I have found a place that builds regulator/rectirier units in the U.S. and they say they are way stronger and less susceptible to faults than origainals...and when I asked they would make them so that failure was covered and not smoke in cockpit...now how good is that...I'm working with them as we speak for my RV regtulator...and they have the Ducati in for inspection...will keep all posted as I know everyone would like a improved unit over Ducati...
Older –
If you are having a company custom make a regulator for you I would suggest you have a few small details made in your favor.
First, make sure the company you are working with realizes that the regulator also needs to have built in rectification that can handle inputs as high as 240 volts AC … see paragraph 20 in the following reference:
http://www.rotax-owner.com/pdf/UNDERSTANDING%20THE%20914%20ROTAX.pdf
The total heat sinking capacity of the Ducati regulator is IMHO … anemic. I would have the regulator’s total current capacity designed to be higher than the Ducati regulator. As an example, the Silent Hektik regulator is rated for 47 amps. That does NOT mean the Rotax engine will supply 47 amps because it won’t … according to the Rotax documentation the internal electrical system on the ROTAX 912 only puts out 18 amps, so that is the max per Rotax. However, having the extra current capability worked into the design means the regulator will be designed to dissipate heat easier.
The Ducati regulator uses a potting resin on the base of the unit which does not make an even contact with the mounting surface. I would ask for the regulator to have a smooth aluminum bottom so it can dissipate heat through the bottom mounting surface as well as from the cooling fins … here again like the Silent Hektik regulator which has both a smooth bottom and well over 2X the cooling fin surface area compared to the Ducati unit.
I would also be curious to ask your manufacturer about the feasibility of soldering the high current devices with silver bearing solder which has a somewhat higher melting point. Some of the Ducati regulator photos looked to me as though the circuits were getting hot enough to melt regular solder.
As a side note: It has been suggested vibrations may also be a factor. I purchased high temperature silicone standoffs which I was going to use by piggy backing the regulator on a large heat sink and suspending the assembly on the silicone standoffs. But have not tried them out because I'm in the process of moving the regulator into the radiator's air duct and off the firewall shelf so don't think vibration will be as much of a factor at the new location compared to the firewall shelf or instrument panel base.
Good luck on your endeavors and keep us posted,