As always, well thought out and implemented. Pure jewelry.
Thanks for posting.
Regards Brian
Thanks for posting.
Regards Brian
S2's Build Thread
- Thread starter sswartz
- Start date
Scott
Lifetime Supporter
Neil, I don't understand your point or why you placed motorsport in quotes. The connector in the photograph is manufactured by Deutsch. It's one of their AS (autosport) Series products. Motorsport and autosport are interchangeable marketing terms with respect to describing a connector (i.e., they are minimally more compact and lighter than a mil-spec connector).
I use Deutsch connectors wherever possible. AS Series for critical connections, bulkhead connections or where I need a high pin count and the DT series for everything else.
Neil
Supporter
The quotation marks were because it was quoting from your post. I was not aware of the Deutsch AS series connectors. I have only used their aerospace/MIL connectors in the past. In my opinion, Deutsch makes the best connector in the industry.
BTW, that wire marker color code is the same as for the old carbon composition resistors. Those markers are far superior to the stick-on wrap-around.
Scott
Lifetime Supporter
ENGINE HARNESS: P1
I’ve retained BM-Technik Race Engineering (BMR) out of Germany to design and fabricate the harnesses for my car. They’re a top-tier firm focused on race teams, OEM hyper cars (I’ve seen a few of the projects, but they’re under NDA), vintage F1, etc. and a friend talked them into taking my project on.
The harness is a large project because I have a lot of electronics and, to keep everything as clean as possible, I’m subsuming all of the commercial harnesses (e.g., engine, climate, EPAS, brake booster, active dampers, active wing, parking brakes and exhaust cutouts) with the following bespoke ones:
The design effort was estimated at 300 hours which doesn’t include programming or labor to build the harnesses. To put that in perspective, Allan spent 1,200 hours on his most recent SL-C build which includes paint. Given the level of expertise and effort to design bespoke harness the handful of companies that can this are all bottle necked at the design phase so it’s not easy to get a slot. Stefan, who I introduced in a previous post, spent two weeks with me in Boston figuring out how to layout the harness, mounting the devices and taking careful measurements. He’s also the Systems Engineer for the Inter EuroPol team, which just won the LMP2 class at Le Mans for the second year in a row, so he’s constantly on the road and only able to work on my project between races. The best guys are always oversubscribed and it’s been a privilege to work with Stefan and the rest of team.
Engine harness PDF export from in Zuken E3. Each harness will have its own design sheet.
Stefan used Zuken E3 to design the harness. Apparently, it’s a high-end CAD package which runs about $30k/user/year and is used for everything up to nuclear subs. Fortunately, they provide PDF and Excel exports as well as a free read-only, viewer application to us mere mortals. The application enables you to interact with the design: trace wires, list all of the detailed part numbers, etc. While the chassis harness is the largest, the first harness we focused on was for the engine which connects to the chassis harness via a 128-pin autosport connector.
In some cases it makes sense to not fully populate a layer so that wires can be moved to a subsequent layer. In this situation, non-conducting, white filler wires are added to achieve the perfect twist ratio and provide a smooth base for the next layer. Filler wires contain no copper so they are more flexible and weigh less than standard wires of the same gauge.
Engine harness cross section. The actual layers and filler wires are shown in another view.
Core: The bulky pre-twisted cables were located in the center of the harness to keep it as flexible as possible. There are four composite cables; two shielded twisted pairs and two shielded triple pairs (i.e., a total of eight twisted pairs). To make the core as round as possible, eight standard wires were added before starting the first layer.
Layer 2: With a contra-helical harness each successive layer is twisted opposite to the direction of the previous layer. This maximizes flexibility and minimizes cross section. Note that the filler wire hasn’t been added yet so the outer layer has gaps.
Layer 3: The layer is approximately half way done. Note the identification markers that were discussed in the last blog post. They are the small colored pieces on the wires near the connector.
