While the Twin Cam iterations of both the Softail and Dyna were noticeable improvements over their Evolution-engined predecessors, the Milwaukee-Eight version, in my view, is a huge leap forward.
Words and Photography: Boz
Facilities: V-Twin Custom, Rye – www.vtwinuk.com
Originally published in American-V magazine, Issue 95, February/March 2019
I’ve owned several Softails and Dynas, and while I got what the Softail was about, with its rigid frame looks, the Dyna chassis was just too much of a compromise for me. It worked in spite of itself, rather than because of any sound engineering practices, and was far inferior to the FXR that it ultimately replaced. But the Dyna has gone now, even if its model designations have carried over, and the new Softail chassis is far superior. And not just to the Dyna, but also to it’s earlier namesake.
It’s a brave but necessary move by Harley-Davidson, and it’s pleasing to see – and feel – that function has been less compromised by form.
Having now owned a couple of Victorys I found myself to be far less forgiving of some of Harley-Davidsons previous idiosyncrasies but even with that more critical stance, I have become a big fan of the new Sport Glide.
It’s practical, it handles, has a lovely ride quality and the single front disc actually works very well.
Anyway, enough of this, and on to what this series of articles are about: improving the performance of the M8 engine.
While this new engine has been with us for a couple of years, the aftermarket has been a little slow coming up to speed, and even though a small number of manufacturers are putting some serious R&D into this, many just seem to be trying to alter existing Twin Cam products to physically fit and hope that it works.
Something that I have learnt over the years as an engine builder and rider – and properly quantified in the years that I’ve been dyno tuning – is to assume nothing. What worked on the previous generation of engines, doesn’t necessarily work as well with the next.
This is very important, and perhaps requires an explanation.
Manufacturers are handcuffed to ever-tightening emission and noise constraints, so we are looking as much at the potential performance increases as the power output of successive generations of engines as they leave the factory. And the only common denominator is that they are 45-degree V-twins made by Harley-Davidson, hindered by whatever legislative constraints their respective eras had to deal with.
When the Shovelhead was replaced by the Evolution for the 1984 model year, the bore and stroke (3.498˝x4.250˝) remained the same, for a nominal 80 cubic inch capacity. The Evo, however, had a noticeably improved power potential due to better flowing heads, better combustion, and better cooling.
What worked with the Shovel didn’t always carry across to the Evo.
When the Twin Cam – which had little similarity to the Evo – arrived for the 1999 model year, the bore was increased and the stroke shortened (3.75˝x4.00˝) for 88 cubic inch capacity. But more than that, the inertia mass of the flywheels was substantially altered and the power characteristics were different.
The exhaust ports were in the same place, however, so many existing Evo exhausts would physically fit, and many camshaft manufacturers initially either simply transposed their Evo cam timing figures across to the Twin Cam, or else based the design on a gut feeling and theory.
As it turned out the Twin Cam was more tolerant of higher compression than the Evo, and of shorter cam timing if compression wasn’t increased. And when Harley improved the heads for 2006 and increased the stroke to 4.375, the Twin Cam – depending on year and model – became a 96, 103 or 110-inch, with the 110 having different heads and compression.
Harley’s use of EFI for Big Twins, the rollout of which was completed in all markets by 2007 – 2008 for the Sportster – added far better control over ignition and fueling, as well as helpful data logging when tuning with the appropriate software.
And now with the Milwaukee Eight, we have four valve heads, higher compression, twin plugs, a bigger bore – and a longer stroke also in the case of the 114 – and a different combustion chamber and port design.
So, while potential engine power output and the way it produces that power still remains a function of capacity, combustion efficiency, compression ratios, port flow, valve timing and tuning, other supporting components such as intake and exhaust systems, camshafts, etc. may very well have a different influence on not just the peak numbers, but the way in which the power is produced compared to what we’ve been used to.
And yes, the more efficient M8 does tune differently than its predecessor.
We are dealing with a completely different animal and the only way to learn is to get one and endeavour to understand how it differs from what went before.
And with an open mind, systematically test different components in order to build a picture of what works and what doesn’t.
This is called Research and Development, and I am unashamedly a massive fan of it.
The problem with R&D, however, is that it is very expensive. We need a bike as a test mule, we need to buy an assortment of various components, we need to fit them and we need to test them.
It is also very time consuming – which doesn’t pay the bills – and an appropriate test facility is required.
Sorry, that infamous internet “butt dyno” doesn’t cut it at this level: there is no substitute for hundreds of hours on the dyno as a way of learning and quantifying how the various components interact in order to find a combination that gives the best results relative to the way the bike will be ridden.
