I copied VE tables (and others) from an ANTY chip from a '7747 ECM and use them in an '8746 ECM (started with an ANLU chip).

I am only bench testing the tune in the '8746 right now on the kitchen table with MAP at 0 in Hg (full atmo kPa), TPS at WOT and CTS at room temperature. I generate the ignition pulses from an FGen.

The problem is that the injectors go constant at around 3500 rpm (watching on a scope).

The BPW constant is set at 120, as calculated from the InjectorSizing.xls Excel sheet.

Is the programming between the '7747 and '8746 so different that I cannot simply directly copy the ANTY VE tables? Or is it something else?

Thanks in advance for any insight/advice.
Lou

 

What injector size are you using? Did you perform the exact same test with the 747?

It's not surprising for TBI to go static at lower RPMs at colder temperatures and WOT - They're already borderline when warm at high RPM, then add ~20% open loop enrichment, and another 10% of extra ECT WOT (on top of the normal 20%) from the room temperature ECT, and you lose 30% of the RPM range before static.

 

The injectors will be the stock 454 units (80 lb/hr) running at 15+ psi.
But right now I don't have any on the system, it's on the kitchen table, being bench tested first, before it goes on the engine.

Never had the '7747 ECM, I just have the '8746 (that I have and know how to tune in the first place).

As I said, I calculated the injector BPW constant using the Excel spreadsheet that circulates here (InjectorSizing.xls, I believe). For the 460 displacement, 15 psi pressure, 85% DC, .45 BSFC. The sheet gave me a new BPW constant at 120. That sounds pretty reasonable but maxes out the injectors at 3500rpm. Same with BPW set at 100, the rpm max might be just 100 higher.

I'll try it tonight with CTS set at operating temps, i.e. 180 degrees. 30% more fuel seems like a lot of enrichment. If that's the reason for my problem, I should see about 75% DC then.

Funny coincidence, the engine redline is 5000 rpm and with your words "and you lose 30% of the RPM range before static", that's 1500 rpm less .... which brings us down to 3500 rpm ... sounds very close.

I'll see what I can do tonight with a "warmer CTS".

Lou

 

For enough fuel at 5K RPM you will need enough injector for a BPC of about 90.

BPC: 90
IAT: 82°C
MAP: 100
AFR: 12.8
VE%: 100
BLM: 128
Bias: 396usec
RPM: 5000

5.127 msec, 85 DC% (8 cyl)

Note that I used 100% VE, 100 KPa MAP, and an operating temperature of 180 F. Also the injector bias is set to 0.396 msec.

Which puts the fuel pressure at 30 psi for 80.5#/hr injectors and 460 ci of displacement.

Note that this is just a rough-in valve. At 5K RPM I doubt the VE% will be 100. And the MAP KPa will be falling off due to intake restrictions. OTOH, the injector bias is likely going to be higher, which cuts into the actual 'fuel' time.

RBob.

 

Alright, I'll drop the BPW to 90 and change the injector bias and see how much that helps, thank you.

Also, how did GM ever fuel their 454 with 80# injectors at 9psi if I need 30psi to run the 460? That sounds a little off. Or do they not spin it too fast? Is it a 3000 rpm engine?

Lou

 

They are low RPM motors. By 4K RPM they were done. Other reasons include a higher operating temperature, hot intake systems (water jackets and thermo-vac pre-heat, small intake runners, and so on.

This example would be closer to what a stock 454 is:

BPC: 135
IAT: 100°C
MAP: 95
AFR: 12.8
VE%: 90
BLM: 128
Bias: 396usec
RPM: 4000

6.164 msec, 82 DC% (8 cyl)

The IAT being the CTS in this case.

Note that the values I gave in the previous post is just an example. If you run the 80.5#/hr injectors at 30 psi the injector bias will probably need to be north of 750 usec.

RBob.

 

The 460 has a big cam and made about 340hp with a carb on it.
I plugged in some data into the Excel sheet for calculation of necessary fuel pressure and came up to 15psi to support 350hp (at 93 lb/hr).
Is that too farfetched? Do I really need to further double the pressure?

I am attaching the XLS sheet (in a ZIP file). Is it incorrect/imprecise? Comments? Thoughts?

Thanks a lot for all y'all's input.
Lou

 

Yeah, if assuming .45 lb/hp hr, 85% max DC, 350hp should come out to 93lb/hr required.

My calcs are showing VE at about 66 to 73% depending on the RPM of peak hp (5000 vs 4500 RPM, at 12.5:1, .45 BSFC). If the BSFC is worse than .45, say .5, the VE could be up near 81% (if peak power at 4500 RPM).

I bet that thing is going to get pretty choked up on the small throttle bores.

If it were me, I'd try 18psi first, re-calc the injector constant, put 80% VE at peak power RPM, 95% at peak torque (can that thing make 480 ft-lbs?) and give it a go. Just make sure your fuel pump can keep up - if you're running an un-referenced regulator, it shouldn't drop at high RPM WOT (with room to spare).

 

Alright, the BWP constant of 90 helped and the injectors max out now just at the redline (5k).

However, the injector bias had a very adverse effect. I set it to 396 microseconds and the injectors went static at 250 rpm. I'll say it right now, I've never used the bias constant so I don't know much about it but 396 microseconds sure is wreaking havoc on the injection timing. Should I just leave the bias at 0 as it was before?

Thanks
Lou

 

I'd say that the ECU or XDF file is wrong. The bias is a two byte value and should never have been zero. If 396 usec is entered as the high byte the ECM sees it as 101,376 usec (101 msec).

RBob.

 

Yeah, the high-byte/low-byte situation would make sense (even mathematically, it would explain the static injectors above 250rpm).

Would you happen to know the right offset of the bias info? I'll check the XDF file.

Thanks
Lou

 

The bias is at location $2DA for 2 bytes.

This is the same area that has the min/max hyst PW's for sync2async and back, along with the min & max async PW's.

RBob.

 

I think you might be looking at the $42 bin format.

The $61 format puts the injector PW bias at 0x267 with formula 15.26*X.
The stock ANLU bin has value 0x001A there which corresponds exactly to 396us.

Ok, FOUND IT! The XDF file didn't specify that it is a two-byte constant, that's why it was pulling in only the first byte (which was zero originally) and that's why it was maxing out the injectors because it was setting the high byte, meaning that whatever I set it to was multiplied by 256 by the ECM.

Alright, I'll submit the corrected XDF to Mark to post on his TunerPro webpage.

Thanks for the help!
Lou

 

That I was, too much on the mind and definitely a post reading comprehension issue. Says '8746 right there in the subject line.

Check the other PW value setups in the XDF. Those are also two byte parameters.

Welcome.

RBob.

 

Will do. It makes sense, all the PW constants are store at an odd address.

Is there a reference document where I can find all the data formatting info? That way, I can look over the whole XDF and see what else might be off.

Thanks
Lou

 

Or let me ask a simpler question - does the $61 format use two bytes for any other fields besides the following?
- Prom ID
- Checksum
- Hwymode timer
- VATS frequencies
- PW constants

Lou

 

There is an older hac of the 8746 over on the diy-efi site:

http://diy-efi.org/gmecm/ecm_info/1228746/

A bunch has changed since then. I thought that a newer version of the hac was posted here on TGO. But I can't find it.

A parameter with an FCB is a single byte. A FDB defines a two byte parameter.

RBob.