Part 2 thread of the work on our hombuild project of a turbocharged KTM 500.
You can find part 1 tread in this link https://www.snowest.com/forum/showthread.php?t=325741
This has been a long project and since 2013 we´re learning more every year.
So far, turbocharging has worked quite well and with some final improvements, we think will have something works extremely well and with good reliability.
The last post on the previous thread was of the conrod failure.
The engine is rebuilt and up and running with a stock conrod installed. The boost is turned down to 0,5 bar that hopefully will work in the short term. Our conclutions is that a stock rod is not suitable.
We will order stronger custom made rods this summer. The 2016 bike will be replaced with a 2019 bike that all our future builds will be based on the 2017 engine family. By this me can order several custom rods that will fit 2017 and newer engines. Probably better price if you order a bunch of rods.
Still some issues of the turbosetup on the 2017 bike to iron out.
Have had since the beginning of the build on this perticular bike some issues.
The previous 2016 turbo bike runs perfectly but the 2017 seems to struggle with basically the same setup. The 2017 KTM 500 is a completely reworked engine that clearly makes this turbo build work sligtly differenly.
It runs perfectly on full throttle but in midrange, especially when throttling up or down, it boggs or completely dies. This makes it hard to ride despite lots of power at full throttle.
Have tried to troubleshoot this issue by replacing various components to find the errors without sucsess. The conclusion ended up to that there was something completely wrong with the fuel map.
Connected an A/F guage and a logging unit to get some figures whats happening. Surprsingly was that the fuel map was spot on but the error turned out to be a misfault in reading correct crank rpm. The piggyback efi-unit rpm input is connected to the grey wire on the on the oem fuel injector. This inuput signal turned out to be really bad due to the fact that the throttle enrichment mode not seems to be modulated with duty cycle but instead the number of pulses. The readout log shows that when advancing throtte on midrange, it suddenly jumps from 6.000 to 15.000 rpm within a second which we all know is completley wrong. By this wrong input, the piggyback injector fires wrongly and floods the engine of fuel. That explains the missfiring of the engine when you slightly twist the throttle under load.
The most resonable solution would be to get a better prm signal in order to get the piggyback efi to calculate the correct fuel not matter of boost or rpm.
The problem on most dirtbikes is that they doesn’t have any revcounter or similar output signal.
Tried to connect to an other wire to count crank revolution . With an occiloscope connected, you can monitor the quality of the suitable input pulse. The crank sensor showed a nice sinusoidal wave which unfortunately is not a good signal for a rpm signal to the piggyback efi box. It has to be a square wave for it to work. There are probably some aftermarket filters avaliable than can convert a sinusoidal waves to square but we didn´t choose to go that path.
Tested the wire to the volt regulator. This left a good square wave but due to the number of coils in the stator, it indicated 16,000 pulses already at idle which makes it impossible to map up to 10,000 rpm. I also tried to connect input signal to the ignition coil but it did not deliver a quality pule for the piggyback efi-box.
The easiest and best solution turned out be replacing the current piggyback efi unit to a new box that instead moderates the duty cycle of the stock injector in the throtthle body. The additional fuel injector will be ditched. It will most likely not be able to deliver enough fuel with the oem fuel injector. Duty cycle will probalby be more than 100% at moderate boost levels. In that case I will go for a larger injector. As far as I know, no larger bolt on KTM injector is avaliable but with some modifications, some other suitable after market injector will fit.
This looks to be a better solution when dealing with wierd rpm input signals. We will be able to get better control of fueling and also a simpler build. Less parts simplifies the build and also deliver fuel more accurately.
Thanks to our moderate boost levels, a large single injector will likely have enough flow rate to deliver fuel to a 90+ hp engine.
100+ hp 1 bar boost will likely require an extra injector but also some heavy engine modifications such a lowcomp piston and probably more...
Our goal has since the beginning been a reliable and powerful everyday turbo snowbike.
Have put the rear wheel on, ready for some dyno tests.
Replaced the rear shock with a rigid strut. No suspension is allowed as the turbocharger is very close to the wheel.
