The Top DSM Community on the Web

For 1990-1999 Mitsubishi Eclipse, Eagle Talon, Plymouth Laser, and Galant VR-4 Owners. Log in to remove most ads.

Please Support ExtremePSI
Please Support Rix Racing

1G Mechanical Fuel pumps??

This site may earn a commission from merchant
affiliate links, including eBay, Amazon, and others.

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Less draw yes, safer than taking a chance on 1 of the 3 pumps dieing yes. At least with one high volume electrical pump if something goes wrong it will most likely just stop producing fuel flow. At the point of needing this much fuel you should already or plan on running a EMS with failsafes for the just in case moment.

Jake Montgomery ran that magnafuel 750 pump or the Weldon is also a great pump. When you look at the majority who run this style pump they have 16v electrical systems not the standard 14 volts setups. I personally got rid of my electrical pump setup because 2 pumps died and smoked my head gasket. 2 hks head gaskets were close to the price of one aero mechanical pump. Getting the bracket and all new lines fittings was expensive vs just swapping to a big electrical pump for much cheaper.

I run a kiggly 12 tooth crank trigger and 2G cam sensor no problems with timing drift on the car. I street drove it no problems. At the end of the day I just upgraded to what the end state goal was for the car. That is what you must base this decision off of in my opinion.
 
I had a nice detailed post, but it got lost. Oh well.

Summary. Mech is a pita for a street car. 2 walbros looks like it'll be good for more than 150mph traps (800whp) for me. I watch fuel pressure and it stays good, when they are not enough a pair of 044's will be plenty.

They last a couple years each, replace them yearly for $200, and sell the old ones for $100. Lots less hassle then a mechanical.
 
I had a nice detailed post, but it got lost. Oh well.

Summary. Mech is a pita for a street car. 2 walbros looks like it'll be good for more than 150mph traps (800whp) for me. I watch fuel pressure and it stays good, when they are not enough a pair of 044's will be plenty.

They last a couple years each, replace them yearly for $200, and sell the old ones for $100. Lots less hassle then a mechanical.
Im coming to this conclusion, my goal is 9s..... with 8 1650s and 3 pumps I should have enough fuel to get there
 
With 1 set of 1650s*a walbro 450, and a bap you should get there. I did it with 2 sets of 750s, a half dead 255, and a new 255.

3 pumps and 8 1650s is beyond overkill.
My injector duty was getting up there on e85, I figured Id go to a dual rail and just get another set of 1650s. Ive learned my lesson, im going to be pushing the car this year and didn't want to have to upgrade again.We will see...
 
Less draw yes, safer than taking a chance on 1 of the 3 pumps dieing yes. At least with one high volume electrical pump if something goes wrong it will most likely just stop producing fuel flow. At the point of needing this much fuel you should already or plan on running a EMS with failsafes for the just in case moment.

Jake Montgomery ran that magnafuel 750 pump or the Weldon is also a great pump. When you look at the majority who run this style pump they have 16v electrical systems not the standard 14 volts setups. I personally got rid of my electrical pump setup because 2 pumps died and smoked my head gasket. 2 hks head gaskets were close to the price of one aero mechanical pump. Getting the bracket and all new lines fittings was expensive vs just swapping to a big electrical pump for much cheaper.

I run a kiggly 12 tooth crank trigger and 2G cam sensor no problems with timing drift on the car. I street drove it no problems. At the end of the day I just upgraded to what the end state goal was for the car. That is what you must base this decision off of in my opinion.
How do you wire in the 2g sensor using the kiggly kit? This works with dsmlink?
 
Just cut the crank signal wire run it to the kiggly crank trigger. The power and ground wire are shared with the cam sensor. Ecmlink can only be run with the 2 tooth crank trigger which is still provides better resolution than the cas by itself. Not everyone uses the 97-99 cam sensor though. I think it's due to drilling and tapping the camshaft mostly. I like the fact it's stationary regardless of which option you choose and gives a cleaner look in the bay. I can't speak on performance benefits between the 1G cas and 97-99 cam sensor because even the fast 7 second guys run both.
 
Just cut the crank signal wire run it to the kiggly crank trigger. The power and ground wire are shared with the cam sensor. Ecmlink can only be run with the 2 tooth crank trigger which is still provides better resolution than the cas by itself. Not everyone uses the 97-99 cam sensor though. I think it's due to drilling and tapping the camshaft mostly. I like the fact it's stationary regardless of which option you choose and gives a cleaner look in the bay. I can't speak on performance benefits between the 1G cas and 97-99 cam sensor because even the fast 7 second guys run both.
The 12 tooth crank sensor will give better resolution, the stock 2g crank sensor offers no significant resolution over the 1g cas, they are both 2 tooth sensor setups.
 
There's a whole lot more to it. Most standalones have a hard time with the stock dsm signals. There's a lot of harmonic content on the cams, which can throw off a standalone if they are not set up to filter it. The crank should be a cleaner signal. 12 tooth will not give better RPM resolution, it theoretically could provide better timing accuracy, but less precision.

