Turbo installed. Lasted 4 days. Threw a rod.
#125
Moderator
iTrader: (12)
Join Date: Nov 2008
Location: Tampa, Florida
Posts: 21,026
Total Cats: 3,123
I told you how it could happen. Cam timing being off plus launch control. Or tuning is off.
Something to consider is that the spark timing is tied to the CAS and if that cam is off a tooth or two related to the crank then your very safe tune becomes insanely dangerous. If the tune is only marginally safe then things get to be much more interesting with only a minor cam timing mistake.
It could be the tuner or the engine assembler at fault. Find someone who can download the tune from the MS other than your mapper and post it here for an analysis. We've done this before and can ease your mind as to the underlying cause.
Something to consider is that the spark timing is tied to the CAS and if that cam is off a tooth or two related to the crank then your very safe tune becomes insanely dangerous. If the tune is only marginally safe then things get to be much more interesting with only a minor cam timing mistake.
It could be the tuner or the engine assembler at fault. Find someone who can download the tune from the MS other than your mapper and post it here for an analysis. We've done this before and can ease your mind as to the underlying cause.
#126
Right…
This is clearly one of those scenario's where the internet has been given part of the information and run with what's assumed by that.
I tuned this car
For those in the USA who aren't aware, this isn't my first rodeo and was a standard job - Connor is a customer who I get on very well with and things like collecting the car from him was done due to that relationship, to help him out rather than him finding a trailer etc.
I've tuned more cars than the majority of this forum as a fact. I subcontract as a tuner to very well respected companies locally & internationally and tune everything from low level, high revving NA race cars to 1000bhp+ drag cars, running every different type of fuel available on a range of management systems. I offer E-tuning services internationally and 80% of my income is from tuning. I've tuned a large majority of the UK's higher powered MX5's and have installed and tuned countless turbo setups on stock BP engines. Connor's car was routine.
The car was tuned on a DynaPack hub dyno [Abbey Motorsport, Surrey, UK] with a Plex tuning knock sensor fitted and Motec PLM fitted to the downpipe. As part of my tuning routine I intentionally cause a knock event at low rpm & low load to ensure that my knock hardware is working correctly. This doesn't cause harm to the engine and means that I can be 100% about my tuning. A manual fuel pressure gauge was fitted and checked on throughout tuning.
Connor's car has a HKS hybrid setup with the smallest T25 turbo available - it's fitted with a .3bar actuator, the most restrictive down pipe I've ever seen [read as small bore, down pipe can only be described as 90 degree point into 2.25 tube]. He also has an intercooler that creates .3bar of boost drop at actuator boost levels - meaning that at 6kRPM the ECU actually saw 102KPa when on actuator. The complete setup was supplied to Connor second hand extremely cheaply through a UK forum from a chap that Connor himself has described as being "a lovely guy but a complete idiot". The injectors are 550CC EV14's with spacers fitted but no gauze filters.
Base timing was set before leaving Connor's and again at the dyno.
Connor's car has previously been run NA [numerous posts about issues with this engine by Connor on this very forum, including seizing a cam in the head last month]. It's previously been run with a limiter of 7.4k - I have stock rods in BP running at 7.75k limiters in some of my track cars, but they're all NA applications. Connor's car gets drifted and abused more than most - hence why I'm careful tuning it.
Currently the limiter is set to 7.2k.
With the MS controlling boost through a Pierburg solenoid I was able to make the boost I needed and by running at .3bar the torque curve showed a perfectly reasonable shape for a BP engine [if cam timing was out, it would show here].
Onto the final tune.
The car struggled to make boost up top - initially I thought it was due to the wastegate being blown open but I lockwired the WG shut for a run on the dyno and couldn't get it to generate the boost up top. Based on my notes the final runs looked like this:
160KPa @ 3k, 190KPa @ 3.5k but dropping to 170KPa @ 4.5k, 165KPa @ 5k, 160KPa @ 5.5k, 155KPa @ 6k to redline.
Boost was left controlled open loop due to the wide range of duty cycles that caused PID wind up when left in closed loop [I'm sure with hours of tuning CL would work well, but this was not meant to be his final setup so would be wasted time].
I don't have access to the dyno graphs right now but I'll post AFR, boost and all the graphs over the next couple of days [I'm on the dyno there for the next 3 days].
