Hi All-

I recently installed a 123Ignition dizzy in my '73 2.0. It apparently can be installed in any of 4 orientations, in 90 degree increments and operate properly when wired to the plugs appropriately.

The d-jet version of the 123 dizzy also has two wires that are used to time the two groups of fuel injection pulses (cyl 1+4 and 3+2). These wires can be connected to the ECU in either polarity, and the "correct" polarity to time/phase the injection pulses seems to depend upon both the distributor orientation, and the wire polarity. Instructions from 123 Ignition show a connector sketch that "suggests" a particular polarity, but I think the correct one also depends on the chosen dizzy orientation, as well as the firing order. (The dizzy can be used in non-914 applications. E.g., with a straight 4.)

I'm an EE. I've been through Paul Anders' ( @pbanders ) very helpful writeups on the ECU and d-jet operation. I've got d-jet schematics. And, I've got a generic (Volvo?) chart purporting to show the injection timing WRT the cylinders (which seems to have a lot of discrepancies with a 914 setup with different firing order.) I've also looked at the injection pulse/spark timing with an oscilloscope (with a high-v probe).

It looks like the orientation and polarity can be adjusted to inject fuel at any of the following times:

1) Inject cylinders 1 and 3 during intake stroke and cylinders 4 and 2 during exhaust stroke.

2) Inject cylinders 1 and 3 during power stroke and cylinders 4 and 2 during compression stroke.

3) Inject cylinders 1 and 3 during exhaust stroke and cylinders 4 and 2 during power stroke.

4) Inject cylinders 1 and 3 during compression stroke and cylinders 4 and 2 during intake stroke.

I think the choice of the 4 options above would affect the dwell time of the fuel charge in the intake runners and might also cause some raw fuel to be sent out the exhaust if a pulse occurs during valve overlap.

Which of the 4 configurations is correct? Is my analysis correct, or did I miss something? Any other advice you can give me?

Thanks for your help.

Cheers.

I installed the 123 with stock D jet as shown per the installation instructions and it performed flawlessly. I indexed and installed it as was the stock dizzy...meaning clocked the same way.

I have since sold the car so I never really experimented with it in the aspect that you are suggesting.


Good luck.

Thanks. I think the #1 plug wire on the OEM dizzy is supposed to be just clockwise of the vacuum dashpot and is in the 11 o'clock position in the engine bay when looking down at it from the rear of the car. But, there is no dashpot on the 123. How did you clock it to match OEM? I guess you could align the retaining straps, but that still gives you two orientation choices.

Also, I think it will run "good enough" with any orientation and polarity. But sadly, I'm hopelessly obsessive -- I'm looking for "best", and wondering how it was designed to work.

From a design standpoint, I think you'd want to avoid injecting during the intake stroke, as (I think) that would create asymmetry among the cylinders, and also inject fuel during the valve overlap time. I dunno. Maybe it doesn't matter...

If I recall, DJet squirts a 50% charge in two batches paired as cyls 1+3 and 2+4, which is different than you indicate, but it may not make any functional difference. Lets start there.

https://bowlsby.net/914/Classic/zTN_Man04.pdf See page 7

Thanks, Jeff! I hadn't seen that doc yet.

Tracing the cable numbers 3-6 on page 226 (aka p.8) of your doc to the corresponding pins on the ECU schematic *looks* like it fires 1+4 as a group, and 2+3 as a group. But, it's entirely possible that I'm screwing this up somehow. Also, with a scope, I've been able to confirm a single fuel injection pulse per distributor revolution for each group of cylinders. (It's the timing between the FI pulse and the spark for a given cylinder that I'm wrestling with at the moment.)

Any guesses?

Cheers!

PS -- a second thanks for your "tech notebook". It's been really valuable!

Jeff, after further reading of your doc, I came across this jewel on p. 225:

"Fuel injection valves are parallel connected into groups of two (group 1 = cylinders 1+4; group 2 = cylinders 2+3). Both fuel injection valves of one group inject fuel at the same time. The injection valves of cylinders 1 and 3 inject fuel past the open intake valves, whilst the injection valves of cylinders 2 and 4 inject onto the still closed intake valves while the exhaust gases are being forced out."

This would seem to correspond to scenario #1 in my original post. I can easily configure the 123 dizzy to match this.

Thanks for your help!

Sounds good. Hence the need for two 50% fuel charges per combustion cycle.

Added plus with the electronic dizzy is the elimination of the trigger points....and yes with regard to the cap and plug position. On the stock dizzy the vaccum can was in an awkward position and because the diaphram had failed and replacement parts my reasoning to use the 123.

Good luck

I'm not sure if we're on the same page here or not, depending on what you mean by "fuel charge" and "combustion cycle".

Each *cylinder* gets 100% of its fuel charge (25% of the total) in a single fuel pulse, but the timing is different for each cylinder (1 and 3 during intake; 2 and 4 during exhaust).

Through 720 degrees of crank rotation, there will be a total of 2 fuel pulses (one per "group"), with each pulse delivering 50% of the total charge needed by all cylinders during that 720 degrees. (I think this is what you meant by"two 50% fuel charges per combustion cycle"...)

Do I understand you correctly?

Cheers.

Thanks! Yeah, the 123 dizzy looks pretty sweet. It did force me to choose between vac advance and vac retard, though, as it doesn't have the ability to do both like the OEM one. (I chose vac retard -- profile 'A' -- as otherwise the timing was too advanced at idle to get the idle speed down.)

Cheers.

The trigger points contain 2 switches which each initiate an injection pulse 180 degrees dizzy rotation apart, each switch firing a bank of two paired injectors.

???

Should be using ported vaccum for the advance and no retard or manifold vacuum. I.m wary of your post.

Sure. On a '73 2.0 d-jet 914, the pulses are the "B" ones in that diagram. One pulse per cylinder per 720-degree I-C-P-E cycle. (with 180 degrees of dizzy rotation corresponding to 360 degrees of crank rotation.)

Do you agree?

Thanks.

Bosch D-Jetronic is "batch injection".

Edit: look at the attached file, traces AA-AB, 5-10, upper right corner. Injectors are batched together and two are fired simultaneously. "Batch injection".

Click to view attachment

On a '73 2.0 d-jet 914, the throttle body has two vac ports -- one for advance at low load, and one for idle retard. (See ports "A" and "B" on the diagram below.) The OEM dizzy (039 905 205) also has two corresponding nipples.

The 123 dizzy has a single vac port that can be used either for vac advance (e.g., profile "1") *or* vac retard (e.g., profile "A"). You have to choose.

With profile #1, setting the centrifugal advance to give 27 degrees total (w/ vac hose plugged off) at 3500 rpm gives you an total advance of 5 degrees at idle. I couldn't get my idle speed below 1500 rpm with that much advance, even with the idle screw on the throttle body full clockwise. (Granted a vacuum leak could theoretically be the root cause of the high idle, but I've checked/replaced virtually *everything* and am 99% sure that I don't have a vac leak.)

Using profile "A", there is no vacuum advance, but with the throttle plate closed at idle, there is enough retard at idle to get the idle speed to ~1000 rpm. This should (hopefully) mimic the total advance/retard of the OEM dizzy at idle.

If I'm missing something, or have screwed up somehow, please help.

Thanks.


Click to view attachment Click to view attachment

Thanks! That's consistent with my analysis. On the upper right of the schematic:

Pin 3 controls injector for cylinder 1
Pin 4 controls injector for cylinder 4
Pin 5 controls injector for cylinder 3
Pin 6 controls injector for cylinder 2

(Though I may have pins 5 and 6 reversed...)

Cheers.

Yes....I believe you are missing something.


First.....I.ll assume that you have the -4-R-V-IE.

If this is the case then the advance curves are set by a small dip switch. The 123 curve for each motor and or distributor is listed on their website or in the little book that comes with the dizzy. Don.t know why you are posting the curves for the old dizzy numbers cause they are cross referenced by the 123 matrix to match the advance profile.

