Sharing ideas on how to make a more efficent motor using Flyback (MODERATED)

[verpies]

...presently i am working on a coil shortener using a reed relay triggered by a Hall sensor (+ transistor), hopefully the 15 or so mJ in the cap will not fuse together the reed relay's contacts.

If you find out that the reed relay is not fast enough, you could use a MOSFET to short out the C2 (see the 1^st schematic below)

Parts recommendations:
Q3 should be a 15A,  N-ch MOSFET capable of blocking at least 150V.
Q4 should be a 1A, N-ch MOSFET, capable of blocking at least 50V.
L2 should be approximately a 1mH choke capable of withstanding 500mA of current.
ZD2 should be a 500mA Zener diode featuring Zener voltages from 12V to 18V.

Theory of operation:
The L2 acts as an inverting boost inductor, that provides a gate voltage for Q3, that is higher that +V_1A, because N-channel MOSFETs need a gate voltage that is higher than the Source, in order to turn on. 
This higher voltage could've been delivered by a second external power supply, supplying a voltage that is +10V above +V_1A. However, in this circuit, L2 acts as such power supply, thus a second power supply is not needed.

When the "Discharge C2" signal is high, Q4 closes and the gate of Q3 is pulled low through R3.  Consequently, Q3 opens and currents starts building up in L2 (up to the +V_1A/R3 limit).
When the "Discharge C2" signal goes low, Q4 opens and the L2 current is interrupted.  Consequently an inverted voltage is generated at point E (this voltage is higher than the +V_1A supply rail).  This voltage is clamped by ZD2 and D2, which act to maintain L2 current and protect the gate of Q3 from overvoltage. As a result of this, Q3 closes and shorts out C2 through R2.
Note, that the "Discharge C2" signal is active when LOW !

This circuit should not be operated from a supply voltage ( +V_1A) higher than 20V without alterations.

A MOSFET driver U2 and a diode D3 can be used instead of Q4 and R4 (see the 2^nd schematic) and if this driver has an inverting input, then the the "Discharge C2" signal becomes active when high.

P.S.
A P-ch MOSFET could not have been used for Q3 because its body diode would be always shorting out C2.

[itsu]

Great stuff verpies, i certainly have to try it.


Meanwhile i have redone my measurements as to see if the overunity cap recovery condition still exists.

I tried to incorporate my new knowledge about doing power measurements on selected parts (inbetween cursors) of a wave form, but it also
influence the display of the other channels, so for a running situation (40Hz @ 18% duty cycle) i have 2 screenshots, first one without the selected math,
second one with selected math (ontime only).

I have a 100K potmeter for R2 set for the C2 voltage to go to 0 when the next pulse comes (yellow trace)
Blue trace is the CSR (1 Ohm 1%), purple trace the drain voltage, all signals referenced to point B
Supply voltage is 12.4V from a battery.

The 2e screenshot has the cursors setup for the on time only to calculate Ch2 x Ch3 which gives a mean wattage of 404mW (x 10 because of the 0.1 csr = 4.04W).

If needed i can provide the same for a stopped condition.


Itsu

[itsu]

Here an expanded (zoomed in) version of the ontime with math:

[itsu]

Here the same situation as the one above, but with yellow cursors (and data in upper right corner) and with the math trace (red) in lower amplitude:

[gotoluc]

Well Itsu, I tried the software but I think it's the scopes limitation.
.99 looked into this scopes specs before and was surprised of the low resolution the capture side had.

Below is:

Capture from the Scopes USB port

Capture from Software Screen Capture Tab

Capture from Software Data Capture Tab

I like the what the Data Capture Tab best and you can see more details but no scope data is included.
Do you think a firmware update would help?

Anyone else can reply as well

Thanks

Luc