Speeduino MPFI vs Holley Sniper EFI or MPFI vs TBI(my observations)

[tanis4457]

I figured I'd just throw this info out there based on my experience between two different EFI systems.

I have a 76 Ramcharger that I built a Speeduino MPFI system for on the stock low compression 440 motor. I used a crank mounted 60-2 wheel with a hall effect sensor, a set of LS D585 truck coils, a Edelbrock mopar EFI manifold and a Holley throttle body. I tried both the Speeduino V0.4 and seawolf UA4C based boards. At idle I tuned the motor to a best of 33kPa with around 24 degrees of timing, the engine seemed to run very smooth at idle and low speed cruising. Under some higher load situations I could never tune out some surging issues that I determined to be either speeduino or RFI/EMI related that I could never find with my scope. It also developed some really weird situations where the RPM signal would surge all over the place yet show a perfect square wave in tunerstudio and with scope.

Having also installed a Holley Sniper on my 67 Dart I abandoned the speeduino as I was sick of screwing with it for 4 months. I used the Holley Sniper with the hyperspark distributor. My idle now is around 38-40kPa with around 23 degrees of timing and on some of my initial test drives the Holley Sniper has been flawless in throttle response and drivability on the same route I took the vehicle with the previous EFI system.

At the end of the day I can say the Holley Sniper is a great platform working on 2 of my vehicles. On this low compression 440 in the Ramcharger I can say it responds just as well as a MPFI system. I must clarify that I can't say this a perfect A/B test between systems but after getting familiar with these self-tune EFI systems over the past couple years I wouldn't hesitate to say that Holley developed a great product that stands on its own next to MPFI systems.

 

[67Dart273]

Why did you decide the Speeduino route instead of say, Mega/ Microsquirt?

 

[goldduster318]

I've got a MS3 Pro and the thing is awesome. I don't know anything about speeduino but I'm thinking this isn't a platform that's as developed as Megasquirt or the holley MPFI systems.

 

[tanis4457]

Why did you decide the Speeduino route instead of say, Mega/ Microsquirt?

It was cheaper (~$200) and had the option of doing 4channel batch fire and 4 channel wasted spark which isn't available in the Mega 2 so you would have to step up to a Mega 3. With the cost difference there I figured I'd give it a try and see what a ECU at that price point could do.

 

[tanis4457]

I've got a MS3 Pro and the thing is awesome. I don't know anything about speeduino but I'm thinking this isn't a platform that's as developed as Megasquirt or the holley MPFI systems.

It is more experimental and the software is constantly being updated and improved which is nice, I think it might be more suited at this point as a replacement for a stock ECU on an already developed EFI motor. Having used the Holley Sniper computer based software for 2 years I can say that I like the layout better and the 32x32 map compared to the 16x16 map of tunerstudio

 

[pishta]

Why mess with individual ignition channels when EDIS module does it for you (acts like 1 then does the ignition sequencing)? I suppose the 60-2 wheel may not be compatible with EDIS. That is an interesting board, may have to check it out. Seems many random MS problems were poor grounds or noisy VR circuits.

PS>>> the EDIS is proprietary to the 36-1 wheel as it takes 1 revolution to 'set' the timing of the EDIS sequencer. It counts the pulses (from the missing tooth) and fires the corresponding coil pack every 120 degrees for a 6, every 90 for a V8, etc. The coil pack sends a spark to 2 cylinders at once, wired in pairs so one on the compression stroke and the other on the exhaust stroke, ie. a 'wasted' spark. (1-6, 8-5, 4-7, 3-2 on a V8) You'd think the wasted spark would steal power from the combustion spark but it really doesnt. combustion side is usually 4X what the exhaust side is. It also does the advance and retard by a signal from the ECU, pretty smart little box. Also has a 10 BTDC 'limp home' mode in the absence of the SAW signal.

 

[tanis4457]

Why mess with individual ignition channels when EDIS module does it for you (acts like 1 then does the ignition sequencing)? I suppose the 60-2 wheel may not be compatible with EDIS. That is an interesting board, may have to check it out. Seems many random MS problems were poor grounds or noisy VR circuits.

PS>>> the EDIS is proprietary to the 36-1 wheel as it takes 1 revolution to 'set' the timing of the EDIS sequencer. It counts the pulses (from the missing tooth) and fires the corresponding coil pack every 120 degrees for a 6, every 90 for a V8, etc. The coil pack sends a spark to 2 cylinders at once, wired in pairs so one on the compression stroke and the other on the exhaust stroke, ie. a 'wasted' spark. (1-6, 8-5, 4-7, 3-2 on a V8) You'd think the wasted spark would steal power from the combustion spark but it really doesnt. combustion side is usually 4X what the exhaust side is. It also does the advance and retard by a signal from the ECU, pretty smart little box. Also has a 10 BTDC 'limp home' mode in the absence of the SAW signal.

