I had occasion to review the fuel injection material the club has collected over the years.
As it turns out, a lot of it is internal engineering documentation of the efforts to get the first 12 cars to run right for a press release showing of the 300D FI in Miami. Chrysler had one hell of a time even with their engineers there to support the effort. The cars were barely drivable, but managed to get through the show OK. The result of the Miami show was a list from the engineers of things that required immediate correction to the FI system. Some of the problems were inherent in the design and some were troublesome FI parts. After the Miami effort, they fixed only what they could fix. The engineers concluded FI was not ready for production but management liked the idea of it and production continued.
I moved on to the 3 best test reports, called “Technical Reports” , especially #4511-33 , “A676 Bendix Fuel Injection System Evaluation”, dated 6-1-60. A significant takeaway in 1960 was that there was no net gain in output, despite some wild claims and published graphs early on, and that fuel economy suffered, even with highly tuned systems of closely matched parts. Chrysler allowed Bendix + or – 6% variation in the AF ratio (twice what carbs were allowed), and the FI units could not do that consistently. Note that this injection is 100% open loop, no feedback available. (O2 sensor, or even memory for that matter, was yet to be developed). On the other hand, carburetors are all open loop.
Chrysler had an initial order for 85 FI units with Bendix and a planned program for a second generation of it starting with #86. I do not think 85 were made.
The whole Bendix thing came from outside Chrysler and was aviation based. It had overrun an older internal program for a continuous flow FI that had been ongoing at Chrysler since 1955. The idea of FI was, surprisingly to reduce cost; apparently Carter carbs cost a lot even then.
This older Chrysler FI project was probably like the GM TBI. As the problems with the Bendix unit became apparent, this project was briefly resurrected in combination with new “Ram Manifolds” (late 1958) but it too, had too many problems “that had not been resolved”. All the #86 Bendix and later FI efforts and engineering projects were subsequently cancelled. “Lack of durability” was the main issue but there were also a few quite alarming specifics too (fire for one, on one Electrojector in Miami).
One report is a single page from “Service Division” –it outlines the main issues, for them, (more in a second), but ends with “complete replacement of the system , with dual 4 bbl carburetors is the only solution” for repeated complaints. (their (Service dept) take on it).
There is a hand written note about which parts (with numbers) come from Chrysler to change to 2x4 which are local, and who pays; that swap needs an all new fuel tank too! Says 6.5 hours to do the whole thing . (you gotta be kidding me). Some kind of filter or fuel injection related device was mounted in the trunk behind the back seat, had “to be removed”-- might be a clue to original fuel cars--(holes in floor there ). The replacement was in retrospect, wise. This was not a good thing to have on your car in 1958, despite the aura that has grown around it since then . An aura I frankly shared…less so right now!! Like some other things I have been involved with (Boston Computer Museum ,at one point) this belongs in a display in a museum,--an extremely significant thing, --essentially it was the correct approach , what we have today. . Not really suitable for a driven car in 58, however. .
Some of the major inherent design problems were the injectors themselves were too slow to close, initially taking 4 msec to shut; later , after much truly heroic effort, they got it down to 1.5, (note that the variable pulse width of the squirt was intended to be 1-5 msec or so—so hanging open for 4 is unusable, as is any drip or leak ) ---they were also very variable , individually,-- inconsistent, bringing on cylinder A/F matching issues; even worse, excessively rich cylinders would destroy the rings with raw gas and then that gas got into the oil. This caused rod bearing failures. The seal point or valve at the front of injector started as a nylon tip, they went to metal discs, then very exotic magnetic materials in the solenoid etc etc. trying to speed them up and make consistent. Heavier springs = it destroys itself by hammering. All this while gun at head about release date. And a leak sprayed the distributor (20psi pump) causing a bad fire. Then certain failures would result in a “hydraulic lock” of a cylinder full of gasoline. It tended to lean out at WOT, causing a rich setup at less than WOT . This was never resolved.
Obvious in the tone of some reports, that some not unexpected dissention developed among fans and foes of the whole thing. A statement is made somewhere in it that “management turned away from leading edge R&D things , due to other problems in 1958/59” , when industry apparently suffered a sales drop. So this decision to back away and recall was the right one . A+++ for effort, the D- for result does not take away the A+++ .
Interesting comparison, I had a 60 Vette with Rochester mechanical injection, it was actually trouble free for the ~ 4 years I had it, perhaps as experienced “carburetor only” guys had built it. No electronics. It leaked slightly at a seal at drive cable off distributor, a known but easily fixed issue . That is a continuous flow design with very short rams under that cover ….they must have been watching each other. Note that the 57 FI Vette , the first one, was apparently ok. . Finally I had a Bosch K Jetronic Mercedes, 450 SL another early injection, ~ 1978, kind of half and half, mechanical and electrical,. It never was tuned quite right. Baffled me and others. Lots of little electric add ons, on a mechanical injector , to cover problem areas, mostly around warm up / idle. .
Second Last--- *** There still is not an available wiring schematic of the Electrojector , which is what I was really after ; Jim B mentioned he might have one, but I cannot find his email to copy him on this. Others may be interested in this info too, feel free pass on.
I anyone has the schematic, I would greatly appreciate a look at it. (EE hat on) . Thoughts of a modern injector retrofit dance around.
Last; two papers are mentioned, an SAE paper in Dec 1957, (may be the same as the generic overview paper we have) And a “Chrysler Training Manual , August 1957” ; they mention “negative #8310 -36” as the schematic diagram. Sleuthing needed…….
Best regards, thank you, Gloria, --and soon back to you,
John Grady PE
PS, I like one quote from an obviously harried engineer, on the injector minimum time..: “Remember , very little fuel is needed at idle and low cruise, now it is spread among 8 places, instead of one carburetor jet…. (and injector won’t close fast)
PPS, it looks to me like pulse is no better than continuous injection or carbs for HP output. That basic assumption driving this idea was never really tested, before all this was started ; needed today for emissions…Not HP. 20,000 HP nitro fuel rail is continuous…And, Holleys outperform FI on dyno…due to bulk charge cooling --they think.