Turbo Tech Tip #01 - A short Rant about Primary Tube size

If you enjoy this tech tip, this is only a small glimpse of the type of content that's offered on our Patreon. We have 11+ How-to Video's over there, with our Turbocharging How-To video being close to an hour long, complete with tuning from scratch on v-twins and inline fours. You can become a free member to access posts like this, as well as build updates. Or, consider joining the Journeymen tier to unlock all of the how-to content. Check it out! www.patreon.com/doctormotorcycle

This is a subject that has a lot of voodoo and witchcraft involved in it in the N/A world, and almost just as much in the turbo world. That said, even the most low effort, cursory search will confirm what I'm about to say in this article.....yet people continue to build turbo setups with manifold sizing that's way, way too large.

Much like with turbocharger sizing, one of the mistakes beginners make with turbo setups is focusing too much on making the header replicate an N/A style header, with long primary tubes and a merge collector with a large, 2.25" (or bigger!) pipe that merges together really far away from the turbo, and then snakes to the turbine. 

To be fair, that type of design has it's place (max effort drag racing and land speed, mainly) but on the street ridden bikes you and I build, we should be willing to give up just a hair in peak power if it means we can get the party started another 500, and in extreme cases, 1000RPM or more earlier. Allow me to explain.

When I had the big T28 turbo on the Sportster, there were a bunch of things that I thought were holding it back, one of them being the "log" style merge that I was using. Fast forward to a few revisions later, I had the spare time and willingness to cut the manifold apart and make a proper 2-1 merge collector with equal length primaries to test, with the only changes being primary tube length and merge design. "This is it, it's going to be a freaking beast now" I thought after all that work of re-doing the manifold.

Seat of the pants, I couldn't tell a difference. If it did do anything to peak power, it didn't make enough of a difference to show on timed 40-80MPH tests (which I'd done a zillion of before and after). What it did do is make the bike even more laggy, especially around town where the turbo isn't spinning all that fast out of boost. If you jumped on it in 2nd gear at 2,500rpm, it wouldn't hit full boost until 5,000rpm. Terrible! It wasn't even fun to ride, it had no response whatsoever. By the time the turbo kicked in, the torque was already laying over.

Looking back, I can laugh at how big I made the primary tubing and merge size. The primary tube diameter was 1 7/8" I.D., with the tubes merging into a 2.25" collector (that I almost made 2.5"!). Mate that with a turbo that belongs on a 1,800cc engine, you just get a bike that's kind of lame to ride. I had all of these theories about why I should be running a big turbo and big piping on the bike, but when faced with evidence that was directly contradictory to my beliefs.....I started to reconsider. Why am I so committed to using such large piping?

Pictured above is the Version 2.0 Setup that I re-designed for the Sportster, where I stepped down from a T28 turbo to a GT15 turbo, and went from big 'ole 1 7/8-2.25" collector piping right down to 1 5/8" I.D. piping merging directly into the turbine housing, making the primary tubes only as long as they need to be to get to the turbo, and merging the pipes right on the face of the turbine housing. Old me would've found this sacrilegious! Inch and five eighth's piping? Surely you meant to put this on a Honda Grom!? GT15? That's a little baby turbo! wHaT aBoUt tHe bAcKprEsSuRe rAtIo?

The new turbo setup made more power EVERYWHERE, even past 6,000RPM, which intuitively doesn't make sense. The bike absolutely came alive with the re-designed setup in a way the old setup couldn't hold a candle to, at any point in the rev-range! Not only that, it comes alive right away! 2nd gear roll, go wide open, BOOM, full boost, bike takes off like a scalded cat, the way it should be. Remember the 40-80MPH time I was talking about? Both setups on 9psi, the GT15 was consistently .7-.8 seconds faster on those runs, which is a massive difference. When the old setup was at 73mph, the new GT15 had already hit 80mph.

The question is not how big can you go, the question should be "what's the smallest I can get away with?"

Sizing Suggestion - Make the primary tube diameter no larger than the exhaust port. I'm going to experiment with going slightly smaller and see what happens, but for now, keeping the size the same is a good rule of thumb. If space dictates you have to run a less than ideal merge collector, just do your best to point the pulses at the turbine the best you possibly can.

Much like Carburetor Sizing, turbo sizing is yet another area that many people often over-complicate or overthink. Selecting a turbo that's too big makes for a laggy, boring-to-ride bike that feels slower than it should, and will have you questioning whether or not it's even adding power when on low boost. Here's why.

When you look at a Compressor map for a Turbocharger, it's graphing what's called adiabatic efficiency. It's the measure of the work the compressor can perform that is not lost as heat. So the more efficiently sized the compressor is for the application, the less power you throw away due to the increased heat from the less efficient compressor size. Make sense? 