The filler wire has been added and the layer has been tightly wrapped with a Kevlar/Aramaid lacing cord to keep everything in place. The lacing cord is the single, flat, white thread that is wrapped in the opposite direction to the wires in the picture above.
Filler wire has not been added to the outer layer yet. Service loops were not added to reduce cost and the size of the boot. The mating 128-pin connector and the leaf connectors will have service loops in the advent that something needs to be modified.
A transition
Multiple vices are used to maintain the twist on the transition legs until the lacing cord is applied. The filler wire hasn’t been added to the outer layer yet.
I’ve retained BM-Technik Race Engineering (BMR) out of Germany to design and fabricate the harnesses for my car. They’re a top-tier firm focused on race teams, OEM hyper cars (I’ve seen a few of the projects, but they’re under NDA), vintage F1, etc. and a friend talked them into taking my project on.
The harness is a large project because I have a lot of electronics and, to keep everything as clean as possible, I’m subsuming all of the commercial harnesses (e.g., engine, climate, EPAS, brake booster, active dampers, active wing, parking brakes and exhaust cutouts) with the following bespoke ones:
- Chassis
- Engine
- Front
- Left-Front Hub
- Right-Front Hub
- Nose
- Rear
- Engine Cover
- Left Door
- Right Door
The design effort was estimated at 300 hours which doesn’t include programming or labor to build the harnesses. To put that in perspective, Allan spent 1,200 hours on his most recent SL-C build which includes paint. Given the level of expertise and effort to design bespoke harness the handful of companies that can this are all bottle necked at the design phase so it’s not easy to get a slot. Stefan, who I introduced in a previous post, spent two weeks with me in Boston figuring out how to layout the harness, mounting the devices and taking careful measurements. He’s also the Systems Engineer for the Inter EuroPol team, which just won the LMP2 class at Le Mans for the second year in a row, so he’s constantly on the road and only able to work on my project between races. The best guys are always oversubscribed and it’s been a privilege to work with Stefan and the rest of team.
Engine harness PDF export from in Zuken E3. Each harness will have its own design sheet.
Stefan used Zuken E3 to design the harness. Apparently, it’s a high-end CAD package which runs about $30k/user/year and is used for everything up to nuclear subs. Fortunately, they provide PDF and Excel exports as well as a free read-only, viewer application to us mere mortals. The application enables you to interact with the design: trace wires, list all of the detailed part numbers, etc. While the chassis harness is the largest, the first harness we focused on was for the engine which connects to the chassis harness via a 128-pin autosport connector.
Contra-Helical Twisting
Motorsport harnesses utilize contra-helical twisting to maximize flexibility, reduce diameter and mitigate strain on the wires. With this approach, each successive layer of wires is twisted in the opposite direction of the previous layer. In a typical harness, the core has seven wires, with each successive layer adding seven additional wires (i.e., core=7, L1=14, L2=21, L3=28, etc.). However, the actual number of wires in a layer may vary to accommodate pre-twisted pairs or wires of different gauges. Optimizing which wires go in which layer requires a lot planning. Bulky cables (i.e., pre-twisted shielded pairs) are typically placed in the center to increase flexibility and the other layers are driven by the location of branches/transitions and the devices. Obviously, you’d rather not dig through multiple layers of wires that are continuing straight to pull out wires that are going into a transition to head in another direction. You’d prefer to have them on the outer layer. If you think of the harness as a tree, ideally the wires in the first branch in the outermost layer and the wires in last branch to be on the inner-most layer. However, you might have a wire that uses a splice to service both the first and last branch which complicates things.In some cases it makes sense to not fully populate a layer so that wires can be moved to a subsequent layer. In this situation, non-conducting, white filler wires are added to achieve the perfect twist ratio and provide a smooth base for the next layer. Filler wires contain no copper so they are more flexible and weigh less than standard wires of the same gauge.
Engine harness cross section. The actual layers and filler wires are shown in another view.
