8 stack and Holley ECU's

[Stephen]

Morning Gents,

I'm looking at converting my 419w from Sniper EFI to 8 stack.

I was going to use the Holley Terminator ECU, as I like the Holley tuning software.

I just have a few questions:

Are people using a crank trigger on the front as well at the dual sync distributer? Or just one of the two?
Are you using MAP for idle then transitioning to Alpha N beyond 3%TPS?
Is anyone willing to share a map of a similarly sized Windsor?

 

[Andrew Robertson]

Hi Stephen, I'm running a Holley HP on a LS3 8-stack set-up on my restomod project. Personally I would use a crank trigger for timing because of its much superior stability and just use the distributor for cam sync assuming you want to run sequential. I'm using full Alpha N due to the relatively low vacuum the 8-stack provides. I'm making around 520hp with a 226/238 112 LC cam. If that's in the ballpark for your 418W happy to post you my tune.

I will say that the Holley is good but not great for full Alpha-N - it really has been created as a speed density ECU primarily, in my opinion. Alpha N is very fussy to tune in the first 5% or so of TPS and the software does not give the tuner complete control over some stuff because it tries to make things easier for the average guy. The learn also doesn't work that well down low for Alpha-N either, so tuning for idle and cruise is done via datalogging and manual adjustment. Having said that, if you are prepared to put the tuning work in to get cold start, idle and cruise well dialled in, it will run good.

Cheers, Andrew
Wellington, New Zealand

 

[Brian]

If your manifold has a common vacuum plenum with a port from all 8, Megaquirt has a hybrid speed density / alpha N function where under part throttle, up to 80 kPa (or where you sett eh threshold) it'll run quite nicely in speed density mode, but when you throw it the beans, it goes into alpha N.

I have it on a 383 SBC and it's quite sedate under part throttle, but screams when opened up. It's batch fire MS2, but the feature is available in MS3 if you want to got sequential.

Timing of spark is super critical, but honestly, timing of the injection is really not. In fact, under low duty, the fuel is squirted on the back side of a closed valve by OEMs, and it's closed 2/3 of the time. So if you're using the distributor for the timing, there's no real need to add a trigger. But there are a lot of nice features EFI can bring to spark timing.

The sniper's ECU will not be sequential. Sequential is a big advantage for low load drive-ability as the injector is making one more easily measured squirt per 2 revs rather than 4.

 

[Stephen]

Gents, @andrewrobertson @Luce_

Thanks for the above.

I assumed (wrongly) the Holley would work with speed density and then Alpha N.

So really my only option is an MS3 with both a crank sensor kit and then use the dual sync distributor as a cam sensor.

I currently run a custom timing table using my Holley to trigger my MSD 6AL and Dual Sync Distributor. I'd like to keep this set up, as running coil near plugs looks a bit too modern.

 

[Brian]

One warning about megasquirt. You'll likely need to be your own expert. They have a not deserved bad rep because the low price point lets kids afford and poorly install it, then bring a almost running car in to be tuned. So most tuners will not touch it unless it's already running pretty well. At that point, you really don't need a tuner.

But Tuner studio and megalog make it fairly easy. Tuner studio will tune your VE table, and with megalog, you can make a histogram over the entire rpm/throttle or vacuum map and easily see where your troubles are, and it has an algorithm that can fine tune your VE from the histogram.

Check it out if you're interested. http://tunerstudio.com/

I've had MS2 on my cobra 13 years. Only issue has been the original trigger not being rated for the temps under the hood. In traffic over 100 degrees outside, I start getting sync errors. I know the problem, but haven't gotten around to making a new trigger mount for the newly available hall effect trigger.

 

[Stephen]

One warning about megasquirt. You'll likely need to be your own expert. They have a not deserved bad rep because the low price point lets kids afford and poorly install it, then bring a almost running car in to be tuned. So most tuners will not touch it unless it's already running pretty well. At that point, you really don't need a tuner.

But Tuner studio and megalog make it fairly easy. Tuner studio will tune your VE table, and with megalog, you can make a histogram over the entire rpm/throttle or vacuum map and easily see where your troubles are, and it has an algorithm that can fine tune your VE from the histogram.

Check it out if you're interested. http://tunerstudio.com/

I've had MS2 on my cobra 13 years. Only issue has been the original trigger not being rated for the temps under the hood. In traffic over 100 degrees outside, I start getting sync errors. I know the problem, but haven't gotten around to making a new trigger mount for the newly available hall effect trigger.

I've used Tuner Studio with Speedunio ecu's.

I am concerned with the lack of size of the tables though. It's been a while since I've used 16/16!

 

[JP Verweij]

Megasquirt can be used to run a stock dizzy and a single IGBT coil. I run a MS2E setup (not GT40) which fires the IGBT coil ( a stock coil will fry the MS). My dizzy only distributes the spark to the plugs by cap & rotor. My advance system is not locked, it just phases along the MS2E ignition curve.
Looks are just stock carb & ignition.
I am running crank trigger only and in my case, MAP sensor, TPS & ATS. (Ford Capri mk1 2600 Retroject injection
TunerStudio MS is a must, especially the Ultra is recommended. Makes fine tuning yourself way easy.
Downside of TunerStudio, its Windows based and won't run on Android systems. So its tuning with a Windows Laptop & a carcharger that powers your laptop.

EFI needs a return, to keep fuel flowing through the rails. Fuel cools the rails & injectors. Without a return you will suffer injector heat soak. Makes hot starts a pain in the ass.
Heated up injectors will change the injector impedance (resistance in Ohm) and as a result it will alter opening time (shorter duration) creating a lean mixture.

 

[Brian]

It interpolates between the cells, so while you're limited to 16x16 points to adjust, but it makes an infinite map, and there should be no abrupt changes. And you don't have to space the bins linearly.

If you've played with Speeduino, this sounds right up your alley.

 

[Andrew Robertson]

Hi Stephen, don't get me wrong on the Holley - it will work fine if you commit the time. The main benefits of it is the huge forum support and the user friendly software. The main issues I have with it are the hybrid Alpha N mode (Alpha N for idle and light cruise, transitioning to Speed Density for the load above that) needs more flexibility (I would then use it rather than full alpha-N), the idle control uses set parameters that interfere a bit at 2% TPS and below, and some table axis are fixed and use a linear scale not suited to Alpha-N which requires lots of resolution for the first 20% or so of throttle angle. To illustrate how much resolution and control is required, my car cruises at 65mph and 2000 rpm in 5th gear at 4% TPS, and 0,1,2,3,4% TPS all have materially different fuel requirements. My engine load TPS axis is in single digit increments up to 7%.

Also - in terms of hardware - get a high quality contactless TPS sensor. Something that drifts by even 1% will be a real pain - I found this out.

Cheers, Andrew

 

[Brian]

Great bit of advice about the TPS. Most in the aftermarket realm are crappy and noisy. It took me a while to catch this. When cruising, my Cobra would occasionally hiccup. No rhyme or reason, could be 3 times in a 10 minute span or not for 2 days. It ended up being an occasional blip of the TPS causing an acceleration enrichment that wasn't needed.

 

[Stephen]

@andrewrobertson did you mean speed density transitioning to alpha N?

 

[Stephen]

Last time I used tuner studio for an Itb set up where I had no Map reference I had two TPS tables, with TPS %age as the switch over.

The first table has 1% to 16%, the second had 17% to 100% split as evenly as possible.

It makes good use of the secondary table function.

 

[Stephen]

As for TPS I normally use Penny and Giles contactless sensors.

 

[Andrew Robertson]

@andrewrobertson did you mean speed density transitioning to alpha N?

No - load function starts TPS then transitions to MAP. Holley's use case is a big cammed v8 with erratic and low manifold vacuum at idle and low rpm low load where TPS is preferable as an indicator of load, transitioning to MAP at the point there is a usable signal with decent resolution, say above 1500-2000.

 

[Stephen]

No - load function starts TPS then transitions to MAP. Holley's use case is a big cammed v8 with erratic and low manifold vacuum at idle and low rpm low load where TPS is preferable as an indicator of load, transitioning to MAP at the point there is a usable signal with decent resolution, say above 1500-2000.

Every day is a school day!

So it works in the reverse of what you’d want with ITB’s?

 

[Brian]

I took a cheap chinese 8 stack and made it work really well. There's 2 separate chambers. The one with 4 smaller lines tap into the wall and catch both runners. It's for the MAP sensor. The other one with 8 3/8" copper tubes are for a Ford PWM idle air valve. If I was to have tried a single common chamber, the idle air valve would have affected the MAP reading.

Also, the large 3/8" tube make the balance much less critical as it's enough air that at idle, the throttles are almost closed.

The setup has been very well behaved and the idle air valve holds a solid idle with A/C and or fan kicking off and on.

 

[Brian]

I cut the stacks short on the lathe so I could run a big air filter. Added a Powerstroke vacuum pump since this is in a C3 Vette with all of the vacuum shit.