Second....I.ll assume you have a stock cam. If not that can cause the high idle.

Lastly post a pic for which vaccum port you used on the TB...or confirm from the cross section.

If all of the above is correct then you idle mixture is wrong...the static timing is off....or most likely...you have a vacuum leak.

Yep. -4-R-V-IE, with preset profiles set with the rotary DIP switch.

I currently have the retard port on the throttle body (connection "2" or "B" on the diagram on post #15) connected to the 123 dizzy.

The 123 instructions list two different profiles for the 039 905 205 (AKA Bosch 0 231 174 009) distributor -- "1" with vac advance, and "A" with vac retard.

I posted the original dizzy advance/retard chart to show the original values. From the line in the chart for the -205 dizzy, at idle there should be no (or at least negligible) centrifugal advance, no vac advance (the advance port is "disabled" by the closed throttle plate), and about -10 degrees of vac retard. Under these conditions, a factory advance setting of +27 degrees at 3500 rpm on the original distributor corresponds to a net -5 degree retard at 1000 rpm idle w/ throttle plate closed. I believe that I've reproduced that with the 123 "A" profile and vac retard.

Thanks for your help. Much appreciated.

I have a 123dizzy and wasn’t able to get idle rpms below 1000 using either the advance of retard settings and at least one other person here had the same experience, IIRC. If you can’t get below 1500, something else is wrong. How was your idle before installing the 123?
Have you checked the throttle body air bleed screw? Sounds like a vacuum leak or timing is off.

Thanks for your comment.

The whole idle/retard discussion is kind of tangential to the original FI phasing questions that I started the thread for.

Nevertheless, it's interesting, and may be helpful to somebody else down the road.

I *can* get it to idle just fine at around 1000 rpm with the 123 vac retard connected and profile "A". With profile "1", I couldn't get it below 1500 rpm. I figure that under idle conditions (no centrifugal advance, throttle plate closed) that profile "A" gives a net idle timing of -5 degrees (which matches OEM) and profile "1" gives a net timing of +5 degrees. I attribute the elevated idle with profile "1" to that difference in timing. But, I don't rule out that something else may be in play (e.g. an undetected vacuum leak as has been mentioned.)

FWIW, this is not a stock engine. It's a 2056 with a webcam 73 cam. But, I don't believe this is the primary cause of the high idle with the vac advance profile.

Cheers.

The ported vacuum is not "disabled". That is how the advance vacuum signal should work. At idle the port is just past the throttle plate on the atmosphere side so the signal is 0 Hg. As part load starts the vaccum rises advancing the timing signal. Conversely the retard port is manifold vaccum and is highest when the throttle plate is closed.

I.d make sure the dip switch matches your engine config and make sure everything is set up correctly.

If it is you will get a rock steady idle at 800 RPM.

Thanks!

I think we're on the same page. By "disabled", I meant that the advance port shows 0 in-Hg vacuum and so doesn't produce any advance at idle because it is above the throttle plate (as you said). I'm aware of how the two ports function WRT throttle position.

Cheers.

FWIW, this is not a stock engine. It's a 2056 with a webcam 73 cam. But, I don't believe this is the primary cause of the high idle with the vac advance profile.

You want to use vac advance with this engine. Setting #2 the engine likes to idle between 950 and 1000 RPM with 10° to 12° BTDC at that engine speed. If the RPM's are to high you probably have a vacuum leak or running to rich. Did you try to lean out the idle mixture?

Thanks, I'll try #2 out. I have the Tangerine kit installed to adjust the MPS, and I've tuned the AFR with an wideband O2 meter. (Yes, I know about the idle mixture knob on the ECU. And, I have the TPS adjusted properly...)

In a bit of an update for anybody searching/reading this in the future, I contacted 123 to find the start and end pressures for the "A" (retard) profile as that information is missing from their instructions for all of the retard profiles.

