iBook G4 Fan Mod (rewiring) questions?

[Cox Orange]

I have have a spare ibook G4s and want to use the mainboard, DC-in-Board etc. and put it into a little square custom made box (just as big as the dimensions of the Mainboard.

My question is:
Can I cut the connector of the fan and solder a bigger fan to the connector (maybe with a resistor, which limits it to 5V operation)?

These are the specifications by Delta for the stock fan:
BFB0405HHA
Rev.: -A
Functions: -R00 / -F00
rated Voltage: 5V
operating Voltage: 4.0V to 5.5V
Input current: 0.16A
Input power: 0.80W
speed: 8500rpm
air flow: 0.085m3/min. / 3.00hcfm
static pressure: 14.33mmH2O / 0.564 inH2O
noise: 34.5dB(A)

Can the fan pull more from the logic board, than the stock fan is rated?

 

[Intell]

Can the fan pull more from the logic board, than the stock fan is rated?

Not likely. The power requirements inside laptops is very tight. Any changes may cause things to blow or stop working. It also may not be a good idea to run the iBook outside of its case. The case bottom and the topcase/palmrest are designed to channel incoming air over its heatsink and exhaust it out the rear. Without this, the heatsink will have to rely on ambient airflow to cool itself.

 

[Cox Orange]

Not likely. The power requirements inside laptops is very tight. Any changes may cause things to blow or stop working. It also may not be a good idea to run the iBook outside of its case. The case bottom and the topcase/palmrest are designed to channel incoming air over its heatsink and exhaust it out the rear. Without this, the heatsink will have to rely on ambient airflow to cool itself.

Oh, I didn't know that the topcase is so thoughtfully designed.
I thought it is just the metal shield keeping the heat in (along saving it from interferences) and with the heatsink and fan combo.
May it be, that the way it is packed so tightly ensures that the fan can suck the air? (It seems to have a high static pressure, too).

I actually have to admit I had it running with the topcase off for a few days and the side panel off (which is around the ports) and the metal shield loosend where the HDD and the optical drive is. Maybe I just had luck, that room temperature was 19-21°C.
I had put an external sensor in the HDD bay, where I had the SSD, and it never climbed above 51°C. Always while encoding in a 16h duration. But I also have to admit, that it shut down twice. I guess that was a precautionary heat shut down, then?

Would you think I'll be able to pack it tight enough, if I put a metal on top of the existing metal shield, when I put it in the box?

Hm, I guess I'll better scrap the fan swap idea, to not destroy the logic board, but I could give the box "design" a try.
There is an additional problem that comes to my mind: if I pack the opt. Drive above the Mainboard (just like in a mac mini). I will get to much heat over the CPU/VGA chip, when burning a DVD, right?

 

[Intell]

The iBook's fan draws in cool air from the side vent, then it goes over the hard drive and between the top of the heatsink where its fins are and the topcase. It is then sucked by the fan and blown out the rear. Without the topcase, the fan just blows around random air that doesn't cool anything.

 

[jrsx]

The iBook's fan draws in cool air from the side vent, then it goes over the hard drive and between the top of the heatsink where its fins are and the topcase. It is then sucked by the fan and blown out the rear. Without the topcase, the fan just blows around random air that doesn't cool anything.

Very true. I've also noticed by just removing the keyboard (not RAM or cover), the whole system generally cools down faster, probably because you're getting more air from right by the heatsink, sucked through the heatsink and RAM, and out the back.

 

[Cox Orange]

I got to revise, what I had said above about having had two eventual thermal shut downs. I discovered it was a flimsy DC-in-Board (when I did not walk carefully through the room).
I monitored the temperatures during a h.264 encode over 16hours and the fan would kick in (not spinning constantly or for hours), but would still lower the temperatures and stop spinning later.
I had the top case off, the airport card and shield off, the big metal shield (over the logicboard) was on, though screws unscrewed. Optical drive taken out. SSD (50°C inside) in HDD Bay.
If you think about it, when the ibook is closed, the fan sucks heated air from the HDD and uses this to cool the CPU/GPU.

