y8s Hydra Nemesis Parallel Install Thread
I'm going to be doing this because I want to pass OBDII plug-in tests.
Don't come in here and tell me why I shouldn't do it. That's not this thread. I'll start with a list of ECU functions and their pins and if they should be ported back over to the stock ECU. I already have the male and female harness connectors and pins. I can wire them all up like a boomslang and intercept what's needed and send it off to the hydra. Most of what I'm looking for is answers to questions that come up about sharing sensors, piggybacking signals, etc. Next post starts the hard dataz. Matt |
Stock ECU functions
These are basically just pasted from my Hydra pinout spreadsheet.
Definitely going to the stock ECU: Code:
|
Need to be Shared with Stock ECU:
Code:
|
OK So I plan to have the stock ECU do the following functions (you can find them above)
I'm NOT sure the stock ECU will care if the hydra will do these:
Also: It's a pretty easy task to piggyback a digital signal since they're just pulses, right? Finding bias resistor values or making a voltage follower for the stock water temp sensor will be kind of a pain since the hydra may have an internal bias resistor. Might be easier to just put in a second sensor. |
Definately put in a second coolant sensor. Much easier and less headache.
The VVT is questionable, as the stock ECU will constantly be trying to put that cam in a certain spot buy the HYDRA will put it elsewhere. The stock ECU will never win. That could cause a CEL, but I dunno. Is there a CEL for VVT related problems? Knock sensor won't matter I do believe. I unplugged mine for a while and the stock ECU never cared. The MAF sensor you are on your own. That has to be given a signal. If you can fake it out then I guess you'll be fine, but there is a potenital for CEL from this if your fake out plan isn't just right. Someone needs to build a circuit to show the "A-OK" code for passing OBD2. I know, it's probably not legal, but... It would be simpler. :) Also2: where are you gonna put the Hydra? Right next to the stock ECU? I did that with MS2 and hated it. Now that I'm standalone Mine is velcro'd to the carpet behind the passengers seat. Easy to get to and no mess under the dash. You might consider remote mounting it. |
Originally Posted by patsmx5
(Post 304170)
The VVT is questionable, as the stock ECU will constantly be trying to put that cam in a certain spot buy the HYDRA will put it elsewhere. The stock ECU will never win. That could cause a CEL, but I dunno. Is there a CEL for VVT related problems?
Also, when my knocksense went haywire, stock ecu returned P1609 Powertrain Control Module (ECM CPU) Knock Sensor Circuit or something like that. Not sure if Hydra will do the same. But this says there is a KS MIL |
Very interesting thread. I wish I could keep the stock ECU running the idle control and A/C. Especially since the wideband with the Hydra doesn't do so well at tuning the idle controls. Good luck.
|
subscribed! Very interested in doing this as well!
|
Originally Posted by patsmx5
(Post 304170)
Is there a CEL for VVT related problems?
Likewise, you've for P1512 and P1518 for VTCS valve stuck (vacuum switch not changing in response to solenoid command) plus P1569 and P1570 for VTCS solenoid undervoltage / overvoltage (solenoid fault). In other words, leave the damn things connected to the stock ECU. Knock sensor won't matter I do believe. I unplugged mine for a while and the stock ECU never cared. Someone needs to build a circuit to show the "A-OK" code for passing OBD2. I know, it's probably not legal, but... It would be simpler. :) |
thanks joe. I hadn't yet looked at the list of fault codes.
AH wait, now I remember. Tec3 used its own knock sensor. FYI to all, I did run the Tec3 in parallel. It only ran fuel and spark. I think the thing that's going to suck is if I can't run VVT externally, I will have to fool the ECU twofold... 1. The AFM needs an input signal somewhat proportional to load and rpm. 2. The AFM and RPM signals drive the VVT settings. So you can see that if I want to control VVT in parallel with no CEL, I'm going to have to do it in a roundabout way. But I can (sorta?) do it. |
oh yah, also...
if I do end up with a parallel install and can't successfully get the hydra to do all of the things it is awesome at... it might become a liability. |
Originally Posted by y8s
(Post 304222)
it might become a liability.
|
Originally Posted by compy
(Post 304228)
ms ftw.
Originally Posted by y8s
(Post 304221)
thanks joe. I hadn't yet looked at the list of fault codes.
http://alt1040.com/wp-content/upload...tter-whale.png I think the thing that's going to suck s if I can't run VVT externally, I will have to fool the ECU twofold... 1. The AFM needs an input signal somewhat proportional to load and rpm. 2. The AFM and RPM signals drive the VVT settings. So you can see that if I want to control VVT in parallel with no CEL, I'm going to have to do it in a roundabout way. But I can (sorta?) do it. |
Originally Posted by Joe Perez
(Post 304241)
Why are you so dead-set against letting the ECU run the VVT system? You're never going to be able to control VVT directly without throwing codes.
|
Originally Posted by y8s
(Post 304283)
\It's just easier to know what the cam is advanced to when I put "+6 degrees" in a box and not "airflow duty cycle 33% at 2 psi and 3400 rpm"
For standalones, it's easy- pretty much anything is "better than nothing." But since you're already condemning yourself to a life of parallel operation, why make it even harder than it has to be? Get the MAF-substitution map built and let the ECU do its thing. |
an alternative is a quick harness swap to a less-than-ideal situation to keep the CEL happy and then switch back for max control and power.
|
If you let the factory ECU control VVT, and if VVT advance at different RPMs is a function of airlow, you can therefore modify VVT by diddling your AFM cheat signal. Problem is, other things would be affected too, like EGR.
