i use a top 2049 and i reflowed it using my smd reflow unit last week i do this with laptops mainboard and ps3 xbox etc so far i think ive fixed about 300 total failures units no boots beeps or irratic behavoir if you have a strange fail main motherboard sit it flat as you can do the underside away from big package first wait 20 mins till cold turn it over and add a 5p 1cent or something small in dia coin on top of the bga/pga chips in there center to add some protection from harsh heat and a little weight and hot air them back on @375 deg c dont wait around with the air spread it about evenly at the sides or overall of packages and green rims of bga bridge controllers and video chips {also cards chips !!!!! if you have a failed card !!}
also do the support ic's wait again 20 mins and re-assemble
all else fails instead of talking just blow hotter air at it
so....the top... i used the latest software !!! from www.top2048.com
{with win xp you must go to the device manger and the windriver tab and replace the driver with the one in the driver folder C:\Top\Topwin6\driver\xp seems there installer has trouble to update drivers even if you remove and reboot first the dos window states files already exist and are locked so doesnt overwrite them so there is a need to manualy add them
like unplug the programmer
now go to the control panel choose system >> hardware >> device mangager >>jungo>> windriver >> driver tab >> update driver >> install from a list or specific location >> dont search i will choose the driver to install..>>have disk.. browse for it in C:\Top\Topwin6\driver\xp {ok for win98 -xp} or C:\Top\Topwin6\driver\vista {ok for vista and windows 7} as needed
it will eventualy show you 'windriver' with no date
choose next the machine will set a restore point and install it
reboot !!! and then once the o/s is running plug in the programmer and wait 30 seconds then run the front end
this will fix the driver
even is you uninstall the older one and reboot some files from it remain locked and this step is needed there is also a note of this in the top site...
i noticed this top2049 there is very little solder on some ic's!!!!! hand solder of some buffers was needed using an eye loop X20-mag two dry joints on psu caps found
ive written many times about eproms lately ive gone back to the top and find with the latest software its much better with eproms 100ns / 120ns cmos eprom types set delay = 100us 150ns eprom types or above types set delay 200us 255ns eprom types or above types set delay 1ms dont forget to chill the eproms in the freezer to 0c or so... for best fix of electrons... cell weighting and also MOST IMPORTANT ~I THINK FOR ANY VPP IC if a write fails... sometimes !!!!! using c1000 but not with failed good known types {bad delay doesnt need a reset as below the software as is fails to reset propely all part ie the software drivers buffer and programmer
so... try this or further programming will also fail i find close the software disconnect the programmer
wait 25 seconds reconnect the programmer start the software and set it all up again with a new eproms erase the eprom that failed and try it at a lower or higher delay
as for the willem it writes most but the top at this point has written more of them the willem has trouble with 1001 ic also 010 in some delay ranges { i think its the current each uses from vcc 6.5v
and dont forget to add the ,vcc=6250 for cmos however ive been using 6500 for eproms period and it seems aok so far only 15 fails in 100 runs
so not bad the ones that failes i managed 10 further and found the other 5 had dirty pins so used a glass fibre burnish pen to clean the pins
so im please to report 100 out of 100 in the end for the top2049 with eproms for 27cXXX types all is fine
i take it back
youll find in previous posts ive made a listing to overwite the ??? range with eproms and the top2049.cfg file read these post i feel although this programmer isnt well documented this can be fixed and make many people with top's happy
with just a set of friendly .cfg files ill post a set of updated for all for eproms as a zip i think it will only need updated once every so often
tonight i programmed over 100 eproms with 35 fails 10 ic's gubbed somehow and 25 bad delay or bad erase vcc limits...i think a further erase restored another 19 so a further 5 programmed on the willem leaving 1 {dont know}
but i found the 0} erase for 5 mins 1} burnish the pins of all ic's using a glass fibre burnish pen 2} chill to 0C and blank check 3} choose the delay based on a magnitude of the chips refresh ie -10 is 100ns so 100us delay in the config panel 4} after a blank check or other loading etc usage let a little time 10 seocnds for the app to settle before programming 5} unplug all non needed usb units using the same controller as the programmer {each controller usualy has two ports most mainboards have 2 4 or 6 ports so 1 2 or 3 controllers so avoiding mice and keyboard on the same one of two is a must
and thats it
nmos seems to be very similar to cmos and seems need twice the delay so for a 100ns chip you need a 200us delay as for cmos for a 100ns chip you need just 100us delay set in the config menu setting any nmos ive used seems quite happy to flash at 1ms delay
i hope i helped
ill prepair another paste in epr section for all .cfg files in the top software
i know my eproms so its no sweat !!! and simple script
machine toshiba equium laptop usb only type p5b asus main machine thinking of adding dual lpt pci card both used for top and lpt via usb where i can
old p2 programming grumbling tank with xp sp3 for true serial and lpt
all machines bios set to enable bios plug and play aware o/s and lpt spollers services and drivers removed legacy detection enabled