 

[Intell]

The air from the hard drive bay is still cooler than the air around the heatsink. Then not all the air goes through the hard drive bay before going across the heatsink.

 

[4JNA]

almost all laptops and small form factor computers are designed this way. pull in from multiple sources, and route from coolest to warmest components before exit.

take a mini apart, air is pulled in below the logic board, routed over the passive chipset heatsink and hd,then pulled in and blown across the cpu before being blown out the back.

i won't go into thermal dynamics or design, but this is by design. cold/cool air over the warmest component straight away leads to condensation in extreme situations, and water is bad juju for delicate electronics.

i've done a project like the OP, custom all aluminum enclosure just a bit bigger than the ibook logic board. i used a CAD program to design a two piece enclosure. the top case had spots for copper blocks over the cpu and other chip, attached the copper blocks to the top with arctic silver, and used the same on the chips. on the outside of the top case, i used 'fins' to keep the entire thing cool, looked a bit like a heatsink with cables plugged in. i was going to call it the 'Wise Terminal' but thought Wyse might not get the joke and send in the law squad. silent except for the original ibook fan which i kept just to pull some air over the rest of the logic board for safety and piece of mind. used a CF card for the drive, worked really well. i designed and built it for the 'MacMod' competition which died when machtech bought macmod. RIP

i long ago sold the device, but will see if i can find either pics or the dude i sold it to to see about getting pics. easy come, easy go, always another project and all that.

best of luck OP, pics after you get done. always fun to see other peps projects!

 

[Cox Orange]

Congrats on your project, though (even, if you do not have pictures anymore). Your project sounds as the best to be desired heatwise (Aluminium case + copper bolts etc.). Didn't the case itself get quiet hot? (I thought about making contact with a CPU to a huge alluminium plate for a desktop - like a PowerMac G4 - but then decided that this would be unwise (static charging - after walking around, accidentally put fingers to the metal that conducts the electricity - grounding, etc.... I do not understand enough of these things anyway.)

Let's see, what comes out of my project. I will update here.
But still no solution for adding a bigger fan...
...(I guess I will have to keep the original fan). I have to put a bit of solder iron on the little "arms" that go to the DC-In-Board behind the dc-jack (to make it not "shaking" anymore). Would it be possible to grab some power there for a bigger fan?

 

[4JNA]

...But still no solution for adding a bigger fan...
...(I guess I will have to keep the original fan). I have to put a bit of solder iron on the little "arms" that go to the DC-In-Board behind the dc-jack (to make it not "shaking" anymore). Would it be possible to grab some power there for a bigger fan?

i would not add additional power requirements to the original fan circuit as it was designed with specific fan/power specs in mind, just my two cents. if you really want a bigger fan, power it from one of the USB ports. you loose a port but gain a fan, and if you are designing your case anyway, include a little space for a powered USB hub to turn the remaining port back into 4. you could power the hub from the input board with no problem as there is no battery to charge. presto! big fan and twice the ports.

mine did get a bit warm under load, but nothing like the original ibook, or not even close to a 1.67 powerbook under load, those things could light your pants on fire if you were not careful.

was doing a lot of industrial/military work at the time, and got the idea from using something like THIS case. these were all peecee things, but had an ibook with a dead screen and faulty keyboard, etc, and thought why not turn it into something cool. bit of late night design, i happened to be friends with a machine shop owner, a couple (several) free beers later, PRESTO!, cool desktop ibook in a custom case.

can't wait to see what you come up with. best of luck.

 

[Cox Orange]

Thank you for the idea for the USB/fan etc. I will think about it. I might keep the fan as it is for easiness. I am not a fan of shared bandwidth on USB... but the 2 stock USB ports are shared anyway, aren't they?

This was for the 1,2GHz G4 late 2004, 12" and now something different. Because...

I got some new parts. Along minor translucent ibook G3 and PowerBook parts, I got 2x ibook G4 1,33GHz complete, which I actually planned to harvest the DC-In-Board from for my working own preexisting everyday in use 1,33GHz G4 that has a sleep problem.