Engine Harness Construction
While Stefan is having fun at the races, his younger brother Gustav has been working on the engine harness and he sent me the following pictures.Core: The bulky pre-twisted cables were located in the center of the harness to keep it as flexible as possible. There are four composite cables; two shielded twisted pairs and two shielded triple pairs (i.e., a total of eight twisted pairs). To make the core as round as possible, eight standard wires were added before starting the first layer.
Layer 2: With a contra-helical harness each successive layer is twisted opposite to the direction of the previous layer. This maximizes flexibility and minimizes cross section. Note that the filler wire hasn’t been added yet so the outer layer has gaps.
The filler wire has been added and the layer has been tightly wrapped with a Kevlar/Aramaid lacing cord to keep everything in place. The lacing cord is the single, flat, white thread that is wrapped in the opposite direction to the wires in the picture above.
Filler wire has not been added to the outer layer yet. Service loops were not added to reduce cost and the size of the boot. The mating 128-pin connector and the leaf connectors will have service loops in the advent that something needs to be modified.
A transition
Multiple vices are used to maintain the twist on the transition legs until the lacing cord is applied. The filler wire hasn’t been added to the outer layer yet.
Attachments
Scott
Lifetime Supporter
ENGINE HARNESS: P2
Continued from former post...
Crimping tools must be properly adjusted to achieve a proper crimp. The best way to achieve this is to crimp a sacrificial pin to a piece of scrap wire and validate it with a wire crimp pull tester. The pin is placed in an appropriate slot in the rotating turret, the wire is placed in the pulling mechanism and the operator slowly and evenly pulls on the lever until the crimp gives up or the wire snaps. The maximum applied force is displayed digitally. The motorsport spec is 9.8kg (21.6 lbs.). If the desired strength isn’t achieved, the crimping tool/die is adjusted and the process is repeated. For this reason, you want multiple spare pins so that the operator can dial in the settings. For example, my fuel injectors don’t utilize motorsport connectors, but after a couple of attempts Gustav was able to achieve 9.4kg (2% under the motorsport spec).
Raychem DR25 heat shrink tubing has been applied. It’s surprising that 180+ conductors, filler wires, multiple shields and the jacketing fit within a 19 mm (0.748”) diameter and remain flexible.
90-degree transition completed and yellow labels applied
Continued from former post...
Crimping tools must be properly adjusted to achieve a proper crimp. The best way to achieve this is to crimp a sacrificial pin to a piece of scrap wire and validate it with a wire crimp pull tester. The pin is placed in an appropriate slot in the rotating turret, the wire is placed in the pulling mechanism and the operator slowly and evenly pulls on the lever until the crimp gives up or the wire snaps. The maximum applied force is displayed digitally. The motorsport spec is 9.8kg (21.6 lbs.). If the desired strength isn’t achieved, the crimping tool/die is adjusted and the process is repeated. For this reason, you want multiple spare pins so that the operator can dial in the settings. For example, my fuel injectors don’t utilize motorsport connectors, but after a couple of attempts Gustav was able to achieve 9.4kg (2% under the motorsport spec).
Raychem DR25 heat shrink tubing has been applied. It’s surprising that 180+ conductors, filler wires, multiple shields and the jacketing fit within a 19 mm (0.748”) diameter and remain flexible.
90-degree transition completed and yellow labels applied
Next Steps
The engine harness has been completed and pin-to-pin tested. Stefan is finalizing the design of the chassis and nose harnesses.
Great job!
Scott
Lifetime Supporter
One of the challenges with building a custom car is that when you change one thing it often forces other changes, some of which you might not realize for years. When I decided to go with a fully-built supercharged engine I knew that I’d have to modify the rear hoop, modify the firewall and fabricate a custom induction tube. A year and half after fabricating the induction tube I realized that the throttle body was going to block the fuel fill tube. After some scrambling, I discovered that then new LT5 throttle body was markedly smaller than the LS7 throttle body despite having a larger throat. So the induction tube went in the scrap pile, I designed and CNC’d two custom weld flanges, and fabricated new induction tube. I also had Solar Performance port and make a bunch of performance modes to the throttle body. Problem solved, right? Not so fast.
