[time-nuts] FE-5680A performance - replacing fried chip

[C. Turner]

I, too, am one that cleverly managed to cook the '240 chip:  Too many 
clip leads in too small a space allowed the +15 to get into the lock 
indicator pin while I was evaluating the unit.  It, too, worked OK after 
doing that.

Rather than order another 74ACT240 or even just an HC/T240, I rummaged 
around and found a 74LS240 and replaced it, making good use of my hot 
air rework tool:  Removing the old chip and installing the new one was 
far easier than it would have been had it been an un-socketed DIP!

I suppose I should be slightly concerned about mixing logic families, 
but I can't seem to get myself too concerned about it:  The lock 
indicator and 1pps outputs work fine and I doubt that I'd really care 
about a few extra nanoseconds here and there and the 'LS doesn't really 
hog too much current compared to the entire unit, anyway...

(FWIW, I can see the 1PPS on my ancient '465B scope, but having dim room 
lighting and turning up the trace intensity to full brightness helps:  
It *really* is hard to see, but it's 1uS wide, just as it should be.)

* * *

On random things, on one of my units it had the section of the board 
populated with the +5V switching regulator - but it was not powered up:  
A bit of sleuthing around with an ohmmeter and with the help of the data 
sheet for the regulator chip revealed which two jumpers needed to be 
installed and I was in business.  If someone else has the "10 MHz only" 
5680A with the un-powered 5 volt switcher on board and wants to know 
which jumpers to install, I could make that information available if it 
isn't to be found elsewhere.

The other "identical" '5680A unit had that portion of the board 
unpopulated, but a few minutes with a soldering iron, 7805, a pair of 
0.1uF monolithics and some wire-wrap wire brought that one online as 
well with the tab of the 7805 soldered to the ground plane in the area 
where the parts weren't and with a bit of heat sink compound on both the 
body of the regulator and the same section of the insulating sheet 
allows at least some of the heat to be transferred to the bottom panel.  
Since then, I've modified another in the same way and also had good results.

I was initially concerned about dumping a couple more watts of heat into 
the box, but attached to any semblance of a heat sink, it's no big deal 
- but I suppose that the extra 100-ish mA above what a switcher might 
pull is a bit annoying...  (When warmed up, the steady-state current at 
15-18 volts stays constant at 800mA for each of the units with the 
in-built 7805.)

73,

Clint
KA7OEI

EB4APL<eb4apl at cembreros.jazztel.es>  said:


> John,
>
> You are not alone in this trip.  Another member of this list tells me
> that he connected  the power to the lock indicator pin when he received
> his unit and he was eager to test it.  His 74ACT240 also got cooked but
> the rest of the unit continued working normally.