Pierre's 170W in 1600W out Looped Very impressive Build continued & moderated

[listener192]

Rotor trimmed, now covers 6 slots instead of 9. Would work better at 5 slot width.

This was running with all relays working (initially).230V -26V 50Hz transformer. 26VAC RMS rectified by two parallel 35A FWBR connected directly to 150uF/400V cap. Relay feed directly from this cap. 

Super caps supplied via 4 Ohm 50W resistor, recovery direct to super caps.

Recovery current pulses 10A approx.
Even with 0.1uF 400V caps across each relay, the contact resistance slowly increased and setup has become useless again. 

Before failure the output was driving a 60W lightbulb with about half intensity.
Unable to perform additional tests. 

Code original, adjusted for 30 slots.
The first run also attached, see how the coil drive current is consistent compared with the later shot which is at twice the clock rate.

A couple of high side 100V 0.1uF caps across relay contacts failed. These were replaced by either 400V or 1000V caps, which were OK.
Difficult to see how Pierre got this to run for any length of time on the same boards with the songle relays.

RegardsL192

[listener192]

I cannot see how the relays could be run fast enough with the current code. The output increased until a point was reached where they could not switch fast enough.Perhaps inverting the DC supply to the second half of the relays would double the output frequency and get us close to 50/60Hz albeit with a waveform that would be a quarter sine.

I would also mention that the lack of the 6th coil decreases inductance and output.

By using a low value cap for the 400V cap, half wave DC or a pulsating waveform could be obtained for switching via the relays.
See the attached link.

http://www.nptel.ac.in/courses/IIT-MADRAS/Electrical_Machines_II/animation/sinwave3-9.htm
A regular 3 phase induction motor has two moving flux components, a traveling wave and a pulsating wave due to the AC.
By using a low value cap for the 400V cap, half wave DC or a pulsating waveform could be obtained for switching via the relays.
If both waveforms are in frequency/phase synchronism, then they may reinforce each other.

Regards
L192

[listener192]

The caps charge up to about 30V without the device running, this is higher than 25V shown in the videos,so instead of 26VAC RMS I think Pierre's transformer must be closer to 22VAC RMS
L192

[gmolina]

Hi L192, 25V in ultracaps bank and 7V in 4Ohm resistor.

GM

[listener192]

Hi L192, 25V in ultracaps bank and 7V in 4Ohm resistor.

GM

The 4 ohm resistor just limits the charging current to the cap bank. There is no volt drop across the resistor once the caps are charged, (in the non running state).
L192