So what can a DW400 really do?

[engineermike]

This isn't my car, but a close friend's...but he decided to see just how far he could go on a single deadheaded DW400. The basic setup is a 2019 GT, M6, Whipple, 3.0" griptec pulley (~18 psi), and e85.

This is the Whipple-supplied DW400. Other mods were a 22v BAP, power wire upgrade, and DW95 injectors. It has the stock GDI system, but it was tuned to get the max from it. The DW95 injector data was borrowed from the 2018 CJ, which had them factory installed. The stock 2018 CJ data is shockingly close to ID1050x data. Also, the lift pressure was borrowed from the CJ as well and set to 55 psi. The key here, so far, has been tuning the lift pump PID feedback loop to control it through the stock FPDM using PWM functionality. He happens to be a control systems engineer and has spent a good deal of effort perfecting the feedback control of the pump.

In the log below you can see the lift pump duty cycle tops out around 47.2 x 2 = 94.4%, while the injector duty cycle hits a max of about 9.8 ms. The injectors are only at about 60% duty cycle in the traditional terms, but actually 80% in gen3 coyote terms due to the logic DC limits. In other words, there's still more on the table to account for colder ambient temps. Max flow of 730 lb/hr is not bad, especially on the stock fuel line, bucket, and supplied whipple rails.

Sponsored

 

[Angrey]

Wait, so you're getting 730 liters/hour out of a single DW400 at 55 psi?

Could you do the split calculations to see how much is being bled off to the siphon system? Cause 730 l/h is pretty significant power. I'm assuming that the motor isn't using all of that, whether it's going to the low side or high side, at 18 psi the ID fuel flow calculator puts engine consumption somewhere around 510 l/h.

 

[engineermike]

Hey @Angrey it’s 730 lb/hr, not l/hr.

 

[Pistol_91]

50% blend compared to 90% seems to help huh?

 

[engineermike]

50% blend compared to 90% seems to help huh?

With gdi blend, more is better up to about 90%. We are commanding 90% here but with soi, eoi, and pressure limitations 48% is all we could get out of the stock gdi system. Most e85 setups are in the 30’s blend range. This has ve and knock advantages, but in this case the advantage is that it takes some load off of the pfi injectors.

 

[Pistol_91]

With gdi blend, more is better up to about 90%. We are commanding 90% here but with soi, eoi, and pressure limitations 48% is all we could get out of the stock gdi system. Most e85 setups are in the 30’s blend range. This has ve and knock advantages, but in this case the advantage is that it takes some load off of the pfi injectors.

So theoretically with larger port injectors, the gdi blend should be closer to commanded 90%

 

[engineermike]

So theoretically with larger port injectors, the gdi blend should be closer to commanded 90%

No, the other way around.

Actually the gdi system is priority and has all sorts of limits. It attempts to follow commanded gdi blend but when it hits eoi or max angle limits, the remaining fuel requirement is sent to port. I always set the gdi blend to 90% at all loads over about .8. It will automatically reduce blend as it hits the limits. The bigger port injectors just allow more to be handled by the port side. Either way, the single dw400 still has to carry the load of both.

 

[Pistol_91]

No, the other way around.

Actually the gdi system is priority and has all sorts of limits. It attempts to follow commanded gdi blend but when it hits eoi or max angle limits, the remaining fuel requirement is sent to port. I always set the gdi blend to 90% at all loads over about .8. It will automatically reduce blend as it hits the limits. The bigger port injectors just allow more to be handled by the port side. Either way, the single dw400 still has to carry the load of both.

Got cha. Do we know of the longevity of this pump at 22v or is this testing still too new? On a side note, "you need a fuel system for e85" - every tuner. LOL

 

[engineermike]

Got cha. Do we know of the longevity of this pump at 22v or is this testing still too new?

DW has supposedly “approved” it for use at 20 volts. Technically, the 22v bap only actually supplies 21 volts to the pump due to apparent input current protections. We did bench test a dw400 at a true 22v for minutes at a time and it didn’t burn up. A stock pump burnt up in seconds at a true 22v. And one last data point, there’s an article from fordmuscle circa 2019 that shows where they tested it up to 22 volts.

On a side note, "you need a fuel system for e85" - every tuner. LOL

Makes their job easier though!

