2X044
-
Posts
41 -
Joined
-
Last visited
-
Days Won
8
Content Type
Profiles
Forums
Events
Posts posted by 2X044
-
-
Surprisingly my wife didn't want to drive my car at WOT for extended periods while I measured this stuff.
I hit up a mate today to drive while I fought the oscilloscope.
I saved this from one of the runs, ignore the peaks as they aren't real. Using the period for the 5 waves gives a PWM frequency of 40 Hz and the signal is the full car battery voltage. The PCM grounds one of the wires just like for injectors etc. Easy!
-
On 6/3/2023 at 8:46 PM, BeerTurbo said:
i actually asked turbosmart if they could read the falcon canbus for the rpm for there boost controllers & read the can bus for tq reductions to fire there electronic bovs but they said it wasn't worthwhile. that's was easily 2 years ago now.
are you able to look into getting a controller to read the canbus instead of reading the sensors directly ? cause thats is a dickload of wiring.
I am taking RPM direct from the canbus.
This thread is a little complicated as there is two (almost) entirely separate ideas being developed as time and parts permit.
1 - An engine protection system that shuts down the engine if low oil pressure or low fuel pressure is detected.
2 - A boost controller for the NA + T crowd.
The sensor wiring I have mentioned previously is, at this stage all for an engine protection system.
The challenge that I have currently with the boost controller is that I have not found all the data required to mimic the factory closed-loop control published to the canbus. I can request the information from the PCM through the canbus but that creates other problems. Regarding the boost controller, I am going to get it working as an open loop controller that uses the available information (rpm, gear, throttle position, wheel slip) and see how it goes.
Next thing I need is for someone to drive my car while I see what the factory boost control system is doing on an oscilloscope (what PWM frequency is being used, what voltage is being used etc.). The weather is terrible though so it will have to wait.
-
I'm sure that I could spoof the IAT sensor using the same technique as the CHT and EOT but I wasn't sure it was necessary. I would think bringing in a low rev limiter would do a pretty good job of killing the boost. If you are making significant boost by 1500 rpm you need a bigger turbo.
I'm mindful that there are a lot of wiring that is being introduced. Listed:
- Cable to the CHT sensor
- Cable to the EOT sensor
- Cable to the Boost Control solenoid
- Cable to the Fuel Pressure Sensor
- Cable to the Oil Pressure Sensor
If I keep adding more it will be another loom. I'm already thinking that I need to run one large cable (20+ cores) through the firewall before breaking it into the separate cables. Easy enough to get it through the clutch line grommet (especially for an auto).
-
The parts arrived today to snoop on the boost control solenoid valve. I also have all the parts for the final oil/fuel pressure measurement system. Once I have tested both of those I will finalise a prototype PCB that will have all the current I/O.
I have also changed the micro-controller to the ESP32-S3 which has both BLE and WiFi.
Regarding boost control, I can still not see a way of fully mimicing the factory closed loop system but the open loop system that the system should be able to do straight off includes adjustments or completely separate maps for the following:
RPM
Selected Gear
Throttle Position
Wheel slip???
It is also possible to have many more points in the maps (say every 250 rpm).
Any thoughts on whether an open loop boost control system with this functionality would be wanted?
-
I found a good use of the time I spend sitting outside my daughters dancing lessons. Only slightly anti-social...
Spoofing the engine oil temperature sensor, I can implement a rev limiter as shown in the screenshot below. I was only revving to 1800 before bringing in the limiter to 1500 so it isn't exactly obvious, I'll do a more severe test tomorrow. It didn't feel like a hard limiter though, will have to look into how it is doing the limiting.
With this we can implement either an engine shutdown and/or a limiter. From a safety perspective (human safety not engine) I prefer the rev limiter to the shutdown.
-
Sorry for the miss information, it is pin A17 and is pink/black.
I'll go back and correct the original post if it will let me.
-
1
-
-
Hi All, this is not a question but rather my attempt to make this information searchable (googlable) for people following a similar path to me.
Unlike the BF/FG the BA Mk2 does not use pin B42 to actuate the reverse lockout solenoid on a t56/tr6060, rather it uses pin A16, which, on an auto is used by a transmission shift solenoid. Considering this, I it may be that this pin can also be used by the BA Mk1 if you are performing a T56 conversion - testing required! If you do this testing feel free to post your results here.
Please note that you will need to have a BA Mk2 T56 strategy loaded in the PCM - I am using the F6 strategy with diff ratio correction.
Cheers,
Peter
-
On another thing, I fixed my boost creep problem but I don't fully understand how.
I was running the F6 tune but had retained the standard XR6T boost maps. I changed to a turbosmart actuator and still had the issue and would see 12psi and boost cut. Putting the full F6 tune I never see boost creep.
Not sure how the same turbo can creep on one tune to 12psi and not on another (holds steady at 7-10psi) when the only change was to the boost maps. Only thing I can think is that the boost maps caused a feedback to the spark timing...
-
Hi all, sorry for being a phantom. Since I last was here I started a company, rolled that into another company and more. Good news though, I've finally got some time to scratch myself so I have dusted off this project.
I'm thinking through the closed loop boost control and to fully mimic the factory system I need the following inputs:
- RPM (CAN packet 0x207)
- TPS (CAN packet 0x207)
- EngineCoolantTemp (CAN packet 0x427)
- MAP (CAN packet ???)
- Manifold Charge Temp (CAN packet ???)
- ?
I can't see where the MAP and Intake Air Temp is available via CAN without sending 0x7E0 requests. Any ideas?
Do tuners retain the manifold air temperature table or just simplify this? I can't see how it can be retained without an OEM lab.
Cheers!
-
For anyone interested. I found my boost control issue. Now I need to decide what I'm going to do about it and how that affects the testing for this system. Thinking of getting a Pulsar turbo as it isn't just the turbine housing but the CHRA is a bit second hand too. I can't see any reference to anything but a 15 psi spring in the Pulsar turbo internal gated options.
This is the wastegate fully closed...
This is the wastegate fully open.
The worst of both worlds, laggy and then not able to control the boost. Funny thing is I can't see how this ever worked, it hasn't failed, with the auto I just didn't notice as much as I do with the manual.
Genuine OEM quality.
-
My apologies for no updates of late. I have been busy with my day job. I've got the parts to be able to spoof the EOT sensor and will share what I get from that soon.
-
On 12/18/2022 at 6:27 AM, hjtrbo said:
Way to slow to respond and had a 10deg offset from my previous logs. Calling it a fail. Have relocated to the turbo feed and put temp sensor back to where it was in live oil flow. Just made a npt fitting in the lathe and tigged it onto the banjo nut.
Thanks for sharing your experience, nice for the rest of us to be able to learn the "easy" way.
On 12/20/2022 at 4:32 PM, SpeedLabs said:Hey everyone,
I just came across this project and am really impressed with what you guys have managed to do with the factory PCM. I'm keen to get involved with the project once I get up to speed with it. Much more reading to go on my part yet! My day to day is dyno tuning and race care electronics, mostly with standalones, based in NZ. I'm currently looking to get a new shop project car, recently decided on a Barra and just the other day came across PCMTEC and have decided it's definitely the right platform for me!
Here's my 2 cents on the current topic, hopefully it's of use to somone. I use the Honeywell MIP series for every race car we build, measuring oil, fuel and coolant pressure, with great success. But they are sensitive little things that can be prone to failure if they don't get just what they need. The datasheet is a great resource https://www.farnell.com/datasheets/2897784.pdf The big one that kills them is vibration. They don't like to be directly bolted to an engine, they are much more likely to survive when they are remote mounted with a rubber isolated mount. Haltech do a cheap remote mount kit https://www.haltech.com/product/ht-039103-pressure-sensor-extension-kit/ The other big trap is water getting into the connector causing corrosion. Right angle boots like this are a big help https://msel.co.nz/right-angle-rubber-boot-for-pressure-sensors/
The failure mode on these sensors tends to be intermittent and erratic too which is a big headache. Looking at the data you'll see numbers that don't make sense or brief spikes to 0. It's worth proceeding with caution if you're looking to cut the engine on a street car. I typically will only implement a limiter to ~1500rpm on loss of oil pressure on a street car to allow the driver to limp the car out of the path of potential danger. An emergency over ride switch or button is another option. 90% of the cuts we see are from wiring faults or sensor failures. But the other 10% has saved many race motors!
Also you can get the sensors for a good price with free shipping here
https://nz.element14.com/honeywell/mipan2xx100psaax/pressure-transducer-seal-gauge/dp/3364907
Octopart.com is the go to find the best place to buy anything electronicThanks for the advice on the Honeywell sensors, I hadn't considered vibration to be an issue for them and was planning on engine mounting the sensors. I might still engine mount for my testing purposes, but it also shines more attention on the need to build a strategy of dealing with component failure.
