s-l640.jpg

Edited by Sam44
Link to post
Share on other sites

I'm making custom gauge and boost controller dash mounts today using apoxy resin and cotton sheet. 

First to make the molds. (Base and cover) 

IMG_20200506_145753.jpg

IMG_20200506_145940.jpg

IMG_20200506_145833.jpg

IMG_20200506_162930.jpg

Cover 

IMG_20200506_193332.jpg

Now front face pannel and to use body filler to refine the shape and edgings. 

Getting there. The red shaded area will have around 12 warning lights fitted with symbols: coolant level, fuel pressure, boost pressure, charge temp, engine block coolant pressure ext.

dash.jpg

The yellow dot on the image will be a shift light. 

IMG_20200819_163724.jpg

Edited by Sam44
Link to post
Share on other sites

I've recently fitted these plugs after my NGK platinums heat range 6 required changing for a heat range 7. Because of the effects on emissions on the NGK platinum 7 I opted for a different plug and I'm running these that are dealing with the cylinder temps alot better and still have the reach of a heat range 6 plug. 

https://e3sparkplugs.com/

Edited by Sam44
Link to post
Share on other sites

This has arrived today. Its off a paseo 1.5ltr 5efe it's an additional support for the gearbox. 

I found this on another Toyota site were there is a good mix of cars from the ep91 era and was recommended to me. This was after my rear gearbox mount was ripped out of the casing recking my c160 box the other week (2nd gear launches from standing)

I've had to fit the ep91 UK 4efe box,, which I could not swap over the Cusco lsd unit (it's does not fit).

I've got a g6  box on its way. 

 

IMG_20200508_120606.jpg

Edited by Sam44
Link to post
Share on other sites
On 5/15/2020 at 6:46 PM, RoyalDutchie said:

Nice progress, I do like the more scientific approach keep it up.

Definitely interested in the closed back impeller mod for the water pump. 

Thanks.

I've hit a tipping point now were a bigish mod or mods are required to push forward in power, this I don't think will be of any  benefit to speed\proformance. 

Time to start working on torque levels and spool up. 

Parts to come: copper .6mm head gasket, thermal protective inlet manifold gasket, under chassis fuel cooler, light green 3sge 372cc injectors, twin plate Audi throttle body, corrola tubular inlet manifold modified with a bigger plenium chamber and finally anti lag induction system. 

The aim now is to fill the rpm range with as much torque as I can, I'll be happy with 5e power levels from the hp happy 4e. 

I'll be removing the water pump soon to install a electrical pump and restrictor plate. I'll post up some pictures of the closed back impella. 

Edited by Sam44
Link to post
Share on other sites

Bargain buying all second hand and cheap cheap cheap. 

372cc 3sge denso injectors

Alloy flywheel 5e fitment, 200mm clutch 

Mega squirt V3 standalone ems.

Low milage lsd unit to be reconditioned and installed into the c64 lotus box.

The lot came to under £400. 

Hks blow off valve. 

Injectors just came back from cleaning and testing 

£60

IMG_20200531_063707.jpg

Edited by Sam44
Link to post
Share on other sites

Interesting stuff again,

I am planning on using a corolla manifold on my turbo build as well (design seems best compromise for n/a performance to 6k rpm ceiling and equal length runners with central feed to all 4 cylinders - good for equal cylinder filling on boost), was just wondering what the rationale behind increasing the plenum volume was?

You mentioned in another post that the 5efe manifold you were using has a bad balance across cylinders- "cylinder 1 being the weakest down nearly 20% at 6psi boost levels against cylinder 3 at the 100% reference strongest cylinder".

Was this tested on a flow bench? Or do you have pressure monitoring on all 4 inlet runners? If so sounds cool, pics please! 

Also the Audi twin plate throttle, is this still necessary to help fill cylinders evenly now you've switched to the corolla mani? Or is there another purpose you've yet to reveal!

Good work so far

Link to post
Share on other sites
On 5/31/2020 at 4:42 PM, Claymore said:

Interesting stuff again,

I am planning on using a corolla manifold on my turbo build as well (design seems best compromise for n/a performance to 6k rpm ceiling and equal length runners with central feed to all 4 cylinders - good for equal cylinder filling on boost), was just wondering what the rationale behind increasing the plenum volume was?

You mentioned in another post that the 5efe manifold you were using has a bad balance across cylinders- "cylinder 1 being the weakest down nearly 20% at 6psi boost levels against cylinder 3 at the 100% reference strongest cylinder".

Was this tested on a flow bench? Or do you have pressure monitoring on all 4 inlet runners? If so sounds cool, pics please! 

Also the Audi twin plate throttle, is this still necessary to help fill cylinders evenly now you've switched to the corolla mani? Or is there another purpose you've yet to reveal!

Good work so far

Cylinder balance testing. Basically cutting off 1 cylinder at a time and measuring the loss in engine proformance or crank rotational speed. Along with other tests regarding volumetric efficiency of the cylinder. 

