PDA

View Full Version : Popping a cherry



AlphaWolf
08-30-2007, 01:48 AM
My first prom removal ever.

http://www.dealdatabase.com/forum/attachment.php?attachmentid=6424&stc=1&d=1188451844

How you like them apples? :D That big capacitor there was a bit ominous looking, but with the technique I used it turned out to not be that big of a deal.

But this was more or less a simulation, as I did this on a dead tivo, and have no way of verifying that I didn't actually damage anything. That said, I have a few questions for the EE's here.

First of all, I did this using chipquik, which I applied using some soldering paste and a Weller model WM-120, which is a 12 watt soldering iron with a very fine tip, and the product page claims it has an 800F tip temperature.

I basically globbed flux all over the prom, then melted it into the pins with the iron. Then I globbed the chipquik all over the pins with the iron, stuffed the tip of the iron in between each pin and held it there for roughly 3 seconds each, ensuring to allow for plenty of time for the chipquik to absorb into the pins.

After that I took an ordinary hair dryer, and held it over the prom for about 20 seconds, then I pulled the dryer away and gently lifted the prom up with my fingers while it was still hot, and it slid right off. Then I just cleaned up the mess using the hair dryer and a solder pump (lost my solder wick.)

The first question is, is there anything inherently wrong with the method described above? Anything stupid I did that might have damaged something? Anything I need to be careful of?

Also, the dryer made the whole tivo (motherboard, chips, case, and all) very hot to the touch (hot enough to burn my hand,) is an ordinary dryer capable of damaging IC's?

captain_video
08-30-2007, 06:59 AM
Congrats! It looks like a pretty clean removal from what I can see. I haven't had the guts to try one myself but a lot of folks recommend using some aluminum foil to mask off the area and act as a heat shield to protect any components around the peripheral of the chip.

mrpenguin
08-30-2007, 08:18 AM
Looked like a clean removal. I use the tin foil heat sheild, but then again, I use the heat gun only psycho removal method instead of the clean and neat method you used. Hair dryier probably does not get hot enough to even melt the plastic battery holder, or damage the battery. That would be the only place I'd worry about it. Look at the holder, did it melt? I know the heat gun sure can :o

Speaking of the battery, just because it happened to me, I'd just like to point out if you remove the battery during this task, if it does not have a strong charge, your tivo may seem dead, but change the battery to a fresh one, it may boot up.

AlphaWolf
08-30-2007, 10:12 AM
Nope, battery holder looks completely intact.

One more question though, what are the odds that I can damage the prom itself by using the soldering iron I am using? Even though it is 12w, it claims an 800F tip.

I would figure that I would leave scorch marks all over the motherboard before I ever damage the prom, but then again the tip is so small that it was quite easy to avoid touching the board.

BTW, it turns out that my extra prom chip is an SST37. :rolleyes: Anybody know what is the right mouser part for the SST39 that can be programmed using the S1 dtivos? There are many PLCC32 variations and I am not sure which one it will accept. I saw another thread mentioning a specific part nomenclature but I can't seem to find that one listed

Jamie
08-30-2007, 10:59 AM
...

Anybody know what is the right mouser part for the SST39 that can be programmed using the S1 dtivos? There are many PLCC32 variations and I am not sure which one it will accept. I saw another thread mentioning a specific part nomenclature but I can't seem to find that one listedI've used mouser stock # 804-39VF0107INHE on Series2s and Series3s. Never done an S1, but I have flashed these in place in a Series2, IIRC.

mrpenguin
08-31-2007, 05:38 PM
I used SST39cf010-70-4i-nhe, mouser # 804-39vf0107inhe also as mentioned in this thread (http://www.dealdatabase.com/forum/showpost.php?p=261682&postcount=235) (may have more info there). I have also flashed them in the tivo as well, they work. too bad my IDE controller does not seem able to flash the 37's I would have spares.

Omikron
08-31-2007, 06:16 PM
AlphaWolf, drop me a PM with your shipping address and what version of PROM you need programmed for what TiVo and I'll send you one free of charge. It's the least I can do for all the help you've given the community. ;-)

AlphaWolf
09-01-2007, 02:06 AM
Thanks! But, I already made my mouser order for two pairs of SST39s and sockets though, so that may not be necessary. I just successfully extracted the prom off of my S1, and am currently in the process of figuring out how to seat the one socket I do have (this looks like it will be more difficult than the prom removal is.)

I am for all intents and purposes expecting the possibility of failure here, as the possibility exists that I toasted the SST39 during its removal (Granted this is a 12w soldering iron we are talking about here, I honestly have no concept as to how hot these chips can get before they are damaged.) The good sign is that thus far there are zero scorch marks visible on the chip or on the board, but from what I understand you may not necessarily see a scorch mark to have a damaged chip.

If it is damaged, I might take you up on that offer, as I would have no way of programming the prom.

As far as getting the socket onto the board (which is very hard to even fathom doing, with or without the middle peice) has anybody tried tinning the bottom of the pins first (or the pads first,) and then heating them from the top to melt the solder to the board on the bottom? If not, how do the pros out there think that would work?

ScanMan
09-01-2007, 10:24 AM
This is what I've done. Tape the socket down on two corners with (blue) painter's tape. I knock-out the center, but that's a matter of preference. Then apply no-clean flux to the pin(s) ready to be soldered with a toothpick. I initially do one or two on each side to stabilize the socket, then I can remove the tape. I find applying the flux to the pin first helpful, it will help the draw the solder in as well as clean the joint. Then get a tiny, tiny bit of solder on the tip (use the smallest diameter solder you can find) - I'm talking not much bigger than the size of a pin head here - and touch the pin for a couple of seconds. Should give you a nice clean joint. Then take your time, flux and do the rest; if you get a little too much solder on adjacent pads, just touch a piece of desoldering braid to wick it off and repeat. I tend to have an old toothbrush around with some isopropyl alcohol to clean a pin if I have to desolder and for final brush clean. I've done quite a few using this basic technique - just take your time and use a steady hand...oh and one of those magnifier/light things is real handy too! Good luck!

crashHD
09-01-2007, 07:33 PM
<WARNING>I'm absolutely not an expert at this</WARNING>

I could not do it with the center in place in the socket.

When I did one of my R10's, I tinned the bottom of the pins, lined the socket up precisely over all 32 pins, and held it there steady while I put a dab of hot melt glue at each corner of the socket. Then I heated the tips of the pins one at a time until they stuck to the pad. When I was done, I took the center of the socket (the piece I had knocked out), and stuck it on the pcb in the bottom of the socket well (after testing for shorts/opens),with a tiny dab of hot glue. When you're done, the glue that held the socket in place during soldering can be pulled off with a tweezers. I went the other way, though. After I was done, I went all the way around the socket with a fine bead of hot glue. I figured that way if I ever have to pull that prom back out, I should have less chance of pulling a pin loose.

I just want to reiterate, I am no expert at this. I just got my hands on some R10's for little to nothing, dared to mangle them (did actually ruin one), and found a method that worked for me. If there's any experts out there that see something I should have done/not done/done better, I welcome your input.

AlphaWolf
09-01-2007, 11:13 PM
****ing hell. I just lifted a pad while trying to put the socket on. I accidentally put too much solder on, tried to remove that extra solder, and the pad came with it. Shitty. At least I only did this to an S1 dtivo that I don't use anymore, but still, makes me not want to try it on my tivohd.

There is a possibility that I don't need to repair it though. Last night when I was verifying continuity, I couldn't trace that particular pad to anything to test it against, leaving the possibility that it is maybe unused?

Laying the pad out with a plain text diagram, | and - being pads, and \ being the diagonal cut corner on the board, the X being the damaged pad, it looks like so:



| | X | | | | | |

- -
- -
- -
- -
- -
- -
- -

\ | | | | | | | | |


Anybody know if that one is critical? If not then I won't bother with it. Otherwise, it is going to be difficult to find where that one goes as I can't follow the trace coming out of it.

EDIT: Bah, just figured it out, it connects to the CPU at least indirectly, so it probably is critical :/ I could either make a bridge wire going from the CPU to the prom directly (which would probably bridge it to whatever else it goes at the same time) or I could go from one of those little tiny eyelet things on the motherboard, of which I have identified two that go to it, one of which I think I can use. Trouble is figuring out which kind of wire can fit in that.

What is probably the best way to go about this?

Omikron
09-02-2007, 02:25 AM
If you're looking for which wire to use, I might recommend using some 30 AWG Kynar wrapping wire as it is exceedingly thin and easy to route into small vias.

Now, if you're careful and have a steady hand, you could try scraping off some of the mask from the trace that was connected to the pad. It will probably be easier and safer to solder a wire to the trace than trying to connect it directly to the CPU. Worst case, it's still screwed, right? ;-)

P.S. If it's a TiVo HD you're trying to mod, you might drop Jamie a PM to see what cool things I did to his two TiVo HD units.

Jamie
09-02-2007, 11:18 AM
P.S. If it's a TiVo HD you're trying to mod, you might drop Jamie a PM to see what cool things I did to his two TiVo HD units.I PROM mod'd several tivos in the past, but when I did my S3, I pulled a pad, much like you (AlphaWolf) did. I was able to repair it, but it was a pain. I don't think I'll be pulling that prom from the socket again, for fear of breaking the repair. Now I'll happily pay someone else with more smd rework experience and better tools.

