rlm328

Don't run any resistors and when you burn up your DME you will have an excellent excuses to go and buy a stand alone system.

951_RS

I run 72 lbers with no resistors and have had no problems so far

Bri Bro

The injectors for the mid 80s 911 and 944 are the same, so they are 4.5 ohms, not 2.5 ohms. Link from paragon, look at the year and model info:
http://www.paragon-products.com/Fuel...80.150.158.htm

If you look at the calculation on post 14 you will see that four stock injectors use 8 amps so 6 injectors will use 12 amps. Well below the limit of the driver transistor which is 16 amps.

You can run without resistors The resistors just add a little margin to the DME drivers . This is more important for cars on a track that see extended periods of wide open driving.

Bri Bro

Now that is interesting, I was thinking the injector flow rate changed between the N/A and the turbo, and the impedance stayed the same. It doesn't. That brings up an interesting point, the 944 has always used lower impedance injectors then the 944T.
http://www.clarks-garage.com/shop-manual/fuel-12.htm

I am looking at a model 90, 1989 911 and the DME as a separate wire to each injectors, six in total. It looks a lot different then a 944 DME. I will see what the other workshop manual has in it.
http://www.cannell.co.uk/964_Worksho...g_Diagrams.pdf

Found it, page 73, 86-87 911
http://www.cannell.co.uk/911_72-89_W...p%20Manual.pdf
Driving six 2.5 ohm injectors would take 24 amps, the driver transistor spec I found has a 20 amp max
http://www.datasheetcatalog.org/data...1/218016_1.pdf
Inside the 911 DME. Note the injector driver is a T220 package, not a TO3 one would expect from a 20 amp NPN darlington driver
http://dorkiphus.net/porsche/showthread.php?t=16863
And you think we have questions, look at this 911 post.
http://forums.pelicanparts.com/porsc...jectors-8.html
And the series resistor R409 is a 0.1 Ohm and it has a feedback into the control IC. It could be a current limiting sense resistor for the control IC S400.


A test would be to put a variable load resistor (large wattage) on the injector outputs of the DME and vary it from 1 to 0.4 ohms and see if the driver IC (S400)current limits the output. One would have to trigger the S400 IC and measure the current pulse while varying the load. Another way would be to pull a injector plug off and place a load on it while measuring the current pulse. Anyone want to volunteer their DME/ wiring harness for a little destructive testing?

The only way the S400 driver could current limit the output transistor is to increase the voltage drop across the driver transistor. This in turn would increase the power dropped across the transistor. This causes more heat to be generated in the DME. This also means that current limiting the injector current also lowers the voltage going to each injector.

If the DME current limits the injector output, and it is looking like it does, the way to reduce the DME heating is to drop the current externally in a resistor so that the DME doesn't have to current limit injector current. So the circle turns.

Bri Bro

They look the same to me.

spyder348

Thanks for the responses. Wired in the 1.5 resistors with 16-18 gauge wire. Total resistance seems almost identical to the stock setup. i geuss they are the right resistors to go with. so for others info. i geuss upgrading to 55s or 72s with the common low imp. 2.4 2.5 resistance, 1.5 ohm 25w resistors are the way to go. no worrys about losing my dme or running to high of a duty cycle.

Bri Bro

This is the way I did the math for the external resistor.

Stock injector are 4.5 ohms. Four in parallel = 1.1 ohms.
Aftermarket injectors are 2.5 ohms. Four in parallel =0.62 ohms

So, you need to add 0.48 ohm to the 0.62 ohm to get back to the stock 1.1 ohms.
Since the external resistors are in parallel, it takes two 1 ohm resistors to get 0.5 ohms. So 0.5 added to 0.62 = 1.12 ohm. Real close to the goal of 1.1 ohms.

Using two 1.5 ohm in parallel give you 0.75 ohm. Added to 0.62 = 1.33 ohm. This is only 0.23 ohm higher so it should be close enough.