I'm going to talk about the dsm system here, but make some simplifications to make it clearer, one simplification is that the dsm crank sensor signal is 2 sets of 90* on and 90* off.

Generally the ecu uses the processors clock to decode the cam sensor signal, in the case of the DSM ecu it's 2Mhz. That means there is a counter that increments every 0.0000005 seconds, every time the crank sensor changes state (tooth edge passes) the counter is reset, and the value stored. So in the case of a DSM with a 2 tooth sensor (4 edges) at say 6000RPM, we get an edge every 0.0025s - that's 5000 clock pulses - the counter contains 5000 when the next edge arrives. Essentially the ecu calculates RPM by 15/(counter*.0000005).

Hopefully by now you've realized that the ecu can only measure time in increments of 5x10^-7 seconds. Lets look at the implications of this. We saw at 6000RPM, we would have 5000 in the counter. What if the counter has 4999, 6001.2 RPM, and for a counter of 5001 we see it's 5998.8. So we have an RPM resolution of 1.2 @6000. At high rpm it gets worse as the counts come down, and 1 count becomes a larger portion.

Now lets look at it if we were able to run a 12 tooth (24 edges) on a dsm ecu, it's now RPM = 2.5/(counter*.0000005), and 6000RPM is now 833.3 counts, we will use 6002.4 RPM and 833 counts. So lets look at 832 - 6009.6RPM, and 834 - 5995.2 RPM, a resolution of 7.2 RPM.

Unless you increase the clock speed of your ecu, increasing the number of teeth does not improve resolution.

More teeth has the same effect on ignition timing...Lets look at the DSMECU at 6000 with 2 teeth. The ecu only knows degrees per count. It looks at the count from the last edge transition and sees it was 5000, and knows there is 90* between each edge transition, so we have 90/5000 = .018* per count. the 0* edge just transitioned, and we want to fire the plug at 20*, 20/.018 = 1111.11 when the counter hits 1111 fire the plug. Lets look at what timing each increment gives us 1110*.018 = 19.98, and 1112 = 20.016, obviously we have a timing resolution of .018*.

Now lets look at the 12 tooth on the dsm ecu. 90/833 = .108!! Again, 20*/.108 = 185.18, fire the plugs at 185! 184 gives us 19.872*, and 186 gives us 20.08* - resolution of .108*! Again we see that the 12 tooth does not improve resolution.

Where the 12 tooth may help is in accuracy. All the above numbers assume the crank spins with constant velocity, it doesn't, and that can really mess up the timing calculations when for the first portion it is traveling at 2x the average speed. We thought we were really moving at .018*/count, but we were really moving at .036*/count, so the crank moves farther than we thought it did, and we get retarded timing. By going to a 12 tooth we get a position update much more often, and thus the our .108*/count is an average over a much shorter period of time, and there is less time for things to get off. Thus we end up with more accurate timing. In the real ecu there are a lot more filtering and smoothing calculations going on to midigate this, and it ends up not being that terrible, but it's still present.
 
There's a whole lot more to it. Most standalones have a hard time with the stock dsm signals. There's a lot of harmonic content on the cams, which can throw off a standalone if they are not set up to filter it. The crank should be a cleaner signal. 12 tooth will not give better RPM resolution, it theoretically could provide better timing accuracy, but less precision.

I'm going to talk about the dsm system here, but make some simplifications to make it clearer, one simplification is that the dsm crank sensor signal is 2 sets of 90* on and 90* off.

Generally the ecu uses the processors clock to decode the cam sensor signal, in the case of the DSM ecu it's 2Mhz. That means there is a counter that increments every 0.0000005 seconds, every time the crank sensor changes state (tooth edge passes) the counter is reset, and the value stored. So in the case of a DSM with a 2 tooth sensor (4 edges) at say 6000RPM, we get an edge every 0.0025s - that's 5000 clock pulses - the counter contains 5000 when the next edge arrives. Essentially the ecu calculates RPM by 15/(counter*.0000005).

Hopefully by now you've realized that the ecu can only measure time in increments of 5x10^-7 seconds. Lets look at the implications of this. We saw at 6000RPM, we would have 5000 in the counter. What if the counter has 4999, 6001.2 RPM, and for a counter of 5001 we see it's 5998.8. So we have an RPM resolution of 1.2 @6000. At high rpm it gets worse as the counts come down, and 1 count becomes a larger portion.

Now lets look at it if we were able to run a 12 tooth (24 edges) on a dsm ecu, it's now RPM = 2.5/(counter*.0000005), and 6000RPM is now 833.3 counts, we will use 6002.4 RPM and 833 counts. So lets look at 832 - 6009.6RPM, and 834 - 5995.2 RPM, a resolution of 7.2 RPM.

Unless you increase the clock speed of your ecu, increasing the number of teeth does not improve resolution.