I'm waiting for the TiVoHD's to come back, but Omikron added ttl-to-rs232 conversion circuitry for the serial port, and put a connector on the third sata header for me. It remains to be seen if that third sata header is useful for anything, and from what Omikron tells me, the labor to solder a connector onto it is non trivial due to other nearby surface mount components.

crashHD
09-02-2007, 12:44 PM
Trouble is figuring out which kind of wire can fit in that.


I used a piece of wire I stripped out of an junk ide cable, and connected my lifted pad to the motherboard eyelet.

AlphaWolf
09-02-2007, 02:17 PM
I PROM mod'd several tivos in the past, but when I did my S3, I pulled a pad, much like you (AlphaWolf) did. I was able to repair it, but it was a pain. I don't think I'll be pulling that prom from the socket again, for fear of breaking the repair. Now I'll happily pay someone else with more smd rework experience and better tools.

Hmm....its not that I am against paying for it (in fact I have already spent more money trying to do it myself than I would have to pay somebody else, and I went into it full well knowing this) but I just like being able to do these things for myself. But now it looks like I am going to break expensive things in the process, which I don't want to do. *shrug*

I seem to have perfected the prom removal part though, and I think I know exactly where I went wrong (and what NOT to do in the future) when I lifted that pad. Basically what happened (I know, I shouldn't have done this) is I used a solder pump to remove the excess solder. That in itself wasn't the problem though, the problem was that while I was melting the solder on one pad, I was simultaneously touching another pad, and overheated that adjacent pad when I hit the pump. That is what sucked the pad off. I should have just used solder wick, and I would have avoided that completely.

Soapm
09-02-2007, 03:45 PM
Hmm....its not that I am against paying for it (in fact I have already spent more money trying to do it myself than I would have to pay somebody else, and I went into it full well knowing this) but I just like being able to do these things for myself. But now it looks like I am going to break expensive things in the process, which I don't want to do. *shrug*.

"Every man has to know his limitations" - Clint Eastwood

lrhorer
09-06-2007, 05:31 AM
First of all, I did this using chipquik, which I applied using some soldering paste and a Weller model WM-120, which is a 12 watt soldering iron with a very fine tip, and the product page claims it has an 800F tip temperature.
Unless the soldering iron is heat controlled (this one isn't), there's no way to actually tell the tip temperature, but that iron with a normal tip sitting in room temperature air should reach about 800F. If you were to place the iron in an insulated environment (like wrapping it in asbestos), it could hypothetically reach well over 2000C. Unless the energy is being converted to some other form (like in a battery or a motor) the temperature of any object into which energy is being dumped will increase until the heat lost to radiation, conduction, and convection into the environment exactly equals the input energy. In the case of a tungsten filament inside an evacuated glass envelope of a 12 watt quartz-halide light bulb, that's about 5500C. In the case of a 12 Watt heating pad sitting in the open, it's about 40C. It all depends on the surface area and how it comes into contact with the environment. In the case of your 12 Watt soldering iron, if the iron is left in contact with a semiconductor device for too long, it could indeed destroy it, but it would take a pretty good time period - probably much more than a minute if the component is soldered on a board. An 800F lead temperature applied to any integrated circuit for less than 1 minute should not damage the component.


Also, the dryer made the whole tivo (motherboard, chips, case, and all) very hot to the touch (hot enough to burn my hand,) is an ordinary dryer capable of damaging IC's?
Absolutely. Again, it's all a matter of how much heat, how long, and over how large an area. With a 1500 Watt hair dryer, it is certainly possible to get a modest sized surface area to reach well over 1000C. Once again, however, it's going to take a while, and just being hot enough to burn your hand is not necessarily hot enough to cause electronics components to fail if they are only subjected to that temperature for a relatively short period. A temperature of 100C (boiling water at sea level) can cause a very bad burn, indeed, but exposure to 100C for an hour or so won't damage most electronic components. I don't recommend it, however. 'Try and keep the heat as localized as possible, and air cool the entire board with a fan if your work caused it to be too hot to touch.

lrhorer
09-06-2007, 05:42 AM
One more question though, what are the odds that I can damage the prom itself by using the soldering iron I am using? Even though it is 12w, it claims an 800F tip.
You want that tip as hot as possible. In general the best approach to soldering is to use the smallest iron which will get the job at hand done quickly, but to make sure the tip is as hot as the iron can get it. Do not place the iron in an air stream or lay the tip against anything. You want the solder on the component lead to melt as quickly as possible. If the tip temperature is only 500F, then it could take several minutes to heat the lead up to the point where solder will flow (about 450F), and the component will get very hot, indeed. If the tip temperature approaches 1000F, however, then the lead will reach molten solder temperatures very quickly, and you can get in and out before the body of the component gets very hot, at all. If the tip temperature gets much higher than 900 -1000F, then you are going to have problems with solder oxidation and shortened tip life, but within reason, the hotter the tip, the better.