The information on this web page represents the efforts of a number of other people who went to significant lengths to reverse-engineer the original Bevel Ducati Voltage regulator. It's a way to gather the information in one place in (hopefully) a semi-coherent manner.
At 04:04 PM 10/11/06 -0700, Jordan wrote:Malcolm Moore's schematic is good, but it's not enough for a dangerous man with a soldering iron to do anything useful with. Namely, what is T1? A coil(s) of some type, but the configuration? Also Q1, Q2 values? ditto SCRs? JordanFrom: Malcolm Moore To: BevelHeads Date: Thu, 12 Oct 2006 03:42:33 -0700 Hi, I was intending to respond to your query on Tuesday under the "'69 450 desmo regulator" thread so I'll do both here. If anyones eyes glaze over, go to the next message :-) Firstly regarding my schematic on Willy's site: T1 is a transformer with four separate windings (coils), A B C & D. The dashed line is intended to indicate the windings are all on the same core. Where isolation is required, triggering SCRs via a trigger transformer was (is) a common industrial control technique in the days before optocouplers were readily available. To understand how the circuit works it is best to start with Q2. If you imagine Q1 to be erased from the drawing, Q2 is a simple oscillator. The coil in it's collector circuit (T1-B) is magnetically coupled to the coil in it's base circuit (T1-A). When power is first applied Q2 will start to conduct. This causes an increasing current to start flowing in coil T1-B and so a changing magnetic field is produced in the transformer core. The coil T1-A acts like a good little transformer secondary and a voltage is produced at it's terminals. The coil is connected such that this voltage acts to turn Q2 off. The now reducing current in T1-B then produces a voltage in T1-A of opposite polarity to before and this turns Q2 on again. The end result is Q2 rapidly turns on off on off etc. The other two coils on T1 are also secondaries and produce pulses of current that are applied to the gates of the SCRs. An SCR will start rectifying when it has BOTH the gate triggered AND the anode is +ve with respect to (wrt) the cathode. For each SCR this condition occurs on every alternate opposite half cycle. To trigger the SCR, the gate must be +ve wrt the cathode. This can be repetitive pulses or a continuous current (see below). On each half cycle, once the SCR has started conducting, it will continue doing so until the anode-cathode current falls to zero again. Therefore the gate pulses after the start of conduction are redundant in that half cycle, but still continue. The above has got current flowing into the battery. However when the battery is charged we need to prevent it overcharging by stopping the SCRs from turning on. If we stop the oscillator the trigger pulses will cease. This is easily achieved by clamping the base of Q2 to it's emitter and that is what Q1 does (the emitter has the arrow head, the base is the line at right angles to the thick bar and the collector is the other terminal). As the battery voltage rises zener Z1 will start to pass current, which will turn Q1 on. This clamps the base of Q2 to it's emitter and so charging ceases. After the battery voltage later falls, Q1 stops conducting and the oscillator starts again. I can't tell you the transistor type numbers in the one I dismantled because the metal cans of the transistors had lost their markings. There didn't seem to be any imprint left on the epoxy that encased them either. However, the site previously posted here with the Moto Morini regulator led me to the Dutch Moto Morini club at: http://www.motomoriniclub.nl/#English%20indexpage their technical page is http://www.motomoriniclub.nl/tech.html which at the bottom of the page includes a hand drawn schematic of their later type for bridge rectifiers. http://www.motomoriniclub.nl/latereregelaar.gif The control circuit is nearly identical to my drawing (although it is drawn as a mirror image). They quote Q1 as 2N2222A and Q2 as 2N3053. They also give information about the oscillator which indicates it ran at ~24kHz. Someone on Bevelheads once posted an equivalent type for the press in SCRs but I can't find it at present. The transformer core in my unit was also an FX2236, as in theirs. They also indicate the number of