Saturday, September 16, 2017

Drone Warfare

There is a deep primal urge within the lizard-brain of every car enthusiast. It usually surfaces when said enthusiast gets a new vehicle, but it can appear also when routine maintenance is performed on the exhaust system. The desire to undo countless hours of fine engineering work and replace the OEM muffler with something a bit louder is as strong as it is irrational. Rarely is anything besides a few decibels to be gained in doing so, yet we persist and do it over and over again.

One of the greatest risks in changing the muffler is that after spending a little time and often, a lot of money, the exhaust will drone. Drone is an unpleasant experience that happens when the sound waves from the newly loud exhaust resonates the body structure of your vehicle at it's natural frequency. It makes you feel like Manuel Noriega in your self inflicted NVH torture chamber.

I recently did this to myself. Not with Grace, but with my Jeep XJ daily driver, the TURD. The original 1998 exhaust system was on its way out and I got that urge. I picked up a glasspack from Summit racing for about $30, since my friend in high school had owned a glasspack equipped XJ and I remembered it sounding good. I paid an exhaust shop to bend up a new exhaust system and install the glasspack with a new catalytic converter.

Maybe I've just gotten old and maybe teenagers are immune to drone but as soon as I drove the Jeep, I knew I'd made a terrible mistake. The drone at 1500 RPM and 2800 RPM was unbearable, and it sounded so ricey I was a bit embarrassed to drive it. Remember, this is a beat up Jeep with a badge on the rear that says TURD, and the sound was the embarrassing part. I used an app on my phone called Sound Analyser to look closer at the problem areas, and found my biggest issue to be sound levels at 145 Hz.

Sound level readings with glasspack installed. Red areas show unpleasant drone.

After a bit of research, I came across the Flowmaster Super HP2 muffler. It would fit right in place of the glasspack and had a pretty good tone, from what I could find on YouTube. It wasn't cheap, but I figured I'd give it a try.

Top: Flowmaster HP-2   Bottom: Summit Glasspack

Inside Summit Glasspack

Inside HP-2 Muffler

The HP-2 or Hushpower muffler is not the traditional Flowmaster chambered "mousetrap" muffler design, it's closer to being a "straight through" muffler, without actually being straight through. It helped overall sound levels, as well as tone. While it took away the embarrassment of driving my Jeep, it didn't quite fix the droning when at 1500 and 2800 RPM. From my sound level 

Tailpipe as made by exhaust shop
Tailpipe after my modification

Over  the past couple years, I've done a bit of research on different types of noise cancelling resonators for exhaust systems. Most people thing of absorptive resonators when they think of an exhaust resonator, but those don't cancel noise, and they only absorb high frequencies. My issues were centered around 145Hz, which is a relatively low frequency. Of the noise cancelleing resonators, the most common is a Helmholtz resonator. A Helmholtz resonator is like an empty soda bottle that you blow across the top of. A wave resonates inside the bottle, and comes out amplified. If sized correctly and included in an exhaust system, this can actually cancel out sound waves of a given frequency. These are actually included inside many OEM muffler designs. 

A quarter wave  or branch resonator is very similar in function and can be similar in form. It is simply a capped pipe that is teed into the exhaust system at 90 degrees. At certain frequencies, depending on the speed of sound (influenced by temperature inside the pipe) and the length of the pipe, noise is cancelled. After doing a few fairly simple calculations and educated guesses, I chose 30" to cancel my 145 Hz drone. Link to my 1/4 wave resonator calculator. The construction of the resonator was simple and can be seen below.

This is where the resonator tees into the main exhaust system.

I was pleasantly surprised when I went for a drive after finishing. My calculations were correct, or at least correct enough to do what I wanted. The drone at 2800 RPM was completely eliminated, and the drone at 1500 RPM was actually reduced. This can be seen in the plots below. The first plot shows an acceleration without a resonator, the second shows acceleration with a resonator. In the color plots, you can see that the red spot is turned light green, and in the upper chart, you can see that the 145 Hz region was reduced from 74 dB to about 61 dB. A 13 dB reduction means that exhaust sound energy was reduced 20x. The average level (RMS) was reduced by 11 dB. That is an unbelievably large improvement.

I can't say enough about how effective this method is in reducing unwanted exhaust noise.  I was beyond pleased with the result. I've finally found an effective method to please my adolescent lizard brain with a loud exhaust while simultaneously using physics to appease my ever advancing old-man "get off my lawn" requirements.


