In this chapter, I discuss the installation of a Holley HP EFI system on my personal hot rod, a 1972 Olds Cutlass with a 6-71 Roots-blown big-block Chevy. This is commonly referred to as a draw-through installation as the fuel is drawn through the supercharger. In addition, the setup I run uses dual throttle bodies. When I selected this system, it was the only TBI system that was designed for such an application.
This Tech Tip is From the Full Book, EFI CONVERSIONS: HOW TO SWAP YOUR CARB FOR ELECTRONIC FUEL INJECTION . For a comprehensive guide on this entire subject you can visit this link:
SHARE THIS ARTICLE: Please feel free to share this article on Facebook, in Forums, or with any Clubs you participate in. You can copy and paste this link to share: http://musclecardiy.com/muscle-car-tech-tips/how-to-install-holley-hp-carb-to-efi-conversions/
This system has dramatically increased the drivability, predictability, serviceability, and performance of my Olds. To date, it’s by far my favorite upgrade. Everyone who rides in the Olds is simply amazed at how docile it is when cruising and how nasty it is when you’re in the throttle. Isn’t that the balance every enthusiast is after for a street-driven hot rod? It puts a big smile on my face each time I get behind the wheel and it always will.
I spent a considerable amount of time on the front end of this project selecting the components. MSD, Holley, and MagnaFuel really helped to fine-tune my list to ensure that everything worked harmoniously together. I also upgraded the charging system on the Olds accordingly. The installation couldn’t have gone more smoothly, except for a few AN adapters that I needed and didn’t order and a few that I did order and didn’t need, but that’s just par for the course.
Although the HP EFI is a self-learning system, it’s really best to do the tuning on a high-horsepower boosted application in a controlled environment such as a chassis dyno versus trying to do it on public roads. Mike “Zippy” King (then of) and Greg LaFontsee of Automotive Diagnostic Specialties (ADS) performed the tuning.
Generally speaking, the drivability is excellent and I’ve put several thousand miles on the Olds since the conversion. Tip-in from idle is a bit of a challenge as there are four return springs to overcome (one for the primary of each throttle body and two standard return springs). Opening 900-cfm worth of throttle blades (the primaries of each throttle body) requires a bit of finesse to keep from accelerating from a stop too aggressively, but I got used to this fairly quickly. The Olds has a TH400 with a manual valve body and shifting through the gears is nice and smooth. It cruises nice in town and on the highway. The engine management works just like that of a modern vehicle, except for a knock sensor that I elected not to use because of the harmonics of the Gilmer-style belt drive for the blower.
Most carbureted Roots-blown applications have a pretty good “blower surge” at idle. This is caused by a lean/rich condition that is common to draw-through Roots-blown applications. This surge gets worse as you increase the drive ratio of the supercharger with respect to the crankshaft. The Olds has none. Truth be told, I do like a little blower surge, but I’d prefer not to keep two feet on the brake pedal at a stop sign to keep it from pushing through the intersection. The lack of blower surge with my setup is primarily because the A/F ratio is set to 14.7:1 in the idle area and the blower is underdriven by 13 percent. I’m cool with that.
Throttle response is instantaneous, and I do mean just that. You really have to be careful in a street-driven vehicle with this kind of power on tap.
Diagnosing EFI Running Problems
After you install your EFI system, you may encounter various problems that require some diagnosis and problem solving and mechanical and tuning work to resolve. Certain mechanical problems can manifest themselves as an EFI problem as well. The following are three specific problems I had with my Olds, which is equipped with a big-block Chevy and Holley HP EFI system.
About six months after the installation, the Olds began to run just awful. I pulled out the laptop, connected it to the ECU, and began looking at the Learn Table. Something just wasn’t right as it was populated with a bunch of positive numbers in the idle and cruise areas. This indicated that the ECU was adding a bunch of fuel to achieve the desired A/F ratio, which was the outside of 3-percent margin of acceptability.
As I dug into the problem, I pulled off the driver’s side valve cover and found that I could compress the lifter on intake number-1 with very little effort. Uh-oh. A major top-end teardown revealed that the hydraulic lifter for this valve had a very tiny hole worn through the base of it, which rendered it useless. Further inspection showed that 9 of the 16 lifters were badly worn and several cam lobes wiped. When the engine was originally built it had a hydraulic flat-tappet cam and even though I used an oil with a zinc additive it appeared that wasn’t enough. It was time to rebuild the top end completely.
My buddy Frank Beck designed a complete top-end package for me built around a custom-ground hydraulic roller camshaft that he spec’d to Comp Cams. Engine builders are typically secretive about their camshaft profiles. But let’s just say the old profile had .550 inch of intake lift while the profile Frank spec’d has .664 to .669 inch of lift, more duration, and less overlap.
In addition, I pulled the cylinder heads (iron closed-chamber 3964291s) so that Frank’s team could work on them and entirely rebuild them to accommodate the new and lighter 11/32-inch valves and spring package. They upsized the valves to a 2.25-inch intake and 1.90-inch exhaust, from the stock 2.18- and 1.72-inch sizes (yep, 1.72 inches; weird, I know). Given the collectability of these heads, I wouldn’t let them cut the intake runners to match the intake, but they did a bunch of work in the chambers that is really hard to detect. When I picked up the heads and new parts from Frank he told me, “It’s going to be a radically different animal now.”
Now, these are major changes. How would the EFI system adapt? Would I have to take it back for more tuning with the new parts? Once the engine was fully re-assembled, it started right up. The first thing we did was to set the timing via the distributor so that it read identically with a timing light and in the software (per Holley’s instructions). No problem.
Incidentally, you really want to be sure that your pointer is dead on so that when you do this, you’ll be spot on. This is easily accomplished with a piston stop.
The idle in open loop was nasty with a super-nasty blower surge; it was just the kind of stuff that makes horsepower guys smile and do a high five. The exhaust tone was radically different, so much so that stuff on the shelves in the garage and in the house was rattling. Nice.
We took it out for a drive and I had accidentally disabled the closed-loop feature in the software. It ran well and sounded even better with a radically louder exhaust tone. We noted that the ECU was making no changes to the Learn Table and showed the A/F ratio to be in the mid-12s as we drove.
The following day, I rectified this and began driving it in closed loop with the learn parameters set at 100 percent. Within 10 or so miles, the ECU had a pretty good handle on all the new parts. Just 10 more miles and it was better yet. Even more drive time was required to get the Fuel Table optimized to achieve the desired A/F ratios in the idle and cruise areas but it was pretty close at this point.
The fact that the ECU was able to optimize the Fuel Table to perform well with all these new parts so quickly was pretty mind boggling. By now the blower surge was gone. Now it was time to find a nice stretch of road and let the ECU do the same for the part of the tune with some boost. My buddy Bill Surin (owner of the Olds Cutlass featured in Chapter 6) sat in the passenger seat with the laptop, monitoring the A/F ratio as I attempted to ease into the throttle at 25 mph or so. The shift light came on instantly (6,200 rpm) as the car blew the tires off so violently it banged the rev limiter, which was set at 6,700 rpm. I was out of the throttle as quickly as I got into it.
Bill scolded me, “Dude! I said easy! What are you doing over there?” I just had to laugh as my attempt was maybe half throttle . . . maybe. I tried a few more times but we had the same results. We decided the best thing to do was to spin the blower slower and try again another day. I changed the upper pulley to slow the blower from 9 percent under to 11 percent under and drove it a bit like that.
I then took a few friends for a ride, being ever so cautious not to blow the tires off it. No matter how I tried, the new combination overpowered the tires (33×22.5x15s) nearly instantly. It was as if the pulley change never happened.
I took my buddy Marvin for a ride. He’s a big-block Chevy guy all the way and he was blown away by how powerful it was. In my best “hold my beer and watch this” moment, I said, “Let’s see what it feels like with a few pounds of boost.” I feathered into the throttle in first and it blew the tires off so fast they went up into tire shake. I short-shifted it into second, and boom! something broke.
We pulled over and found the problem. I had sheared the wheel studs off the passenger-side rear wheel (1/2-inch Moser wheel studs). Luckily, the wheel did not damage the quarter panel. We were able to get the Olds home on a trailer.
After pulling the axles and upgrading the wheel studs to 5/8 inch, spinning the blower even slower (13 percent under), and accidentally banging the limiter a few more times, I lost a pushrod cup in my 1980s vintage “gold” rocker arms.
Hmmm . . . Frank told me to replace them when we rebuilt the top end and I elected to run them to save a few bucks. “Dude, throw those out,” he told me. (Words of wisdom as it turned out.)