More teeth has the same effect on ignition timing...Lets look at the DSMECU at 6000 with 2 teeth. The ecu only knows degrees per count. It looks at the count from the last edge transition and sees it was 5000, and knows there is 90* between each edge transition, so we have 90/5000 = .018* per count. the 0* edge just transitioned, and we want to fire the plug at 20*, 20/.018 = 1111.11 when the counter hits 1111 fire the plug. Lets look at what timing each increment gives us 1110*.018 = 19.98, and 1112 = 20.016, obviously we have a timing resolution of .018*.

Now lets look at the 12 tooth on the dsm ecu. 90/833 = .108!! Again, 20*/.108 = 185.18, fire the plugs at 185! 184 gives us 19.872*, and 186 gives us 20.08* - resolution of .108*! Again we see that the 12 tooth does not improve resolution.

Where the 12 tooth may help is in accuracy. All the above numbers assume the crank spins with constant velocity, it doesn't, and that can really mess up the timing calculations when for the first portion it is traveling at 2x the average speed. We thought we were really moving at .018*/count, but we were really moving at .036*/count, so the crank moves farther than we thought it did, and we get retarded timing. By going to a 12 tooth we get a position update much more often, and thus the our .108*/count is an average over a much shorter period of time, and there is less time for things to get off. Thus we end up with more accurate timing. In the real ecu there are a lot more filtering and smoothing calculations going on to midigate this, and it ends up not being that terrible, but it's still present.


If you were referring to my reply, of course there is and I of course know that, I was letting the other poster know that the Dsm stock crank timing plate has no advantage over the 12 tooth as he stated it does. I run full standalone.
 
If you were referring to my reply, of course there is and I of course know that, I was letting the other poster know that the Dsm stock crank timing plate has no advantage over the 12 tooth as he stated it does. I run full standalone.


Let's clear this up the OP has a 1G 6 bolt motor which does not have a crank sensor on the crank physically. He asked how do you wire the kiggly crank sensor kit and can it work with ecmlink not link because I think you are confused I'll help you out. I gave him the information that it would with a kiggly 2 tooth setup. When I installed a kiggly 2 tooth on my old ecmlink v3 setup it did clean up the RPMs in my logs or should I say made it much cleaner/consistent based off my real world experience. I never compared the stock plate to the 12 tooth or mentioned a 12 tooth setup you in fact made that statement.

I'm not going to clutter the OP's post up. Best of luck
 
Last edited:
Let's clear this up the OP has a 1G 6 bolt motor which does not have a crank sensor on the crank physically. He asked how do you wire the kiggly crank sensor kit and can it work with ecmlink not link because I think you are confused I'll help you out. I gave him the information that it would with a kiggly 2 tooth setup. When I installed a kiggly 2 tooth on my old ecmlink v3 setup it did clean up the RPMs in my logs or should I say made it much cleaner/consistent based off my real world experience. I never compared the stock plate to the 12 tooth or mentioned a 12 tooth setup you in fact made that statement.

I'm not going to clutter the OP's post up. Best of luck
Where we got into the 12 tooth is where you said "I run a 12 tooth Kiggly setup" and then the poster asks: "how do I run a Kiggly setup?" he was likely assuming that you were inferring that a 12 tooth Kiggly sensor would work on a stock ecu, you didnt state that you had anything other than a stock ecu, you then explained to him that he needs to run a 2 tooth trigger setup after he started asking about a Kiggly trigger, your the one that brought up a 12 tooth sensor if you go back and read. You also stated that the 2g crank sensor offers better resolution than the cas by itself which I explained that it does not, the 2g crank sensor uses the same sensor style that is in the cas, the sole reason that the 2g has a separate crank sensor is that with the roll-out of obd2 the mandate was to be able to detect misfires to clean up emissions and protect the cars emmisions equipment (catalyst) the way that Mitsubishi chose to do this is by installing a crank trigger separately that could detect a phase difference between the cams and crank when a misfire is occurring, this is why some car owners end up with a misfire code when installing a 1g cas on a 2g ecu.
 
Last edited:
Yes I did state I run a 12 tooth with no problems with timing drift because it does work better than running just a cas especially with timing drift when referring to mechanical pump setups. So I would definitely say this is an advantage. If its not why are you running a 12 tooth crank trigger? There is a reason every standalone recommends running it haltech, link vr4, aem ,pro efi, fuel tech ect. Now I'm not going to act like I know the scientific breakdown like bastard posted above because I don't. For me obviously from my ecmlink logs the rpms were much more stable so in theory that is due to the ECU getting a better signal or how I put it resolution over the 27 year old black top cas. Just posting my own experience. This thread is going way off subject though and that is where I am leaving this.
 
Last edited:
Has anyone run a rear mounted mechanical pump with a cable drive to the front?
 
Support Vendors who Support the DSM Community
Boosted Fabrication ECM Tuning ExtremePSI Fuel Injector Clinic Innovation Products Jacks Transmissions JNZ Tuning Kiggly Racing Morrison Fabrications MyMitsubishiStore.com RixRacing RockAuto RTM Racing STM Tuned

Latest posts

Build Thread Updates

Vendor Updates

Latest Classifieds

Back
Top