Dyno graphs on the same engine using many many different setups foucing mainly on the torque curve. Will tell you alot about inlet, exhaust, cam configuration. 

Reference to the 4efe units running the Corolla inlet against the 4efe running the 4efte in!et 

Both hp figures are around 85 to 90hp but the 4efte inlet torque curves are very different and there figures tell a different story. also the hp curve. You can clearly see the different in the inlets low rpm gas speed and high rpm volume delivery. With the torque drop off on the corolla unit happening sharp there is a very real problem with the stored air volume (plenium chamber) size. This drop off happens in the power band were when racing  you will be gearing. 

Other graphs to look at to help here are the gen one 4efe EE and ae corrola and starlet gi. The main difference between these setups from what I've been told is the throttle body size the gi runs a large turbo size unit and corolla runs a tiny 35mm unit. (Both using the 4efte design power inlet).

This will also affect the turbos response because the turbo is driven off the amount of air being pumped threw the engine at any point in the rpm. So low rpm gas speeds help fill the cylinder with more air at lower engine rpm. This then brings me to the throttle body design the twin plate design helps increase low rpm gas speeds and also helps increase high rpm volume delivery, plus brings a very smooth delivery (throttle response and control felt on the foot and seen on the graph). better driver control of the engines rpm, small throttles increase foot sensitivity/lower rpm torque, and large throttles give top end hp, its a balancing act.  

Another area of concern is the pre and post throttle body pressure differences at some point of the throttle opening the pressures will equalize this can result in a gas stall (flat point) and also at a set point of throttling you will see a vacuum on the inter-cooler side of the throttle body this is telling us there's not anuff air there at that rpm to feed the cylinders, reference the anti lag comment at the start of the build and the reason for the second dump valve. 

The corrola manifold is by far the best of the Toyota option for road performance forced induction. utilizing a big turbo to amplify it's poor upper engine rpm power/torque/speed, as well as the manifolds ability to increase very low rpm gas speeds (spooling a large turbo sooner) Again reference the many Dyno graphs on this manifold. The manifold also brings great equal cylinder power delivery. The 5e manifold the second best option. But there are many many ways to tune. the turbo cam or 5ehfe cam on the corolla inlet, with a moderate turbo like tfo35 or ct9 hyb using a big exhaust system will be equally impressive and could make higher torque figures because of the big cam Vs the big turbo with boost peak on the 2 turbos the key factor. So this could boil down to the compression ratio and fuel mix. Reference Brad's build on here this lad is tunning his MK1 GT to be a road raping machine. Remembering that the more boost you run the stronger things need to be. This is were money becomes a very real factor. 

The difference between the two setups will be mpg (engine efficiency) and controllability. With a hp cap of around 250hp. But as much torque as we can get our hands on. I'm certainly hoping 5e levels here you can reference the big difference with torque out from the 4e and 5e. From all the research I've seen that has been handed to me and researched my self to confirm the claim, it's looking possible. 

Edited by Sam44
Link to post
Share on other sites
6 hours ago, RoyalDutchie said:

Levin ae111 twin pots can "easily" be installed to boost braking performance, if you still looking for the an upgrade.  Just need some corsa c sxi discs and a ring to reduce the center hub dia of the the disc.

 

Thanks for this information. Im going to start to look into front and rear brake options very soon. Now my acceleration times are getting very good. 

Link to post
Share on other sites
14 hours ago, Claymore said:

Interesting stuff again,

I am planning on using a corolla manifold on my turbo build as well (design seems best compromise for n/a performance to 6k rpm ceiling and equal length runners with central feed to all 4 cylinders - good for equal cylinder filling on boost), was just wondering what the rationale behind increasing the plenum volume was?

You mentioned in another post that the 5efe manifold you were using has a bad balance across cylinders- "cylinder 1 being the weakest down nearly 20% at 6psi boost levels against cylinder 3 at the 100% reference strongest cylinder".

Was this tested on a flow bench? Or do you have pressure monitoring on all 4 inlet runners? If so sounds cool, pics please! 

Also the Audi twin plate throttle, is this still necessary to help fill cylinders evenly now you've switched to the corolla mani? Or is there another purpose you've yet to reveal!

Good work so far

Its was the turbo manifold that produced these figures not the 5e. The 4efe and 5efe did show an imbalance in power balance but a marked improvement on the 4efte unit but again the plenium size was impacting topend hp abilities. Also when boosting the corrola and 4efe starlet and paseo 5e manifolds you will see increases in cylinder temps as you try and push past its flow peak potential which has been estimated around the 270hp mark. The 4efte is around the 300hp mark. 

The twin plate throttle is like combining the gen 1 4efe gi and ee\ae corolla. The difference in these setups is due to the weight and payload of the vehicles with the corrola requiring torque and gi more able to access topend rpm power levels. 

In the later corollas with the TUBULAR manifold design we see the use of the 6speed gearbox and waste spark coils to help access better speeds.