You'll notice percentage's are plotted on a range next to the "efficiency islands", and that the really tiny island in the middle says 72%. This turbo would be 72% efficient flowing 12.5lb/minute of airflow (125 crank HP) at a pressure ratio of 1.75 (11psi of boost at sea level). So this turbo would be perfectly sized for someone with a 60-80HP N/A bike. It could also support an absolute max of 200 crank HP. How do we figure this?

As a general rule, 7.5psi of boost will increase HP by 50%. We can use the factory rated HP multiplied as .1 as a proxy for the corrected airflow in lb/minute for our stock machine. So if your motorcycle has 70hp N/A, 70*.1 = 7 lb/minute of airflow stock. If we take that 7lb/minute airflow and multiply it by the 50% power increase we'll get when it's turbocharged, we get a 3.5lb/minute increase in airflow for a total of 10.5lb/minute airflow. If we plot 7psi of boost at 10.5lb/minute on the compressor map, we can see we're right at 70% efficiency, which means this compressor is pretty much perfect.

But wait a second, the compressor map doesn't plot boost, it plots pressure ratio. What is that?

Pressure ratio sounds complicated, but the formula is simple. You take the boost pressure reading that you'd see on the gauge (how much boost you're running), add it to ambient air pressure (14.7:1 at sea level) and then divide by ambient air pressure. So 14.7 + 7.5 PSI of boost divided by 14.7psi = 1.51 Pressure Ratio. 

Your goal when selecting a turbocharger is to stay as close to right-in-the-middle of the compressor map as you possibly can. If it ends up being closer to the right side, the turbo is too small. If it ends up being too far to the left side of the map, it's going to be too big, and that's something you want to avoid at all costs, unless you're running a very high RPM 4 valve per cylinder engine, which will throw off the calculations somewhat. If you're going to mis-size the turbo, do it so that it's on the smaller end of things. A turbo that provides boost quickly and runs out of breath up top is a much better scenario than a turbo that doesn't come alive until midway through third gear, and is a total dog down low. 

Turbocharger Sizing Recommendation - Select a turbocharger that puts your engine closest to the most efficient portion of the compressor map. 

If you enjoyed this article, consider subscribing to our Patreon. We have over 11 how-to videos, and our Turbocharging How-to is a complete breakdown of turbocharging carb'd motorcycles. Draw-Thru, Blow-Thru, Pitot Tubes, Prepping the carb for boost and so much more! 

www.patreon.com/doctormotorcycle

This is a subject that in my eyes is pretty straightforward, and if you're someone who's pragmatic, you'll likely come to the same conclusions as me. If there's any one piece of advice that you heed from this - proper control of timing is absolutely paramount on a turbo application, and the only way to develop a map is by trial and error. A dialed ignition curve is the difference between night and day, especially with a draw-thru. And not just reliability either. Cold start, warm start, driveability, fuel economy, longevity; it affects everything.

The reality is, aggressive ignition timing, especially on pump gas, can blow up your bike in a single pull if the motor is tired enough. More than likely, you'll blow out the headgasket first (it's almost always the front cylinder, as it usually runs 1-1.5 AFR leaner...more on that later). 

Contrary to what some will say, detonation occurs after Top Dead Center (TDC) and after the spark plug has fired. However, detonation is combustion that's sporadic, and outside of the flame kernel. Think of how an explosion looks - it starts from the ignition source and moves outward - that's normal combustion. Detonation is combustion that's outside of the flame front. This abnormal combustion collides with the flame front, causing a spike in cylinder pressure. It's why the headgasket blows out; cylinder pressure spikes to abnormally high levels.

You'll often hear of people claiming that the "composite gaskets can't handle boost", which isn't necessarily the case. They actually can be ran up to relatively high boost levels - it's the detonation that causes them to blow out. How do you know the headgasket is blown? It's usually accompanied by a chirping sound during cold start that is very noticeable as the engine is turning over. Often times though, people don't catch it, and since the front cylinder is now sucking in air and running lean, it's a matter of time before the front piston is melted. 

So how do we avoid this? Let's first cover the way's you definitely shouldn't go about it;

Leaving the timing stock - This is one of the quickest ways to blow up a turbo bike. Don't do this. The only time you can get away with this is on low C/R, inline four's on low boost (8psi). Think GS750 with 8.7:1 compression. And even then, it will still ping a bit on hot days, a situation you want to avoid.

Here's some alternative methods - 

Retard the base timing - At minimum, you need to retard the base timing. If you have a mechanical advance, you can weld a stopper into the advance mechanism, or lock it out completely. Though this is a bandaid solution, it's better than nothing. BUY A TIMING LIGHT. You need to know for certain how much total timing you're running. All service manuals list the initial and total timing. 

