Replacing the final power amplifier transistors on high power XTL5000

[MattSR]

Hi Guys,

I think I know the answer to this one already (aka send to the depot) but how on earth is one supposed to change the finals on an XTL5000? The heatsink/transistor combination looks like its soldered down fairly heavily and would need a proper IR SMD rework station (i.e. with preheat etc) to remove the heatsink and transistors without damaging any other components...

I'll post a picture for those un-initiated.. but in the mean time, am I out of luck? Is this board ready for the bin?

Cheers,
Matt

 

[Mars]

I don't mean to distract from your question or hijack your thread, but once again, this proves why the entire XTL line is poorly-designed and nearly impossible to troubleshoot or easily repair. The uniboard design is terrible and ill-conceived. Long live the Spectra!

Matt, I sincerely hope you're able to get this resolved.

 

[MattSR]

Oh no problem at all, I think you're spot on with that assessment..

 

[com501]

Terrible and ill conceived, ah, yes, but it's cheap. When the entire cost of an XTL5000 mobile at Moto engineering is around $200, it is FAR easier to throw circuit boards at it than design field replaceable componentry. I think the amount of shops that don't even do board level work is steadily diminishing, too.

 

[PRC148]

It's disposable, like everything else today.

 

[freerider-aust]

if this is the same as the high power mcs 2000 it looks like a special jig that is heated to remove what looks like a heatsink.it ended up being to hard.
another way i have removed hard to get out parts is to destroy the part carefully without damaging pcb or other parts.then desolding each pin.this method is a last resort
unfortunately it sounds like a throw away part
good luck with this one

 

[com501]

The trick on these boards is a heated bed, and a hot air focus head on the top. Pricey equipment for a dubious return on investment. You STILL have to have the correct autotune software and service monitor to even get the PA bias close to correct. The manual method just plain sucks.