I went through and checked grounds, changed locations, double checked the routing of cables still didn't solve the issue, I suppose the LS coils could've been causing an EMF issue that was getting picked up by something but I couldn't ever verify it.

 

[my68barracuda]

It was cheaper (~$200) and had the option of doing 4channel batch fire and 4 channel wasted spark which isn't available in the Mega 2 so you would have to step up to a Mega 3. With the cost difference there I figured I'd give it a try and see what a ECU at that price point could do.

Coped the paragraph below from the Speeduino Website

Speeduino Goals

• Simplicity of development

In short, the main goal is to be in all places as simple as possible. No weird build environments, no knowledge of assembly needed, favour simplicity over performance where needed and make as low a barrier to entry as can possibly be achieved. Both the hardware and software/firmware sides of the system are covered with all being covered under open licenses.


Favoring simplicity over performance, lowest possible costs,, could result in some drivability issues. But I applaud them for working the low price point market.

 

[TT5.9mag]

What type of crank and cam sensors were you using? If they are VR sensors you absolutely must use some type of shielded wire with the shielding terminated to ground. Sounds like you had a noise problem with the speeduino. What kind of noise canceling adjustments can you make with their software?

 

[goldduster318]

What type of crank and cam sensors were you using? If they are VR sensors you absolutely must use some type of shielded wire with the shielding terminated to ground. Sounds like you had a noise problem with the speeduino. What kind of noise canceling adjustments can you make with their software?

Agreed, the MS3 Pro wiring harness had these shielded, I have LSX round coils with sequential spark and injection, the wire for the crank sensor runs quite near the coils (and especially the electrical wiring feeding the coils) and I have no issue. I have an 04 3.0L Ford Ranger VR Crank sensor. I feel like they actually spent some time making noise filtering.

A distributor makes plenty of electrical noise too given the spark jumping between the rotor and cap, I would think the LSX coils are actually less electrically noisy overall. If you were running 2 wire coils where the coil drivers are in the ECU, that's a different story

 

[tanis4457]

What type of crank and cam sensors were you using? If they are VR sensors you absolutely must use some type of shielded wire with the shielding terminated to ground. Sounds like you had a noise problem with the speeduino. What kind of noise canceling adjustments can you make with their software?

I was using this crank sensor https://www.newark.com/zf-electronics/gs100101/hall-effect-sensor/dp/05M6387 and I used shielded wire and had the shield grounded at one end (the end near the ecu). There's a none/low/med/high trigger filter which also slows the RPM processing I tried it at med and high and that didn't help. For all I know there was some noise getting coupled at the board level somewhere, yet I never saw errors coming in on any of the other sensors. In the end the headache wasn't worth anymore of the time involved troubleshooting.

 

[67Dart273]

A distributor makes plenty of electrical noise too given the spark jumping between the rotor and cap, I would think the LSX coils are actually less electrically noisy overall. If you were running 2 wire coils where the coil drivers are in the ECU, that's a different story

It is not necessarily high voltage that makes noise, that affects low level circuits. For example, the EM field of the PRIMARY wiring driving the coils is a very "big" factor. This gets induced into other wiring just like a transformer

Google up "Faraday shield"---shielded, balanced sensor wiring with one end of the field only grounded

 

[TT5.9mag]

That Newark sensor is Hall effect so it is not as susceptible to noise as a VR sensor. What were you using for a cam signal? Or maybe you weren’t using one since you were batch firing. I guess it really doesn’t matter since you shelved this system and are happy with the sniper.

 

[TT5.9mag]

Del, I remember reading that only one side of the shield gets terminated to ground but never understood why. Any thoughts?

Google up "Faraday shield"---shielded, balanced sensor wiring with one end of the field only grounded[/QUOTE]

 

[tanis4457]

That Newark sensor is Hall effect so it is not as susceptible to noise as a VR sensor. What were you using for a cam signal? Or maybe you weren’t using one since you were batch firing. I guess it really doesn’t matter since you shelved this system and are happy with the sniper.

No cam signal being used. My whole point of the thread was to point out the effectiveness of these new TBI EFI units and in comparison to MPFI are pretty close. My guess is that the MPFI might provide 1mpg better in around town part throttle driving. At wide open throttle neither provides an advantage unless we are also talking individual MPFI injector tuning etc and that is outside the scope of my original intention and also outside the capabilities of off the shelf units like the edelbrock pro flo.