You can find part 1 tread in this link https://www.snowest.com/forum/showthread.php?t=325741
This has been a long project and since 2013 we´re learning more every year.
So far, turbocharging has worked quite well and with some final improvements, we think will have something works extremely well and with good reliability.
The last post on the previous thread was of the conrod failure.
The engine is rebuilt and up and running with a stock conrod installed. The boost is turned down to 0,5 bar that hopefully will work in the short term. Our conclutions is that a stock rod is not suitable.
We will order stronger custom made rods this summer. The 2016 bike will be replaced with a 2019 bike that all our future builds will be based on the 2017 engine family. By this me can order several custom rods that will fit 2017 and newer engines. Probably better price if you order a bunch of rods.
Still some issues of the turbosetup on the 2017 bike to iron out.
Have had since the beginning of the build on this perticular bike some issues.
The previous 2016 turbo bike runs perfectly but the 2017 seems to struggle with basically the same setup. The 2017 KTM 500 is a completely reworked engine that clearly makes this turbo build work sligtly differenly.
It runs perfectly on full throttle but in midrange, especially when throttling up or down, it boggs or completely dies. This makes it hard to ride despite lots of power at full throttle.
Have tried to troubleshoot this issue by replacing various components to find the errors without sucsess. The conclusion ended up to that there was something completely wrong with the fuel map.
Connected an A/F guage and a logging unit to get some figures whats happening. Surprsingly was that the fuel map was spot on but the error turned out to be a misfault in reading correct crank rpm. The piggyback efi-unit rpm input is connected to the grey wire on the on the oem fuel injector. This inuput signal turned out to be really bad due to the fact that the throttle enrichment mode not seems to be modulated with duty cycle but instead the number of pulses. The readout log shows that when advancing throtte on midrange, it suddenly jumps from 6.000 to 15.000 rpm within a second which we all know is completley wrong. By this wrong input, the piggyback injector fires wrongly and floods the engine of fuel. That explains the missfiring of the engine when you slightly twist the throttle under load.
The most resonable solution would be to get a better prm signal in order to get the piggyback efi to calculate the correct fuel not matter of boost or rpm.
The problem on most dirtbikes is that they doesn’t have any revcounter or similar output signal.
Tried to connect to an other wire to count crank revolution . With an occiloscope connected, you can monitor the quality of the suitable input pulse. The crank sensor showed a nice sinusoidal wave which unfortunately is not a good signal for a rpm signal to the piggyback efi box. It has to be a square wave for it to work. There are probably some aftermarket filters avaliable than can convert a sinusoidal waves to square but we didn´t choose to go that path.
Tested the wire to the volt regulator. This left a good square wave but due to the number of coils in the stator, it indicated 16,000 pulses already at idle which makes it impossible to map up to 10,000 rpm. I also tried to connect input signal to the ignition coil but it did not deliver a quality pule for the piggyback efi-box.
The easiest and best solution turned out be replacing the current piggyback efi unit to a new box that instead moderates the duty cycle of the stock injector in the throtthle body. The additional fuel injector will be ditched. It will most likely not be able to deliver enough fuel with the oem fuel injector. Duty cycle will probalby be more than 100% at moderate boost levels. In that case I will go for a larger injector. As far as I know, no larger bolt on KTM injector is avaliable but with some modifications, some other suitable after market injector will fit.
This looks to be a better solution when dealing with wierd rpm input signals. We will be able to get better control of fueling and also a simpler build. Less parts simplifies the build and also deliver fuel more accurately.
Thanks to our moderate boost levels, a large single injector will likely have enough flow rate to deliver fuel to a 90+ hp engine.
100+ hp 1 bar boost will likely require an extra injector but also some heavy engine modifications such a lowcomp piston and probably more...
Our goal has since the beginning been a reliable and powerful everyday turbo snowbike.
Have put the rear wheel on, ready for some dyno tests.
Replaced the rear shock with a rigid strut. No suspension is allowed as the turbocharger is very close to the wheel.
Last edited:
