T520 Quad Core Support

[esaym]

I just bought a lightly used T520 with i7 quad core cpu from ebay this week. This was an upgrade from my T61 that had just gotten too sluggish and lacks the ram for what I use it for (software development projects with Linux as the OS)

So far the T520 is awesome. I had no need for the dedicated nvidia card (and it doesn't work well under linux anyway) in it so I disabled it in BIOS. But to my horror, I discovered that the DVI ports on the docking station are connected to the now disabled nvidia card. I tried to use only the nvidia card but did not have much luck getting it to play nice in Linux. I debated on just sending the laptop back to the seller but I had already spent the whole day disassembling it and cleaning it, and I do like it overall. Plus don't know what other laptop I'd get that has good linux support and is also quad core...

I dug around on google and noticed a few forum posts from people with the bare bones version of the T520 installing quad cores bought off ebay or swapped from other laptops and it seems to run fine.

So what I've done is grabbed a used T520 motherboard without the nvidia card off of ebay. The listing said it had bios version 1.36, my current T520 has bios 1.32 so I think this board must be newer and so compatability shouldn't be an issue (I've heard that early T520s did not support i7 quad core).

So if all I want is to use integrated graphics and have a quad core i7 cpu, will this motherboard swap work? Other than being a pain in the rear, everything I read seems to point to yes. But just checking or wondering if there will be any surprises....

[esaym]

The only other issue I can think of would be fitting my current heatsink to the non nvidia board.

The board with nvidia:
http://tinyurl.com/hhtqy63

The board without:
http://tinyurl.com/j3s7gus

Everything seems like it should line up, hole-wise. But to the left of the cpu socket on the board with nvidia, there are some ram chips that the heatsink touches.
On the non-nvida board, there is some chip looking thing there as well but it looks like it is thicker than the ram chips on the nvidia board.

It seems though that I can cut off the part of the heatsink that will interfere though. I assume I shouldn't try to cut through the heatpipe, I guess its full of juice or something
http://tinyurl.com/jugxt32

Yes I have a dremel

[hhhd1]

There is a whole thread about using the nvidia heatsink on the integrated motherboard , because it provide much better cooling,, and the integrated heatsink is designed for dual cores only too, the good news is that it will work, but be cautious, i suggest to not cut the heat pipe itself, rather just cut the extra pads that cover the GPU RAM, and bend the heatpipe upwords slightly so it is clear from the board, which should be enough.

And be extremely careful when handling that part, it is extremely fragile, and easy to bend and break without any tools, or with slight intervention of pliers ( i learned that the hard way )

There is nothing liquid inside the pipes of the heatsink you posted pictures off, .. if there was, then it may have evaporated without me noticing ..

before doing so, there is also another thread here about getting the nvidia working with debian or something close to debian, .. and it was successful, try to look it up, it was relating to either a w520 or a t520, which should be the same for linux support.

[esaym]

So the swap went alright, took all saturday though. There were a couple of surprises. I'll post some pics and talk about those in another post.

But the issue right now is the battery is not charging
Everything was fine after the swap and I used it until battery was down to about 40%. I plugged it in and went about my business. I have custom battery thresholds to charge once below 50% and stop at 70%. This morning, I'm pretty sure the battery was at 71%, I used it for about an hour and it was at 66%. I tried to force it to charge while I ate breakfast, so I set the thresholds to 80, 85%. But when I got back it was still at 66%. I used it some more on battery. It is now at 42% and still not charging. I've rebooted a few times. Tried plugging it in while turned off and even tried seeing if it would charge on the docking station but the results are always the same. The battery light on the back of the LCD flashes while plugged in, so it is if the charging circuit is working, but just not charging.... anything else to check ? I'm going to try next to put in the original hard drive with windows 7 on it and see if that does anything. Next I'm going to see if I can rig together the old motherboard and plug it in with the battery to see if I can get it to charge on the old board.

