CARBURETION and your Large/Small port 16V 4AG

[dr.occa]

jamiemirror Wrote:dr.occa, MjLj=MegaJolt. Ignition only management "box". Fully mappable, needs MAP or TPS signal. Can be bought as kit or assembled device.
When you go ITB style - you can't use AFM (vane flap-based device). Using TPS, static advance or advance based on RPM only is practical mostly for circuit racing. So you can try to deliver good vac. signal to MAP sensor and use MAP/RPM igniotion maps (this is the MegaJolt task).
About fuel pumps - any carburated Honda or Subaru is good donor (for 2-nd hand pumps) but FACET brand pumps are available new too. Choose one having in mind your maximum power (and fuel consumption).

Appreciate the clarification jamiemirror on MjLj . Had a feeling that's what it stood for. Using the MSD timing curve "computer" was more economical both in inducing the timing curve as well as fuel consumption (over static advanced timing). Running static advance can lead to overly reach conditions during certain portions of the rpm band which in turn can thin any "good" carbon build up in the cylinder walls and leak pass rings, etc...etc...

As for used fuel pumps from carbed Hondas and/or Subarus; as long as the fp they provide is no more than 7psi and hv I don't see why they wouldn't work as well as for the Facet fp's that you also suggested (they look oddly familiar to a lot of generic off brands that most domestic & import tuners here in the US stay away from).

But again, my initial intention was only an essential listing. Recommendations you made still are good options to choose from. Because there are quite a few products out there (I'm not familiar with everyone of them by a long shot!) I left those specific choices upto the enthusisasts' to decide on. Thanks for the additional information either way.

[firehawk]

BenR Wrote:You can still use the Stock ECU to control timing if you continue to provide it with a semi stable map and TPS signal, its not redundant.

For now, I'm planning to use the stock ECU/distributor for ignition on my car with 44 PHH Mikunis and have been studying how it all works. I've read Jeremy's excellent pages (see here) several times over and asked a bit of advice from him as well. Apparently TPS voltage signal is not needed used for ignition at all and the basic advance map uses RPM and MAP for choosing the advance angle. The IDL signal from the TPS is used to select a simple advance map used for idle/engine braking. If IDL is left open, the ECU will always use the basic map. The intake air temp sensor (THA) will also be missing, but the sensor can be replaced with a 2kohm resistor to make the ECU see a constant 20 celcius temperature.

However, as a longer-term solution, I'm planning on making a simple mappable ignition-only system which replaces the microcontroller inside the standard ECU and uses the standard distributor, igniter/coil and remaining sensors. It'll be a long-term project so don't expect any results yet but hopefully I'll have something to test on my car in the autumn.

[dr.occa]

firehawk Wrote:

BenR Wrote:You can still use the Stock ECU to control timing if you continue to provide it with a semi stable map and TPS signal, its not redundant.

For now, I'm planning to use the stock ECU/distributor for ignition on my car with 44 PHH Mikunis and have been studying how it all works. I've read Jeremy's excellent pages (see here) several times over and asked a bit of advice from him as well. Apparently TPS voltage signal is not needed used for ignition at all and the basic advance map uses RPM and MAP for choosing the advance angle. The IDL signal from the TPS is used to select a simple advance map used for idle/engine braking. If IDL is left open, the ECU will always use the basic map. The intake air temp sensor (THA) will also be missing, but the sensor can be replaced with a 2kohm resistor to make the ECU see a constant 20 celcius temperature.

However, as a longer-term solution, I'm planning on making a simple mappable ignition-only system which replaces the microcontroller inside the standard ECU and uses the standard distributor, igniter/coil and remaining sensors. It'll be a long-term project so don't expect any results yet but hopefully I'll have something to test on my car in the autumn.

great info firehawk. thanks.

just a reminder, this is just an essentials thread I started. this will help get the carb project off of jack stands and on the road sooner than later. ecu management isn't necessary BUT one of many options that would fall under ignition/engine management (i.e. megasquirt, haltech, etc...). highly impressive work by JMR and yet another reminder of how many talented and un-recognized tuners are in our midst: very humbling.