1 hour ago, SpeedLabs said:I've been thinking about this one some more and at the risk of throwing a major spanner in the works, one option that could be worth considering for some people is using a Maxx ecu in a piggyback configuration https://www.maxxecu.com/ (not related to Maxx the workshop). It's a highly customizable ECU with advanced CAN IO configuration. Just about anything you can imagine doing it will do, and if the feature doesn't yet exist they will often write new firmware for you, sometimes within 24 hours. That would give you a lot of the functionality you're looking at implementing without having to go through the long process of bringing new hardware and software to market. Plus you'd get onboard high speed data logging, built in lambda control, multi-level engine protection, ability to control extra injectors, fuel pressure deviation tracking, staged fuel pumps, fully configurable outputs etc etc. There could still be a business case in selling it with a pre-loaded configuration and plug and play harness if you wanted to make a business out of it, implementation is everything.
Please don't get me wrong though, I'm definitely not wanting to derail your project! Just putting some ideas out there from my own background that might be of use to someone. The car world needs more people like you guys thinking of novel solutions to these problems
The Maxx is the path I'm considering going down as I want to experiment with running dual fuel with methanol as the secondary fuel. There's no E85 at the pump here and methanol is cheaper than 98, where as ethanol from a drum is $4 litre
I agree there are many ways of skinning this cat and the best system for an application will depend on many variables. @Roland@pcmtec suggested early on that the system should be installed quickly and easily without any tools. So far the design on working on will get close to this goal.
-
54 minutes ago, hjtrbo said:
Agree with Puff. 150psi is a better choice to cover all bases.
Here is the Honeywell oil pressure sensor plumbed in for reference. Just common fittings available from any industrial supplier.
Must be nice working on a clean motor out of a car. Messing with the EOT and Oil Pressure sensors on my car is a painful experience.
I checked if the number-plate Valdez was available for my car due to the oil leaks...
Do you have experience with the EOT sensor in that location, I suspect it'll be fine but it is away from any "live" flow, might just make it slower to respond. I've left the EOT sensor alone and put the oil pressure sensor in a tee off the turbo feed (AN fittings for braided lines).
-
33 minutes ago, hjtrbo said:
@2X044. How quick have you managed to get the engine to shut down? Have you tried instantly I.e. set auF11960 set to 0. Would be very good to have it shut down that quick if fuel or oil pressure drops dangerously low.
I need to increase the logging speed on some of these parameters to get a more accurate number, however based on one of the logs, The throttle starts being rapidly shut in approximately 0.5 seconds (about how long it takes for the CHT to rise), the injectors are being cut at a similar time and the engine is no longer spinning about 4 seconds later.
What I have found is that auF11960 controls the cylinder shut down, but as soon as the CHT goes above ~125 deg C other parameters are immediately acting to reduce torque to 50% through fuel cuts and throttle position before the cylinder shut-down parameter kicks in.
In general I am hesitant to make any parameter 0s, as this gives no room for electrical noise. Nobody wants there engine to cut-out when they turn on the radio/wipers. In general, I believe there is a reason Ford didn't make all these parameters 0s. I'm cautious about whether I have already gone too far.
at
-
A small update, I've just ordered the connectors to be able to interface with the engine oil temp sensor the same way that I am interfacing with the CHT sensor. Depending upon what we are trying to achieve with the engine, I've found some EOT parameters that may be more suited to our needs. Let's hope that I don't also need to interface with the Intake AIr Temp sensor too.
-
16 hours ago, hjtrbo said:
@2X044 perhaps. I have 2 Honeywell MIP series selected for fuel pressure and oil pressure duty. Data sheet supports anything automotive and as such I am confident that this is a good purchase. Roughly $130AUD each Pressure Sensor 100PSI Honeywell (efihardware.com).
Thansk @hjtrbo, I hadn't found those sensors in my search. I like the idea of a Honeywell sensor over some no-name sensor. I see they make a 100 psi and a 150 psi sensor, most companies I've looked at tend to use the 150 psi but the unit you linked is 100 psi. The SAAS sensor that I ordered is also 150 psi. Can anyone see a need for a 150 over 100?
-
More testing done today and have the engine shutting down near enough immediately on sending a high temp signal. I've recorded some log files for anyone interested. First one is a slow sweep from 147 up to 220 degrees, this shows that the shutdown is happening in two steps, first a reduction in torque to 0.33 multiplier, then 10ish seconds later, a further reduction to zero.
24-11-2022 01-30-56 PM Log.teclog
By changing auF11960 to 1 second, this brings the shutdown on nearly immediately.
This log file shows the reduction in torque to zero occurring almost immediately.
24-11-2022 02-05-10 PM Logzerotqafter1s.teclog
Finally this log is a bunch of running tests, the first being to 220 deg C, the next ones to 195 deg C (below critical temp) and the final ones to 147 deg C. Need to do more tests between 126 and 137 deg C to see what happens between CHT overtemp start and CHT overtemp torque reduction.
24-11-2022 02-15-10 PM LogRunning147deg.teclog
A few things I noticed but could be looking in the wrong spot, I haven't seen any "enrichment" to the contrary the lambse goes to 1.05.
The torque reduction due to ECT (throttle) is active and drops from 255 to approx 5. Suggesting that the throttle is being used by the system not just fuel-cut.
The torque reduction due to injector cut is also active and drops to 0.333 before dropping to zero when above critical. When below critical it drops to 0.5 and never recovers or changes or seems to have any effect, I suspect this will go back to normal when the temperature drops below 136 deg C.
The torque source changes from "driver demand" to "target RPM" immediately on high temp but I can't seem to see what the "target RPM" is.
I've looked at a couple of aftermarket systems to see how they implement engine protection and one uses a significantly reduced rev limit with a fuel cut limiter. They claim this is safer than a complete shutdown due to allowing a car to be moved to a safe place (either on track or on the freeway). It seems like we could do something similar but I haven't yet found the parameters.
I've also ordered a 3 wire oil pressure sensor that can serve as the oil pressure input to the system though for testing I'm going to have to simulate the readings as I don't fancy running low on oil pressure intentionally. If someone has an ex-taxi that they don't mind destroying we can test it on that. The sensor is 0-150psi, has anyone ever used an oil pressure sensor to measure fuel pressure? I can't see why you can't unless they aren't E85 resistant.
Any ideas on parameters to achieve a rev limiter would be appreciated.
-
2 hours ago, Puffwagon said:
Yes, if you can limit the engine's ability to rev up then that will give the user time to react.
There are 2 types of users to think about, the "I put petrol in and it goes vroom" type that can't catch a preventable failure and the others that are alert and aware enough to react to a torque reduction event. Hopefully we can accommodate both but we can only do what we can do.
Torque reduction for fuel safety would work as long as all the injectors were 100% switched off for long enough for the engine to drop boost pressure. You wouldn't want a partial cut as it'll still have 3 cylinders firing and they can still be damaged by running lean.
We have torque reduction during a gearshift in an auto car, mostly it's done by reducing ignition timing. Perhaps this can be used for something.
Rate of change for oil pressure below a certain threshold should work fine.
Thanks for the feedback. I'll give it more thought and see what can be done with the PCM.
-
1 hour ago, Puffwagon said:
I'll have to reply tomorrow, ran out of time today.
No worries. One thing to add to the thought process is, we can shutdown the engine before it hits 0 psi oil pressure, perhaps using both a threshold and a rate of change parameter, for example 25psi and dropping faster than 5 psi a second.
-
2 hours ago, Puffwagon said:
How's it going @2X044?
Slowly,it was my birthday on Monday and I went away for the weekend. I am planning on doing more testing tomorrow to try and work out why the engine isn't shutting down.
I need to get better with the PCMtec logger as I'm interested in working out how many cylinders are shutting down (injector pulsewidth per cylinder) and also what the idle air flow is doing. Once I can get it shutting down from idle, I will have the fun of trying to get it to work at higher loads/rpms.
I've found a few more parameters that I can play with, such as the maximum requestable torque under fail safe cooling conditions.
@Puffwagon in your experience and considering that the engine is going to keep turning even with all cylinders shut-down due to the clutch/converter still being driven by the back wheels, would cutting requestable torque to SFA be enough to get the driver to realise they need to put the car in neutral and kill ignition ASAP.
-
12 hours ago, dat111 said:
very intrigued by this, i really dont know what is going on with 90% of it. but would this be a viable option for the NA+T guys?
as that seems to be a growing community. being able to use computer with an add on like this for boost control, would be great.Hi @dat111, there are two goals of the system that I am working on, the first is a boost control system that mimics the factory PCM turbo strategy. This is quite simple from an electronics side but quite tricky (for me anyway) for the software side. The second function is a protection system for the engine to protect from low oil and fuel pressure. This is quite simple software wise but a little tricker electronics wise.
I've been spending my time recently working on the protection side of things as my test hack has a boost creep issue (stuck/sticky wastegate) that would prevent me from doing any testing on the boost control system. I don't want to pull it apart to fix as that would stop my testing on the protection system while it is apart.