LS7 throttle body (top) and stock LT5 throttle body (bottom)
Final version of the induction tube
When designing the engine harness we discovered that the LT5 throttle body uses the SENT protocol. SENT (Single Edge Nibble Transmission) is a high-resolution, unidirectional, single-wire serial communication protocol used primarily to transmit automotive sensor data (e.g., mass airflow and throttle position). The harness guys in Germany, my two MoTeC experts and my engine tuner had never heard of it…
If I were using a GM ECU, I would have been completely unaware of which protocol was being used because it would have just worked out of the box.
Fortunately, my engine tuner, Ed Senf, dug into the MoTeC M150 documentation and found that SENT was supported. Since documentation doesn’t necessarily equal reality, he purchased a stock throttle body and bench tested it. As can be seen in the video below, the test was successful. Ed really likes the throttle body and will use it on future projects because it saves two precious analog voltage input resources (main and tracking) and uses a single universal digital (UDIG) input that often goes unused in many applications.
LS7 throttle body (top) and stock LT5 throttle body (bottom)
Final version of the induction tube
When designing the engine harness we discovered that the LT5 throttle body uses the SENT protocol. SENT (Single Edge Nibble Transmission) is a high-resolution, unidirectional, single-wire serial communication protocol used primarily to transmit automotive sensor data (e.g., mass airflow and throttle position). The harness guys in Germany, my two MoTeC experts and my engine tuner had never heard of it…
If I were using a GM ECU, I would have been completely unaware of which protocol was being used because it would have just worked out of the box.
Fortunately, my engine tuner, Ed Senf, dug into the MoTeC M150 documentation and found that SENT was supported. Since documentation doesn’t necessarily equal reality, he purchased a stock throttle body and bench tested it. As can be seen in the video below, the test was successful. Ed really likes the throttle body and will use it on future projects because it saves two precious analog voltage input resources (main and tracking) and uses a single universal digital (UDIG) input that often goes unused in many applications.
Ken Roberts
Supporter
I'm pretty sure all the LT1/LT4/LT5 engines use this digital protocol (2014-up).
Last edited:
Scott
Lifetime Supporter
Stefan, Ed, Abe and I have been busy installing the harnesses and working to get all of the devices working. I have a bunch of posts to catch up on, but I reached a major milestone so this one will be out of sequence.
We had a successful first start! My dog took off in a panic because she’s been watching me work on the car her whole life, she’s seen the shirt and apparently doesn’t trust that everything went together properly… smart dog. That said, she prefers the exhaust sound to Stefan’s German techno music.
The first start sounded awful because two of the cylinders weren’t firing. The fix was simple. There is a dedicated PDM output for every two cylinders and one of the outputs wasn’t configured properly. The second start sounds much better. With a couple of tweaks we were able to get it to idle at 750 RPM and not loud. In the video below, note that the sound is smoother when the exhaust cutouts are closed. They are configured to automatically open when the throttle position is above 60%
The stepped, equal-length, 180-degree crossover headers combined with the titanium X-pipe result in the sound that I was looking to achieve… absolutely nothing like a LS. Stefan thinks it sounds like a lower-reving BMW LMDH (a flat-plane V8 hyper car).
The next big milestone is to get it on the dyno.
We had a successful first start! My dog took off in a panic because she’s been watching me work on the car her whole life, she’s seen the shirt and apparently doesn’t trust that everything went together properly… smart dog. That said, she prefers the exhaust sound to Stefan’s German techno music.
The first start sounded awful because two of the cylinders weren’t firing. The fix was simple. There is a dedicated PDM output for every two cylinders and one of the outputs wasn’t configured properly. The second start sounds much better. With a couple of tweaks we were able to get it to idle at 750 RPM and not loud. In the video below, note that the sound is smoother when the exhaust cutouts are closed. They are configured to automatically open when the throttle position is above 60%
The stepped, equal-length, 180-degree crossover headers combined with the titanium X-pipe result in the sound that I was looking to achieve… absolutely nothing like a LS. Stefan thinks it sounds like a lower-reving BMW LMDH (a flat-plane V8 hyper car).
The next big milestone is to get it on the dyno.
That’s awesome Scott. A long time coming.
Congrats, on a job well done.
Regards Brian
Congrats, on a job well done.
Regards Brian
Scott
Lifetime Supporter
The chassis, engine and front harnesses arrived from Germany and there’s a lot to connect. According to the Bill of Materials (BoM) there are 115 connectors, 1,489 meters of wire and well over 1,000 pins.
These numbers don’t include the rear, nose clip, engine cover, climate, steering column or door harnesses. Keep in mind that some of the wire length is due to the fact that small gauge motorsport/aerospace Tefzel wires are used in lieu of larger gauge commodity wires because they match the capacity of the autosport connector pins, facilitate contra-helical twisting and improve harness flexibility. For example, a 20-amp circuit would be implement with four 22-gauge wires. So a motorsport harness will have more wires, but smaller diameter, lower weight and more flexibility that a functionally equalivent OEM harness.
Since all of the oil, fuel, cooling, etc. lines were already installed some thought was put into the best way to install the harness. The first step was to spread the 1:1 drawing on the bench, lay the harness on top to match and temporarily zip tie certain sections together to make it easier to feed those sections though tight spaces.
Laying out the engine harness on its 1:1 3’x 5’ plan
The bulkhead connectors were mounted to 0.60” laser-cut and CNC-bent stainless steel brackets. Depending on the shell size, they use M2, M2.5 or M3 x 6 mm long screws which are fiddly — easy to drop and hard to find. Fortunately, there are specifically designed nut plates so there is no need to deal with tiny nuts. They have distorted threads to prevent the screws from backing out and when bonded to the backside of the bulkhead flanges the screws are much easier to install.
A single 128-pin connector connects the engine harness
Stefan did a great job measuring and designing the engine harness. The lengths were spot on and the wires pretty much disappear after the big connector.
Several of the sensors are difficult or impossible to access without pulling the engine. To improve reliability, all engine sensors were potted, extended with Tefzel wire / Raychem DR-25 heat shrink and terminated with motorsport micro connectors. The micro connectors are compact as can be seen in the picture below.
The autosport micro connector in the lower left corner is a #3 shell which can have up to six pins. The pins, nut plate (note that the connector isn’t a bulkhead) and M2 screws are for a #3 shell. F1 teams use even smaller #1 shells for low-current applications. For size comparison, the top left connector is a 2-pin DT series and the connector between it and the penny is a 6-pin DTM series.
All of the engine sensor connectors were mounted to stainless steel plates. I 3D printed breakout boxes in other locations. but simple plates worked best when mounted to the engine.
The dual lambda-to-CAN device was modified with a bulkhead autosport connector and mounted to a vibration isolation bracket
Crank sensor mounted to the valve cover
These numbers don’t include the rear, nose clip, engine cover, climate, steering column or door harnesses. Keep in mind that some of the wire length is due to the fact that small gauge motorsport/aerospace Tefzel wires are used in lieu of larger gauge commodity wires because they match the capacity of the autosport connector pins, facilitate contra-helical twisting and improve harness flexibility. For example, a 20-amp circuit would be implement with four 22-gauge wires. So a motorsport harness will have more wires, but smaller diameter, lower weight and more flexibility that a functionally equalivent OEM harness.
Since all of the oil, fuel, cooling, etc. lines were already installed some thought was put into the best way to install the harness. The first step was to spread the 1:1 drawing on the bench, lay the harness on top to match and temporarily zip tie certain sections together to make it easier to feed those sections though tight spaces.