[BenR]

dr.occa Wrote:

BenR Wrote:You can still use the Stock ECU to control timing if you continue to provide it with a semi stable map and TPS signal, its not redundant.

If not, why bother with all this hard work for sorting a crap ignition curve when you cna just buy an ignition only management system?

And it wouldn't be more work to rig up a "semi stable map" signal along w/ jury rigging the TPS just to be able to use the efi ECU (which you won't be able to manipulate w/o extra work and extra money)? I'm not arguing that it isn't do-able. Anything can be rigged up. The information provided above is to facilitate not to convolute the process in converting to a carbureted platform. What I've outlined above is straight forward.

As for a "crap ignition curve" I would ask, how would it be "crap"? I guess it would be schite if compared to let's say a mega-jolt, haltech, electromotive, etc...stand alone system. But I would argue that they all would produce a turds worth of timing in unskilled hands and conversely. Maybe then, our japanese counter parts using the archaic kp distributor adapted to the 4AG haven't a clue that they're using "crap" to achieve only a "crap ignition curve".

Do this, go through what's been outlined above and show where failure would occur in the process drawn out. If there's an even quicker, affordable(a method accessible to all that wouldn't necessitate illegal activity) and easier option w/ an =/> effect produced then by-all-means please share.

Sorry, am i loosing the plot or is using the stock ECU a simpler and easier affair than buying aftermarket ignition components only to be lefth with an rpm based advance curve like something out of the 70's?

Tap the inlet manifold on each runner, run to a vacuum manifold to feed the vac and the brake servo. Keep the dizzy feeding the ECU with speed information, adapt the tps sensor to the carb spindle (not really an issue if your considering jigging a KP dizzy up to fit a vage head). And to have someone remap it would cost what, a couple hundred euros to tweek the tps value settings and ignition map to suit the rapidly increasing map signal with throttle opening?

In my eyes this seems like the fastest, easiest and cheapest way to get a 3d ignition curve that would provide more torque and economy everywhere over a 2 d curve that platues at 3000rpm:S

Or, if your feeling fruity, 380 euros will buy you an igniton only ecu, then put on a toothed front pulley, a tps sensor, away you go, map your own curve and have infinate adjustability for any further advances in engine specification you have.

As for the japs, every JDM import, not matter how 'mint' requires days of work to fix the bodges they love so much. Just because its from japan doesnt mean its good, its like anywhere else, a few good outfits and the rest at monkeys.

[alimonos]

BenR wrote
"And to have someone remap it would cost what, a couple hundred euros to tweek the tps value settings and ignition map to suit the rapidly increasing map signal with throttle opening? "

Is it possible Ben because here in GREECE we hear that our ECUs are not reprogrammable... Any suggestions ?

[dr.occa]

BenR Wrote:

dr.occa Wrote:

BenR Wrote:You can still use the Stock ECU to control timing if you continue to provide it with a semi stable map and TPS signal, its not redundant.

If not, why bother with all this hard work for sorting a crap ignition curve when you cna just buy an ignition only management system?

And it wouldn't be more work to rig up a "semi stable map" signal along w/ jury rigging the TPS just to be able to use the efi ECU (which you won't be able to manipulate w/o extra work and extra money)? I'm not arguing that it isn't do-able. Anything can be rigged up. The information provided above is to facilitate not to convolute the process in converting to a carbureted platform. What I've outlined above is straight forward.

As for a "crap ignition curve" I would ask, how would it be "crap"? I guess it would be schite if compared to let's say a mega-jolt, haltech, electromotive, etc...stand alone system. But I would argue that they all would produce a turds worth of timing in unskilled hands and conversely. Maybe then, our japanese counter parts using the archaic kp distributor adapted to the 4AG haven't a clue that they're using "crap" to achieve only a "crap ignition curve".

Do this, go through what's been outlined above and show where failure would occur in the process drawn out. If there's an even quicker, affordable(a method accessible to all that wouldn't necessitate illegal activity) and easier option w/ an =/> effect produced then by-all-means please share.