Their response: "For vacuum retard, the customer needs to rotate the ignition by 10 degrees to retard the ignition. The ignition has the following vacuum map: The retard starts at 80mmhg (3.2 in-Hg) and ends at 150mmhg (5.9 in-Hg) with 10 degrees."

When connected to the retard port on the throttle body, I think this should operate the way I want -- retard 10 degrees at high vacuum at idle, and transition to zero advance/retard as the throttle plate opens and the throttle plate moves below the retard port, causing low to zero vacuum at that port.

But, I wonder if there is some venturi effect at the retard port under part-throttle conditions. If so, the 123 "A" profile may be causing some undesired vacuum retard above idle, but under low load.

Nevertheless, I think that those here chanting "Vacuum Leak" were right after all. I've replaced every hose and gasket, and removed every component to pull a vac on it and test. So far I haven't been able to find it. I'm gonna hook it up to a smoke machine next week and see if I missed something. (Maybe I have a hairline crack in the intake plenum or one of the peened fittings there is leaking...)

My plan now is to find a leak, then switch back to profile #1 (or #2) with vac advance.

Thank you to all who responded.

Cheers.

[quote name='Frank S' date='Jul 16 2020, 11:55 PM' post='2834494']
FWIW, this is not a stock engine. It's a 2056 with a webcam 73 cam. But, I don't believe this is the primary cause of the high idle with the vac advance profile.

You want to use vac advance with this engine. Setting #2 the engine likes to idle between 950 and 1000 RPM with 10° to 12° BTDC at that engine speed. If the RPM's are to high you probably have a vacuum leak or running to rich. Did you try to lean out the



And so you would ajust the static timing and run the vac advance.to the 123 and not manifold.vac


I asked if it was a stock engine a while ago.

Yes. Clearly.

When running an advance profile on the 123, I've connected it to the advance port on the throttle body; when running a retard profile, I've connected it to the retard port. Unused ports have been plugged in all cases.

Sorry if I didn't make this clear earlier. I thought it was understood.

THIS THREAD IS MASTER CLASS!

Update for anybody interested:

A special thanks to @JeffBowlsby ! Using his information, I've been able to set up the 123 pulse phasing to match the OEM design, and confirmed with an oscilloscope. It feels like there may be a slight difference in driveability depending on how the FI phasing is set up, but it may be entirely my imagination. The effect seems subtle, if it matters at all.

I did a smoke test to look for vacuum leaks, and it appears that everything is tight. So, I'm now suspicious that the PCV valve is worn and flowing more air than it should (even after cleaning), causing the high idle problem with the 123 "1" profile.

The OEM PCV valve is impossible to find (for me anyway...). And, I'm reluctant to just plumb the PCV to the airbox as was apparently done on later-year models. So, I'm currently in the process of adapting a modern Toyota PCV valve as a replacement. If the Toyota PCV works, I'll start a thread about it as it might be useful to others.

Cheers.

IMO, if there's a choice on the 123 between vacuum advance or retard, you've got to use the retard. If you don't, when you set the mechanical advance to the stock setting, there's too much advance to bring the idle down to 1000 rpm and the throttle plate closed, even with the bleed screw turned all the way in. I know this is a modified motor with a different cam profile, but I'm pretty sure it would still have the same issue.

Vacuum advance (according to the Bosch Blue Book) is required for low load conditions, where the A/F mixture burns more slowly. According to other references I've seen, this additional advance was incorporated not for better acceleration or transition, but to reduce emissions. Apparently, the improvement in emissions wasn't significant in actual practice, because in later 914's, the vacuum advance port wasn't present on the throttle body and the cell was capped off.

If the vacuum advance is really providing significant driveability effects, it should be a simple thing to test on either stock or modified motors where both the vacuum advance and retard ports are present on the throttle body, and the stock distributor is installed and both cells are functioning properly. Simply clamp the advance vacuum line with a fuel line clamp and see if it makes any difference when driving. I'll try this out with my car when I get a chance.