[esaym]

Tried booting into windows and using the normal lenovo power manager tools to see if that would get the battery charging. No dice. The stats say the same thing as in linux, charging but the current rate is always 0. I wasn't able to rig the old mother board up into a state where it could charge a battery. Can't seem to get it to turn on since it has no cpu,ram,etc and don't feel like messing with it.

I'm debating on buying another mother board. There are still many used ones on ebay, but I see some "new old stock" ones from China for not much more. I kind of like the idea of having a new board, but being from China makes me feel that it might just be a reject. Any thoughts?

I'm still not 100% sure this isn't the battery. I'm not sure how to test it, and I don't have any others to plug into the laptop to see if they charge.

So thoughts? Buy another motherboard? Buy another battery? Or buy another laptop? lol

[Hans Gruber]

Reinstall your OS and see if that makes a difference. Do not install any drivers and see if it charges on it's own.

[ajkula66]

So thoughts? Buy another motherboard? Buy another battery? Or buy another laptop? lol

Another battery - a genuine, *known good* one even if it's used - would be my choice.

Good luck.

[jaspen-meyer]

Yes, those pipes are filled with 'juice', see
https://en.wikipedia.org/wiki/Heat_pipe

[hhhd1]

the battery from your t61 should be electrically compatible with the t520's one, try to line up either battery with either laptop, see if it works.

this document suggests that they are electrically compatible,
https://support.lenovo.com/eg/en/documents/ht037009

your problem IMO sounds like a motherboard issue, but can't be sure without testing like you should.

[esaym]

the battery from your t61 should be electrically compatible with the t520's one, try to line up either battery with either laptop, see if it works.

this document suggests that they are electrically compatible,
https://support.lenovo.com/eg/en/documents/ht037009

your problem IMO sounds like a motherboard issue, but can't be sure without testing like you should.

Couldn't get it to line up. Went ahead and bought a used battery and a supposedly "new" motherboard from ebay. Should both be here on Friday and then we'll go through this again

I did notice after the laptop tried to "charge" the battery for 12+ hours, it starts flashing the battery LED orange. If I take it out and reinsert, it goes back to blinking green. So could just be the EC saying there is something wrong after not seeing current, or it could be the battery is cooked, don't know

[esaym]

Yes, those pipes are filled with 'juice', see
https://en.wikipedia.org/wiki/Heat_pipe

Yea, I went ahead and cut the one off I didn't need, sprayed out the inside with electronic cleaner, and put a little JB-Weld over the opening

[esaym]

Hmm weird. I noticed earlier that the current battery charge percent had "increased" from 36% to 37%. Figured maybe it just crept up from sitting. Charge current was 0 at the time. But then a few hours later I noticed it was at 54%. The battery state was 'idle' instead of charging. Unplugging from AC caused the battery led to flash orange and stayed that way until I unplugged/plugged the battery. It sat for about a minute 'charging' but the rate was 0. Then all of a sudden a minute later the current went from 0 to 200ma,then 800,then 1200, then back to 0 where it sits for a few seconds and then bounces around again before going back to 0


edit: right now it is holding steady at 3400ma

[esaym]

So I decided last night to go ahead and try to discharge the battery as much as possible in order to maybe 'calibrate' it since it was now charging. Figured since I had another battery on the way and another motherboard that I'd throw it through the works and really make sure its working. With prime95 and some gfx tools running, I manged to get the battery stats to says it was discharging at 69 watts I let it run that way for about an hour until the battery was down to 10%, I gave it 20 minute break and then ran prime95 some more until the battery was at 1%, then I plugged in the AC and suspended the laptop and went to bed (actually while it was charging I gave the laptop a good shakedown/beating to make sure nothing was 'loose' or anything. The battery stats showed a steady 3400 charge rate during this process so it seems nothing is loose or shorted inside).

About 8 hours later it was reporting the battery was at 96%. The status was still 'charging' but the battery led was solid green (instead of blinking) which I thought was weird. I unplugged/plugged in the battery and played with thresholds but could not get it to show any charge current. Discharged it to 66% while working this morning but still same results, status is 'charging' but current is 0

[esaym]

I should add that I got my new 'used' battery in the mail today but it just gave an orange flashing battery led and all the stats for it are 'undefined'

[esaym]

Oh also forgot to add that while digging through all the battery stats under '/sys/devices/platform/smapi/BAT0/' (at least in linux world) that there is one called "remaining_percent_error"

I have no idea what that is. The docs just state that it is the estimated error of the remaining_percent. Goggling doesn't help much, but it does show a few people that posted their values and most of the time it seems to be below 5. Mine has been dancing around between 5 and 10 but right now it is 23 (up from 22 from this morning). Maybe perhaps with an error rate that high, the EC doesn't know what the actual charge percentage is and thus doesn't issue any charge current... don't know.

[esaym]

Oh also forgot to add that while digging through all the battery stats under '/sys/devices/platform/smapi/BAT0/' (at least in linux world) that there is one called "remaining_percent_error"

I have no idea what that is. The docs just state that it is the estimated error of the remaining_percent. Goggling doesn't help much, but it does show a few people that posted their values and most of the time it seems to be below 5. Mine has been dancing around between 5 and 10 but right now it is 23 (up from 22 from this morning). Maybe perhaps with an error rate that high, the EC doesn't know what the actual charge percentage is and thus doesn't issue any charge current... don't know.

In follow up to this, I went ahead and left the laptop on until it completely died. So now the battery is completely dead and won't charge. Really makes it a pain when needing to move from room to room

[esaym]

I have a replacement 'used' battery on the way. I also went ahead and bought an oem 6 cell off of amazon. The problem is I go out of town for a week for the US python conference. Really need to have a working battery and I am out of time. I have another motherboard I can try, but I am not convinced that it is the motherboard. I flashed my BIOS as well. I went from version 1.36 to 1.45 which also included an EC update so it would seem if there was any bad state in the EC that the new firmware would have flushed that out but I still have the same problems.

To make matters even more bizarre I find two reports of similar issues and then after some time, the charging circuit magically 'heals' itself:

The same issues I am having, worked fine, then stopped charging battery so bought another one but that one wouldn't charge either:
http://www.justanswer.com/laptop/6xp8s- ... years.html
After leaving the laptop with "no power" for 10 days, it started charging both batteries just fine

And from this forum, wouldn't charge so bought an external charger and used that for a month and then the laptop started magically charging again:
http://forum.thinkpads.com/viewtopic.php?f=29&t=78391

So in light of all that, I just went ahead and grabbed an external charger off of ebay, it came with a 90w adapter as well, I was wanting a spare anyway. The external charger isn't too big so I'll just take it with when I go out of town.

I'll give this laptop a month to see if it will magically heal itself, after that I guess I'll swap the motherboard again

[esaym]

Now the plot really thickens. After having a completely dead battery for 2 days, it just now decided to start charging... Really weird.

In other news I was looking schematics most of the evening for fun. Looking at the similar x201 schematic: http://plan9.stanleylieber.com/hardware ... matics.pdf

The charger chip is probably the TI bq24741 http://pdf1.alldatasheet.com/datasheet- ... 1RHDR.html

I know thinkpads have an embedded controller, but there are apparently two of them,
The main one (which I think is flashable) is the R4F2112VBG25HV. But it has only one connection to the battery charger circuit, which is 'current select'. But also has the data and temperature connections from the battery.

The other chip ROHM BU77700KVT-GP has several connections going to the charger circuit, including the main 'BAT_CRG' which turns on the charger.

My theory is the charger is either mostly kept in the on state and the charge current is set to 0 by the R4F2112VBG25HV unless charging is called for, or the R4F2112VBG25HV might interface with the ROHM via SPI so it could send messages down that pipe to turn on the charger, but that seems overkill.

