Re: Re: Can someone please explain this.......

Hi Jon and Greg

(I'm going to chop this up some)

>    From: Jon Elson <elson@...>
> Subject: Re: Can someone please explain this.......
>
> Greg Jackson wrote:
> >Roland Fresitad has quite a reputation, but I am more than a bit confused

<S> >2)

> >The output rating of the transformer is stated as 64 VAC at 8A, consistent
> >with 0.5 KVA.  

I have not read the last draft of this article but the transformer is 48 volt
AC which works out to about the 67 volts DC that we measured based on his
shop ac which measured as high as 123 volts. (48*1.414=68.9) Greg you are
not the first to think this way. We should have written DC tn there more
forcefully.

> >With the 1.414 factor to DC, this puts the DC supply at 90
> >volts.  

> The transformer is rated at 64 V, but he doesn't note anything about how
> that
> number is specified.  He doesn't state 64 VAC, RMS, or anything.  I'm sure
> it is not, really, a 64 V AC RMS, but something lower, that gives 64 V DC.

Jon is right here.

> >3)
> >The dropping cap is listed as 500K Ohm.  Maybe I did my calculations
> > wrong, but that looks like it will drop the cap from 90 volts in 11
> > hours.  I would think 60 seconds would be a nicer number to go for.

As I said to a private questioner, I've never seen a commercial CNC
controlled machine that had a dropping resistor whose purpose in life was to
drain away power from the supply. Such a thing would be great if the estop
broke the power between the supply and the drive amps. In that case, put a
big wattage, low resistance across the power supply wires when that switch or
contactor broke the power to the drives. Then you just have to replace
contacts on a regular basis.

> Yup, 500 K Ohms on a 27000 uF cap is useless!  You want it to dump the
> stored
> energy in a couple of minutes, at the worst.  The time constant works out
> to 3.75 Hours, meaning it reduces the voltage to 37% of the original
> voltage in that time.

We found that we got about two steps from the motors if they were running
when we killed power. At an idle the full voltage was gone in less time than
it took to blink at the meter. But there is a well deserved caution here.
You don't want to build and test this supply with nothing connected. We left
that little resistor in there in order to be certain that the cap remained
near zero while we wheeled this thing around the shop and connected it to
motors.

> >4)
> >There are 3 drives in the system shown, each rated to 7 A.  While they will
> >be current limited through induction at high speed and go into current
> > fold back when at rest, there are actually times when all three drives
> > will be operating at 7 A.  Even if the DC bus is at a reduced voltage due
> > to the limits of the transformer, we're still talking about at least 1500
> > watts DC being delivered through a 0.5 KVA transformer.

 
When we did a full out three axis start up, the dc sagged a bit but as
Greg says induction rapidly reduced the demand. We did find some heat from
the transformer (105-110 f) after running the Bridgeport mill for an hour or
so on a start stop program loop. If you build this into a closed cabnet you
will want to allow for some kind of fan and filtering. You should also note
that we allowed for two of the transformers as well as two of the caps.

I did notice that it created a lot less heat that the older commercial CNC
that was running only two axes of the next Bridgeport.

I would think that supply demand would be more of a problem with some of the
newer very low inductance steppers. The Bridgeport steppers are intended for
more amps than a Gecko can deliver. Something like 56 volts and 8 amps comes
to mind. With the Bridgeport power to them, and them idling, you could heat
quite a bit of shop. We found none of this with the Gecko drives. The
motors ran cool, the Gecko's ran cool. For as low a cost as we built it for,
we though it was cool!

> No, this is not true.  These PWM (Pulse Width Modulation) drives are
> essentially
> power in = power out + losses.  The internal losses in these drives is
> in the tens
> of Watts, maximum.  The voltage dropped in the motor windings at idle is
> very
> low, around 2 V maximum.  With 7 A rated per coil at 2 V, that is 14 * 2
> =28 W.
> So, the drives will need something like 38 W per axis.  38 x 3 axes is
> 114 W,
> not 1500.

Thanks for explaining this so well, Jon, and for dragging my fat out of the
fire once again.

Ray