Re: Mariss' power supply circuit

Les,

I think there are 3 energy sources that need to be dissipated. The
kinetic energy you say is 25J and the capacitor energy is 72J.
Guessing that your motor inductance is 10 mH, at 30A that would
contribute an additional 4.5J of inductive energy.

I am not sure your 25J mechanical component is a factor regarding
capacitor voltage assuming you have a brush DC motor. It can be
modeled as a voltage source in series with armature inductance and
resistance. The voltage would be proportional to motor speed and can
never exceed your power supply voltage.

Adding it all up you have about 102J to get rid of. Making things
easier, the kinetic energy will decay exponentially so it can be
treated as an equivalent capacitor. The "capacitor" would be 102 / 72
times 20,000 uF or 28,333 uF.

Now assume you want the energy discharged in 1 second. If 3 time
constants constitutes a discharged supply, calculate a value for R7
from T / C or 0.333 / 0.028333, or 11.75 ohms. Call it 12 ohms as the
closest standard value. Peak current would be 85 / 12 or 7.23A.

102 joules (or watt-seconds) have to be dissipated. A 10W wire-wound
can tolerate 100W for 5 seconds (500 watt-seconds), so it should do
nicely. Obviously a 6 ohm resistor would dump the energy in 0.5
seconds at a peak current of 14.5A. That would be 50% of the IRF540
max rated Ids of 27A. Interestingly 10W would still be OK because the
energy is the same.