Laying out the engine harness on its 1:1 3’x 5’ plan
The bulkhead connectors were mounted to 0.60” laser-cut and CNC-bent stainless steel brackets. Depending on the shell size, they use M2, M2.5 or M3 x 6 mm long screws which are fiddly — easy to drop and hard to find. Fortunately, there are specifically designed nut plates so there is no need to deal with tiny nuts. They have distorted threads to prevent the screws from backing out and when bonded to the backside of the bulkhead flanges the screws are much easier to install.
A single 128-pin connector connects the engine harness
Stefan did a great job measuring and designing the engine harness. The lengths were spot on and the wires pretty much disappear after the big connector.
Several of the sensors are difficult or impossible to access without pulling the engine. To improve reliability, all engine sensors were potted, extended with Tefzel wire / Raychem DR-25 heat shrink and terminated with motorsport micro connectors. The micro connectors are compact as can be seen in the picture below.
The autosport micro connector in the lower left corner is a #3 shell which can have up to six pins. The pins, nut plate (note that the connector isn’t a bulkhead) and M2 screws are for a #3 shell. F1 teams use even smaller #1 shells for low-current applications. For size comparison, the top left connector is a 2-pin DT series and the connector between it and the penny is a 6-pin DTM series.
All of the engine sensor connectors were mounted to stainless steel plates. I 3D printed breakout boxes in other locations. but simple plates worked best when mounted to the engine.
The dual lambda-to-CAN device was modified with a bulkhead autosport connector and mounted to a vibration isolation bracket
Crank sensor mounted to the valve cover
Wow, what craftsmanship going into this build. I always look forward to your updates. You set the bar very very high.
Congrats !!!!!
Regards Brian
Congrats !!!!!
Regards Brian
Scott
Lifetime Supporter
The engine was on a dyno when it was built, but since then I’ve completely changed the front dress to accommodate the A/C compressor and the electric water pump. I also wanted to reduce belt slip on the supercharger pulley. The only way to know if it was going to work was to tune the car on a chassis dyno… see the video at the end.
So I towed the car up to Cohesion Motorsports who shares a facility with Kachel Motor Company (KMC). They have an inground SuperFlow AutoDyn dyno which features a pit — very useful so long as you don’t fall in. Nick and John from Cohesion (who’ve helped with some of the electrical system) prepped the car and Ed Senf flew in from Atlanta to tune the car.
The safety bar in the front is to prevent someone like me from falling into the pit
During one of the early pulls, we lost all boost because the supercharger belt popped off. A quick inspection revealed that the tensioner had failed. This was a surprise, because with the optional ceramic bearing upgrade, it runs about a grand. Even with the body off, servicing the front dress on a mid-engine car is a nightmare. After removing the induction tube, machining one of Nick’s SnapOn sockets to fit the allotted space and a fair amount of profanity, we were able to remove the tensioner. It has a massive spring which slipped out of slot in the billet housing. Our best guess is that the slot was created to make it easier to install the powerful spring, but that also makes it easier for the spring to pop out. Looking at pictures on their website it appears that the slot doesn’t exist anymore. The manufacturer is sending a new one without even asking for proof of purchase, which seems to support our notion that they changed the design.
We were able to hack together a temporary fix for the tensioner and get the belt back on. A couple of small pulls later and the belt popped off again. Repeated the process again with the same outcome. After pulling the tensioner again we noticed that the M10 bolt had a slight bend in it. It’s a really long bolt that’s supported by my brackets, so we’re hoping that the issue was primarily caused by the failed tensioner and the not my design or fabrication of the brackets. I’m going to upgrade the bolt, with an ARP 12.9-Grade stud, a precision ground washer and a jet nut.
If that doesn’t work, I’ll need to pull the engine and redesign the serpentine system. The only changes that I can think of are to:
The stock supercharger serpentine had 8 ribs and drove the water pump as well. The new serpentine only drives the supercharger, has 10 ribs, more wrap and a better tensioner (in theory).