 

[tanis4457]

It is not necessarily high voltage that makes noise, that affects low level circuits. For example, the EM field of the PRIMARY wiring driving the coils is a very "big" factor. This gets induced into other wiring just like a transformer

Google up "Faraday shield"---shielded, balanced sensor wiring with one end of the field only grounded

It is not necessarily high voltage that makes noise, that affects low level circuits. For example, the EM field of the PRIMARY wiring driving the coils is a very "big" factor. This gets induced into other wiring just like a transformer

Google up "Faraday shield"---shielded, balanced sensor wiring with one end of the field only grounded

you are correct lots of EMI from coils charging and discharging.

 

[my68barracuda]

No cam signal being used. My whole point of the thread was to point out the effectiveness of these new TBI EFI units and in comparison to MPFI are pretty close. My guess is that the MPFI might provide 1mpg better in around town part throttle driving. At wide open throttle neither provides an advantage unless we are also talking individual MPFI injector tuning etc and that is outside the scope of my original intention and also outside the capabilities of off the shelf units like the edelbrock pro flo.

Yea but,,, a throttle body will always have wet intake manifold runners, just like a carburetor.
The dry runners that are inherent with MPFI guarantee better fuel delivery. Probably not available on the basic MPFI system that the OP used as his base line, most quality MPFI ecu’s allow for individual injectors duty cycle setting to compensate for higher cylinder operating temperatures that can occur on certain cylinders due to air or coolant flow differences.
So yes, I would agree that throttle body can compare to a very basic MPFI system. That speaks more to the limitations of that basic MPFI system than the quality of the throttle body system.

 

[tanis4457]

Yea but,,, a throttle body will always have wet intake manifold runners, just like a carburetor.
The dry runners that are inherent with MPFI guarantee better fuel delivery. Probably not available on the basic MPFI system that the OP used as his base line, most quality MPFI ecu’s allow for individual injectors duty cycle setting to compensate for higher cylinder operating temperatures that can occur on certain cylinders due to air or coolant flow differences.
So yes, I would agree that throttle body can compare to a very basic MPFI system. That speaks more to the limitations of that basic MPFI system than the quality of the throttle body system.

I agree, In my opinion I think the take away should be unless you’re going to commit to a quality MPFI ecu setup (more $$$) and take the time to properly tune it (ie: dyno time) then save your money and get a TBI like the sniper.

I think for most, not all people weekend warrioring, daily driving a previously carbed car or participating in timed events a product like the Sniper is a damn fine EFI system.

I will say a HUGE thing the sniper has going for it is the access to pc software. It makes dialing in the base fuel and timing maps way more effective plus the ability to tune the iac, coolant based modifiers, air temp modifiers, compensation limits etc....and a built in datalog/playback that can capture a number of inputs.

The handheld ECU graphs only do so much for tuning ability but like any of these EFI systems the self learn makes them work good. Fine tuning with advanced feature sets will really make them shine.

 

[67Dart273]

Del, I remember reading that only one side of the shield gets terminated to ground but never understood why. Any thoughts?

Google up "Faraday shield"---shielded, balanced sensor wiring with one end of the field only grounded

[/QUOTE]

I don't remember the bits and pieces of theory, but grounding both ends can/ does create a 'ground loop' often, and allows current to flow through the shield, which then allows currents to induce interference into the active conductors. Remember, some of these sensors and triggers use VERY small signal levels---sort of analogous to microphone hum for lack of better

Another part of the puzzle is "twisted pair." I'm sure you've seen such things as "Cat-5" network cable, which are twisted. This prevents cross talk between the pairs as well as helping to cancel induced interference from near by cable and wiring. This is "balanced line." Another term to Google is "common mode" interference. The opposite idea, for example, is using a clamp -on ammeter. Take a parallel lamp "zip" cord, two conductors in parallel. You can not measure the current of that circuit by using a clamp-on ammeter around BOTH wires. The current in them is equal and opposite, and the ammeter picks up nothing. The same idea holds true using twisted, balanced pair in sensitive circuits. OUTSIDE currents and magnetic fields have a difficult time causing current flow in the balanced circuit of the small signal device in question.

Telco circuits are also balanced, even though most, today, are not analog circuits.

 

[pishta]

Nailed it. UTP (unshielded twisted pair) is my bread and butter. The twist enables the signal to be very quiet. you split a pair (ie tip of 4 and tip of 5) and youll get dial tone but it will be noisy as the signal cant 'neutralize' itself. Same effect as when you get an imbalance (bridge tap on one side) you get noise. I found some twisted and shielded pair at my central office: perfect for a VR or hall effect run. Its called twin-axial or bi-axial. 2 twisted cables under a braided shield. ground the braided shield at the ECU board side only.

 

[TT5.9mag]

Thanks guys that sums it up well.