On Wednesday 24 September 2014 06:30:24 Juergen Gnoss did opine
And Gene did reply:
> Short update on that.
> 
> Gene was right, the Y driver was "tired". (Didn't know such thing
> happens)
> 
> When running in MDI with simple straight Y | XY | YZ moves or jogging,
> all axis is moving correctly. Giving a lot of small segments Y didn't
> make it.
> Slow down the machine by means of max_Velocity and max_Accel on Y
> improved that and machine runs without hick-up's. Later on, changing
> the Power IC's and Cap's on the Y-driver made Y run as bevor with
> original velocity and accel.
> 
> We never stop learning.
> 
> Thanks at all.
> 
> Ju

Thank you for the flowers Juergen. We all appreciate being told we are 
right. ;-)

>From my experience of 65 years in electronics, those low capacity, lower 
(below 150 volts rated) capacitors have been doomed to an early failure by 
an internal chemical mechanism that causes "deforming", and that can only 
e prolonged by operating them at a voltage high enough to maintain the 
forming of the oxide coating on the alu foil.

That is by running them with an applied voltage high enough to "keep the 
battery charged" so to speak, which is usually at least 30% over the label 
voltage.

Running them in a 5 volt logic situation isn't anywhere near enough 
voltage to do this, regardless of the labeled voltage.

What that simplifies down to is that replacing the caps was likely all it 
needed.

The exception to that rule might be the power semi's, which can degrade 
over time if they spend too much time in the analog (not on & not off) 
domain, which can happen if the hexfets gate drive is weak, taking too 
much time to charge and discharge the gate capacitance of the power 
hexfet. Poor driver capability can lead to several kilowatt power 
dissipation in the device, for a microsecond at a time.

Minor digression here:

In case its not widely known, these "power" hexfets are not one humongous 
transistor, but in 99.9999% of the devices, are in fact tens to hundreds 
of them all in parallel under all that epoxy B of a TO-247 or TO-220 
package.  So they can blow one transistor at a time, over time, and 
gradually degrade to where they cannot do the job anymore. This poor drive 
is usually the result of the failure of the driver circuits power rail 
bypass capacitors, either from deforming or just as bad, a large increase 
in their ESR, preventing the gate drivers from being able to use their 
stored energy as output drive, so the slew rate of the applied gate drive 
is slowed.  This translates to the shorter life of the output device 
because of that heat pulse every time they are switched.

The ideal scenario for the 1 u-f and under stuffs is to have enough real 
estate to put in a low voltage mylar film cap, but they are generally 250x 
the cubic volume, aka way the heck too big physically.  So we are stuck, 
not because a better part isn't available, but because that better, lasts 
forever part, simply cannot be shoehorned into the available space.  Not 
to mention the increase in the price caused by the higher insurance rates 
the bean counters would need because of the cost of all the heart attacks 
they have when they see the differences in the cost. :)

Your electronic trivia chapter for today. ;-)

Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Genes Web page <http://geneslinuxbox.net:6309/gene>
US V Castleman, SCOTUS, Mar 2014 is grounds for Impeaching SCOTUS

------------------------------------------------------------------------------
Meet PCI DSS 3.0 Compliance Requirements with EventLog Analyzer
Achieve PCI DSS 3.0 Compliant Status with Out-of-the-box PCI DSS Reports
Are you Audit-Ready for PCI DSS 3.0 Compliance? Download White paper
Comply to PCI DSS 3.0 Requirement 10 and 11.5 with EventLog Analyzer
http://pubads.g.doubleclick.net/gampad/clk?id=154622311&iu=/4140/ostg.clktrk
_______________________________________________
Emc-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-developers

Reply via email to