CD Ignition Interface for ZX750E1 Page 1 of 8
How to use an aftermarket CD (Capacitive Discharge) ignition system on the ZX750E.
Art MacCarley Nipomo, California, USA February 2010.
Background
If you are only interested in the solution, not the background and explanation, please jump to the last section of this posting.
While aftermarket electronic ignition systems and upgraded ignition coils are commonly used on the ZX750E, I could not find any case in which a Capacitive Discharge Ignition (CDI) system was used while retaining the stock ECU (Electronic Control Unit). With apologies to the experts on this forum that probably know everything I discuss below, this is for the benefit of those that, like myself, that had to figure it all out experimentally and come up with a solution. My personal motivation to use an ultimate ignition system was my conversion to methanol, which misfires and runs rough until fully warmed up. The higher energy output and multi-fire features of a CD system could possibly improve this. A typical inductive ignition system delivers about 100 mJ per spark, and electronic “points” and improved coils alone can only marginally improve upon this. A CD system can deliver theoretically much greater ignition energy, limited only by the size of the discharge capacitor, the charging voltage, and ultimately the internal breakdown voltage of the ignition coil. The output of the ARC-II is specified as 189mJ using a 500V charge voltage.
My experience is based upon using the Dynatek Arc-II CD ignition system on my 1984 ZX-750E1. This CDI is advertised as having “the highest spark energy of any CDI on the market”. It works fine. The problems dealt with here have nothing to do with it; they are related to the way that the injection system is triggered via the stock ignition system on the ZX750E. The EFI (Electronic Fuel Injection) system on the ZX750E is reflective of the state-of-the-art in automotive technology in the 80’s, but I haven’t done CDI conversions for other EFI vehicles of this vintage to be able to say whether this discussion applies to anything other than the ZX750E. And I’m still evaluating how much of an improvement was gained, since I’m still playing with other engine variables like the AF ratio schedule and valve timing that was altered slightly when I increased the compression ratio.
I have waited six months to post this, so I could be sure that the solution I came up with had been fully tested and refined before recommending that anyone else try it. The simple interface circuit I suggest below is in its second generation, with additional circuit protection components added to make its possible range of applications more general. I have now accumulated about 2,000 miles with the Arc-II and my interface circuit, so I think it’s safe to share the info with others now.
How the Injection System on the ZX750E is Triggered
The injection system on the ZX750E is triggered by the output of the electronic ignition system, using a voltage detector circuit that will not trigger an injection event unless it detects the presence of high voltage inductive spikes on the primary terminals of both ignition coils. It will NOT trigger directly from the output of the electronic ignition module without both ignition coils connected to it. Each ignition coil fires two cylinders simultaneously and is triggered once per 360 degree revolution (the second firing is between the exhaust and intake cycles on the other cylinder, which is benign). The injection system, however, injects only once every two crank revolutions (720 degrees), with all injectors strobed together at the same time. This provides one injection per intake cycle, although the alignment of the injection with the firing cycle is different for every cylinder. Since the injection system requires inputs from both ignition coils to fire at this 720 degree rate, it can only be concluded that the injection system “counts to four” and injects once each fourth ignition pulse provided by both ignition inputs,. The photograph below shows a dual trace oscilloscope display of the injection pulses (lower trace) aligned with ignition from CD Ignition Interface for ZX750E1 Page 2 of 8
cylinders 1 & 4 (upper trace) at 1400 RPM. Note that cylinders 2 and 3 are ignited together between each of the ignition pulses from cylinders 1 and 4.
Approximately 1400 RPM
Ignition, Cyls 1 and 4.
Injection, all cylinders
For automobiles, the logic behind such a “no ignition, no injection” design would probably be protection of the 3-way catalyst from the possibility of raw fuel being passed to it. The unburned fuel oxidized in the catalyst could very quickly overheat and destroy the catalyst matrix, as well as contribute high HC emissions and poor fuel economy. This problem is minimized by making injection dependent upon verification of ignition, at least up to and including the primary side of the ignition coil (a plug wire could still be disconnected or plug fouled). Perhaps the same logic is appropriate for protection of the turbocharger on the ZX750E, or it was just a retained artifact of the automotive-based design of the injection system. This motorcycle was one of the first, if not the first, production motorcycle equipped with electronic fuel injection, so its major components are typical of automobile components.
Why CD Ignition Systems are a Problem with this Injection Triggering Scheme
For the stock electronic ignition system and all aftermarket electronic inductive (non-CD) ignition systems, the injection system works correctly since it will detect the presence of a single positive- going high voltage spike once every cylinder revolution. This, however, is not the case for an aftermarket CD ignition system for at least one of two reasons:
1. Most, possibly all, aftermarket CD systems include a multi-fire feature that is active particularly at low engine speeds. After the primary spark, there’s a series of additional sparks typically separated by one millisecond. On the ARC-II, this is varied with engine speed, but the delay range is not disclosed. Although higher in initial voltage and energy, the burn time of a CD spark is much shorter than that of an inductive spark. Less spark duration means less exposure to the fuel-air mixture, increasing the chance of a misfire. Lacking such a multi-fire feature, a CD ignited engine would be prone to misfire at lower engine speeds, CD Ignition Interface for ZX750E1 Page 3 of 8
and rough idling. However, this multi-fire feature creates a problem for the injection system on the ZX750, since each spark triggers an injection event.
2. The peak voltage of the primary ignition spike from a CD system is usually higher than an inductive spike, since it is not limited by the inductance of the ignition coil. Also, the polarity may be reversed depending on the way the DC/DC converter in the CD system was designed (it doesn’t matter for ignition purposes). Aware of this in advance, I chose to not take the risk of damaging my (irreplaceable) ECU, especially aware that it wouldn’t work correctly anyway due to (1) above. So I haven’t actually tested if the ECU would accept or survive triggering from the ignition coil, and don’t intend to try.
