Wade

iTrader: (1)

Posted this to the rx-7 list earlier, thought it might be useful here.

*** Apex'i Power FC Injector Staging ***

Who cares about the staging algorithm in the PFC? Well, what we know about the staging of the PFC fits exactly with what we know about the staging of the stock ECU, and I think they either use very similar staging methods (if not exactly the same). And yes, this supports some people's beliefs that the PFC for the RX-7 is based on a modified stock ECU or modeled closely after a stock ECU, and at this point I am inclined to agree with them.

Thanks to those ingenious guys in New Zealand, we now have access to much more information about the Power FC.

Here is a picture (from the Datalogit manual) of the PFC Base Map. The values are milliseconds.
http://www.wvinter.net/~flanham/wlan...el/basemap.jpg

The Power FC uses this basic Load vs. RPM map to determine how much fuel to deliver. There are several corrections, such as water temp, air temp, etc. including the main Fuel Correction map that people use to tune their Power FC. The values in that map are simply percentage corrections (e.g., 150 or 1.50 means the injector on time is multiplied by 1.50).

You might notice that the ms values of the Base Map are ridiculously high. Add large correction factors (like most PFCs have) such as 150, and these numbers are even more ridiculous. In higher RPMS, the millisecond on time is larger than what is actually available during a revolution. How can this be? I asked Steve.

Enter injector staging. The millisecond values listed in the Base Map are only for the primary injectors. Once they duty of the primary injectors reaches 70%, the secondaries are turned on and the actual millisecond value used is appx. 39% of the Base Map value (...plus corrections, of course). Why 39%? XX milliseconds with the 550cc primary injectors is equal to (550/(550+850)), or 0.3928..., which is the ratio that XX milliseconds would need to be reduced in order for all four injectors (two 550cc and two 850cc) to deliver the same amount of fuel as before the switch. This is what Carlos talks about on Scuderia Ciriani, which we know from experience is correct for the stock ECU.

For example, if 10ms is needed for only the 550cc injectors, if we want the same amount of fuel but with all four injectors firing, we'd need ~3.9ms of fuel.

Two 550cc injectors
--------------------
10ms * 1100cc/min = 0.001s * 1100cc/60s = 0.0183... cc/s

Two 550cc injectors AND two 850cc injectors
--------------------------------------------
0.3928ms * 2800cc/min = 0.0003928s * 2800cc/60s = 0.0183... cc/s

Steve thinks this happens when primary duty hits 70% of available on time, and then the switch back to primaries only occurs when all four injectors drop below 40% of available time. The numbers might not be exactly 70 and 40 but it is moot if they are not.

If you run the numbers from the Base Map * Fuel Correction Map and apply this staging algorithm, you will see that the millisecond values are reasonable and match up pretty well with the millisecond values used in standalone engine management systems, as well as the values observed with the stock ECU and a datalogger such as the EFI PMS. If you wanted to go through the trouble, you could translate maps between Haltech, Power FC, PMS, and just about any other engine management system.

So how do you determine when 70% of the available on time is exceeded?


*** Available Injector On Time ***

(1000 RPM) / (60 s/min) = 16 2/3 revs per second, so maximum on time at 1000 RPM would be 1000ms/16.67 = 60 ms.

RPM Available on time
---- -------------------
1000 60ms
2000 30ms
3000 20ms
4000 15ms
5000 12ms
6000 10ms
7000 8.5ms
8000 7.5ms

This is why we only run out of fuel at very high revs, it is relatively easy to need the injectors on 100% of the time. While down at low RPMS we have a huge amount of on time available per revolution.

Using this information and the Datalogit Base Map (IGNORING all corrections, including the Fuel Correction Map), we can figure that staging will occur at roughly:

RPM BOOST
----- -----
3000 15psi
4000 8psi
5000 3psi

This corresponds similarly to when we know the stock ECU stages. It'd be great if someone would enter into a spreadsheet the Base Map and the default Fuel Correction Map and come up with the exact location of all the staging points.


*** What we know and what we don't know ***

All my information about the PFC and its maps has come from Steve. The only thing I might question is the 70%/40% on/off thing, which if someone really needs to know, it could be tested and verified.

FWIW, I can't remember reading anything from Carlos on Scuderia Ciriani that I disagreed with.


*** What does this have to do with our recent List discussions? ***

The Power FC fires the primary and secondary injectors the same amount once all four are online. I believe the stock ECU does the same, as did Carlos. This can be verified by testing if someone wants to do it.

Wade

Wade

iTrader: (1)

Correction:

The switch back to primaries occurs when the switch would result in 40% duty or less on the primaries (not all 4!).

Wade

clayne

^^

MFilippello

iTrader: (2)

while searching for some info on a concern that I have, I came across this thread. It seams to focus on my situation so this is as good a place as any to ask my question.

I am currently running 775 pri and 1200 sec. The reason I chose these sizes is detailed above:775/(775+1200)=.3924 same as for the factory 550/850. I was previously running the 550/1200 setup but upgraded the primaries when I got my gt40r. Stangely enough, I had a stumble problem that occured at switch over with the old setup(550/1200) that was not there with the factory sizes and is now gone again with the 775/1200 setup. It would seem that there is some validity to the theory above. By keeping the ratio stock, I no longer have a stumble at switch over. What I am experiencing however is as follows.

When I get into boost, .8-1.0 bar for now. I am seeing really rich afr's in the low 10's/high 9's once the secondaries come on line. There is no stumble, the transition is smooth but at the higher boost levels the afr's go rich right after transition. Now I have plotted my afr's in the map watch and have started leaning out the cells were I am rich. ie: rows p17-18 columns n9-16. The concern is that I am getting to the point were my correction #'s will be less than those of the stock 2.01 base map.

Is this normal when running large injector setups on the PFC??? I thought that all you had to do to compensate for the larger injectors was to put the right #'s in the injector setup screen. That it did not require retuning the Fuel correction map. My setup is as follows


fr pri 71% .04
fr sec .08

rr pri 71% .04
rr sec .08

q pri 775
q sec 1200

pri/sec trans 50%
sec trans ms 1.5000


everything else is the same and so is my base map. I do not have a problem lowering the values in the fuel correction map to below those of the base map, it just seems odd that this is the case. Anybody else experience the same thing? or do I have some other problem that is causing me to be overly rich that has nothing to do with these tunning issues?

Mike