Edited by Sam44
Link to post
Share on other sites

Hey Sam,

Thanks for the very detailed and concise response, it is clear you have spent a long time researching this and thank you for sharing your findings. I am interested as much in the methodology as the results. 

My choice of the tubular corolla manifold for use on my 4efe + t seems the better option for the parameters of that engine, I will most likely use a tf035hm with 6cm housing (I do have a TD04l also that may be tested or could go TF Turbine wheel and 6cm housing with TD04l billet comp wheel and housing?). Hoping whichever option will spool to produce boost at the correct time to carry the corolla manifold torque curve upwards smoothly and plateau in the mid to upper rev range to a reasonable power level.

Looking for a controllable, fast road driven car, reasonable budget that is reliable and doesn't eat gearboxes. I think if I need better power to weight ratio I'll look into weight reduction instead of chasing high hp "grenade spec motors" Lol, Can't afford a crazy build anyway,  always looking to use the best "off the shelf" parts in the interest of availability and reliability. Hadn't really considered the exhaust mani and down pipe design yet, will see whats available.

The Audi twin plate throttle sounds a good solution to the volume vs flow speed balance. I was unsure if its purpose was related to anti lag D-valve you proposed earlier using a sectioned off throttle allowing one plate for normal induction and the second to open to atmos controlled separately regarding air bypass but you mention a second electronically controlled dump valve for this purpose.

I have another question re the larger plenum size which you did touch on; regarding pressure differences before and after throttle plate. With a larger volume plenum would the "flat point" be more pronounced as there is more volume to fill before the pressure difference is equalised? For example at higher revs coasting with closed throttle then opened to allow boost in? Perhaps a part of the reason the fte uses the small manifold for higher rpm, boosted performance.

Thanks 

Link to post
Share on other sites

Your right the plenium size is there as a reserve of air mainly at wot (wide open throttle) to stop the sufercation of the cylinders. this can be a result of the intercooler piping diameter length and intercooler core volumes. And yes a big difference in pressures mainly vacume at low rpm or Cruz high load, and things like a large single throttle plate which is hard to meter exactly using your foot generally will result in gas stall as the pressures sharply equalise. For examples of this look at the size difference of high hp manifold plenium chambers and intercooler piping. Also there was a video up recently of a 4efe running large itb you can see and hear the bad flat spot as he tries to rev the engine leading him to have to open and close the throttles to get the engine to respond. 

The best way I've found to reduce this is firstly use the correct size throttle body for response this gives you the driver control, and you will notice it like any other control on the vehicle. Large throttle bodies are hard to get right the best aprouch I've found is to reduce manifold vacume when throttle is closed by setting up the throttle stop position so again your correct bringing the pressures closer this then allows for sharp throttle application which normally can't be helped with large throttle bodies (not very sensitive). The problems here will be high rpm idle 1krpm + and high fuel usage. As you spend longer than you realize on closed throttle condition's

Your more than welcome, results will come as I get them and move forward. there are other ways to aprouch this like the VW agu design using a ramp covering the bottom edge of the throttle until a certain throttle position is achieved to clear the ramp/obstruction when the bottom edge of the throttle plate starts to flow a great mod here is to trim this ramp down on the vw engine because as always the manufacturers always go to far for safety.

The turbo you are proposing will work great. For the levels of power you have previously mentioned.

The turbo 4efte inlet cam has slightly more lift on the 4efe extending rpm power levels by around 1krpm the setup you have in mind would benefit from a bigger cam arrangment lifting the torque output to tdo4l levels hunting the motorway speeds down better.

The corrola manifold really affects the low rpm power build bringing the cam option into play with the right turbo. Watch turbo surge of power coming on boost with the thin 4efe rods.

The avic boost control is helpful at plotting a boost graph against rpm making things safer. Well placed slightly more expensive parts like these make a big difference. Lsd would be another great mod. 

The tdo4l 13t compressor side even billet race wheels on the tfo35 would be good to see using the corrola manifold and 4efe cam. 

I'd recommend a big large diameter runner exhaust manifold and large down pipe from the turbo to free up any back pressure to the engine with the COROLLA inlet 4efe inlet cam and tfo35 espically if you go hybrid subject to boost levels. 

Edited by Sam44
Link to post
Share on other sites

Cheers for the advice, the piggyback I have is capable of boost control using a PWM table, duty cycle against rpm so was planning on using it to control boost solenoid.

Was going to use a 6psi wastegate spring to bring boost in slow and use solenoid to increase in mid range and stop boost tailing off at the top. And yes I don't want it all coming in low down as I have no desire to meet Rodney or his brothers!

To start with I'll stick with stock N/A cams and C:R, use corolla tubular inlet mani and try TF then TD then Hybrid if I feel the need. (Knowing my luck both the turbos are shagged so I may end up with the Hybrid "opportunity" anyway.)

Take it easy

Link to post
Share on other sites
  • Sam44 changed the title to Starlet Ep91 4efe turbo conversion.

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...