Re: Gecko burned out ( about tinned leads and other things)

Hi,

Keith just told about this thread and I just read all the posts.

1) Do not tin stranded wire for use with any crimp or clamp
connection. No less an authority than NASA specifically forbids this
in their workmanship standards documents.

The problem has to do with solder "fretting" (I hope that is the
correct term) at the contact points in the pressure clamped
connection. Solder flakes at these locations when motion is present
and forms a black residue. This is powdered solder and this powder
forms a high resistance (about 0.1 to 1 Ohm). Current passing thru
this resistance will cause heating of the connection and will destroy
the contact. In severe cases the connection will melt or catch fire.

The solder powder quickly oxidizes and will not re-melt to heal the
joint.

2) The damage to Keith's drive resulted from a loose wire that
touched an adjacent winding phase connector. This conclusion is based
on destruction of the phase A and D half-bridge MOSFETs. I'm guessing
B came loose and touched C or visa versa.

3) Protection circuits. This is a philosophically interesting subject.

Let's assume a complete suite of protection circuitry adds $10 to the
cost of a drive. It protects against everything; short-circuits,
overheating, reveresed polarity, winding disconnect, misphased
windings, load dump, etc.

Let's also assume there are two customers, entirely different from
each other in how they use the drive. Customer #1 buys 3 drives for
his machine, customer #2 buys 1,000 drives which are used in his
product.

Customer #1 has spent an extra $30 to protect his 3 drives. The
protection has worked but he never knows it. Ocassionally a wire
shorts, he clears the fault and the drives keep running. He may
sometimes wonder if he shouldn't have bought some of those cheaper
drives he sees advertised. They have the same ratings after all.

Customer #2 has on average 1 drive failure per 1,000 drives. He has
design engineers that have designed everything surrounding the drive.
Nothing has been overlooked in the design and ultimate use of said
product. He also has production engineers who insure the drive is
incorporated properly during manufacturing.

If he buys the protected drive, he will have spent $10,000 to protect
against that one $100 drive from failing. That is expensive
insurance. His process is well worked out and he may more than wonder
if he shouldn't have purchasing look into those cheaper drives his
design engineer brought to his attention. They have the same ratings
after all.

You are the drive designer. You can design them without protection
circuits or you can add every useful protection circuit imaginable.
It's more fun to design the protected drive but extra parts add cost.
An added kicker; 80% of your company's sales volume are large OEM
customer #2 type accounts. What would you do?


Mariss



--- In CAD_CAM_EDM_DRO@yahoogroups.com, "Steve Stallings"
<stevesng@n...> wrote:

>
> The body diodes in FETs work fine for recirculation
> currents but cannot stop damage due to arcs because
> circuit inductance and the slow speed of the body
> diodes allow the FET to be subjected to damaging
> voltages. See message number 2194 from Mariss in
> the Geckodrive group addressing this issue.
>
> http://finance.groups.yahoo.com/group/geckodrive/messages/2194
>
> Regards,
> Steve Stallings
>
> --- In CAD_CAM_EDM_DRO@yahoogroups.com, Major A <mandras76@y...>
> wrote:
> >
> > Do I detect a major misconception here?
> >
> > The FETs that drive the motor's windings can be regarded as on/off
> > switches. When they switch off, the idea is that no more current
> will
> > flow through the winding. Unfortunately, the motor windings have a
> > self-inductance, which means that every change in current is
> > accompanied by a large voltage across the winding. When the FET
> turns
> > off, this effect would cause a high voltage to drop across the

FET,

> > which would indeed fail instantly. Now all such drivers have,
> however,
> > a diode connected across the FET which clamps this voltage to

<1V,

> and
> > through which the current that is no longer flowing through the

FET

> > can pass and decay gradually.
> >
> > If you disconnect the wire from the drive during operation, while
> > current is flowing, the large voltage will be present indeed

because

> > there is no protective diode across the failed contact between the
> > motor and the drive. This voltage can indeed cause sparks to

appear

> at
> > the terminal, but one thing it cannot do is harm the FET because

the

> > FET is still protected by the diode. The only way you could harm

the

> > FET is by applying a large voltage of the opposite polarity, but a
> > motor winding cannot cause that to happen.
> >
> > By pulling the wires, you CANNOT harm a drive, however cheap it

may

> > be.
> >
> > Andras