No doubt this is why so few places embark on proper R&D programs … and why so many unfortunate owners can make ill-informed and expensive decisions based solely on conjecture and unsubstantiated sales blurb.
1. V-Twin’s test mule and daily driver, the excellent FLSB Sport Glide. It was essential that we started with the bike completely stock so that we could get some baseline runs first, in order to better show the potential gains from the components that we tested.
So, V-Twin has purchased its own Sport Glide to use: not just as a daily driver, but also as a development test mule.
And since we have a dyno, machine shop and engine building facility in house, we can arrive at a parts combination that can be demonstrated to work.
Every day should be a school day for those of us in the business, so that we can then better advise our customers as to what combination of parts, installation and tuning can best suit their riding style.
So, now on to the testing method, and this initially involves testing one component at a time.
The Dyno is used not just to measure the results but to control the engine when used in conjunction with the tuning software. This means that the bike is tuned at each stage to ensure that fueling and ignition timing are optimised, in order to get the best out of each component tested. If the engine management parameters are not optimised, we are making assumptions as to how each component behaves, and the results are at best spurious, and at worst, totally worthless.
Yup, it is definitely time consuming …
This 107 Sport Glide had 2500 miles on the clock, fresh oil and the same brand/grade of fuel (Shell V-Power) was used throughout testing.
The tuning is done with Technoresearch’s Direct Link – now distributed in the UK through Parts Europe – which I have been working with for the last twelve years. By using the dyno to control the engine, and measuring the results, a new ‘map’ is built on the computer and then flashed into the bike’s ECM: after tuning, absolutely nothing is left physically connected.
There is obviously a necessary learning curve here too, as not only does the M8 have a new ECM with more sophisticated software levels than the Twin Cam, but ‘what the engine wants’ in regard to ignition timing and fueling, is not the same as the Twin Cam.
Fortunately, I have already tuned a number of M8s, and the new software – while more complex – retains a familiar layout.
A final point to note is the relevance of what gear is used for the testing: there are two factors at play here: the way the dyno calculates the power, and parasitic transmission losses through the gearbox.
Typically, the higher the gear, the higher the numbers due to how the engine loads up.
And we want to use the gear that gives a one-to-one ratio as fewer gears are used to transmit the power, which with a stock Harley ‘box is sixth gear.
However, with the tall overall primary gearing that Harley use with their six-speed ‘boxes, this equates to a road speed approaching 160mph at 6,000rpm in sixth, which is a bit unfair on an H-rated tyre.
We have to be sensible with this, and while I don’t mind running in sixth on an occasional basis – to get the final numbers with a good parts combination – I won’t take the chance of a tyre letting go by constantly abusing it on the number of wide open throttle testing that this R&D project will entail, so most of these dyno charts are done using fifth gear, and sixth only used to illustrate the final combination as far as peak numbers are concerned.
Fifth gear testing remains totally relevant for this project though, as each test is consistent and we are after comparisons rather than numbers.
I hate the terms ‘stage 1’, ‘stage 2’, blah blah blah, as it tells us absolutely nothing, but since these nonsensical acronyms have become such common place in the Harley market, I will cave in and discuss what I believe a ‘stage 1’ should represent.
This is the easy to get, low hanging fruit, as we don’t have to go inside the engine, and it should consist of a better flowing air cleaner, an exhaust that hopefully adds power and not just noise, and an appropriate remapping of H-D’s excellent fuel injection system.
The first two are the laws of physics at work, and the tuning can only maximise what they give.
A bad combination of parts cannot be rectified by tuning, because as clever and as educated as we might wish to become, we can’t change those laws. Of course, an un-tuned or badly tuned good combination of parts cannot perform as intended either.
So, to recap, this bike was tuned specifically to suit each component – or combination of components – in order to get the best from them, and make the testing valid.
Something that may come as a surprise to many is that I started with air cleaners first, and not the exhaust systems.
Harley’s stock exhausts, while somewhat restrictive – and this varies from model to model – are not necessarily the butt plugs that they once were.
Something that I found with the Twin Cams is that these engines want air more than they want more noise and we need to deal with the restrictions at both ends.
Now, I am fully aware that many owners choose their exhausts mainly for the sound and the appearance but this article is not about that very subjective view, it is about finding out what works taking into account the limitations of parts availability, budget and time constraints.
For the air cleaners I picked three that I had previously had good results with on Twin Cams: the S&S Stealth, K&N RK series and the Screamin’ Eagle Heavy Breather from H-D.