Solution
The ZX750 ECU needs a *single* clean inductive spike from the ignition coil primary (or equivalent), exceeding approximately 100 volts. I couldn’t determine the voltage threshold exactly, since I didn’t have a mechanism to experimentally vary the primary spark voltage until the point that the injection dropped out. The solution for triggering the injection system is a circuit that mimics the inductive spike effect of the stock ignition coil. A simple interface circuit will do it, one per coil, each consisting of less than $4.00 USD in electronic parts. The schematic of the circuit, parts list, and photos of its construction and installation, appear below. The schematic is actually the input layout for the B2Spice circuit simulation program if anyone would like to develop or study it further. The input to each circuit is just the output wire of the stock electronic ignition module that goes to each ignition coil (either green or black). These wires are also used as the trigger inputs to the CD ignition system, so I included a pass-through connection in each circuit to make this connection easy.
I also replaced the stock 2.4 ohm coils with Dyna DC9-1 0.7 ohm (blue) coils. For CD systems, the lower the primary coil resistance the better, and these are the lowest resistance coils they make.
The ignition wires are a modified 8 mm Accel suppression wire set for a four cylinder car. The plugs I used are NGK-BR9EIX with a 0.7mm gap (slightly wider than 0.5 mm stock). I found that, even with this high power ignition system, misfire will occur at high boost if the plug gaps are opened up any wider than this. This was disappointing; I would have expected the ability to use much larger spark gaps with a CDI.
The interface circuit is shown below. Two circuits are required, one for each ignition trigger, connected at each coil primary input (green or black wires at coils). CD Ignition Interface for ZX750E1 Page 4 of 8
Parts List, each interface circuit (two required):
Inductor: Miller 500mH (the 640 ohm resistor in series with the inductor above is just the measured internal resistance of the inductor itself, NOT an additional component) Otained from Mouser Electronics in the USA, Part Number 542-70F501-RC, $2.86 each USD http://www.mouser.com/ProductDetail/Bourns/70F501AF- RC/?qs=CmRDC9GpjEiVxKusHRedrg%3d%3d
Silicon diode: 1N4005 or 1N4006 (1A 600PIV min) Many different manufacturers, for example: http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=1N4005FSCT-ND
Zener diode: 12V, 1Watt. Many different manufacturers, for example: http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=BZX85C12-ND 46 cents.
Resistor: 510 ohm, ½ Watt. Many different manufacturers, for example: Junk or http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=510H-ND 5.8 cents
In terms of building and installing this stuff, it’s only a little more work than installing a conventional aftermarket electronic ignition system.
CD Ignition Interface for ZX750E1 500 mH inductor, with electrical Page 5 of tape 8 wrapped around it to protect the windings from abrasion
I used two 1K ¼ watt resistors in 12V 1W 1N4005 1A parallel since I didn’t have a zener diode 600 PIV single 500 ohm ½ watt resistor. diode
Large shrink wrap tubing with ends sealed by epoxy.
Alternatively, the entire circuit can be encapsulated in epoxy (never to be messed with again). CD Ignition Interface for ZX750E1 Page 6 of 8
Here are the two interface modules installed, next to the ignition coils. Coils are Dyna DC9-1 0.7 ohm.
And here is the complete installation with the Dynatek Arc-2 fitting neatly under the aft end of the fuel tank. Small pieces of urethane foam are placed below and above the unit to protect it from vibration and abrasion against the tank. CD Ignition Interface for ZX750E1 Page 7 of 8
Testing
I constructed a crude display to monitor each cylinder firing and each injection pulse while riding, so that I could watch for any missed injections or ignition firings. It is the cluster of eight small lights in the photograph below:
The upper row is a set of four NE-2 neon bulbs, one for each spark plug. One side of each neon bulb is chassis grounded. The other is connected to a small sleeve of copper tape wrapped around the corresponding plug wire, and then covered with electrical tape, as shown below. This does not interfere with the plug firing. The green LED on the left in the bottom row is connected through a 500 ohm resistor to the wires connected to any fuel injector (all injectors are wired in parallel). The other three LEDs in this row show unrelated stuff. This allows me to observe both ignition and fuel injection while riding, even during an acceleration run, so that I can correlate any perceived misfire with either the ignition, injection, both or neither. The box below this cluster with the bar graph and dual LED display shows the time-integrated injection pulse duration as a surrogate for fuel flow rate, and the oxygen content of the exhaust (used by my home-built fuel injection override computer for A/F ratio control).
CD Ignition Interface for ZX750E1 Page 8 of 8
This instrumentation helped be to get all the bugs out so that I could be sure that it worked as claimed under all conditions. Indeed, the first generation of the circuit experienced intermittent injection triggering due to an inadequate inductor. The version described here seems to be reliable, after quite a bit of daily driving with it (I commute daily on this motorcycle).
Was it worth the conversion to a CD ignition system?
I can’t say for sure yet. I had hoped to be able to use much larger plug gaps, possibly 1.0 mm (stock is 0.5 mm), to improve the ignition of methanol which vaporizes poorly, especially at colder temperatures. But I found that running plug gaps over about 0.7 mm would still lead to misfire at WOT, regardless of the higher power ignition. This was disappointing, and I have no definitive explanation yet. For what it’s worth, the expensive NGK BR9EIX “Iridium” plugs I am currently using don’t seem to support any larger gap than the stock B9ES plugs. I suspect the limitation might be related to the combustion chamber shape, but I would appreciate any advise about this from other’s experience.