Good day all:
I have attached another interesting *read* from our Russian friends............
It seems to go back a few years and was recently brought to me attention by a friend, thanks Harish.
I won't go into much detail as you can read it for yourself if you are interested. The link is to the original posting on a Russian site: http://001-lab.at.ua/index/donald_smit/0-43
The posted PDF was translated with Google, so it has the typical grammatical/syntax errors, but the idea is understandable.
I will note that there are similarities between this authors findings and that of 'Salty Citrus' from the Chinese group. These findings do not contradict Don Smith but rather, in my opinion fill in the missing gaps that Ole' Don always hinted about but could / would never tell us directly. Frequently Don Smith admitted that he purposely omitted pertinent details.
"Abnormally high efficiency for me starts to begin at the pump pulse duration of less than 1 microsecond. The smaller the pulse width,
the greater the amplitude of the pump pulse you can apply, the higher the efficiency." *paraphrased from attached pdf doc.*
Also wanted to update on my own progress: https://youtu.be/3C1eSpL10j8
I have attached some photos of my *SiC_Pack_Driver_v1* pcb. This pcb is the driver board for my SiC_Pack_v1 board, which is in essence a 7.2 KV SiC switch.
This is the first test of the SiC_Pack_Driver_v1 pcb. After extensive testing with the magnetics of the first hand made SiC_driver board, it became apparent that in order to reach the Nano-second precision required for this project, a *real* pcb must be produced.
So we have the *SiC_Pack_Driver_v1* pcb.
As can be seen, (only testing one of the six channels of the pcb) the rise and fall times are respectable for a high speed magnetics based / isolated SiC Mosfet driver. Actually we are at the *spec sheet* limits for the CREE C2M0080120 SiC mosfet. Of course a digital based driver setup could easily half these speeds, but being old school I opted for the magnetics based isolation as it is pretty much bullet-proof.
The single channel was tested to a frequency of over 1.4MHz @ pulse width of 76-78ns. The rise / fall times are 24ns / 20ns approx. with test toroid. Finished pcb will have slightly faster rise / fall times do to shorter lead length on installed toroids.
Inside information from the Chinese replication group made note of a 80ns pulse width...........
All for now.
take care, peace