I tune in Lambda, so I'll just give a rough idea of my targets below:
up to 80KPA - Lambda 1
100KPa - Lambda 0.95
110KPa - Lambda 0.9
130KPa - Lambda 0.825
150KPa+ - Lambda 0.8
I use these as a starting point, but Connor's car was tuned by those.
Connor's car made 181.5ftlb @ 4.2kRPM with roughly 180KPa, it made peak power 157.8bhp @ 4.8kRPM with roughly 165/170KPa. If you compare those figures to other graphs there was nothing out of the ordinary other than the lack of power up top.
So back to his broken rods…
He utilised launch control - which was set at 4800RPM, using spark cut. Soft limit zone 300rpm, soft retard to -20, enable launch above 10% TPS.
I tune launch on the dyno [for this car, where wheel speed is not referenced and I'm able to do so]- I setup a dyno run starting at Launch RPM, ending at redline with a time of 2 seconds. It's held until triggered by pressing F3 [DynaPack tuners will understand this].
This allows me to build boost using the launch limiter and release the launch button at the same time as starting the run - Connor's car made more boost on the launch limiter than it would on a standard pull, which makes this whole routine worth while - this part of the map was therefore checked over and although I don't have the exact figures to hand torque was no-way near the safe limit of the BP engine [On this dyno I do not tune stock BP's past 215ftlbs @ the hubs].
I've pulled my hair out about this over the last few nights as I've been in contact with Connor, it was only late last night whilst I was out socialising that Connor made me aware of this thread. I've been down the workshop and pulled the MSQ from the server and it's attached below - I'm open to hearing your views.
Once the car was off the dyno the conversation was had to Connor that the restriction up top needed to be sorted to make more power, I recommended a GT2560 turbo and changing the downpipe/exhaust setup - with that he'd be able to make 210ftlb to redline and have a noticeable increase in performance. This is the reason that I didn't waste hours tuning the closed loop to deal with the huge duty cycles needed and also why this was so routine.
I personally think that at some point whilst NA the car has slightly stretched all 4 rods (either through over rev or the 7.4k limiter) and once turbo charged, the extra load put through on the launch control has caused that weakness to become apparent. Connor and I get on very well and I've often made tune tweaks to his car free in the past, spent plenty of time offering him advice on parts of his build and will continue to do so. I'm happy with my work and honestly do not feel the issue is due to me - I know I'm going to hear from you all shortly, but read the above, absorb the information and make informed comments.
Daniel
This is clearly one of those scenario's where the internet has been given part of the information and run with what's assumed by that.
I tuned this car
For those in the USA who aren't aware, this isn't my first rodeo and was a standard job - Connor is a customer who I get on very well with and things like collecting the car from him was done due to that relationship, to help him out rather than him finding a trailer etc.
I've tuned more cars than the majority of this forum as a fact. I subcontract as a tuner to very well respected companies locally & internationally and tune everything from low level, high revving NA race cars to 1000bhp+ drag cars, running every different type of fuel available on a range of management systems. I offer E-tuning services internationally and 80% of my income is from tuning. I've tuned a large majority of the UK's higher powered MX5's and have installed and tuned countless turbo setups on stock BP engines. Connor's car was routine.
The car was tuned on a DynaPack hub dyno [Abbey Motorsport, Surrey, UK] with a Plex tuning knock sensor fitted and Motec PLM fitted to the downpipe. As part of my tuning routine I intentionally cause a knock event at low rpm & low load to ensure that my knock hardware is working correctly. This doesn't cause harm to the engine and means that I can be 100% about my tuning. A manual fuel pressure gauge was fitted and checked on throughout tuning.
Connor's car has a HKS hybrid setup with the smallest T25 turbo available - it's fitted with a .3bar actuator, the most restrictive down pipe I've ever seen [read as small bore, down pipe can only be described as 90 degree point into 2.25 tube]. He also has an intercooler that creates .3bar of boost drop at actuator boost levels - meaning that at 6kRPM the ECU actually saw 102KPa when on actuator. The complete setup was supplied to Connor second hand extremely cheaply through a UK forum from a chap that Connor himself has described as being "a lovely guy but a complete idiot". The injectors are 550CC EV14's with spacers fitted but no gauze filters.
Base timing was set before leaving Connor's and again at the dyno.
Connor's car has previously been run NA [numerous posts about issues with this engine by Connor on this very forum, including seizing a cam in the head last month]. It's previously been run with a limiter of 7.4k - I have stock rods in BP running at 7.75k limiters in some of my track cars, but they're all NA applications. Connor's car gets drifted and abused more than most - hence why I'm careful tuning it.
Currently the limiter is set to 7.2k.
With the MS controlling boost through a Pierburg solenoid I was able to make the boost I needed and by running at .3bar the torque curve showed a perfectly reasonable shape for a BP engine [if cam timing was out, it would show here].
Onto the final tune.
The car struggled to make boost up top - initially I thought it was due to the wastegate being blown open but I lockwired the WG shut for a run on the dyno and couldn't get it to generate the boost up top. Based on my notes the final runs looked like this:
160KPa @ 3k, 190KPa @ 3.5k but dropping to 170KPa @ 4.5k, 165KPa @ 5k, 160KPa @ 5.5k, 155KPa @ 6k to redline.
Boost was left controlled open loop due to the wide range of duty cycles that caused PID wind up when left in closed loop [I'm sure with hours of tuning CL would work well, but this was not meant to be his final setup so would be wasted time].
I don't have access to the dyno graphs right now but I'll post AFR, boost and all the graphs over the next couple of days [I'm on the dyno there for the next 3 days].
I tune in Lambda, so I'll just give a rough idea of my targets below:
up to 80KPA - Lambda 1
100KPa - Lambda 0.95
110KPa - Lambda 0.9
130KPa - Lambda 0.825
150KPa+ - Lambda 0.8
I use these as a starting point, but Connor's car was tuned by those.
Connor's car made 181.5ftlb @ 4.2kRPM with roughly 180KPa, it made peak power 157.8bhp @ 4.8kRPM with roughly 165/170KPa. If you compare those figures to other graphs there was nothing out of the ordinary other than the lack of power up top.
So back to his broken rods…
He utilised launch control - which was set at 4800RPM, using spark cut. Soft limit zone 300rpm, soft retard to -20, enable launch above 10% TPS.
I tune launch on the dyno [for this car, where wheel speed is not referenced and I'm able to do so]- I setup a dyno run starting at Launch RPM, ending at redline with a time of 2 seconds. It's held until triggered by pressing F3 [DynaPack tuners will understand this].
This allows me to build boost using the launch limiter and release the launch button at the same time as starting the run - Connor's car made more boost on the launch limiter than it would on a standard pull, which makes this whole routine worth while - this part of the map was therefore checked over and although I don't have the exact figures to hand torque was no-way near the safe limit of the BP engine [On this dyno I do not tune stock BP's past 215ftlbs @ the hubs].
I've pulled my hair out about this over the last few nights as I've been in contact with Connor, it was only late last night whilst I was out socialising that Connor made me aware of this thread. I've been down the workshop and pulled the MSQ from the server and it's attached below - I'm open to hearing your views.
Once the car was off the dyno the conversation was had to Connor that the restriction up top needed to be sorted to make more power, I recommended a GT2560 turbo and changing the downpipe/exhaust setup - with that he'd be able to make 210ftlb to redline and have a noticeable increase in performance. This is the reason that I didn't waste hours tuning the closed loop to deal with the huge duty cycles needed and also why this was so routine.
I personally think that at some point whilst NA the car has slightly stretched all 4 rods (either through over rev or the 7.4k limiter) and once turbo charged, the extra load put through on the launch control has caused that weakness to become apparent. Connor and I get on very well and I've often made tune tweaks to his car free in the past, spent plenty of time offering him advice on parts of his build and will continue to do so. I'm happy with my work and honestly do not feel the issue is due to me - I know I'm going to hear from you all shortly, but read the above, absorb the information and make informed comments.
Daniel
#128
Well that certainly sheds more light on the situation.
Thank you Daniel, for taking the time to break things down.
I'll check out the msq when I get home..
True, but OP also wasn't giving us any info regarding the "mapper" besides saying that he's well known and he doubts that its the tune. That was a very big variable left to just kinda "trust".
Thank you Daniel, for taking the time to break things down.
I'll check out the msq when I get home..
True, but OP also wasn't giving us any info regarding the "mapper" besides saying that he's well known and he doubts that its the tune. That was a very big variable left to just kinda "trust".
#130
Just what it wanted on the dyno once into boost - hence the comments to Connor about a larger turbo and better downpipe/exhaust setup. I'd usually run more timing through past 6.3k than this car wanted which told me at the time that it just couldn't flow at the top.
Thanks for the comments so far
Thanks for the comments so far
#131
I'm a terrible person
iTrader: (19)
Join Date: Apr 2009
Location: Arizona
Posts: 7,174
Total Cats: 180
Well that certainly sheds more light on the situation.
Thank you Daniel, for taking the time to break things down.
I'll check out the msq when I get home..
True, but OP also wasn't giving us any info regarding the "mapper" besides saying that he's well known and he doubts that its the tune. That was a very big variable left to just kinda "trust".
Thank you Daniel, for taking the time to break things down.
I'll check out the msq when I get home..
True, but OP also wasn't giving us any info regarding the "mapper" besides saying that he's well known and he doubts that its the tune. That was a very big variable left to just kinda "trust".
#132
Looks like a pretty safe tune, conservative spark table.
Good luck working it out. That engine needs to be uncorked. Well, at this point, discarded.
You might want to adjust the "Req Fuel" value (basically cut it in half and double the VE table). As it is now, you're leaving half of your available fueling resolution on the shop floor. I've never found MS's built-in "Req Fuel" calculation to work particularly well. Always ends up giving low values in the VE table (each VE cell can be between 0 and 255, so if your max values are ~100, your Req Fuel is too coarse).
Good luck working it out. That engine needs to be uncorked. Well, at this point, discarded.
You might want to adjust the "Req Fuel" value (basically cut it in half and double the VE table). As it is now, you're leaving half of your available fueling resolution on the shop floor. I've never found MS's built-in "Req Fuel" calculation to work particularly well. Always ends up giving low values in the VE table (each VE cell can be between 0 and 255, so if your max values are ~100, your Req Fuel is too coarse).
#133
Senior Member
iTrader: (2)
Join Date: Mar 2014
Location: Meridian, Mississippi
Posts: 628
Total Cats: 44
Looks like a pretty safe tune, conservative spark table.
Good luck working it out. That engine needs to be uncorked. Well, at this point, discarded.
You might want to adjust the "Req Fuel" value (basically cut it in half and double the VE table). As it is now, you're leaving half of your available fueling resolution on the shop floor. I've never found MS's built-in "Req Fuel" calculation to work particularly well. Always ends up giving low values in the VE table (each VE cell can be between 0 and 255, so if your max values are ~100, your Req Fuel is too coarse).
Good luck working it out. That engine needs to be uncorked. Well, at this point, discarded.
You might want to adjust the "Req Fuel" value (basically cut it in half and double the VE table). As it is now, you're leaving half of your available fueling resolution on the shop floor. I've never found MS's built-in "Req Fuel" calculation to work particularly well. Always ends up giving low values in the VE table (each VE cell can be between 0 and 255, so if your max values are ~100, your Req Fuel is too coarse).
Will changes in the required fuel effect other tables like warm up enrichment, Accel enrichment, and other things or is it just the VE tables it screws up?
Keith
#134
Elite Member
iTrader: (1)
Join Date: Jun 2006
Location: Warrington/Birmingham
Posts: 2,642
Total Cats: 42
I didn't realize how bad the "req fuel" calculation was in the megasquirt until I swapped injectors... with a good req fuel calculator and an injector swap you should need minor tweaks to your fuel map at most... but with the MS system you need to do a full re-tune. Either that or I fucked it up
Will changes in the required fuel effect other tables like warm up enrichment, Accel enrichment, and other things or is it just the VE tables it screws up?
Keith
Will changes in the required fuel effect other tables like warm up enrichment, Accel enrichment, and other things or is it just the VE tables it screws up?
Keith
That's not an issue with the req_fuel number, it's merely a characteristic of your injectors.
#135
Senior Member
iTrader: (2)
Join Date: Mar 2014
Location: Meridian, Mississippi
Posts: 628
Total Cats: 44
With the required fuel adjusted for 1000cc/min injectors I had to add +60 in all areas of my VE table and start a full re-tune from that base. If it was a non-linearity problem I would have only had to add a crap load in the idle areas of the map, and leave the rest alone. When I tried that I was dead lean at anything above idle, so did the "better safe than sorry" start rich and let the VE analyzer live lean it out method. Now, if I end up with values that are about the same at cruise I will have to put it down to non-linearity. I can't check the "high load" stuff because high load used to be 100, now it is 180
Keith
#136
Boost Czar
iTrader: (62)
Join Date: May 2005
Location: Chantilly, VA
Posts: 79,729
Total Cats: 4,126
req_fuel is a multipler. the end.
if he doubled his req_fuel, his fuel map would drop in half.
Horton has ~460cc injectors installed? n/m i see 560cc ev14s.
if he doubled his req_fuel, his fuel map would drop in half.
Horton has ~460cc injectors installed? n/m i see 560cc ev14s.
Last edited by Braineack; 09-01-2014 at 10:01 AM.
#137
Elite Member
iTrader: (1)
Join Date: Jun 2006
Location: Warrington/Birmingham
Posts: 2,642
Total Cats: 42
Not so sure on that, people stopped responding in that thread.
With the required fuel adjusted for 1000cc/min injectors I had to add +60 in all areas of my VE table and start a full re-tune from that base. If it was a non-linearity problem I would have only had to add a crap load in the idle areas of the map, and leave the rest alone. When I tried that I was dead lean at anything above idle, so did the "better safe than sorry" start rich and let the VE analyzer live lean it out method. Now, if I end up with values that are about the same at cruise I will have to put it down to non-linearity. I can't check the "high load" stuff because high load used to be 100, now it is 180
Keith
With the required fuel adjusted for 1000cc/min injectors I had to add +60 in all areas of my VE table and start a full re-tune from that base. If it was a non-linearity problem I would have only had to add a crap load in the idle areas of the map, and leave the rest alone. When I tried that I was dead lean at anything above idle, so did the "better safe than sorry" start rich and let the VE analyzer live lean it out method. Now, if I end up with values that are about the same at cruise I will have to put it down to non-linearity. I can't check the "high load" stuff because high load used to be 100, now it is 180
Keith
PW = DT + (ReqFuel * MAP * VE[RPM,MAP] * AirDen * BaroCor * corrections)
With any other injector, you do as you state, chance the number and a minor retune. EV14's behave considerably differently.
I don't see the big deal though, really how often do you change injectors?
Now, if I end up with values that are about the same at cruise I will have to put it down to non-linearity. I can't check the "high load" stuff because high load used to be 100, now it is 180
Okay.
#139
Senior Member
iTrader: (2)
Join Date: Mar 2014
Location: Meridian, Mississippi
Posts: 628
Total Cats: 44
Don't believe it if you like, but req_fuel (called base fuel in other ECU's), is just the starting point in the fueling equation. it's just a number that's multiplied.
PW = DT + (ReqFuel * MAP * VE[RPM,MAP] * AirDen * BaroCor * corrections)
With any other injector, you do as you state, chance the number and a minor retune. EV14's behave considerably differently.
I don't see the big deal though, really how often do you change injectors?
You're disagreeing with me, but haven't even seen an end result yet? You know your injectors are linear, yet you've not actually tuned the table yet to know whether they are?
Okay.
PW = DT + (ReqFuel * MAP * VE[RPM,MAP] * AirDen * BaroCor * corrections)
With any other injector, you do as you state, chance the number and a minor retune. EV14's behave considerably differently.
I don't see the big deal though, really how often do you change injectors?
You're disagreeing with me, but haven't even seen an end result yet? You know your injectors are linear, yet you've not actually tuned the table yet to know whether they are?
Okay.
And sorry for the thread jack, but I am getting answers here that I didn't get in my injectors question thread.
Because I went from having max load 100 (being naturally aspirated) to max load being 180ish now with the turbo it is hard to see how linear or non-linear the injectors are, I will take everyone's word for it.
Now, this non-linearity would effect warm up enrichment since that is at low pulse widths, but not effect acceleration enrichment since that is normally at higher loads... correct me if I am wrong. Should I make changes to the warm up enrichment?
Keith
PS: really am sorry for the thread jack, but the required fuel subject coming up brought this issue to mind.