[esaym]

Yes, those pipes are filled with 'juice', see
https://en.wikipedia.org/wiki/Heat_pipe

Yea, I went ahead and cut the one off I didn't need, sprayed out the inside with electronic cleaner, and put a little JB-Weld over the opening

For those interested in my handy work:

The cutting was easy..but went a little too deep
http://tinyurl.com/he56wdc

But nothing a little JB Weld couldn't fix:
http://tinyurl.com/zmvwh65

I should also add the board with the nvidia chip uses a different fan plug size (4 pin), the normal motherboard has a 6 pin plug. The 4 pin plug will work in the 6 pin socket though, you just have to plug it in with the plug all the way to the left.

[esaym]

Now the plot really thickens. After having a completely dead battery for 2 days, it just now decided to start charging... Really weird.

In other news I was looking schematics most of the evening for fun. Looking at the similar x201 schematic: http://plan9.stanleylieber.com/hardware ... matics.pdf

The charger chip is probably the TI bq24741 http://pdf1.alldatasheet.com/datasheet- ... 1RHDR.html

I know thinkpads have an embedded controller, but there are apparently two of them,
The main one (which I think is flashable) is the R4F2112VBG25HV. But it has only one connection to the battery charger circuit, which is 'current select'. But also has the data and temperature connections from the battery.

The other chip ROHM BU77700KVT-GP has several connections going to the charger circuit, including the main 'BAT_CRG' which turns on the charger.

My theory is the charger is either mostly kept in the on state and the charge current is set to 0 by the R4F2112VBG25HV unless charging is called for, or the R4F2112VBG25HV might interface with the ROHM via SPI so it could send messages down that pipe to turn on the charger, but that seems overkill.

So once again, the charging stopped not quite full, just 93% before the battery went to status of 'idle'. Pulling the plug and letting it run on battery caused a rapid flashing orange battery indicator. I let it discharge some, but it wouldn't charge again.

That was last night, fast forward to about 2 hours ago. I was still wishing I could use the old motherboard to test the batteries, but everytime I plugged in a battery to it and plugged it into the wall with a kill-a-watt meter, I wouldn't get any current drawn. After looking at all the schematics last night, I noticed the power circuit from the docking port seemed to be wired (power-wise) a little different than the main power plug. I was curious what would happen if I docked the old half assembled motherboard into it with a battery so I did just that. Nothing happened at first so I hit the power button on the docking station. Quickly my watt meter started showing 6 watts being draw, then 15, then 25.. etc. It was actually charging the battery. And this was the old battery that I got from ebay that I couldn't get my laptop to recognize. I tried the other battery that was currently plugged into my laptop with state of 'charging' but with a current of 0. It too started charging! Clearly this is a mother board issue and not a battery issue...

I let the 'new' old battery from ebay charge for about 20 minutes, the watt meter was showing 50 watt draw most of the time. I checked for high temps on the old mother board with an IR thermometer since I didn't want anything to get too hot disassembled. I noticed a small chip that was about 100* F. On closer look, I could tell it was the TI charger chip I mentioned in the above post. Tried trying both batteries that were able to be charged from the old motherboard in the laptop but no dice, it still wouldn't charge them.

Figured I'd go ahead and change the motherboard out to this other one I bought since I had some time. But right after removing the keyboard on the laptop I decided to look closer at the charging chip on this motherboard since I knew where it was now located (just left of the keyboard plug and above the speaker plug).

And...it was actually growing a whisker off of one of the legs....
http://tinyurl.com/j7oubpf

So I scratched in between each leg and spayed electronic cleaner on it while sucking the residue up through a straw running to a vacuum cleaner hose. Now:
http://tinyurl.com/zcd7zal

Put the keyboard back on...and now it charges fine

[jaspen-meyer]

Hats off to you on that one. Great work.