This minor problem broke numerous parts, scattered shrapnel into the valley, and required a major top-end teardown to retrieve it all. Solution? Get the suspension geometry sorted out and lower the rev limiter to 6,400 rpm until it is. Oh yeah, and listen to Frank.
Interestingly enough, this new combination makes so much more power than the old one, I drove the car two or three more times after this happened, not realizing it was running on seven cylinders. Then I noticed that the throttle response was less snappy, so I plugged in the laptop and looked at the Learn Table in the ECU. It was again populated with data similar to when the lifter failed.
Pulling the valve cover revealed the rocker arm on the number-3 intake valve was sitting 45 degrees away from the pushrod, the pushrod stuck between the guide and head, and the lock on the stud girdle smashed to bits. I got really lucky, and fortunately none of the new parts were damaged in the process.
Troubleshooting with Holley Software
Most of the above is simply attributed to hot rodding. I think it’s important to note that in each case where the engine began running poorly, simply plugging into the ECU and looking at the data in the Learn Table showed that there was a lean condition detected by the oxygen sensor that the ECU was attempting to overcome by adjusting the fuel. Thus began the search to find the problem. The feedback available via the Holley software was very valuable as it pointed me in the right direction to find the problem each time. I can tell you that I’m now a believer.
The Olds is a major handful on the street and it still needs some work on the suspension geometry; but only a slick is going to cure the traction issues long term. Since slowing the blower, it’s rarely seen more than 5 pounds of boost and it makes that at half throttle. It gets loose fast. Honestly, it scares me and this is really a case where the rest of the tuning should be finished on a chassis dyno.
14-Volt Battery Conversion
Before doing a conversion, you must consider three things: increasing the output voltage of the alternator to be able to charge the battery, upgrading the wiring to the electric fan(s) to accommodate the additional current they consume at this higher operating voltage, and managing the speed of the fuel pump so that you spin it only as fast as required to maintain the desired fuel pressure.
Alternator Output Voltage Increase
Depending on the type and brand of alternator you have, you may be able to increase its output voltage quite easily. I installed a Mechman adjustable voltage boost module on the Iraggi alternator of the Olds to accommodate this voltage increase. The Mechman adapter installed inline on the regulator harness. It has a small potentiometer for adjusting the output voltage. This does so by adding resistance to the sense lead, causing the regulator to think the actual operating voltage is lower than it really is.
In Chapter 2, I discussed choosing an alternator that offers more output than you really need so that you’re not working it at 100 percent and creating a bunch of heat in the process. As you increase the output voltage, some accessories such as fans, pumps, and incandescent lights actually consume more current in response.
Electric Fan(s) Wiring Upgrade
An electric fan or pair of them consume current that is directly proportional to the voltage available to them. As voltage increases, current also increases. Remember the Camaro in Chapter 2? That is the kind of wiring I see most often with electric fans and it’s absolutely not up to the task. I typically only use 10 AWG wiring to aftermarket electric fans with diameters up to 16 inches. Some OEM-style electric fans, such as the big 18-inch fans from Ford vehicles, require as much as 80 amps. Be sure to wire them accordingly.
I run a pair of 16-inch Spal fans on the Ron Davis radiator. I’ve measured them at 26 amps each at 14.4 volts, so I know they’re going to require even more at 16.8 volts. My current wiring safely accommodates 30 amps per fan, so it’s time to upgrade that just to be safe.
Also, I’ve been itching to build an ECU-controlled two-speed fan setup for the Olds, ever since I did the EFI conversion. I’ve just never found the time to do it. So, I’m going to kill two birds with one stone.
The Holley ECUs, like most, have dual outputs for electric fans, and each can be set to trigger at different temperatures. Instead of using one output per fan, I elected to use one output to turn both fans on in low speed at 165 degrees and the second output to switch both fans to high speed at 185 degrees. (See Figure 4.6.)
Of course you can choose any temperature you like. This is a really clever circuit that GM used on the C4 Corvette, so I can’t claim it. This circuit is compatible with any ECU that has dual-fan outputs that can be programmed to operate at different temperatures, which covers most of them.
Refer to Chapter 7 for in-depth information on how to configure the inputs/outputs of the Holley HP and Dominator ECUs.
High-Volume Fuel Pump Speed Management
Whether or not you plan to use a 14-volt battery in your vehicle, you should consider managing the speed of any high-volume fuel pump. For that matter, any electric fuel pump benefits from this. By managing the speed of the pump, you’ll enjoy several benefits, including increased pump life, decreased pump noise, and reduced heating of the fuel that is common in a return-style fuel system.
Increasing the input voltage to the fuel pump has an undesirable effect as it pumps a greater volume of fuel and increases the fuel pressure as a result. Reducing the fuel pressure with the regulator further increases the problem of heating the fuel in a return-style system, possibly to the point of cavitation of the fuel pump. All bad things.
The correct way is to reduce the speed, not the voltage, of the fuel pump to the minimum at which it can operate and still maintain the desired fuel pressure. The best way to do this is via pulse width modulation (see sidebar “Pulse Width Modulation” on page 13 for more details).
In the Olds, I elected to use an Aeromotive fuel pump speed controller (FPSC) for this. This controller is suitable for any electric fuel pump. The FPSC utilizes Pulse Width Modulation (PWM) so that it can manage the speed of the fuel pump by managing duty cycle (the amount of ON time). It does this by tracking engine RPM. The FPSC uses the minimum amount of duty cycle to maintain the correct operating fuel pressure at idle. The duty cycle progressively increases as engine RPM increases until it reaches 100-percent duty cycle, which is full output. After you’ve installed the FPSC, it’s quite simple to set both of these thresholds using the included instructions.
I’ve used the FPSC in my Olds for nearly a year and it’s worked perfectly. My favorite part is that the pump is really quiet under normal idle and cruising conditions; so quiet that you can’t even hear it. Incidentally, if you’d like to see exactly how the FPSC works, watch the following videos on my YouTube channel: “14-Volt Battery Conversion Part 1” and “14-Volt Conversion Part 2: Aeromotive Fuel Pump Speed Controller.”
Fuel Pump Speed Controller Installation
This installation was quite simple and the instructions included with it from Aeromotive were very specific. In addition, the kit comes complete with every single part needed to successfully install it.
Remove the existing fuel pump relay(s).
Mount the FPSC.
Connect the FPSC to the fuel pump.
Connect the PUMP+ output to the fuel pump positive input via the included 10 AWG wiring.
Connect the GND output to the fuel pump negative input via the included 10 AWG wiring.
Note that the negative or ground terminal on your fuel pump now has no connection to ground, only to the GND output terminal on the FPSC. Connecting the fuel pump or GND output of the FPSC to ground can damage the FPSC. If in doubt, follow Figure 4.16 closely.
Connect the FPSC to the battery. Connect the GND input to ground via the included 10 AWG wiring. Connect the BAT+ input to the positive (+) terminal on the battery via the included 10 AWG wiring and the included circuit breaker to protect the FPSC and the wiring. In the Olds, I connected this to my existing ATC fuse panel and used a UCB-style circuit breaker.
Connect the FPSC to the ignition, tach signal, and override switch. Use the included wire to connect the IGN PWR input to the trigger wire that you used to trigger your fuel pump relay(s).
You really should be using the Fuel Pump Output from your ECU for this as I did in the Olds. In some cases, the ECU harness includes a pre-wired relay designed to directly drive a small fuel pump. If that is the case, you can use the fuel pump output of the relay to trigger the IGN PWR input. If the ECU harness only has a negative output to drive a fuel pump relay, you need to wire a relay to convert this to a positive output to power the IGN PWR input.
Use the included wire to connect the TACH input to the tach output of your ignition system.
If that tach output is also used to drive a tach (and most are!), you need to run the TACH output of the box to the FPSC first, and then connect the input to your tach to the TACH terminal on the FPSC. In the Olds, I just ran a pair of wires from the TACH input of the FPSC to the front of the vehicle and connected one to the TACH output of my MSD 6AL2 and the other to the input wire of my Auto Meter tach. Aeromotive says this prevents erratic operation of the tach and/or FPSC. I’ve had no issues.
Use the included wire to connect the OVER RIDE input to an optional dash-mounted override switch. This input is negative, so when it is grounded, the FPSC operates the fuel pump at 100-percent duty cycle.
This is a really great feature: you may want to do this when you make a pass down the quarter-mile. I connected this input to the second fuel pump switch in my Painless panel. It was necessary to invert its output from positive to negative to do this, but that was simple. In addition, I used the Secondary Fuel Pump Output of the HP ECU to do this automatically anytime the TPS exceeds 40 percent (see Figure 4.5).