turns. There are only two SCRs, the other two rectifying devices in the bridge are diodes D1 & D2. I wouldn't recommend slavishly following the original control circuit. The alternative regulator at http://www.italoclassicbikes.de/downl/fs_downl.htm (which is also on the Dutch site as a pdf) achieves a similar result by sending continuous current into the SCR gates instead of pulses. It looks upside down because it uses pnp transistors. Once again beginning with T2, it is a simple switch which allows current to flow from battery positive via D2 & R7 (or D1 & R8) into the gate of the appropriate SCR (remember, to trigger, the gate has to be +ve wrt the cathode). To control the current, T1 can turn T2 off by clamping T2's base and emitter together. This happens when the zener ZD1 starts conducting which allows -ve current to flow via R4 and ZD1 into the base of T1. This turns T2 off and so charging ceases. When charging is again needed, T1 turns off & T2 turns on because current can again flow via R6 into the T2 base. For the half cycle when each SCR is conducting, it's cathode (terminals P4 & P5 on the Moto Morini drawing) is at about -ve 1V wrt it's anode, which is also chassis. That's alright because it actually slightly increases the voltage available to turn the SCR on. However in the other half cycle the SCR cannot conduct (because it's anode is -ve wrt it's cathode) and the voltage on the cathode (terminals P4 & P5) swings very +ve wrt to chassis. D1 & D2 are therefore needed to prevent each winding turning the opposite SCR on when T1/T2 don't desire that to happen. Whether this circuit would scale down to 6V is debateable (I presume that is needed for the original request in the '69 450 desmo regulator thread). With the reduced voltage the value of R7 & R8 would need to be significantly reduced, and the zener value changed as well as probably the R1, R3, R4 voltage divider. Regards Malcolm.
The Ducati 200W units are 2 lead single phase and will work just fine with your regulators, but after market regulators do not work well on the earlier 150W 3 lead alternators fitted to 750 roundcases and a few early 860 squarecases.
This is because the 3 lead alternator is NOT 3 phase. Ducati were real cheapskates, and it was 1970 when SCRs were expensive, so they designed a regulator with two SCRs, cathode to case mounted directly on the heat sink. To feed this they designed an alternator with a single winding plus centre tap. The centre tap goes to battery +ve and the two ends of the winding then get grounded alternately through the regulator. Dumb design, half wave rectified, but the regulator was cheap!
If you connect the three leads to a 3 phase regulator then the red centre tap lead will always be at a voltage half way between the 2 yellows - and will contribute sod all.
Connected to a 2 phase regulator the 2 yellows will give all they've got, which is fine at idle and up to a couple of thousand rpm, but they hit a limit because the resistance of the windings is way too high. The measurements I've taken show the circuit will not run ignition and 55W headlight at ANY rpm (I took it to 7000).
a) replace the regulator with a rebuilt one of the same configuration (Syd's cycles do one), or
b) use 2 diodes from a bridge to ground as a rectifier and a zener to regulate
c) invert the polarity of one half of the winding and use the two halves in parallel rather than series into any 2 phase regulator (I use a rectifier/zener and it's arguable that the 2 windings won't be completely in phase so a 3 phase unit would be a better choice), or
d) replace the alternator with a later (ST2) unit entirely.
Option c) runs a 100W headlight and ignition quite happily - and you can see at night!
Bottom line: do not bother with after market regulators on a roundcase with a stock alternator - it WILL be worse than stock and it isn't the fault of the regulator it's just a weird way to wire an alternator that no rational regulator designer would have thought of doing in the last 25 years!
Testing and tuning up my '74 GT's charging system revealed the following.
No load alternator output is about 32 to 40vac
Charge balance should be about 1500 to 2000 rpm with ignition only (depending on type of coils) and about 3000 with ignition and lights. That's about as good as it gets with stock rotor and stator.
Measure balance with a low resistance resistor ( about .02 ohms but actual value is not important unless current measurements are calculated) in series with the positive battery wire. Connect digital voltmeter across the resistor. Balance is at zero. Change current can be calculated. See this site if you've misplaced your Electronics 101 text book. Electronics Info Page
The SCR diodes in the regulator can fail for a variety of reasons. Age, vibration, etc. These devices were somewhat exotic 25 years ago. They can be tested for leakage / open / short with an ohm meter. The PDF is here: SCR documentation
I found one of mine "leaky" and replaced both. A direct replacement is NTE 5514. The PDF spec sheet is here: NTE 5514 Spec Sheet. Price is about $6.00 each. Go to main page to find a distributor near you.
Drive the old SCR's out with a flat punch. I took the opportunity to bead blast the entire cast aluminum box to remove all oxide and mineral deposit buildup. ( over the years the box got wet and stayed wet.) Install the new SCR's by warming up the box to relieve the interference fit. Gently push in the new SCR's. I used a appropriate fitting socket to not bend the contacts. Re-wire and refresh all solder connections. Scotch-Bright all the slip connectors.
The SCR trigger circuit is completely encapsulated so there's no repair of it. I'd like to see the circuit diagram because that's the only thing we can't diddle with to improve charging. My hunch is that if the trigger point were lowered, charging may improve some. Of course the rotor magnets are not strong like more modern applications. I looked into re-magnetizing but in the end I don't think that would yield the desired results. The magnetic material used just doesn't have enough Gauss. ( I think that's correct term)
Hope this helps.
Best Regards from Texas, Joe
I'm currently using the stock Ducati regulator with only the ST2 rotor on my Sport. I've got about 3000 miles on the system with no indication of problems. The original regulator is an interrupter (or series type of regulator) as can be seen from the schematic elsewhere in this article.
The interrupter type of regulator operates as a switch to apply the rectified alternator voltage to the battery in pulses only when the battery voltage drops low enough to require it, and effectively opens the switch so the alternator runs with no load when the battery is fully charged.
Most modern motorcycles (from 1990 on) use a shunt type of regulator. The shunt regulator operates as a variable load, and dumps current through this load to keep the rectified voltage down to a tolerable level. The shunt regulator converts the excess voltage & current directly to heat, and requires a good heat sink to dissipate the heat without damage to the regulator. This arrangement puts more strain on the stator windings at higher RPM, since they're flowing current continuously, and always heating.
The original series regulator is a fairly effective regulator for the system, but does suffer from lower charging current at low RPM. However, it doesn't convert the excess voltage and current to heat, so it doesn't place a heating load on the stator coils, nor does it require as much heat sinking.
The downside of the shunt regulator is that it will result in higher open circuit voltages at the stator coil terminals, which could lead to insulation breakdown in the stator. It can also generate microsecond-long voltage spikes on the DC side of the system, which could be a problem if you're running any modern electronic devices (such as a cell phone or radar detector) which are not very tolerant over voltage spikes.
Resistive loads, such as an electric vest or higher power headlights, aren't normally affected by these spikes. In the days of points and coil ignition systems, power supply voltage spikes were never an issue. However, once micro-processor controlled fuel injectors were introduced, the power supply spikes could easily kill or reset the fuel injection computer... Thus, the alternator reliability now suffers to enhance the EFI computer reliability.
The original Rita and Dyna after market electronic ignitions were designed to run with the original charging system and are very tolerant of the voltage spikes.
I owe thanks to the following individuals from the Bevelheads List who provided the information presented here over the years.
Malcolm Moore - for the reverse engineered Regulator Schematic. James de Raeve - for information on alternator capacity. Joe Tokarz - for discussions and information on re-magnetizing rotors. Rene & Sue Waters - for maintaining the ducatimeccanica website! Gene Rankin, et al - for maintaining the Bevelheads list!
I take no responsibility for anything you might or might not do as result of information you find on or through this website... Anything you may do as a result of viewing this website is your own responsibility. I merely provide this information to document what others have done!