Mariss

--- In CAD_CAM_EDM_DRO@y..., "Les Watts" <leswatts@r...> wrote:
> I think I see. C1 charges up through R2 linearly thus forming the
sawtooth
> wave portion of the modulator.
>
> I am building a new servo supply for my machine. The current one is
not
> stiff enough and causes following error estops with the high
acceleration
> high speed moves I want to do. I will incorporate this circuit in
the supply
> I think.
>
> I need about 30 amps cont at 85v and the 50 or so pounds of
transformer
> iron laying on the floor says I will get it. I am using stepdown
isolation
> transformers at half voltage so that iron won't be at the full flux
density
> but there is still plenty of room. I will nominally go from 120v to
60
> so I should get 60*1.414= 84.8vdc. My amps turn off at 92v and have
> 100 wv caps in them.
> Since my line voltage is notoriously high I may put a few bucking
turns
> on the transformers. I read why power companies like higher voltage-
> revenue goes up as the square of that voltage! Too bad they can't
control
> it.
>
> As for sizing R7 in your circuit- well lets see.... my 25 joule of
max
> regeneration
> energy could attempt to pump up the 20,000uf supply cap from its
normal
> 72 joule @ 85v to 72+25=97 joule. That would be 98.5v. It would be
some less
> from losses but that is a maximum. With your circuit at 100% duty
cycle
> dumping
> 25 joule in say .05 sec would require 500 watts as a linear
approximation.
> So the value would be about 20 ohms. I use Dale 10 ohm 25w
wirewounds
> to short the servos in my estop circuit and they don't even get
warm
> so 25w should be fine. I could use a lower value.. if I chose the
resistor
> current
> to be the ps rated current it would be about 4 ohms but if I go too
low
> I think I will estop on every power flicker.
>
> As a check for power failure 3*R7*C= 1.2 seconds ; long enough. My
estop
> will short the servos in a sustained failure anyway.
>
> Les
>
>
> Leslie Watts
> L M Watts Furniture
> Tiger, Georgia USA
> http://www.rabun.net/~leswatts/wattsfurniturewp.html
> engineering page:
> http://www.rabun.net/~leswatts/shop.html
> ----- Original Message -----
> From: "mariss92705" <mariss92705@y...>
> To: <CAD_CAM_EDM_DRO@y...>
> Sent: Monday, June 17, 2002 8:09 PM
> Subject: [CAD_CAM_EDM_DRO] Re: Mariss' power supply circuit
>
>
> > Les,
> >
> > I appreciate you taking the time to understand the circuit. The
PWM
> > modulation is actually a linear function of the over-voltage seen
by
> > the supply. The amplitude of the 120 Hz sawtooth (R2,C1) is set to
> > about 2V peak-to-peak, this being the "regulation" range of the
> > circuit. The "trough" of the full-wave rectified reference clamps
the
> > negative excursion of the sawtooth to the difference voltage.
> >
> > The IRF540 has an Rds of about .07 ohms. It always operates as a
> > switch, the current being limited by the 10 to 20W shunt resistor.
> > The resistor must be wire-wound because wire-wound resistors have
a
> > superb overload capability, typically 10 times their rated wattage
> > for 5 seconds, 50 times rated for 200mS.
> >
> > In an undervoltage situation the circuit would attempt to adjust
> > itself to the new, lower voltage. It would be abley assisted by
the
> > load in this endevor.
> >
> > Like all silicon, MOSFETs invariably fail shorted if abused. The
> > consequence would be the shunt resistor would rapidly overheat and
> > fuse open, clearing the circuit. Everything else would fail
> > harmlessly and would leave the basic supply uneffected.
> >
> > Another thing about the circuit is the component value choice for
the
> > +12VDC supply. The size of C3 is chosen so that the circuit
remains
> > functional for .5 second after AC power removal to fully discharge
> > the supply.
> >
> > Thanks again for your comments; my pleasure is designing
minimalist
> > circuits the way others might enjoy working crossword puzzles.
> >
> > Mariss
> >
> > --- In CAD_CAM_EDM_DRO@y..., "Les Watts" <leswatts@r...> wrote:
> > > I finally found the .pdf file and had a look.
> > > As I understand it the circuit will perform shunt regulation
> > > using PWM at 120 hz. Duty cycle will vary from 0-100% over a
> > > couple volts diff between the filter cap and transformer
secondary
> > > peak. I think it will modulate with a raised cosine law. Right?
> > >
> > > Question: what is the on resistance of the Fet you selected?
Just
> > > want to calculate heat sink thermal resistance. Also, what
happens
> > > in a line power undervoltage scenario?
> > >
> > > And as a little FMEA thing, what would be the most likely part
to
> > fail?
> > > Would it be the fet if say the primary mains went very high?
Would
> > it
> > > fail open or shorted? (I currently have a floating neutral
argument
> > with
> > > my Georgia Power supplier and their electrical feed)
> > >
> > > I want to publicly thank you for publishing this as open source.
> > > I appreciate your weekend efforts.
> > > These things cost about $300 from other suppliers.
> > >
> > > Les
> > >
> > > Leslie Watts
> > > L M Watts Furniture
> > > Tiger, Georgia USA
> > > http://www.rabun.net/~leswatts/wattsfurniturewp.html
> > > engineering page:
> > > http://www.rabun.net/~leswatts/shop.html
> > > ----- Original Message -----
> > > From: "mariss92705" <mariss92705@y...>
> > > To: <CAD_CAM_EDM_DRO@y...>
> > > Sent: Sunday, June 16, 2002 3:54 PM
> > > Subject: [CAD_CAM_EDM_DRO] Re: Power Supply Circuit
> > >
> > >
> > > > Hi,
> > > >
> > > > OK, here's my Saturday's project. I posted a "Power supply
> > > > circuit.pdf" in the files section of this list.
> > > >
> > > > It wound up being different than what I started with. Sensing
> > current
> > > > was too crude. Instead this circuit compares the power supply
> > voltage
> > > > against the xformer secondary voltage. It pulse width
modulates
> > the
> > > > dump resistor and has a very nice proportional response. The
PWM
> > goes
> > > > from 0% to 100% as the power supply voltage climbs to 2V above
> > the
> > > > nominal voltage. It has been tested and works very well.
> > > >
> > > > Mariss
> >
> >
> > Addresses:
> > FAQ: http://www.ktmarketing.com/faq.html
> > FILES: http://groups.yahoo.com/group/CAD_CAM_EDM_DRO/files/
> >
> > OFF Topic POSTS: General Machining
> > If you wish to post on unlimited OT subjects goto:
> aol://5863:126/rec.crafts.metalworking or go thru Google.com to
reach it if
> you have trouble.
> > http://www.metalworking.com/news_servers.html
> >
> > http://groups.yahoo.com/group/jobshophomeshop I consider this
as a
> sister site to the CCED group, as many of the same members are
there, for OT
> subjects, that are not allowed on the CCED list.
> >
> >
> >
> > Post messages: CAD_CAM_EDM_DRO@y...
> > Subscribe: CAD_CAM_EDM_DRO-subscribe@y...
> > Unsubscribe: CAD_CAM_EDM_DRO-unsubscribe@y...
> > List owner: CAD_CAM_EDM_DRO-owner@y..., wanliker@a...
> > Moderator: jmelson@a... timg@k... [Moderator]
> > URL to this group: http://groups.yahoo.com/group/CAD_CAM_EDM_DRO
> > bill,
> > List Mom
> > List Owner
> >
> >
> >
> > Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
> >
> >
> >