[ARC5] Vibrator Power Supplies

[J. Forster]

All right, it is a slew-rate limited trapezoid. It's a lot more like a
square wave than a sine. I know what L*di/dt is. I designed a number of
switching DC/DC converters that went into earth orbit.

I, too, can pic irrelevant nits.


-John

==============



> Output of a vibrator supply is not square wave - the input may be.
> Transformers cannot transfer square waves - the output waveform is related
> to the rate of change of current in the primary; at the top of a square
> wave, there is no current change - so the secondary Voltage falls to zero.
> The output typically is sawtooth - with ringing. Please go back and read
> some transformer theory, John.
>
> Ain't no such thing as silicone steel. Silicone a synthetic liquid rubber
> solution that can cure with heat or UV. Transformer core material is
> usually high-silicon steel.
>
> 73 de Brian.
>
>
>
>> J. Forster <jfor at quikus.com> wrote:
>>
>> > Dynamotors typically have a conversion efficiency of no more than 50%.
>> A
>> > vibrator supply can reach 70% easily, and often more, if designed
>> > properly. The modern day development of the vibrator is the
>> switched-mode
>> > power supply, which often reaches in excess of 95% efficiency.
>> >
>> > I suppose when fuel was plentiful and cheap, what did it matter if you
>> > consumed a gallon to do a job that a quart could do?
>>
>> Not so. Specifically, the WS 19 vibrator supply was considered a
>> significant improvement, because it required the AFV's engine to be run
>> far less to charge the battery. Patton's advance in europe was stopped
>> because of lack of fuel.
>>
>> > Then you need to consider the downstream effects - the filtering
>> > components for a dynamotor are much smaller and lighter than for a
>> > vibrator, because of the frequency.
>>
>> Also not so. The waveform of a vibrator supply is square wave like.
>> Energy
>> storage in the filter can be much less than for a sine wave supply. More
>> filtering is required than a properly designed and adjusted dyno, but
>> nowhere near as much as a single phase line supply.
>>
>> > But with SMPSUs, operating at 1 MHz or
>> > so, the filtering components are very much smaller than for a
>> dynamotor.
>> >
>> > 73 de Brian, VK2GCE.
>>
>> In general, the mass of iron required in a transformer is inversely
>> proportional to the operating frequency, if the peak flux density is
>> held
>> constant. A rule of thumb is 40 Watts/pound for typical silicone steel.
>> However, as frequency goes up, so do eddy current (iron) losses.
>> Ordinary
>> E-I cores, and even tape wound ones, crap out at maybe 5 kHz.
>>
>> -John
>
>