And at the time of conducting these tests, there were no slip-on mufflers available for our Sport Glide, so we picked two complete systems to try: the Bassani Road Rage 3 – which I have had good results with on Twin Cams and Sportsters – and the relatively new Two Brothers Racing system which, like the Bassani, is also a two-into-one made from stainless steel.
So, finally, off we go then.
 1: Firstly we measured horsepower and torque with the bike completely stock, in order to establish a base line. This is a fifth gear run. Notice how the speed limiter shuts things down. |
 2: Here’s the stock bike tested in sixth gear. Notice what I said earlier regarding the numbers. Notice also how it shuts off at a lower rpm. This is still due to the Big Brother speed limiter. |
 3: Back to fifth gear, but this time the completely stock bike has been tuned. |
 4: This overlay gives a better idea. Overall power has increased but only by around 2%, although it did feel more responsive. |
DEEP BREATHING EXERCISES
Air cleaners are not just about the size and type of filter element, they are also about how they manage the air. The three that are tested here all have excellent cotton gauze filter elements, but each adopts a different approach to air management.
As we shall see, the stock air cleaner is quite restrictive, and while there may be some benefit to using a better element, its air inlet is quite small and reminded me of some of the stock Evo air boxes.
Another point to note is that intake noise is also a factor when H-D submit their bikes for emissions and noise testing approval.
The cynical amongst us would argue that since the Heavy Breather is listed as EPA compliant, why wouldn’t they use that as standard, as they did on the Dyna Low Rider S with the 110 engine?
Maybe they just want us to spend more money on their accessory range?
5: S&S Stealth with the stock exhaust, tuned. This is relatively cheap power, around 6% through the lower and mid range, with a 10hp peak gain.
S&S Stealth and stock exhaust.
The three pictures illustrate well how S&S have approached this.
The assembly is intelligently made from high strength plastic mouldings with threaded brass inserts. The fitment is both simple and accurate.
The backplate incorporates a radiused entry into the throttle body – air doesn’t like sharp corners – and the outer cover uses a ‘stinger’: a protruding guide to help direct the airflow.
This is quite well priced, although the optional billet face plate, which is a cosmetic addition, is bit pricey.
6: I was a little surprised by this as it doesn’t perform as well as the much cheaper S&S. But as I said earlier in the article, we must assume nothing. This illustrates that point very well as we move on from the Twin Cam and deal with the Milwaukee Eight.
K&N RK series and stock exhaust.
I’ve had some good results with this RK series from K&N on some Twin Cam builds that I’ve done, albeit with a deeper filter element.
This version is quite pricey, but apart from the rubber inlet entry it is all metal.
It also includes the face plate and a black painted cover for the exposed part of the throttle body actuating mechanism – not shown in the pictures as I didn’t want to risk marking it, in case we sell it on.
K&Ns approach is the theoretically superior radiused entry that rolls back on itself. And, of course, they make the best filter elements that last forever when cleaned at regular intervals.
As with the S&S, fitment is perfect and is a very tidy installation.
7: I had hoped, rather than assumed, that this SEHB would perform in the same way on the M8 as it did on the Twin Cam, and it does. This is nearly a 10% gain where we most often ride. I’ve seen crappy noisy exhausts make a smaller power increase than this on the Twin Cams. For those who are happy with the stock noise level, but want more oomph, this air cleaner and a tune is the way to go.
Screamin’ Eagle Heavy Breather and stock exhaust.
I’ve been a big fan of this air cleaner since it first came out. I don’t know if H-D just plagiarised the original Forcewinder for cosmetic reasons, but this has been one of my ‘goto’ air cleaners since I first tested it on the Twin Cams.
It works because of the increased intake runner length which increases torque in the lower rpm range, but at the expense of the peak horsepower at high rpm.
Unless we really are chasing the numbers by riding close to the rev limiter, this has always been very suitable for the way most people ride.
It has always been expensive and is now even more so with the M8 version, but it is of far neater construction than its earlier siblings.
I still can’t get my head around how an air cleaner can be more expensive than a set of forged pistons though.
8: Here are the air cleaner tests overlaid, sorry if it looks a bit busy.
Completely stock // K&N RK series with stock exhaust // S&S Stealth with stock exhaust // SEHB with stock exhaust
So here’s the conclusion for the air cleaners when used with the stock exhaust, and from discussions I’ve had with many owners through the years, few realise the benefits. Hopefully this will address that, as there are decent gains to be had with the right choice.
For the best improvement in the rpm range where most owners ride, the Screaming Eagle Heavy breather wins, but with a hefty price tag.
The S&S Stealth gives the best peak power, and due to it’s price, represents a good bang for your buck, although less so if we add the additional billet face plate for cosmetic reasons.