I honestly don’t know that many or any tuners out there can modify a PID control loop like a professional control systems engineer can. I know I sure can’t.

 

[ben1]

. Do we know of the longevity of this pump at 22v or is this testing still too new?

The DW400 in question has been in use for 3 years. First 16 months with no BAP input to the FPDM seeing 14.7V max. The next 11 months I had it on an 18V BAP with FPDM input seeing 18.3V when BAP kicked in. And the last 9 months I've been using a 22V BAP with 6awg power wire where the FPDM input sees 21.1V when the BAP kicks in. This is my daily driver so the DW400 has seen around 30,000 miles of use.

 

[Pistol_91]

The DW400 in question has been in use for 3 years. First 16 months with no BAP input to the FPDM seeing 14.7V max. The next 11 months I had it on an 18V BAP with FPDM input seeing 18.3V when BAP kicked in. And the last 9 months I've been using a 22V BAP with 6awg power wire where the FPDM input sees 21.1V when the BAP kicks in. This is my daily driver so the DW400 has seen around 30,000 miles of use.

So to say the least it's pretty reliable. Awesome info guys.

 

[SBR70.3]

Side note, David Deatch is based out of Oklahoma City and I had the opportunity to meet him at an HPDE event at Hallet Motor Speedway. He is a great guy and a damn good track driver!

 

[Det_Riot]

No, the other way around.

Actually the gdi system is priority and has all sorts of limits. It attempts to follow commanded gdi blend but when it hits eoi or max angle limits, the remaining fuel requirement is sent to port. I always set the gdi blend to 90% at all loads over about .8. It will automatically reduce blend as it hits the limits. The bigger port injectors just allow more to be handled by the port side. Either way, the single dw400 still has to carry the load of both.

This is interesting, maybe you can shed some light on this. Gen 3 car specifically, 130lb port injectors, di/pi blend set to 25%, over .8 load, over 4k rpm, we're seeing an internal system limitation where once the port duty cycle clips 70%, the ecu thinks it's running out of fuel and begins to blend back in the DI system and airfuel starts to run way lean.

So by the logic you're explaining, if we keep it around say 50%, and let the DI pump "max" first, then it will supplement port injector after that?

 

[engineermike]

This is interesting, maybe you can shed some light on this. Gen 3 car specifically, 130lb port injectors, di/pi blend set to 25%, over .8 load, over 4k rpm, we're seeing an internal system limitation where once the port duty cycle clips 70%, the ecu thinks it's running out of fuel and begins to blend back in the DI system and airfuel starts to run way lean.

I've not run across it increasing DI blend when PFI hits a limit, but it's possible. I wonder if HPT ECM 4235 "DI/PI Blend Warm Upper Limit" could stop it.

As for the port DC clip, I've spent a ton of time trying to get around that and never really conquered it. However, I did learn that there's a parameter defined by PCMTec called auF61009 "Switch to Enable Variable PFI Window Sizing" that can help. It doesn't do what the name implies, but I've found that switching it on will buy you up to 2 ms potential pulsewidth at high loads/speeds. You could at least get to around 80%. I've found that most Whipple calibrations have this switched on, but Roush and at least some stock tunes have it off by default. It's not defined in hptuners.

So by the logic you're explaining, if we keep it around say 50%, and let the DI pump "max" first, then it will supplement port injector after that?

That's what I would do, but my disclaimer is that I've not spent a great deal of time maximizing DI on cars at your power level. You could also consider lowering the EOI clip from a stock value of 100 to about 80, and increasing the DI duty cycle limits to 1.0 in both ECM 7306 and 7310. This will get you some more DI flow.

 

[WildHorse]

93E10 fuel with a specific gravity of approximately 0.761 / E85 fuel with a specific gravity of approximately 0.791
Percent Difference = ((SG_E85 - SG_93E10) / SG_93E10) × 100%
Percent Difference = ((0.791 - 0.761) / 0.761) × 100%
Percent Difference = (0.03 / 0.761) × 100%
Percent Difference = approximately 3.93%
So E85 is around 4% more dense than 93E10.
Since the specific gravity of E85 fuel is approx 4% higher than that of 93E10 fuel, a fuel pump would need to work about 4% harder to move the same volume of E85 fuel compared to 93E10 fuel.

Sponsored