December 25, 2015
The cast stator
As part of building our axial-flux alternator, I cast the stator in resin. The casting itself went well. However, before the casting, continuity tests between all the terminals showed a good connection, but after the casting all failed. I also tried spinning the rotors to create current in the stator and wiring a 12 Volt battery to each pair of terminals in turn, and but neither created any detectable current through the terminals.
Delta wiring diagram
Since the stator was wired in a delta configuration, the failure must have been either systemic or localized in more than one point. If there was a break in one phase or at one terminal, there would still have been a closed circuit between at least one pair of terminals. For example (using the diagram above), if there was a break at terminal S, there would still be a good connection between terminals R and T, or if there was a break in phase l1, the connections between terminals S-T and T-R would still work.
Stator technical information
- 3-phase with a delta wiring configuration.
- 9 coils of 60 turns of insulated 13 gauge magnet wire.
- Each coil was made by winding two wires together for 30 turns, then soldering them end to end to produce 60 turns.
- The stator contains 15 wire end to wire end soldered connections, where the wire ends were twisted around each other, soldered, and covered with a crimped copper sheath.
- The terminals are brass doughnut shaped buttons with a groove along the side that the wires can partially sit in.
- The stator contains 3 two wire ends to terminal connections, where the wires are wrapped around terminals and soldered.
- Mold size 14''x14.25''x0.75''.
- Cast with approximately 1/2 gallon of Bondo Fiberglass Resin, which uses an MEKP hardener.
From my own observations and consultations with other community members, technical help lines, and online discussion boards (such as fieldlines.com) I accumulated several theories as to what could have gone wrong. I then came up with ways to test each theory. If I could identify what went wrong, perhaps I could fix it, or at least design the next stator to avoid the problem.
- Hypothesis: there was a thin layer of resin left on the terminals that was breaking the connection.
- Action: I sanded the terminals again and performed a continuity test between different parts of the same terminal.
- Result: there was a closed circuit between different points on each individual terminal.
- Status: disproved.
- Hypothesis: the multimeter is malfunctioning.
- Action: I used two different multimeters (analog and digital). Both were checked on other circuits and both worked on those.
- Result: both multimeters register other closed circuits, both showed open circuits between each pair of terminals.
- Status: disproved
- Hypothesis: the heat of the curing resin melted the solder in the joints.
- Action: I called Bondo's technical help line for information, looked up melting point of solder, and checked wooden mold for charring.
- Result: they said the curing temperature of a 3/4'' layer of resin is 280° Fahrenheit. Solder melts at 361° Fahrenheit. No charring on the mold.
- Status: disproved
- Hypothesis: the resin's hardener contains methyl ethyl ketone peroxide, which is reactive with copper, and might have caused the copper wire to fail.
- Action 1: I called Bondo's technical help line for information.
- Result 1: they said that the resin should just harden around the wire.
- Action 2: I cast two copper wires in resin, one of which had the insulation burnt off.
- Result 2: a closed connection registers across both wires.
- Status: disproved
A wire cast in resin
- Hypothesis: multiple soldered connections were broken, either by resin seeping in between soldered wires or between the wires and the terminals, or by the heating stresses mechanically pulling wires apart.
- Action 1: I looked at the connection points through the semi-transparent resin.
- Result 1: no visibly disconnected wires.
- Action 2: seven new joints were soldered and cast in 3/4'' thick resin. Some were soldered the same way as the joints in the stator, some as skillfully as we knew how, and some intentionally poorly.
- Result 2: all seven showed a closed circuit.
- Status: unlikely, but possible
The new joints; some soldered well, others poorly
The new joints cast in resin
The first 4 theories were ruled out, but the last one remains possible. There are plausible reasons why the soldered joints in the stator might have failed while the 7 joints I soldered didn't.
The only thing that I can think to do to shed more light on this problem is to cut away the resin from the coils and examine the joints directly. However, that would take a large amount of work, and it's not certain that I would get definitive results. I also wanted to get the next stator cast before the winter set in and made casting difficult (casting resin requires warm temperatures, and our workshop isn't heated).
Three coils wound from a single wire
On the assumption that the soldered joints were the points of failure in the first stator, I designed the next one to not have any soldered joints that would be inside the casting. Each set of three coils was wound using a single piece of wire, and the 6 wire ends (that connect to the 3 terminals) were left sticking out of the casting. I cast this stator and got positive results for continuity tests between all pairs of terminals.
The new cast stator, with the 6 wire ends protruding from the casting
I might never know for sure what went wrong with the first stator. To find out conclusively, I would want to cast several more stators, with the stator and the casting method being identical except for one variable that is changed each time. But that would require a significant investment of time and money. Given the scale of this project, it is more important for us to focus our resources on building an axial-flux alternator that works, rather than discovering precisely what doesn't.