Holset HX50 Supermoto Jetbike

[enginewhisperer]

Do those ignition boxes need the spark gap? That is probably a huge source of noise compared to just running the spark plug directly.

[racket]

Hi Mark

Ah yes, Russ Moore , thats a long time ago :-)

Cheers
JOhn

[pitciblackscotland]

Apr 30, 2020 19:49:32 GMT -5 racket said:

Hi Mark

Ah yes, Russ Moore , thats a long time ago :-)

Cheers
JOhn

Hi John,
Yeah many years ago,can't find the vid of the bike when it was finished. I still have some parts that i purchased from him, he was selling different kits to help people to start their build of a turbo jet.

Cheers,
Mark.

[slittlewing]

Thanks for the comments guys, I mostly sorted the ignition EMC woes after experimenting with ferrites, flyback diodes, second power supply etc... it turned out to be EMC from the spark plug lead. I was using copper core leads (for more oomph), after changing to carbon core (suppressed) the issue is reduced alot albeit not perfect. Longer term I could shield the HT lead with stainless braid and ground it.

So I tried to fire it up, but was unsuccessful. I already had done 3 spools of the scuba bottle yesterday, and did 3 more today but the last was pretty weak.
Here is the first one:




I initially tried with a fuel pressure target of approx 2 bar (which is what the engine ran at with 1 bar P2 previously). This didn't work so I then tried 3 bar pressure target. This didn't work so I tried 1 bar pressure target and a little petrol added. 
I had to drain the combustor after each attempt so I know the engine is getting fuel. I didn't get a single pop or any smoke of ignition.

As i see it, there are several possibilites:
1) despite the noise, as I am not seeing any P2, maybe I don't have enough airflow!? Could i be getting restricted from the hydraulic QD?
2) my spark plug is too far from the fuel (its around the tertiary holes approx 10" vertically upward from kero injector), or the airflow isnt sufficient to carry the fuel up there.
3) my spark isnt "hot enough".. although I did see constant sparking in my gap tube during each attempt (its external and visible).

I am thinking of trying a glowplug next. It should ignite the kero if impinged upon, and it gets away from the EMC issues.

Failing that, I could try leafblower starting again, or worst case have to move the spark/glowplug boss. This will be difficult though, as previously when I had it protrude axially through a hole in the the flametube wall, that hole allowed a small circle of gas to "shroud" the plug, so I ended up having to weld the spark plug boss to the inner reducing cone (pre turbo) and sealing/welding to the outer wall. If i had to do this closer to my "top cap", I wouldn't be able to get the flame tube out again.

Any thoughts welcome!

Cheers

Scott

[turboron]

Scott, I have to clean my sparkplug after every run due to carbon bridging. My fuel is propane, I use one leg of a Custom Coil ignitor which is about 18,000 volts I believe. I also had to run a large diameter cable directly from the sparkplug to the battery negative terminal.

Thanks, Ron

[jetric]

  Hi Scott,
   Where have you drilled your air start port on the compressor housing? as watching the video it seems to take a long time for the compressor to get up to speed using the scuba tank, It should almost be instantaneous with your size of turbo compared to the larger size of mine fitted to my kart, The air port needs to be drilled tangentially at about 5mm in from the compressor exducer tip as you saw on my engine. Are you still using gas to start your engine before turning on the Kero? If you are using straight Kero now you will not ignite it with a simple HT spark! you will have to use a Cracker box and igniter plug, I demonstrated to you how I start my engine straight on Kero using a Cracker box and igniter plug and I think I offered to show you where to source the components from to make a Cracker box up from the design that me and Andy morris developed, the components are readily available to build them. I have tried the glow plug idea when I made my first engine and it was a bit of a hit and miss if it would ignite the fuel, you had to leave the glow plug on for at least a couple of minutes before you attempted a start in order to make sure it was glowing red hot, then if any raw fuel hit it, it would rapidly cool down and not ignite the fuel, also if there is a large amount of airflow past the glow plug on spool up this also rapidly cools it and prevents it igniting the fuel. Also be very careful on the type of quick connect you are using for the scuba air line connection onto the engine as some makes can be very restrictive to high air flow.

  Rich.

May 1, 2020 10:46:28 GMT -5 slittlewing said:

Thanks for the comments guys, I mostly sorted the ignition EMC woes after experimenting with ferrites, flyback diodes, second power supply etc... it turned out to be EMC from the spark plug lead. I was using copper core leads (for more oomph), after changing to carbon core (suppressed) the issue is reduced alot albeit not perfect. Longer term I could shield the HT lead with stainless braid and ground it.

So I tried to fire it up, but was unsuccessful. I already had done 3 spools of the scuba bottle yesterday, and did 3 more today but the last was pretty weak.
Here is the first one:




I initially tried with a fuel pressure target of approx 2 bar (which is what the engine ran at with 1 bar P2 previously). This didn't work so I then tried 3 bar pressure target. This didn't work so I tried 1 bar pressure target and a little petrol added. 
I had to drain the combustor after each attempt so I know the engine is getting fuel. I didn't get a single pop or any smoke of ignition.

As i see it, there are several possibilites:
1) despite the noise, as I am not seeing any P2, maybe I don't have enough airflow!? Could i be getting restricted from the hydraulic QD?
2) my spark plug is too far from the fuel (its around the tertiary holes approx 10" vertically upward from kero injector), or the airflow isnt sufficient to carry the fuel up there.
3) my spark isnt "hot enough".. although I did see constant sparking in my gap tube during each attempt (its external and visible).

I am thinking of trying a glowplug next. It should ignite the kero if impinged upon, and it gets away from the EMC issues.

Failing that, I could try leafblower starting again, or worst case have to move the spark/glowplug boss. This will be difficult though, as previously when I had it protrude axially through a hole in the the flametube wall, that hole allowed a small circle of gas to "shroud" the plug, so I ended up having to weld the spark plug boss to the inner reducing cone (pre turbo) and sealing/welding to the outer wall. If i had to do this closer to my "top cap", I wouldn't be able to get the flame tube out again.

Any thoughts welcome!

Cheers

Scott

[racket]

Hi Scott

Your spark plug is too far from the fuel spray , positioned between Primary and Secondary holes has always worked for me, you need a nice rich mixture , by having the plug near the Tertiary holes the mixture won't be ideal, the fact that you never got any sort of ignition yet needed to drain the combustor of fuel sorta points to a problem of getting a combustable mix between the plug gap .

Even with propane injection you still need the plug near the injwction point to get a good mixture.

Cheers
John

[slittlewing]

Thanks for the advice guys, so today I repositioned the spark plug just above where the kero cone should impinge on the wall. Great news - the kero is lighting from it.

I have done maybe 6 or 7 fireup attempts. (I had to do a 60 mile round trip to work to recharge the scuba bottle after 3).

The bike is very close to working/idling. I make tweaks to the code between each one. I was getting had over in the first 3 because the arduino was not accepting the new code updates and I didn't realise. I had to change to a spare arduino which fixed this. I need to work on the transition in my code between "starting" (when the spark is on and fuel pressure demand is fixed at 2 bar) and running (spark off, fuel demand dependant on throttle). Sometimes it has been going back to "ready" state (all fuel off) which I need to figure out. I am not sure if it could be at all related to EMC still, I have stainless braid ordered to cover the HT lead.

Hopefully with some more code tweaks I can get it to run! One strange thing, at one point I spooled it with the scuba and it didnt really "take", then i reduced the scuba air slightly and it immediately took. Would this be a sign that it needs more fuel, as it was too lean? As soon as it takes (albeit with less air that time) it accelerates fast however!? Which makes me wonder if it is lean or not. I could try increasing fuel pressure demand to 2.5 bar?

I am nearly out of kero and my supplier is shut due to Covid19, so I will need to use diesel petrol mix soon. I hope this still lights the same way as the kero (and maybe 10% petrol) has been.

One other thing, my P2 pressure sensor isnt showing any boost. I tried it using an airline and compressor into the sensor and it works fine. Its currently in a factory drilling in the compressor housing where I previously had an analog gauge on the sled, which worked fine, but I will move it to a 6mm takeoff on the combustor cap and see if this helps (also where I drain the engine and where my bleed valve is)

Videos:







Comments very welcomed, thanks for your support all!

Scott

[slittlewing]

Had another thought.. maybe at full flow the scuba start is blowing out the flame... will try increasing valve slowly instead of going straight to wide open on the ball valve I have fitted.
Cheers
Scott

[racket]

Videos are "Private" :-(

[slittlewing]

May 2, 2020 17:18:11 GMT -5 racket said:

Videos are "Private" :-(

Sorry, try again now! Not sure how or why youtube did that haha

[slittlewing]

Apr 30, 2020 19:39:18 GMT -5 enginewhisperer said:

Do those ignition boxes need the spark gap? That is probably a huge source of noise compared to just running the spark plug directly.

Yes, the spark gap stops the spark from happening until the voltage gets high enough to arc across it. I think its about 15 to 20kv at the moment. Andy and Richards previous work told me that you needed a really fat/hot spark to ignite kero which led me to experiment with home made spark gaps! I should have probably potted the spark gap tube inside the box to reduce EMC but I wanted to be easily able to tell if the unit was sparking. But you are right, it probably isnt helping with the EMC issue!! Maybe I will make a steel shroud to cover the tube later on.

[slittlewing]

May 1, 2020 12:00:16 GMT -5 jetric said:

  Hi Scott,
   Where have you drilled your air start port on the compressor housing? as watching the video it seems to take a long time for the compressor to get up to speed using the scuba tank, It should almost be instantaneous with your size of turbo compared to the larger size of mine fitted to my kart, The air port needs to be drilled tangentially at about 5mm in from the compressor exducer tip as you saw on my engine. Are you still using gas to start your engine before turning on the Kero? If you are using straight Kero now you will not ignite it with a simple HT spark! you will have to use a Cracker box and igniter plug, I demonstrated to you how I start my engine straight on Kero using a Cracker box and igniter plug and I think I offered to show you where to source the components from to make a Cracker box up from the design that me and Andy morris developed, the components are readily available to build them. I have tried the glow plug idea when I made my first engine and it was a bit of a hit and miss if it would ignite the fuel, you had to leave the glow plug on for at least a couple of minutes before you attempted a start in order to make sure it was glowing red hot, then if any raw fuel hit it, it would rapidly cool down and not ignite the fuel, also if there is a large amount of airflow past the glow plug on spool up this also rapidly cools it and prevents it igniting the fuel. Also be very careful on the type of quick connect you are using for the scuba air line connection onto the engine as some makes can be very restrictive to high air flow.

  Rich.

May 1, 2020 10:46:28 GMT -5 slittlewing said:

Hi Rich, just to say thanks for your information - my tube is impinging right on the edge of the turbo compressor exducer. It could have probably done with being a bit closer to the ID as no doubt some air is going past the outside of the blades. I am not using propane anymore, just purely kero. Andy did warn me about glow plugs and them cooling down etc from airflow and kero on them, luckily it looks like my DIY ignitor boxes are working (I wanted to stick with a standard spark plug as I had welded on an M14 boss and didnt want to strip the engine). Today I managed to refit the sparkplug without stripping it though, albeit without boss and just lightly torqued within the 3mm wall of my combustor!
I am using a Hydraulic ISO A QD but might switch to an air line fitting once I get the thing to start!! 
Many thanks for your help,
Scott

[racket]

Hi Scott

Yep , the 3rd video looked very promising, another 10 seconds and you probably would have nailed it :-)

I agree with your assumption about too much starter too quickly , at this early stage , probably OK once its sorted ................a more gentle spoolup might be worth trying , you'll need your P2 gauge to check how your spoolup is going , a couple of psi with initial fuel ignited then increase fuel as the P2 psi rise , you won't need anymore than ~5-7 psi P2 for a safe idle .

Cheers
John

[slittlewing]

Thanks for the info John, I awoke with plenty of ideas and a list of code updates and things to try.
- I verified without any scuba air and constant spark that EMC was still causing problems. In particular it was knocking out the MAX6675's (Thermocouple measurement PCBs), either making TOT read high (Arduino thinks engine is started then), or TOil Read high (arduino increases PWM on oil pump to compensate for reduced oil viscosity when hot). 
- I tried to make a metal shield to cover ignition box but the spark then went straight to the shield and caused arduino reset
- I sealed all the gaps and the lid on the box with silicone, which looked to reduce EMC and I no longer had issues with a constant spark
- I modified the code so that during startup at throttle >5% there is constant fuelling, at throttle >20% there is also spark

I was lucky that my boss at work let me take the high pressure compressor home, so I can now recharge the scuba bottle in my garage.
I tried several fireup attempts and got close again, but still when its almost at sustain it just starts banging (exploding/relight) to the upset of several locals!
I think I may have then started barking up the wrong tree, and thought maybe this was due to the engine leaning out when P2 increased. I tried 2.5bar and 3.5bar fuel pressures. However I still remember with the engine running before it idled at P2 1 bar with 2bar fuel and nothing has changed.

Andy had the good idea of filming the digital dash to see what might be happening. Sure enough it caught the arduino resetting on the last attempt, although it appears to reset after the bangs when there is still fuel on. However I tried this attempt with constant spark (to give it every chance of lighting) so maybe after 20 seconds or so it did cause enough EMC to give an issue.

In short, I am still a bit confused about the engine going out and then relighting. The videos from yesterday (the last one) show it almost being fine at 2bar fuel.

I plan to try and fully quash the EMC issues by fitting stainless braid to HT lead (when it arrives). I will also try a resistor type spark plug cap. Finally I may see if the kero lights without the spark plug gap, as this will reduce the voltage and EMC. Another thing I can do is alter the arduino code so that it goes directly into "starting" mode when it switches on, therefore after any resets it will fuel the engine again - but this is a bodge really.
My last resort is to try an aircraft ignitor plug and bang box, being that they are 2kv I would think they should have less EMC, but this may not be the case with them having a copper spark plug lead and at least 1Joule of energy per spark!

If anyone has experienced the engine lighting and then going out repeatedly, please chime in! I had it with the "jet sledge" on propane when running lean with too small a drilling but thats all I can reference.

Videos of today:






Trials and tribulations, I will keep at it! It seems so close!

Cheers

Scott