-
More testing done. Changing the CHT time constant works really well as per graphs below which compare a 0.1s time constant to the original 1.2s time constant. Takes a 12th of the time to change as expected.
With the 0.1s TC and starting at a CHT of 90 deg C and a step change to 220 deg C, 150 deg C is reached in 0.1s and 200 deg C is reached in 0.2s.
After logging the A/D counts I changed the appropriate values to push up the temp even higher. When I changed the upper end of the dual resistor curve to 700 deg F I just got a CHT fault. More investigation needed, I noticed there was a CHT "calibrated range" which may be in play.
Another issue was experienced too. I did a "running test" and found that even with a CHT of 220 deg C, the engine didn't stall every time.
The PCM is configured to shut down 12 cylinders when critical temp (204.4) is reached. I might try playing with this number to see if 12 means something else like 1 cylinder every 12... My car shakes about a bit while doing this though so don't want to spend too long testing this bit.
I was also a little concerned that the gauge in the dash might get damaged from trying to move so quickly, but it just meanders up there.
-
On 11/16/2022 at 12:39 PM, Roland@pcmtec said:
Short answer:
CHT is filtered by auF11645 Time constant for filtered CHT. which is set to 1.2 seconds by default. Change this to make it move instantly.
Datalog MID33881 for CHT Fault - This will be set if the AD count exceeds auF10305 (1018) or goes below auF11617 (10)
Long Answer:
Next the following may assist you. This is the CHT transfer function for A/D counts per degree. There are two transfer functions auF1361 and auF2557
it uses the "dual resistor" model when the inferred temp is above auF12365 Use dual resistor system above this temperature. which is set to 80°C by default.
Modify the following scalars to adjust how the over temperature modes function. They don't have proper descriptions yet but I have named them in this post to assist.
There are 3 over temp levels.
Level 1 P1285 is set and the overtemp lamp will come on.
Level 2 P1299 is set and torque reduction will occur by shutting down the number of cylinders set in auF10294
Level 3 Number of cylinders specified in auF1613 will be turned off
I believe Level 2/3 are set up to shut down all cylinders by default, however you can set it up to be staged. Eg run on 3 cylinders, then full engine shutdown.
auF1544 CHT Level 1 temperature setpoint (indicator lamp turns on) 126 °C
auF10952 CHT Level 1 P1285 Temperature Hysteresis (falling edge, eg must be lower than auF1544 ) = 110 °C
auF2280 CHT Level 2 Temperature setpoint = 126°C
auF1870 CHT Level 3 Temperature setpoint (cylinder shutdown) = 204°C
auF11550 Level1 P1285 Temperature Setpoint
The count setpoints are I think are in multiples of how often the code is called. So this may vary on BA vs BF etc. The max number is 250. It might be milliseconds.
auF11550 Level 1 P1285 Count Setpoint
auF12210 Level2 P1299 Count Setpoint
auF12214 Level 3 CHT Critical Level Count Setpoint
Datalog MID06769 for Level 1 current count
Datalog MID06770 for Level 2 current count
Datalog MID06771 for Level 3 current count
Datalog MID33881 for CHT Fault - This will be set if the AD count exceeds auF10305 (1018) or goes below auF11617 (10)
When testing make sure the engine has been running for this long otherwise fail safe cooling is not enabled
auF12786 Time allowed in run mode before entry into fail safe cooling is permitted. = 10 seconds
You can manually force fail safe cooling by setting auF1549 = 1. This is if you want to see what actually happens in fail safe cooling mode.
auF11665 Time since power up to keep fail safe cooling light on for bulb test.
This is how long fail safe cooling must be active before the coolant lamp is turned on.
auF12949 Frequency of the engine coolant light flashing on/off. = 1 second
This is the interval at which the engine coolant light will flash
Thanks Roland,
I can see a few options to make the response quicker.
1 - Reduce the time constant.
2 - Reduce the level 3 setpoint (150 deg C instead of 200).
3 - Increase the temperature corresponding with low A/D counts.
The change-over between the single and dual resistor ranges also appears to add a delay. It may be that the system can only work quickly when the CHT is in the higher range. The cold start rev-limiter crew might not be protected.
-
A huge thanks to the PCMtec team.
In response to their generosity, I have committed to commercialising this product, or if I can't, then I will release everything I've done under an open source license so that others can freely use my work.
Thanks again!
Feeling motivated!
-
2
-
micro-controller using CAN bus data for ancillary control
in General
Posted
Started on a basic GUI.
I need to do a final check on the CAN comms with the ESP32 S3 and I will get some PCBs made for testing the boost control side of things.