Start the engine and verify that the FPSC works correctly. You should notice the tach LED flashing rapidly to indicate the presence of a tach signal.
Adjust the MIN VOLTS per the detailed instructions provided in the manual. This allows you to reduce the duty cycle and the average operating voltage to the minimum the fuel pump can be driven with to maintain the recommended fuel pressure at idle.
Adjust the SET TACH per the detailed instructions provided in the manual.
This allows you to set the RPM point at which the FPSC permits the fuel pump to operate at 100-percent duty cycle.
Volt Meter Upgrade
Like most vehicles, my Olds had an old-school voltmeter that was not going to be able to accurately display the voltage of the alternator once it was increased. No problem; a new Auto Meter volt meter fit the bill perfectly.
14-Volt Battery Installation
At this stage, all of the work has been completed to ensure that the 14-volt conversion goes off without a hitch. I now need to remove the 12-volt batteries, install the 14-volt batteries, install the Mechman Adjustable Voltage Boost Module, and set the output voltage of the alternator.
I adjusted the output of the alternator according to the directions included with the Mechman Adjustable Voltage Boost Module. I set the output of the alternator so that I measured 17.0 volts at the batteries after cold starting the engine. This couldn’t have been easier. Be sure you connect your DMM to the battery directly when setting the output voltage. See the sidebar Increased Voltage to Electric Fans on page 66 to see how the voltage normalizes after the alternator recharges the batteries after a cold start.
Other accessories such as ignition boxes and the ECU consume less current in response. For Olds, here are the measurements before and after the 14-volt battery conversion.
- Outputcurrentof the alternator at idle at 14.1 volts, at operating temp, all accessories on: 91.3A
- Output current of the alternator at idle at 15.9 volts, at operating temp, all accessories on: 84.7A
The XS Power AGM batteries require substantially less current to maintain a surface charge than did the Diehard batteries they replaced. This reduces the work load of the alternator.
Low-RPM Rev Limiter Interface
Using a low-RPM rev limiter is common in drag racing. You may have one set up to trigger when you hold down the trans-brake button. Most rev limiters intentionally misfire the cylinders so that you can hold the gas pedal to the floor and sit at a programmed RPM.
Misfiring the cylinders means that there can be unburned fuel entering the exhaust. When operating in closed loop, the oxygen sensor picks that up as a rich condition and the ECU begins to make changes to correct this.
So, you have a few options:
- Run the system in open loop when drag racing
- Run the system in closed loop and reduce the Learn percentage to +/- 5% or less
- Run the system in closed loop and use the low-RPM rev limiter built into the ECU, if it has one
Choosing an ECU with a built-in low-RPM rev limiter is really the way to go. As it is, the Holley HP EFI system has this option and the transmission in my Olds has a trans-brake that I’ve never used. (Yeah, like I need that.) Regardless, the interface is simple because you can program this input for positive or negative polarity (see Figure 7.15 page 133 for more details).
Regardless of the specific brand of ECU you elect to run, this interface is simple. Now, when you depress the trans-brake button, the ECU does two things: engage the low-RPM rev limiter (pedal to the floor!) and switch the ECU into open loop.
According to Holley, some transbrakes can create quite a bit of back electromotive force (EMF) when the magnetic field collapses after the trans-brake button is released. This can cause damage to the input of the ECU. At the time of this writing, Holley is working on a module to address this problem entirely.
How cool is that? So, yes, racers really can have their cake and eat it too with EFI.
Wideband A/F Gauge Installation
I also elected to add a wideband A/F gauge to the dash of the Olds. A dash-mounted gauge allows you to easily keep tabs on the actual A/F ratio while driving. I chose an Auto Meter Sport Comp gauge as it matched the gauges already in the dash of the Olds.
The gauge comes complete with an oxygen sensor, weld ring, ring plug, detachable wiring harness, etc. I had a pre-existing plugged oxygen sensor location in my exhaust from the dyno tuning sessions and I elected to use it for the oxygen sensor for the gauge.
I always use Auto Meter gauges in my projects because I’m a stickler for accuracy. I’ve always found their gauges to be incredibly accurate. Good data is paramount when assessing feedback from the engine and you really shouldn’t be looking at a laptop screen when cruising around.
Written by Tony Candela and Posted with Permission of CarTechBooks