Re: [CAD_CAM_EDM_DRO] AM Radio and power supplies...

Jon, & all,
I fixed the exact same problem on a similar product using the Allegro
3977 chips. All motor leads on each axis were passed thru an iron powder
filter bead (1 turn, no wrap) with great reduction in EMI.
The bead must be close to the edge of the board.
These beads are available from many sources, but can be salvaged from old
computer cables. The molded plastic lump near a connector contains a bead.
RichD

Jon Elson wrote:

>
> turbulatordude wrote:
>
> >Hi all,
> >
> >I have a simple 4 axis unit, 4 Gecko's, power supply and such.
> >
> >My problem is that the radio whistles and whines when the Gecko's are
> >connected.
> >
> >The pitch raises and lowers over a minute or less.
> >
> >I was wondering if I am just stuck with the problem or if there was
> >something I was doing wrong.
> >
> >The Gecko's are in a steel enclose with plastic window for seeing the
> >status and air vents.
> >
> >The cables are all shielded between the Geckos and the carrage, and
> >the last foot is not shielded.
> >
> >
> The problem is the Gecko output stage is driving full DC supply voltage
> onto the motor leads at 25 KHz, with a rise time of about 100 nS. This
> is a prescription for EMI (Electromagnetic interference). For cost reasons,
> Mariss had to leave out EMI filters on the output of the drives. You could
> put your own filters in, but every wire going in and out of the control
> box would have to be filtered, and probably shielded, with the shield
> terminated right at the box as it enters.
>
> I built my own servo amplifiers and put in filters right on the board.
> These amps are not in a shielded enclosure. The filters are so effective
> that I can listen to AM radio with slight interference, and FM with none
> at all. My amps run at 100 KHz, otherwise very similar to the Gecko output
> stage.
>
> A filter would have a small inductor between the Gecko and the output wire,
> with a capacitor from the output wire to the enclosure. The inductor would
> be something like 50 - 100 uH, and the cap would be .1 to .2 uF and very
> low equivalent series resistance. The inductor would have to be capable of
> carrying the motor current. It is possible that the clamp-on ferrite EMI
> filters that just clamp over the cable will help quite a bit. These can be
> found at the local computer store.
>
> >My goal would be to listen without the noise, but am also afraid that
> >just because I can hear the whine, it might effect other things in
> >the shop. My PC's seem to have no effect. Am I worrying about
> >nothing ?
> >
> >
> I wouldn't worry about this level of interference causing computer
> crashes, etc. Computers generate LOTS of EMI, and the cabinets
> have to be designed to hold it all in, or the radio/TV reception gets
> trashed by the computers, too!
>
> This is actually a problem, as the FCC requires anybody who causes
> interference to fix the problem or not use the equipment. Some European
> countries are even more severe than the US.
>
> Jon