Sorry, am i loosing the plot or is using the stock ECU a simpler and easier affair than buying aftermarket ignition components only to be lefth with an rpm based advance curve like something out of the 70's?

Tap the inlet manifold on each runner, run to a vacuum manifold to feed the vac and the brake servo. Keep the dizzy feeding the ECU with speed information, adapt the tps sensor to the carb spindle (not really an issue if your considering jigging a KP dizzy up to fit a vage head). And to have someone remap it would cost what, a couple hundred euros to tweek the tps value settings and ignition map to suit the rapidly increasing map signal with throttle opening?

In my eyes this seems like the fastest, easiest and cheapest way to get a 3d ignition curve that would provide more torque and economy everywhere over a 2 d curve that platues at 3000rpm:S

Or, if your feeling fruity, 380 euros will buy you an igniton only ecu, then put on a toothed front pulley, a tps sensor, away you go, map your own curve and have infinate adjustability for any further advances in engine specification you have.

As for the japs, every JDM import, not matter how 'mint' requires days of work to fix the bodges they love so much. Just because its from japan doesnt mean its good, its like anywhere else, a few good outfits and the rest at monkeys.

...hmmm, inspite of my repeated efforts in reminding that this is an bird-brained essentials list (almost like carbed 4age for dummies) i guess it's highly important to change the initial intention behind the thread. obviously you're in the wrong class Professor. this is carbed 4age 101, the class you're looking to head up is 201.

BenR Wrote:Tap the inlet manifold on each runner, run to a vacuum manifold to feed the vac and the brake servo...

though i agree that this is an easy enough task (although i only used cyl 4 for the brake booster) w/ the TPS and ECU mod, it's going to take longer and more money. Simple, "caveman" technology can be had for practically a steal. 70's technology, hmmm...but yet there are some classic cars still running around with this rubber band & paper clip invention that would leave almost every 86 behind on a 1/4 mile.

again, this falls under the heading of "IG-NI-SHUN". it is upto the individual what product they go with. i made a simple suggestion that i know works better than static full advance. if the end user wants to get more elaborate than by all means, they are free to do so if it is within their abilities.

BenR Wrote:As for the japs, every JDM import, not matter how 'mint' requires days of work to fix the bodges they love so much. Just because its from japan doesnt mean its good, its like anywhere else, a few good outfits and the rest at monkeys.

HUH?! ummm, ooooook...the Corolla is from Japan...
if it's good enough for "backwards", stick utensil-using, karate-chopping tuners, it's good enough for a fellow flip-flop wearing, judo flipping enthusiast in the states. i'm not about to express any of my bigotry towards how others accomplish the tune. if they accomplish it, they accomplish it. whether it's better than my methods or not. you sure are taking a ridgid stance when there's more than one way to skin a cat. learn to be more flexible especially in an area that has no bearing on the eternal.

when enthusiasts are ready to go for an upgrade they'll do so when they're "ready".

Let's keep things simple first everyone. noone's any less of an enthusiast just because you're still trying to drive around in a 20+ year old car with an aged suspension :wink:

[BenR]

I fear you are confusing simple solution, with simple work.

To adapt a new ignition system or another dizzy is FAR more work than to tap each manifold runner and supply the map sensor with a map signal.

Even if you do not run a TPS (which is nothing more difficult than a 2 bolt clamp affair if you want to bodge it), the Ecu will still send out a more advanced ignition curve than a dizzy with advance weights will.

The whole point of carbs, or rather 1 throttle per cylinder is increased torque and bhp correct. So what is the point in taking electronic ignition off and replacing it with something that will reduce the area under the torque curve.

It all just sounds silly to me, not that i am being rigid, maybe just sensible.

Plus i dont see how you can even skip over the whole 'free' part of using the stock ecu.

There are indeed plenty of ways to skin a cat, but if you have an automatic cat skinning machine, whats the point in taking out flint stone and start making a splint that is sharp enough?

[dr.occa]

BenR Wrote:I fear you are confusing simple solution, with simple work.

To adapt a new ignition system or another dizzy is FAR more work than to tap each manifold runner and supply the map sensor with a map signal.