Unfortunately, the rather pricey K&N RK, while nicely made, and a better performer than the stock air cleaner, doesn’t stack up on the Milwaukee Eight in my view, other than for very subjective cosmetic reasons.
Now we move on to the next level, which is testing the exhausts, but I’m not bothering with either the stock air cleaner or the K&N.
EXHAUSTIVE ANALYSIS
It’s well known that Harley engines have traditionally benefitted from a two into one exhaust, but I suspect that some manufactures have gone down this route for a currently fashionable racing look. Just because it’s a 2-1 doesn’t mean that it will perform, there’s more to it than that, and the Remus system for the XR1200 and the Roland Sands Slant for the Twin Cams and Sportsters are two examples of how function can be sacrificed for the purpose of form. I’ve had good luck with Supertrapp’s various offerings since I first used one on my FXR way back in 1989, but they have nothing available for the M8 Softails at this time.
Two Brothers Racing exhaust with S&S Stealth and Screamin’ Eagle Heavy Breather
I’ve not used one of these before, so was very intrigued to try it.
The TBR looks well made, if the finish is a little industrial looking, and – unusually, has the collector as part of the muffler – with the individual headers using slip joints and springs.
The M8 Softails have less room around the front cylinder exhaust port due to the proximity of the oil cooler between the front down tubes, so removing the exhaust flange nuts is more awkward than we have been used to with the Twin Cams and Evos.
I would strongly recommend anyone with an M8, to replace these nuts with stainless steel ones, used with copper grease, even if the stock exhaust is retained. The nuts corroding to the exhaust studs are going to be more difficult to deal with in the future than the Twin Cam’s are because to this access issue.
This FLSB, of course, is new enough for corrosion not to hinder the exhaust swap.
Fitment was good and there is a supplied bracket that allows attachment of the saddlebags’ ‘cotton reel’ front mount, but having removed them for testing I can’t say how well that fits.
And the front header sits high, suggesting that there will be no ground clearance issues when cornering hard.
 9: Stock |
 10: Stock |
 11: TBR and SEHB |
Road Rage 3 and SEHB on the FLSB. Although the front header appears to hang a little low, it doesn’t cause any issues. The fitment to the bike is excellent.
Bassani Road Rage 3 with S&S Stealth and Screamin’ Eagle Heavy Breather
I’ve been quite a fan of this pipe since it was released a couple of years ago.
I first used it on a Dyna Low Rider S 110 and was impressed, and I’ve also had good results with their Sportster version and more recently a Fatbob 114 M8 with Harley’s torque cam.
No surprise, then, that this pipe was on my list.
Fitment is impressive, as is the finish: better on both counts than the TBR.
 12: Stock |
 13: Stock |
 14: And here’s the Bassani Road Rage 3 with SEHB overlaid against the S&S Stealth. As we’ve seen all the way through this testing, the S&S has the peak horsepower edge, whilst the SEHB produces more torque in the lower and mid range area. |
Hopefully, if all of these charts haven’t proven too confusing, we can now see the validity of systematic testing in order to arrive at a parts combination that best suits our riding style.
By only testing combinations, rather than the individual components, we wouldn’t have shown which component was holding up any particular area – or, indeed, which component is contributing to the gains – which is why we started simple, isolating the air cleaner first.
So, if we are chasing a peak horsepower figure at high rpm, then we can see that the S&S Stealth paired with the Two Brothers Racing is the combination required.
15: Where do we want our power produced?
Stock // Two Brothers Racing and S&S Stealth // Bassani Road Rage 3 and Screamin’ Eagle Heavy Breather
Look at the area under the torque curve between 2000rpm and 4000rpm. It’s easy to see which combo give the best result if this is where the bike is most often ridden. Conversely, look at the hp curves above 4000rpm. If this is where the bike is ridden, then we can see the better choice here.
On the other hand, we have seen that the Bassani Road Rage 3 and the Screamin’ Eagle Heavy Breather produces a bigger gain in the lower and mid range.
Finally, I wrote earlier that the testing was carried out in fifth gear to protect the H-rated rear tyre from excessive road speeds, and that the lower gear produces slightly lower numbers because of the way that the dyno works, combined with power losses through the gearbox.
Here is the overlay of the bike completely stock and what I believe to be the best real world combination, in sixth.
16: 6th gear runs
Stock // Bassani Road Rage 3 and Screamin’ Eagle Heavy Breather
A huge improvement for real world riding, and very responsive.
Next time we will continue this testing, but with different camshafts, and then with the cylinders bored out to give 117” conversion. Should be fun …
Comments 1
Very interesting; may have to pay a visit to Rye 🙂