Problems Using A CDU with Trix N switch motors and possible solution

Hello Everybody. Looking for some advice from some of out Euro followers or those whose interests lie in DC electronics. I power most of my N gage Atlas Under table point motors using the Capacitor Discharge Unit describe by Pete Thorn in his 1974 book. It was built by my father withing a control panel he built for me for Christmas years ago. This circuit shown in the attached diagram works almost flawlessly with the Atlas Machines! I believe the current flows as shown but I am not sure. However due to my track design i need to use two Trix curved turnouts. Using the momentary push buttons in the circuit with the Trix motors connected to the CDU the behavior is extremely erratic. At first I though the pulse from the CDU was a bit too powerful, as the Trix points “bounced” off the stock rails, leading me to try some high Watt resistors in the return (black wire) installed at the terminal block. The operation was still erratic, and I noticed that the little red lever on the turnouts would actually “pull out” of the switch machine such that pushing the same color button actually caused the points to close in the other direction. Thinking that there must be some unknown electrical features of the Trix motors, I removed the motors and operated them from both the 17-20v DC and AC terminals on a MRC powerpack. Touching the route leads to make momentary contact produced smooth and reliable switching making me think that these solenoids work better with a smooth, rather stable current rather than a hard pulse. Unfortunately such a change to the internals of my control console to provide this is difficult. So my next though was that a voltage regulator between the CDU and JUST the Trix switches to provide the needed even current. I asked an AI if this would work and it verified that this could work and provided me with a design and calculations to choose the components. I told it that I had no data on the timing and the Trix motors, so I asked it to search the web for the needed calculation–the end result was the AI exceeded it’s page limit and would not let me copy or screen shot the design without signing up for a more formal account for $$$. Never the less I have enough information from other sources to build something but I still have some major questions.1) has anyone come up with a simpler solution to improve the operation of the Trix motors?? 2) unsure whether the CDU is using an AC or DC power supply-The CDU is physically attached to a separate coil connected to the main power supply and I don’t know if it is just a step down or an actual AC-DC transformer. The output terminals on this supply are inaccessible to measure. 3) The most efficient way I can affect JUST the Trix turnouts would be to attach the voltage regulator in line with the solenoid load on the return side of the circuit-on the black wire. Does a voltage regulator need to be ahead of the load to work? Really would appreciate some help as there is very little information on Trix point motors on the Web.
!
Turnout-CDU circuts|177x238