Even if you do not run a TPS (which is nothing more difficult than a 2 bolt clamp affair if you want to bodge it), the Ecu will still send out a more advanced ignition curve than a dizzy with advance weights will.

The whole point of carbs, or rather 1 throttle per cylinder is increased torque and bhp correct. So what is the point in taking electronic ignition off and replacing it with something that will reduce the area under the torque curve.

It all just sounds silly to me, not that i am being rigid, maybe just sensible.

Plus i dont see how you can even skip over the whole 'free' part of using the stock ecu.

There are indeed plenty of ways to skin a cat, but if you have an automatic cat skinning machine, whats the point in taking out flint stone and start making a splint that is sharp enough?

the question is which method would actually be the one making the flint stone...

again, there are plenty of answers that will fill in the blank to the ignition problem. if by all means you say a hamster and your sisters bra knock yourself out. i don't know if i would agree that your solution would be any quicker than the few that i used as suggestions. do me a favor and tell me if you can do what you suggested with tapping each runner, and setting up the TPS to working and running order within 2 hrs it took to install a 6al, tach adapter, coil blaster and timing "computer" plus the @5 minutes to set the timing set. this included soldering and running the wires in a neat and orderly fashion. either way works i'm sure. i never disbelieved either that the gts ecu could produce a better curve.

this is boring.

[BenR]

Frankly i'm shocked that this suggestion is opposed so drastically.

There are no extra wires, there is no setup time, there is no modification of anything, everything works as it does stock :S

Tapping 4 holes in a manifold takes what, all of 2mins?

You dont HAVE to have the TPS connected as its been said already.

A vacuum manifold is what, £8 from pirtek, and you have a large bore takeoff for the brake servo aswell, seeing as you had to tap anyway for the brake servo.

I dont know how anybody can argue with electronic ignition, maybe you like less torque?

[firehawk]

The easiest way of making ignition to work with carbs is indeed using the standard ECU. It can be done in two different ways;

1) Leave the sensors (MAP, TPS, air temp, idle pot) disconnected. This will cause the ECU to drop into "limp home" -mode, which means a constant 10deg BTDC ignition advance. By turning the distributor, one can get a locked, say, 30deg BTDC advance over the whole RPM range. Not an optimal solution but it gets the engine running. The "check engine" -light will be constantly lit.

2) Create a MAP signal and fake the rest. For this, you'll need to tap into each runner on the intake manifold, draw vacuum from them into a small plenum and connect the MAP sensor there. Since the injector holes on the head need to be plugged when using carbs, it would be easy to make plugs which have a small hose barb on them. By joining the four hoses (with, for example, three Y-connectors), one could use the purple TVIS vacuum canister (which is no longer needed) as the plenum.

The rest of the sensors are not used for ignition, so they can be faked with a few resistors. Air temp sensor can be replaced with a 2Kohm resistor, TPS voltage can be connected to +5V via 1Kohm and to ground via 10Kohm resistor, and idle pot (VAF) to +5V via 1Kohm and to ground via 4.7Kohm resistor. This method allows one to use the full 3D ignition map of the stock ECU and use the standard revlimit.

Both solutions are relatively easy and cheap to do and require only very basic fabrication skills. For customizeable maps and revlimit, one would need to either "chip" the original ECU (which is what I'm working on), use an external piggyback ecu or go for a completely standalone system such as the MJLJ. All of these are more complex solutions and require a lot more technical knowledge, tuning time on a dyno and a variable sum of money depending on which solution is used.

Distributors using mechanical advance methods are technically simple solutions but adapting them for 4A-GE requires quite extensive fabrication skills and equipment (lathe, CNC grinder) unless a commercial solution is used (finding one may be difficult and costly). Additionally, in most cases the standard ECU would perform better thanks to the factory-tuned 3D ignition map.

Edit: Oh yeah, the brake booster. My TRD manifold has a threaded hole tapped to the #4 runner and the long tube from the stock manifold screws into it. Even the original vacuum hose fits after I shortened it a few centimeters from the check valve side.