The brackets should be sturdy enough
The good news is that nothing leaked and, other than the tensioner, nothing broke. The better news is that the equal-length 180-degree cross-over headers and X-pipe sounds great… a bit like the BMW LMDH. The muffler/catalytic converter bypass is configured to open when the throttle position is exceeds 80% at which point things get really loud. The predominate whine is from the straight-cut gears as opposed to the supercharger. Upshifts were awesome, but downshifts were terrible because we weren’t achieving good blips. Looking at the data, Ed believes this was due to the issue with tensioner which resulted in the supercharger restricting rather than increasing air flow.
Here’s the video…
So I towed the car up to Cohesion Motorsports who shares a facility with Kachel Motor Company (KMC). They have an inground SuperFlow AutoDyn dyno which features a pit — very useful so long as you don’t fall in. Nick and John from Cohesion (who’ve helped with some of the electrical system) prepped the car and Ed Senf flew in from Atlanta to tune the car.
The safety bar in the front is to prevent someone like me from falling into the pit
During one of the early pulls, we lost all boost because the supercharger belt popped off. A quick inspection revealed that the tensioner had failed. This was a surprise, because with the optional ceramic bearing upgrade, it runs about a grand. Even with the body off, servicing the front dress on a mid-engine car is a nightmare. After removing the induction tube, machining one of Nick’s SnapOn sockets to fit the allotted space and a fair amount of profanity, we were able to remove the tensioner. It has a massive spring which slipped out of slot in the billet housing. Our best guess is that the slot was created to make it easier to install the powerful spring, but that also makes it easier for the spring to pop out. Looking at pictures on their website it appears that the slot doesn’t exist anymore. The manufacturer is sending a new one without even asking for proof of purchase, which seems to support our notion that they changed the design.
We were able to hack together a temporary fix for the tensioner and get the belt back on. A couple of small pulls later and the belt popped off again. Repeated the process again with the same outcome. After pulling the tensioner again we noticed that the M10 bolt had a slight bend in it. It’s a really long bolt that’s supported by my brackets, so we’re hoping that the issue was primarily caused by the failed tensioner and the not my design or fabrication of the brackets. I’m going to upgrade the bolt, with an ARP 12.9-Grade stud, a precision ground washer and a jet nut.
If that doesn’t work, I’ll need to pull the engine and redesign the serpentine system. The only changes that I can think of are to:
- Replace the automatic tensioner with a manual one.
- Reduce the diameter of the super damper and the supercharger pulley. This will reduce the speed of the pullies, but it will increase the potential that the supercharger pulley slips. Both of the pullies are custom and ATI has a 120-day lead time on custom super dampers.
- The intake air was hotter than expected and intercooler coolant temp was too low. We ran the pump and noticed it was only drawing about a quarter of expected current indicating that we had too much air in the system. We cracked the bleeders on the heat exchangers, ran the pump and bled a bunch of air. Problem solved.
- We maxed the MAP sensor’s range. I had installed a 2-bar sensor because I’m targeting about 16 psi of boost. Do’h…. I forgot to account for ambient pressure. I’ll replace it with a 2.5 or 3-bar sensor.
The stock supercharger serpentine had 8 ribs and drove the water pump as well. The new serpentine only drives the supercharger, has 10 ribs, more wrap and a better tensioner (in theory).
The brackets should be sturdy enough
The good news is that nothing leaked and, other than the tensioner, nothing broke. The better news is that the equal-length 180-degree cross-over headers and X-pipe sounds great… a bit like the BMW LMDH. The muffler/catalytic converter bypass is configured to open when the throttle position is exceeds 80% at which point things get really loud. The predominate whine is from the straight-cut gears as opposed to the supercharger. Upshifts were awesome, but downshifts were terrible because we weren’t achieving good blips. Looking at the data, Ed believes this was due to the issue with tensioner which resulted in the supercharger restricting rather than increasing air flow.
Here’s the video…
Similar threads
