Here's what I posted off site, looks like some other folks may be asking the same questions. This has a separated battery pack, but you may find the information about the control side insightful.
I got a pretty good run on this and have this PDF to discuss. One of the things to note on this schematic is that the ESC batteries are in no way tied to the control electronics. The ESC’s common ground plate is also isolated. I show a separate battery to run the receiver. My thinking on that is an External BEC is typically a switching voltage regulator. These devices work by producing short bursts of current through an inductor to translate that charge to another voltage/current. They produce a lot of noise, that’s why most of the come with those little “rings”, which are ferrite toroidal cores. The power is wrapped around them to suppress the noise these devices create. Since a dual motor configuration is by nature, a noisy electrical environment, it doesn’t make sense to use a “noisy device” into the Receiver, the very thing we are trying to make noise immune.
Also, The ESC’s depicted on this drawing are LV ESC W/BEC types, meaning they have an internal BEC. This is very common when running ESC’s common to the 4S battery range, which is probably what you guys are running per motor. If so, then the batts are 4S type each, for a total of 2 in the system. Each ESC is using its internal BEC to bias the output of the opto-isolator associated with it. It is the ONLY thing these BEC’s are used for. As you can see, the grounds are tied for the ESC’s power input at the common “plate” of the Cap Bank. If you choose to use wires, then realize you are degrading the ripple reduction effort by as much as 4X. I’m going to try and explain this. Wires are round, and as such, any time an electrical current flows through them, they can easily create a circular magnetic field around the outside of the conductor which has a closed path. By establishing this magnetic field, you are turning that “wire” into an inductor. The only way to remove the inductive effect is to remove the closed, circular path of the magnetic field. This inductance greatly adds to the ripple problem. Now a Flat Plate, not being round in nature and being designed with the correct length/width size, makes it very difficult for that magnetic field to close in upon itself, or wrap around the whole plate. What the fields end up doing is to form many, little circular fields on the face of the plate. These consecutive fields, end up cancelling each other and viola, the magnetic field’s effect disappears! Read that over a few times, it took me a long time to understand this.
Also note, that the receiver’s power (+) is not required to bias the inputs to the dual opto-isolator. The power from the receiver battery is basically all being used by the rudder servo, which is good, leaves more power for servos with higher torque outputs. I always recommend putting a small, electrolytic cap on the receiver directly. That may have to get soldered into the wire leads if you don’t have enough free ports on the receiver. Rudder servos in boats pull big-time current pulses, and this can cause a power fluctuation in the receiver and glitch it into performing incorrectly. (Like steering into a stone wall or the shore).
The most probable cause of ESC failure is power disturbances on the control side, which glitches the microprocessor inside the ESC. It will lose track of where the correct motor position is and will no doubt over amp the ESC as it tries to get back on track. To relate to this better, you may search for threads written by airplane guys where they report blowing the ESC when landing the airplane and the prop grounds out on touch down and stops rotating. This stops the prop and the same problem occurs, ESC loses true position, guesses wrong and BLAM! So it’s imperative that the ESC has a clean source of controller power (on the 3-terminal wire).
Well, let me know where you agree and disagree. If you guys need part numbers and such let me know. I can look some stuff up.
Note: This is NOT the new schematic I'm doing in response to Larry and Jesse. But it shows a good loop isolator scheme on the control side. These are LV ESC W/BEC's
Bookmarks