Hobbs Switch With Dyna/Similar - This may surprise you, but if you are using the provided retard mode curves that are supplied with the dyna 2000i on a Harley, you're actually worse off than if you'd have just ran curve #4 and backed the base timing off 6* and not ran a hobbs switch. The retard mode curves are suitable for a high revving inline four, not a Big V-twin that makes peak torque at 4,000 RPM - right where the retard move curves have all the timing in!? You need the timing to be soft around peak torque and the area below - imagine a curve closer to curve #4 on the Dynatek, and then adding a hobbs switch on top of that which pulls 4-7 degrees at 3-4psi. That would be much closer to what would be a proper ignition setup.

EFI/ECU Timing Control with 2-3 Bar MAP sensor - This is the gold standard for turbocharging, and one of many reasons why turbo's with EFI are much more reliable; you just have so many more tools at your disposal. Let's discuss a simple and effective rule of thumb for setting up your spark tables.

Starting with a stock advance map that's currently in the bike, I'd highly advise to first ensure the existing map doesn't ping at all - if you're getting detonation while the bike is N/A, it's going to be a nightmare to solve once it's turbocharged. Detonation will show up on the spark plugs as specs of black or aluminum. The black is carbon deposits, the aluminum is generally pieces of the piston crown, and the top compression ring. Every time you have a detonation event, think of a hammer chipping away at the internal parts of your motor, as that's essentially what's happening.

Now lets say we're turbocharging a Harley with an M8 114. The stock Compression ratio is 10.5:1, with the stock HD early closing camshaft, and around 200psi of cranking compression. And lets say we're running a GT20 turbo on 8psi of boost, and no intercooler. You'd want to pull at minimum, 1.5* of timing per psi of boost at and  below peak torque, and 1* per psi of boost after peak torque. This will be a good starting point for a safe tune. Please keep in mind that the Thundermax ECU that everyone seems to use on Turbo Harley's gets rid of the knock sensors/timing retard function, so if it pings, it won't pull timing. 

On an Inline four, you can just pull 1 degree per psi of boost, and half a degree per psi after you get past the peak torque area. Inline four's are generally much more tolerant of ignition advance. I actually do pull 1* per psi on inline four's, but I run 10psi on customer builds and 12psi on straight pump 91 on my personal bikes. Just know you're playing with fire at those boost levels, especially if you're doing freeway pulls. 

What you'll find universally on air cooled bikes is that if you build an ignition map around the worst case scenario (hot day in Arizona, for example), the bike will be a turd with no power. You can only shed so much heat on an air cooled bike, and once you get to 50,75, or even 100% more power than stock, your ability to keep it from detonating on straight pump gas will be very limited, partly due to heat soak, but mainly from the fact that pump 91 in this context just kind of sucks. When the dynamic Compression Ratio (Static compression + intake close of the cam + elevation + boost level) gets over the mid 11's, you'll struggle to keep the bike from detonating once it's heatsoaked. You can get around this by running excessively rich Air Fuel Ratio's, though that's a topic in and of itself. But briefly - Many OEM's, the Mitsubishi Lancer Evolution for example, have a 9:1 AFR target at high boost from the factory. In fact, many stock turbo cars have super rich AFR targets at WOT. It's cheap insurance. Just know power drops off rapidly below about mid 10's on pump gas...if you can keep the spark plugs firing. Often times you'll just blow the spark out, bike will feel like it's hitting a rev limiter. 

Quick Tangent - Run 1 heat range colder spark plugs on a turbo build and close up the gap to .028". On my GS400 EFI turbo, on 15psi of boost, it would blow the spark out above .025" gap....and that's with LS coils. On 20psi it needed a .022" gap. On my Twin Turbo Virago, I had to close the gap up to .018"!

If you enjoyed this writeup, please consider joining the Patreon. We have over 13 how-to videos over there, Including - 

Comprehensive Turbocharging Video - Complete breakdown, Draw-thru, Blow-thru (pressurizing carbs), vacuum routing diagrams, turbo sizing etc. Almost an hour long, complete with draw-thru specific tuning on an inline four 

Complete EFI conversion walk-through - Motorcycle specific EFI conversion with a Microsquirt ECU - If you're taking the EFI plunge, this video will be extremely helpful! Full setup including Crank trigger for full Fuel and Spark Control

A dozen more videos ranging from cable modifications to how to make a seat, and more. Regardless if you join or not - thank you for reading and checking out my work, I appreciate it. Check back periodically for more tech tips in the future!

www.patreon.com/doctormotorcycle