PS: This blog post was done, and it sat in my drafts box for a couple weeks. I didn't want to post it without a video of what it sounds like. With the new (old) house I just hadn't had time to do it. Today I finished up the exhaust job and had time to make the video. First I'll show the downpipe fab and then show the video.

I had an exhaust shop do the rear part of the exhaust because it was pretty straightforward. At the front, it was going to be a little more complicated. There were two items to deal with:

  1. I didn't want to mess with a two-bolt flange on the header because the bolts always rust and then the nuts are impossible to get off. Instead I wanted to have a V-band connection. To do that, the header needed to come off
  2. I also wanted a large diameter, non-crushed downpipe. As you'll see in the pictures below, the downpipe was significantly crushed. It's a common issue on XJs, I guess as the springs sag, the front driveshaft's front u-joint smashes the downpipe and restricts flow. I figure mine was crushed to about 50% of it's original area
Now for the pics:

Crushed downpipe

Left to right: OEM downpipe size, crushed downpipe, new downpipe size

Expanding the collector to mount v-band flange

V-band flange mounted to collector

Manrel bent XJ downpipe
The new downpipe did just what I hoped it would and increased top end power. Previously, the power would die out above about 4000 RPM. Now it pulls very strong from 4000-5000 RPM. Mission accomplished.

For your viewing pleasure: This is what it sounds like. I like it. I think it brings out the inline-six sound nicely. It's not too loud, but it lets you know it's there.

PPS: This is my Michigan Garage V1.0. It's an almost 2 car garage. The structure is musty and starting to lean, so it will probably be torn down and replaced with V2.0 next year. Since most of my attention will be focused on the house for the next year or so, it'll do.

Sunday, August 6, 2017

The 5 Year Summer

Five years and one month ago I packed up my dad's pickup with my twelve belongings, and with my Ranger in tow, I headed for my new home in Bellflower, California. It was a long, strange trip complete with an exploded battery in the 106° Nevada desert. This past week I moved, this time to Diet Canada Michigan. I can't remember if I've mentioned it here, but my employer is moving my position from California to Michigan, and in the interest of having a job, Jen and I have decided to go. 

Grace has always been a cruel mistress, so it's fitting that we'll be leaving the world class canyon roads I built her for without ever driving her on them. Instead, we'll get to experience the potholes of Michigan together.  In reality, she'll be on the back burner for a bit, until our house is in order... By the way, we bought a house!

I like old cars, Jen likes old houses. Our house, named Old Sid (from a name that was on the property deed) was built approximately the same time Henry Ford was down the road cranking out Model Ts in any color you wanted, so long as it was black.  Old Sid will be getting a new kitchen, new bathrooms, some landscaping work and at some point (my favorite point) a new garage. So yeah, Grace will have to get used the the fact that she's not an only child anymore and Old Sid is mom's favorite.  

Today, I'm just going to try to get current with the work that was done before we moved.


Brakes, in their ideal form, will bring a vehicle in motion to a stop and very little else. Of course that would be far too simple, so my the brakes on my Falcon sometimes attempt to also bring the rear of the car into the direction of travel.  For some reason, the rear brakes had far higher brake force than the front. A tail happy brake system can lead to creases in the seat upholstery from the strong sphincter puckering action it can induce. In the interest of upholstery longevity, I began looking for a solution. 

Applying Occam's Razor, I attacked the simplest explanation first. It is my assumption that the brake pads that came with the suspension and brake kit were incredibly cheap. TCI, who sells the kits, most likely gets the cheapest pads available in anticipation of the end user purchasing a high performance pad after a short amount of time. I picked up a set of Wilwood Polymatrix E compound pads, which are a high performance street compound. A race compound would not work well when cold, and would have much more wear than the street compound.


In my last post, I wrote about the fuel system issues we had on the dyno. Grace ran out of fuel pump at about 350HP. I think the reason was two-fold. First, the pre-pump filter was not large enough, and second, the Bosch 044 pump was pulling fuel too far. 

To address these issues, I bought a larger filter, capable of flowing 400 GPH, up from the old filter's rating of 125 GPH, and I installed a Walbro 255 right at the tank as a feeder or lift pump upstream of the Bosch 044. 

Walbro 255 mounted to the front of the fuel tank.

400 GPH 40 Micron fuel filter from Robb MC Performance
Filter and pump installed

I also wanted to make sure that both pumps got full voltage from the battery, so I rewired the pumps to provide larger gauge wire and shorter runs of wire feeding them with fresh, high-grade, organic, grass-fed, amps and volts. The relays are in the trunk and have fuses built into them.


Cars without interiors are pretty miserable to ride in, and Grace was no exception. I felt it was about time to start working on the miserableness aspect of this car. The first big step was to install carpet. There are a lot of vendors that sell carpet for Falcons, but it seems they all have one thing in common: they all sell the exact same carpet. This is likely true for most makes and models, since there is one major manufacturer of aftermarket carpet, namely Auto Custom Carpets. I found a wide range of prices but the best price was through Rock Auto. I think I paid about $200 for a molded, mass-backed carpet.  

The mass backing adds a little shape to the carpet, and should help reduce interior noise a little bit.

mass backing and felt backing

Due to my raised transmission tunnel I knew no carpet kit would fit exactly. The rear carpet fit quite well, but when it came to the front, I had to do quite a bit of cutting to get it in. This is one of those projects that I'll definitely revisit. The front carpet will be modified to fit the trans tunnel properly, but for now, the noise levels are drastically reduced, so I'm happy.

rear section installed
felt backing installed on mass backing


When things get going too good, Grace is always willing to bring you back to reality. While on the freeway breaking in my fancy new brake pads in, for no apparent reason, the head gasket decided to let go while boosting at about 10 psi (or 275 hp on this engine). Keep in mind that about a week and a half ago on the dyno, this car was making up to 350 HP on E85 and 320 HP on 91 octane. 

At this point I had a follow up dyno appointment set to finish up the tune, now that the fuel system had been improved. I figured I could get the head gasket changed (I'm pretty quick at it by this point, I think I've done head gaskets or rebuilds on this engine about a dozen times by now... seriously, what's wrong with me??) and still make it to my dyno appointment, so I set to work.

Coolant bubbling out the #2 spark plug hole
Cylinder #2 pushed the gasket out

Shiny clean cylinders
But seriously, so shiny!
Luck was not with me, because even though I made good time thinks still went wrong. Any time you see someone using an EZ-out in the garage, step away, say nothing, avoid eye contact, and don't make any sudden movements. while trying to lower the cylinder head onto the engine, it slipped and I broke an oil fittting adapter, which wasted about two hours of my time that I did not have.

I wanted to make sure the head gasket didn't let go yet again so I went to the trouble of getting the correct assembly lube for my ARP head studs. After reading through an ARP technical presentation a few times, I was convinced that this was a necessary step if I wanted to actually achieve the correct fastener preload, or in other words the correct clamping force.

In the end it was all for nought, because the engine was still burning coolant after the head gasket was replaced. With the move coming at me like a freight train, I didn't have time to do anything but load a sickly car on a trailer and have it shipped across the country.

So here I sit in my barren house, waiting impatiently for my car and tools and real mattress to arrive (I've had enough air mattress for the next few years). I'll try to keep updating the blog when I can, but it might be a little less often until Old Sid gets the love he needs. Thanks again for reading, keep your stick on the ice!


Sunday, June 11, 2017

The Day of Reckoning

I had a friend in high school named Blake. Blake was one of the nicest guys you'll ever meet and also drove the most stereotypical redneck truck imaginable (1979 Chevy pickup, lifted 6 inches with 33" tires, Flowmasters, and a carburetor that was way too big). Blake also got really nervous when he went on dates. So nervous in fact that he would sometimes have to excuse himself during the date so that he could go puke. I'm not a terribly nervous person by nature, but there are a few things in life will give me the Blakes. Those things include but are not limited to: getting married, drinking the water in Mexico, and going to the dyno.

My appointment with Church Auto Testing was at 2PM on Friday, so I worked a half day. Even though Grace has run reasonably well for the past 2 years almost, that whole morning doomsday scenarios ran through my head. I was about 50% sure that the head gasket would go, or the intake manifold endcaps would pop off, or the driveshaft would rip itself in half, or... 

The time eventually came and I met my buddy Vaughn for lunch and after a good 'merican burger we drove down to Church. Church is just off the 110 freeway in the Port of Los Angeles. As a side note, I love ports. I could just hang out there for hours and watch ships come in. My wife thinks I'm crazy, but who doesn't love cargo ships and shipping containers?? I've got a friend who loves them so much, he's building a house out of one!

We arrived a little early and Shawn, the tuner, was still out to lunch, so we pulled in and looked around. Inside the shop sat more than a few performance cars. Besides the cars in storage there were just two dynos in the shop. Church has been around for 15 years and does nothing but tuning. They use DynaPack dynos which is one of the reasons I chose them to tune my car. A lot of shops use DynoJet dynos, which are an inertial dyno. This means that the dyno doesn't have the ability to hold the engine at a steady speed or load, and as such, is not useful for tuning part throttle or part load. Dynapack dynos connect directly to the hubs of the car, which meant that we had to remove the wheels to mount the car on the dyno. Unfortunately the rear wheels on my car are a huge pain in the arse to remove and even worse to re-install, so prior to going to the dyno I mounted a set of skinny trailer tires to the rear that I had borrowed from Vaughn.

Once we were mounted up, the tuning began. Shawn started tuning for with the 91 octane gas that was in the tank. First he roughed in the low load and part throttle area of the map, and then got the acceleration enrichment dialed in. This took about about 30 minutes. Now we could start some full power runs so we started at the base boost of 10 PSI. When this came out at 242 HP / 238 lb-ft, the nervousness was gone and I knew that this was going to be a good, puke free dyno day!

After a few runs we worked up to 17 PSI, this is the video for the run:

I suppose we could have kept going, but I wanted to keep the the tune conservative, and I figured 320 HP was sufficient. Vaughn and I then drained the fuel tank of all the 91 and added 12 gallons of e85. The flex fuel sensor read about 75% ethanol. This pretty normal for e85, which is rarely actually 85% ethanol.  When we restarted the engine it ran a little rough, but after a few minutes of tuning all was well.

Shawn set the boost to 20 PSI and attempted a pull. At about 4500 RPM he shut it down because it started to run out of fuel flow. This was a little strange because the fuel system should have been able to keep up. Shawn had me attach a fuel pressure gauge to the pressure regulator and watch as he made another pull. I saw that as RPMs got above about 4000 RPM, the pressure began to drop. This told us that while the injectors should be large enough, the fuel pump was not providing enough flow to keep up.

Despite the fuel flow problem, Grace made 348 HP and 378 lb-ft by about 5000, which I considered to be freaking rad. With sufficient fuel flow Shawn thinks 380 HP / 400 lb-ft is easily achievable. He dialed the boost on e85 back to avoid any damage, so at about 10 PSI, it's making 275 HP and 270 lb-ft.

But enough about the numbers, let's talk about the shape of the torque and power curves. I had initially intended on sweeping cam timing to find the best setting. After the first few runs, I got a pretty good idea of what the curve looked like.  For any given boost level, peak torque is at about 4600 RPM, and peak HP is reached right around 6000 RPM. This isn't a terribly wide spread in peaks, but once full boost is reached, the torque curve is reasonably flat. I have to also remember that this engine was designed the same time John Fogerty was hitchhiking home from Woodstock, so it's not going to have the same mesa shaped torque curve as a variable-timing twin-cam motor designed in a new-fangled CFD program. That said, I was pretty happy with the horsepower peaking at 6000 RPM, so I didn't feel a need to mess with the cam timing.

My turbo, a Borg-Warner EFR series turbo, marketed as a responsive, and early spooling turbo. I run the 6258, which is the baby of the bunch topping out at only 450 flywheel horsepower. As you can see on gasoline, it reaches full boost (17 PSI) by about 3300 RPM. On e85, the spool happens a bit quicker, but with the higher 20 PSI max boost, it is reached at about the same 3300 RPM. For a Ford 2.3L that should ultimately put down 380 HP to the wheel, I think full boost at 3300 RPM is pretty awesome.

I arrived just before 2PM, but didn't get tuning until about 2:30. We spent about an hour and a half on the dyno which cost me $550. Dyno tuning is never cheap, but with a good tuner it is money well spent. All in all, I'm very happy with the result: Grace made plenty of power, didn't leave any ventilation holes in the side of the engine block, and I didn't puke.

That face when your engine doesn't blow up on the dyno, and makes more power than you expected