Question: When you say 1.0 are you going to update it? If so, what improvements are we looking at?

]]>I mean VFT. Pfft.

]]>I never thought I’d like to see a more theoretical approach to a problem. I must be getting old…

I thought I was (were?) pretty down to earth and “practical” compared to the guys from our theory/math department (theoretical solid state physics/quantum whatever), but it seems that compared to an electrical engineer I’m the one throwing too much theory at the problem ðŸ˜‰

In my current job I’ve developed a tendency towards “do it properly, or don’t do it at all.”

In this case I just didn’t like being fed with the over simplified formulas. It makes it harder to really understand I think. Start with the basic differential equations for the circuit and properly deduct/approximate.

Anyhow, I just ordered a KIT and will be happy to do some soldering. Even if the product is based on dirty math ^.^

]]>It took me 1 hour to realize that all exponential functions have been approximated away by linear current ramps.

Basically V/R(1-exp(-R/L*t)) ~ V/L*t for small times t (inductor turn on phase). And then t was again replaced with D*T or D/f. This also gives Ipk = V*D/(f*L). And then again sneak in a factor of 1/2 to get the average current (still a linear ramp).

If 1/(1-D) is replaced with Vout/Vin (see the wikipedia link on the clock’s ‘make’ site), you end up with Iin_avg * Vin < Iout*Vout. If it were (=), that would make sense (e.g. the average power you get out of it must be the average power you put in). Conservation of energy of some sorts.

I’m so glad I didn’t study EE ðŸ˜‰

Dirty math all over the place.

]]>