Now my thinking is, I would like to use one of the "defective" 1,33GHz one for my "box-mod".
But there is a problem:

Exemplar A) *1 has defective display data cable (I think). I tried testing two different LCDs, 3 data cables and 2 inverters. Still the same symptom: LCD gets fully lit up, but no picture. (1 of the 3 data cable even gave a black screen and got really hot).
So, this would be fine, since I only use the external mini-VGA-out, which is working and shows a picture.
(Though it doesn't matter, nevertheless, I would still be interested what is going on with the lit up LCD with no picture).
But there is another problem. *2 The socket for the cable that is connected to the Power-ON-Button was loose from the beginning and I had to press the left side of it to the logicboard. When I got to testing the third LCD data cable, I finally, accidentally ripped it off.
So, question: can the socket be soldered or the two wires of the PW-On-Button be soldered to it directly? I am able to start it by bridging it with a flat screwdriver (though it takes several attempts everytime).

Exemplar B): has no Pw-On-cable socket anymore and I can't start it with a screwdriver. No idea, what to do there. (The DC-in-Board is ok on both btw. and battery can be charged, too).

--------

Of course I can take the working 1,2GHz model from the beginning of this thread, but I would rather like to use a 1,33GHz one (it shortens iMovie encodes by 30-60minutes, from previous test and also it would have a better GPU (ATI 9550) than the 1,2GHz (9200, same as a mac mini).

 

[Cox Orange]

I made a picture of the socket.

The socket (black square) has two little "needles", where one would seat the connector on (little black dots in the picture).
For shorting it with a screwdriver I only need to connect, what I have painted blue and is on the board.
I wonder, what the little strands are for (red in my picture) since they come from the socket housing and would connect with two small leads(?) on the logicboard.

So I have:
- ibook G4 1,2GHz from the beginning of this thread (working)
as a plus the HD cable would be flexible and I would be able to move it on the back of the logicboard, and have place for an SATA-IDE adapter to put in a 2,5" SATA HDD.
- ibook G4 1,33GHz (exemplar B) which I do not get to boot (no socket, screwdriver trick does not really work)
- ibook G4 1,33GHz (exemplar A) which is bootable, but I have to use a screwdriver to start it. Answer needed, if can be soldered. Display lights up, but no picture (which is ok for the mod). External VGA-out works.

Any ideas on the power-button, resoldering, despite taking the 1,2GHz model?

 

[4JNA]

the black socket can be re-attached, but you need some decent solder skills to pull it off. the 'other' little legs are the support/mount points, they are supposed to help hold the thing on the board, they carry no signal. metal posts soldered to mount pads, you can ignore them and just worry about the two signal leads. i've had lots of luck using 'super glue' to hold the socket to the board, and then a quick touch to solder the leads. another idea would be use a couple short bits of wire soldered to the leads on the logic board (where you have been touching the screwdriver) and then solder the socket onto the ends of the wires, it can be connected and taken apart carefully, and tucks into the modem area with no problem. pretty common problem, i've seen many, these are a few different ways i've dealt with it.

on the boards, both 1.33 boards sound like problems waiting to happen to me, i'd pick the 1.2. difference in speed in everyday use is not noticeable in my experience. best of luck.

 

[Cox Orange]

Thank you!

On the display, though I will discard it, do you now what is the case with the displays etc.? I am just asking out of curiosity. (both displays get lit up, but no picture, though I get the desktop on the external monitor). I tested two inverters and three data cables, thirs data cable seems indeed defective.

On the soldering thing. I am not sure, it seems it only starts, if I match the ends of the silver rectangles on the board (see the upper arrow in dark green) or should it also boot, when I touch it in the middle? (blue arrow). The red stripes is what I thought is the original position where the two tiny wires that come from the socket were.

 

[4JNA]

you have it right, the red lines are the 2 circuit lines you need to worry about, the outside blue bits are again just socket mounts and don't do anything. when i use the superglue trick, i cut the blue mount things off and make sure the bottom of the socket and the circuit board are clean/clear and flat so it holds well.

you have a basic 2 wire circuit. complete the connection with a switch, or a screwdriver, or a tricky bent paperclip, whatever, it passes the 5v onto the logic board as a go signal. wires, socket, however you wish to fix it is fine, especially since it's not going back in the ibook plastic. you could find a front panel control switch from an old peecee case and solder the two wire to the logic board and it would work fine as an example.

video thing has two possibilities based on years of supporting/fixing these things, either the LCD panel is defective, it would still 'light up' as you say, but would never show anything, or just random lines and crap, or the video chip on the logic board is starting to separate, the dreaded ibook video failure. last but not likely guess was going to be the signal cable, but sounds like you already swapped that with no change in symptoms, so i'll go with my two guesses.

swap the LCD if you have a known good panel to check with, or boot it as is and put just a couple fingers worth of pressure on the video chip to see if that changes anything on the LCD. best of luck.

 

[Cox Orange]

I got the thing about the blue bits, in your first post, but thank you for clarifiying on how you did it (cleaning/clearing these foots)!
Also thanks for hint on the "PC" power-on button, I already thougth about that and then thought, whether I might use a PowerMac G4 front panel (with the LED that chines behind the button).

For the red thing, I mean, shall I connect whatever thing (Paperclip, screwdriver, wires) to the very top end (see 1) of the rectangles or should usually every point (see 2) on the rectangles do?

I found a picture on the internet, where I could paint in what I mean (my camera is not as good). I feel, that it only starts, if I exactly hit the position on top (1) and not just anywhere on these rectangles (2).
(PS: maybe I am miss-naming it. What you can see in the picture, are these actually contacts, or is it the soldering lead that I look on, where the ends of the original "wires" had been under. So that I have to melt that before and put it away?)

I plan to have a translucent Apple icon (coloured in rainbow like) in the top of the box, that is lit up by an led. Could I attach the led-wires to the power on socket as well? (sorry, if it gets childish with my many questions).

---
On the display. Well, the external VGA-out works, so I think it should not be the GPU, or could a case exists, where a GPU is loose and the external still works, but not the internal LCD?

I could try the parts of the 1,2GHz model for testing what part of the display-chain is wrong. That is the only LCD+etc. I know, that is working. For the inverter, I thought, if the inverter is dead, it is black, with pictures in the dark.
I found a post about a microfuse on the backside, that has something to do with the inverter. http://www.ifixit.com/Answers/View/8912/Backlight+LCD+display+not+working,+how+to+fix+it

 

[Cox Orange]

Ah, I think I found the answer.

The "rectangles" are referred to as power pads (what you just said, sorry). The answer to my question is: you can connect it everywhere on the two pads. Like someone has done it with copper, here: http://s291.photobucket.com/user/spudnuty/media/Wiresoncircuitboardpowersocket.jpg.html

Pictures and and additional info from here: https://discussions.apple.com/thread/1807365?tstart=0
http://www.ifixit.com/Answers/View/11119/Can+I+fix+the+power+button+Connector

4JNA, I would like to say, that you perfectly described what to do and I use your info. I only gathered the links together, due to that I was looking for a picture.

For future reference, if someone digs out this thread in the future I will list what I have found along searching for a picture
- https://forums.macrumors.com/threads/409231/
- the following picture is interesting, too, it gives another option to repair it (mentioned on the above apple discussions thread on page 3) –->http://s291.photobucket.com/user/spudnuty/media/iBook13312powersocketnew.jpg.html

 

[4JNA]

you got it. your picture with the blue and white wire drawn in is exactly correct, and again if it's not the ibook plastics and original power switch, then anything will do.

the second post with the two internet pics, the first with copper, the copper is being used to replace/repair the damaged circuit trace. a lot of times when these sockets get yanked off the board, they pull the circuit board trace off as well. no trace, no connection, no longer going to work. you can use a small bit of wire, you can use copper like the pic, you can also use solder if you have the skills to pull a small trace (start with a 'puddle' at the trace end or the connection end and 'pull' a thin line of solder where the trace used to be). i'd pick wire most of the time because it was quick, easy, and wouldn't fail under load or heat. just my two cents.

I plan to have a translucent Apple icon (coloured in rainbow like) in the top of the box, that is lit up by an led. Could I attach the led-wires to the power on socket as well?

should be no problem, most pull a small enough amount of power that it should not matter or effect anything. check the rating on the one you are using of course, but should work well and look amazing.

happy to help (i hope), can't wait to see the final product. best of luck.

 

[Cox Orange]

These are the specifications by Delta for the stock fan:
BFB0405HHA
Rev.: -A
Functions: -R00 / -F00
rated Voltage: 5V
operating Voltage: 4.0V to 5.5V
Input current: 0.16A
Input power: 0.80W
speed: 8500rpm
air flow: 0.085m3/min. / 3.00hcfm
static pressure: 14.33mmH2O / 0.564 inH2O
noise: 34.5dB(A)

Can the fan pull more from the logic board, than the stock fan is rated?

I got to say something about the fans again. I already noticed a few years ago, that the 1,33GHz model had a sunon fan and in another ibook I found a bigger fan (longer fins, smaller motor diameter) and back then I thought, someone had eventually put a fan from a 14" model in it.

The quoted fan is, what I found in the 1,2GHz ibook.
Now I had a closer look at the two half-defective 1,33GHz logic boards. They have the following:
1x Delta BFB0405HHA 0,33A (what I read on the fan, but the manufacturer lists it as 0,16A at 5V)
1x Sunon Maglev Fan GC05401DVH-S (or 8?) K5A13J, 5V, 1,3W (which should be 0,26A, I could not find that on the internet).

So it seems, we now have a range of 0.16-0.33A. I wonder, where Apple set the maximum.
Also, I wonder which one is the better one. (I should do a test, if only both logicboards worked.)

---
Thinking about the design...

I had in my mind from the beginning on to do the outside like this:
http://awesomewallpapers.files.wordpress.com/2009/05/dark_wood_mac_by_adapt02.jpg?w=400
I guess that is something like rosewood (will find something similar).

The other idea would be to leave the rainbow Apple (which would be in the center) and just put a little Apple of the size of the tiny Apple on the front of the Sawtooth G4s. But I would use a white or a bright wood (poplar/cottonwood or maybe a birch/beach without red tint) Apple. (Maybe combining it with the function as a power-on knob).

No to the serious stuff:

I am thinking about where I will put the ODD. (the box will be with the long side looking away from me):
1) On the bottom front (under the HDD) it would most likely be good for heat reasons, but problematic given how near media would come out just over the table. May look stupid, too.
2) On the top above the GPU and HDD etc. at the front would be classically like the macmini. I play to put an additional heatsink block of 2-3cm on top of the original heatsink, so it would not be directly above it, also I will put part of the bottom metal shield under the opt. Drive, I am not sure, if that helps for electromagnetic interference or so?, but it might not hurt anyway.
3) on the right side over the CPU/GPU, the slot-in coming out over the ports. I think that would be the most unfunctional. (or under the ports).
4) slot-in on the left side at the bottom under the CPU/GPU. Hm, may be ok (just as number 2)).

I am thinking if bottom or above may automatically lead me ending up in a thicker or thinner box (actual size). Then there is the design (not as in construction, but in look), it may be that one or the other will (just) look (subjectively) thinner. Hm...

In the meantime I started looking around for a thinner (and heat-controlled) soldering tool, but I guess I will find something (was thinking about a small-budget Weller tool).

Also, in the meantime, I came to the conclusion, I want to keep the socket and not solder the wires directly to it. So the LED would have to go to either the USB port or the battery charger, but I will need guidance on that. Well, I know, it sounds a bit confused and not very clear, but I will come back to that, after I have finally started soldering the socket back to the logicboard.