The other way of doing this is to build a box that fakes the cam sensor signal phase angle by an amount specified by a Hydra output. If the box puts a lag in the signal, the ECU will compensate and advance the cam.... |
right. I mentioned that above and how much it would suck.
basically I'd have to pull on my left ear to advance timing and rub the shift knob to retard it. |
Ya don't like my box?
BTW said box is way easier to design if you only need to advance the timing over factory. Would that be the case? |
I suppose if stock is optimal, it doesn't matter. Considering the effect was largely RPM based and not boost based...
i just have to find your black box and recall the wiring colors :) |
Originally Posted by y8s
(Post 304308)
an alternative is a quick harness swap to a less-than-ideal situation to keep the CEL happy and then switch back for max control and power.
Let's see... In order to do a swap to get through inspection I'm thinking I'll need to swap the following: ECU Injectors o2 sensor(s) MAF (put it back in somehow...) I could pop the wastegate off in order to keep the boost to a minimum, pull the hose off of the BOV so there's no drivability issues... Thoughts? y8s, have you progressed any further on your parallel install? |
does figuring out it's easier to use a second knock and coolant temp sensor mean "progressed"? because that's all I've done. ... oh ... that and highlighted a copy of the ECU harness connectors in colors corresponding to my lists above.
in theory to get through the inspection you may not need to pull anything. if you can set an output map to provide MAF data, the rest can be shared. the stock ECU wont throw codes as long as you're providing good tuning. |
miniature update:
I bought a GM temp sensor and pigtail for $15ish. why? because: 1. it's easier to drill and tap a known thread size I have a tap for 2. Hydra themselves said they use the DELCO (GM) setting for the miata temp sensor anyway, so at worst, I'm improving the accuracy of the system (hah) 3. no opening of ECUs. also, I got some data from JasonC SBB on the stock AFM. TBA. |
who doesn't like pictures?
I took jason's magic AFM equation and popped it into excel in something resembling a 3D map. and made.... a 3D graph: Here's the raw equations up to 8k rpm and 200kPa MAP. http://y8spec.com/nemesis/afm-signal-uncap.gif clearly you'll get an overvoltage on the ECU input for the AFM so ... Jason suggets capping the voltage at whatever the AFM reads at WOT in a stock vehicle. lets assume for the sake of easy pictures, that number is 5. http://y8spec.com/nemesis/afm-signal-capped.gif Similarly, there's a minimum voltage that the ECU will want to see, so that needs to be measured too. I've got an OBDII scanner I can graph with so perhaps that's the magic bullet. Or maybe it's "stick a voltmeter across the AFM pins"? And here's what the 01 service manual says: MAF Voltage: Ignition switch ON: 0.9—2.0 V Idle: 1.7—2.4 V (after warmup) MAF value: Idle: 2.2 - 3.2 g/s (warm, 750-850 rpm) 2500 rpm: 6.5 - 8.5 g/s I suspect I can plug in the OBDII and take readings at various engine speeds and then plug in the hydra and try to put in values that the ECU interprets as the same as if it were a real AFM. |
I've updated the excel spreadsheet to match Jerem/FM's "PWM Map 9 Scale" spreadsheet since the PWM maps do not show units or values, just percentages X or Y. I will simply match the one for VVT in terms of X and Y values and place my duty cycle values in there.
As it is now, I have created a nested IF statement that takes the first constant value of the AFM equation and makes it a variable. this allows me to set a minimum value for the first IF() so that I can match the test requirements for the AFM signal at 0 RPM, 100kPa (key on, car off) as well as the warm idle value (700-1000 rpm, 20-35 kPa). Then the second IF() statement is based on the function compared to a max value. if it exceeds the value, it uses the set max value. Wanna see it? Of course you do. I have created 3 variables: coef (the constant value) min (the minimum value) max (the maximum value) and I can drag the cell corner to set the 3D region matching the PWM map because of the $ in the cell location names.
Originally Posted by excel!
=IF(IF((coef+0.00023577*$B9+0.00907392*E$3-(0.0000000181099*($B9^2))-(0.000021516*(E$3^2))+(0.000002085015*$B9*E$3))>ma x,max,(coef+0.00023577*$B9+0.00907392*E$3-(0.0000000181099*($B9^2))-(0.000021516*(E$3^2))+(0.000002085015*$B9*E$3)))<m in,min,IF((coef+0.00023577*$B9+0.00907392*E$3-(0.0000000181099*($B9^2))-(0.000021516*(E$3^2))+(0.000002085015*$B9*E$3))>ma x,max,(coef+0.00023577*$B9+0.00907392*E$3-(0.0000000181099*($B9^2))-(0.000021516*(E$3^2))+(0.000002085015*$B9*E$3))))/5
NEEDLESS TO SAY, it will take about fifty years to pgup and pgdn this into the hydra software... :) |
All times are GMT -4. The time now is 11:15 PM. |
© 2024 MH Sub I, LLC dba Internet Brands