Re: [CAD_CAM_EDM_DRO] Re: Lost steps

Ozzie@... wrote:

> Okay we have some discussion of wire sizes, power supply, decoupling
> caps!
> The main power supply is two transformers, 10A each in parallel,
> feeding to a rectifier and then a big capacitors, all supplied by
> Camtronics, with knowledge of what I was building.The wires supplying
> motor voltage to the boards is 18 Ga stranded wire, +and-.

Uh, oh. 18 gauge sounds fine for the current, but it may require a much
heavier
wire to suppress unwanted voltage from the inductance.

> It's not shielded.

Well, you might try shielding these, too.

> There are two capacitors right at this power connection on
> the board.

Ah, that's good, at least.

> The signal wires from the DB25 junction board are 22Ga stranded, home
> shielded with foil. The shielding is wrapped with some bare wire and
> then grounded to the chassis.

Do you have an individual ground wire accompanying each signal wire from
the driver
to the DB25 junction board?

> The 5V power supply is built as instructions given to me by Dan via
> e-mail. The main power supply run through a resistor, a 5V regulator,
> and two capacitors. The 5V then goes to the boards through 18Ga
> wires.

Hmm, it is possible for 5v regulators to oscillate. Some of them
require capacitors
very close to both the input and output terminals, with the other end
going to ground.
This is from the manufacturer's own literature.

> All of this is redundently grounded to the chassis/case. The power
> cord is grounde, the rectifiers are grounded, the 37 Volts is
> grounded, the 5Volts is grounded, and each board is grounded on all 4
> corners to the ground trace which also grounds both power supplies.

Hmm, someimes there can be TOO MANY grounds, causing all sorts of
circulating currents, known as ground loops. You might take a voltmeter
and
try reading both AC and DC between the computer case and the driver
cases.
If there is much voltage difference, it could easily explain the
problem.

> The driver side of the DB25 cable has all unused wires grounded and
> the port end I don't know about, but it is a very standard port card,
> screwed to the housing/case in the normal manner. It is a commercial
> cable and the pin housings show continuity to each other

ALL unused wires grounded? There are some additional outputs that you
may
be shorting out. Pins 18 - 25 are the official ground connections, but
I don't
know if the others should be grounded. This could be reducing the
voltage
available for the other outputs.

Jon