Mediator Power Problem Solved

[stopthegop]

I think I figured out what was causing the mysterious power problem with my A4000 Mediator. I suggest owners of Mirage towers from Elbox check their systems for
extremely tight tiewraps (about 6 of them in total) on the wire bundles for the AT-ATX power converter PCB that comes stock in the Mirage case. Mine were on so tight they were constricting the PSU from outputting the correct voltages. This caused the ATX logic to assume something was faulty at power-on then the PSU just shut itself down. I removed all six tiewraps. Now the system stays powered on as it should and I noticed the fans are now spinning at normal speed, rather than sounding "strained" as they did before. I recommend to remove those tiewraps.

[Fester]

Hi,

Ah the wonderful flow of electrons! Perhaps you have a bad connection on one of those wires. Try playing with those wires, moving them around a bit. I find it hard to believe that the ties were too tight, unless the tightness actually dammaged the wires.

Fester

[adolescent]

stopthegop wrote:
Mine were on so tight they were constricting the PSU from outputting the correct voltages.

:crazy:

[stopthegop]

Mine were definately too tight. The wires have deep claw print impressions on them. Jiggling the wires is actually what led me to suspect the tie wraps. With MB out of the case and a PSU hooked up I gently pushed down on the wire bundles to mimick the effect of pushing it back into the case against the IDE, SCSI, and floppy cables. When I pushed the one going to the A4000 motherboard, it shut down immediately. I checked all the connections and they looked good.. Thats when I thought it might be the tiewraps. I removed them. Now when I push on that wire bundle the Amiga stays powered on. Its been up almost an hour now whereas before the most it would stay on was 30 seconds. Its sounds kind of stupid (tiewraps of all things), but apparently thats what it was. At least I hope so. :-?

[Homer]

Yes, that is odd.
Crosstalk or noise is not usually a problem with static DC feeds, but bad connections are.
I work on trains and it is quite normal to tightly ty-wrap cables together - in fact they use a torque set tool to do them all to the same tightness, which does indeed mark the cables if the ty-wrap is released. This includes unscreened bundles of low voltage DC cables (110VDC), and separate bundles of screened signal cables.
I therefore also agree that this is not normal, and the root cause for your problem is yet to be found !
I think the pushing of the wire bundle to the motherboard may point to the true fault here, which is likely to be a socket/pin interface, or even a fractured wire into the back of a crimp connection.
Ah well, have a :pint: and think about it :-D

[keropi]

I have an Elbox adaptor on my 4000D , it gives me troubles too, sometimes it power-on and after 0.5secs it shuts down again! could this be the prob? the strainers?

[stopthegop]

Crosstalk or noise is not usually a problem with static DC feeds, but bad connections are.



Crosstalk isn't a factor here as these lines just carry an unmodulated DC voltage. No "talking" to cross. :) However, if these cables are crimped to such an extreme that a 16 guage wire effectively becomes a 12 guage wire, it could have unintended consequences. For one thing, if it is a wound cable, as these are, it could create inductance, which in a DC circuit acts like a resistor. This would reduce DC voltage, though not necessarily the DC voltage as output from the source, i.e. the power supply. Thats just to clarify what I said in the start of the thread when I said the binding of the wires had an effect on the PSU's output voltage, which is obviously not correct. It had an effect on voltage, yes. But not on what the power supply itself was generating. This same thing can happen with network cables when they are strangled too tightly -- throughput starts heading south.

[adolescent]

Edit: Post deleted.

[Homer]

stopthegop:
I kind of disagree with your inductance theory, as crushing the wire would not reduce the cross sectional area of the wire, only change its shape, and I've seen some pretty flattened cables on trains (after removing ty-wraps) that never gave a bit of trouble. Also they are a straight twisted cable, not wound in a manner that causes excessive inductive effects with a static DC feed.
Inductors in a DC circuit act only to slow down any change in current (Voltage stays the same but rate of change of current is limited - this is not a resistive effect), and mutual inductance would be unlikely here due to these being static DC feeds.
Conclusions:
1.These cables are too short to suffer from the effect you are thinking of, and are not wound into coils to increase the inductive effect, and carry a static feed, so no complex AC capacitive or inductive reactance is involved.
2.Natural laws of physics may not apply in NYC :lol:
3.You have not found the real problem :-P
4.Does it matter, since you have it working :pint:
5.Homer should shut up, as stopthegop is quite happy :lol:

[stopthegop]

True. Leave the theories for academics. My miggy works. :)

[Homer]

I love it when a plan comes together :pint: