3G slot - is it full PCI-E?

Hi, I was wondering if the 3G slot on the current (8.9") Aspire One is a true PCI-E slot, or just the USB part as used by most 3G cards, etc.?

I am interested in being able to use a PCI-E slot in whatever netbook I buy (possibly to hack in an Expresscard device) but it seems like mostly the 3G slots aren't true PCI-E. I'd prefer to not have to replace the wi-fi card with a USB device in order to get a PCI-E... (Maybe I should go back to looking at netbooks with Expresscard slots... though the options there don't look so hot either...)

 

Maybe these two threads will help you.

http://forum.eeeuser.com/viewtopic.php?id=24012

http://forum.eeeuser.com/viewtopic.php?id=39580

 

I appreciate the effort but I'm really not sure why either of those links would be any help to me at all.

First of all, both of those threads are on EEE User. Not Aspire One User.
Second, neither one answers my question. The first says that the Wi-Fi card slot (PCI-E) on the EEE (what model? Probably the 7", though maybe all of the rest of 'em, too...) supports the PCI lane but not the USB. The second says that someone successfully got a 3G card working in the 3G card slot - which means nothing beyond the fact that the 3G slot includes USB lanes.

So the first link says nothing about the 3G slot.
The second says nothing about whether the 3G slot provides PCI-E.
Neither thread applies to the Acer Aspire One.

I can appreciate if you simply don't know the answer to my question. With the manufacturers leaving the connector off to cut costs it's hardly surprising that most people haven't even tried putting the wi-fi card in the 3G slot, let alone some other PCI-E device... Thanks anyway...

 

Well, this is a good one.

I just read through some materials and checked the so-called System Block Diagram provided in AA1 Series Service Guide. Now I have the same suspicion.

On that diagram, the connection between ICH7M and 3G module is labled as USB, while the connection between ICH7M and WIFI+WMAX is PCI-E/USB, which might be indicating that the port for 3G module is USB only, although as we know the definition of a complete mini PCI-E port includes both PCI-E 1X and USB DATA+/-.

Moreover, that diagram also labled other three PCI-E connections from ICH7M to 2 Card Readers and the Ethernet. Then I decided to go over the Datasheet for ICH7 Family, which can be found in www.Intel.com. It reads explicitly that ICH7-M has four PCI Express root ports. Then it becomes clear. The four PCI-E ports of ICH7M in AA1 are all occupied, 2 by card readers, 1 for ethernet, and 1 for WIFI.

Maybe that's why we've never seen someone hacked the "Mini PCI-E" port on the bottom of AA1 with any other card working on it, only the 3G card which uses USB connection instead of PCI-Express.

 

Wow, that was some good information... I wonder if it'd be practical to replace the second card reader with some other PCI-E device? Probably be a lot of work and damn hard to fit it into the case... I guess it's back to the drawing board... Thanks.

 

So you're saying the unused PCI-E port on the back side of the motherboard, is USB based? if thats the case, i say thank you because ive been looking for a third unused USB port on the motherboard. I used the top-side PCI-E usb slots, and the unused 9-pin port on the underside, but wasnt sure whether the USB on the unused PCI-E port was active or not.

 

Well, I'd say that's a good idea, theoretically

Practically, I really doubt someone could neatly recover the PCI-E port linked to one of the cardreaders, and make use of it with some other Mini PCI-E device without leaving the surroundings a mess inside AA1. But, there's still hope, maybe it will be done anyway, just like last time someone hacked 110 with SATA port regained.

 

I think that is what the System Block Diagram indicates.

As for USB, that diagram shows 7 USB ports have been settled around ICH7M in AA1, 3 for external USB connectors, 1 for Mini PCI-E socket for WIFI, 1 for 3G module, 1 for the camera, and 1 embeded in the connection for SSD together with the IDE channel. The last one is somehow puzzling to me, since it seems no one has ever figured it out. And according to what Intel says, ICH7M has 8 USB ports integrated, so I guess the one with the 9-pin port on the underside next to the southbridge chip is the 8th, and due to its designed vacancy, it does not have to appear in Acer's diagram.

However, there is still another possibility. The one with the vacant 9-pin port could also be the one bundled together with IDE/ZIF. If that is the case, there should be one more unused USB port, either the circuit for it is not established, or it is still undiscovered.

 

What unused 9 pin port?

 

the CN11 port found here: viewtopic.php?f=43&t=4236

 

Are you sure the card readers are PCI-E devices? I looked online for "lspci" results from the AAO and I didn't see the card readers - at least not in the section with the other two PCI-E devices (wired and wireless LAN)...

At this point I doubt the 3G slot has a PCI-E lane - but if the "lspci" data is correct, there's still two open PCI-E lanes implemented, if not connected...

Of course, on the other hand, I've heard the card readers don't show up unless there's cards plugged in - maybe that affects the "lspci" results...

 

I'm now using Windows XP so I can not run "lspci" at this moment, but I just checked this with Everest UE 4.50. As reported, the two cardreaders are indeed mounted on PCI-E ports, #1 and #4, network controller Realtek 8139 takes #2 and the WIFI controller takes #3.

 

Well, that cinches it, then. Thanks again for the info.

 

You're welcome

By the way, I checked the two cardreader controllers on the MB when I opened my AA1 for some mods yesterday and just found out it would be extremely difficult to detach them from the PCI-E bus. In fact, I failed to figure out how they are connected, which made me more convinced that it would just be "mission impossible".

 

I have removed the Card Reader from the righthand daughter board. This was to make room for a USB ATSC tuner. So now the PCI-E ports for the card reader are available for the 3G connector. If the PCI-E lines are in the cable from the motherboard to the daughter board, then rerouting the PCI-E lines should be straight forward.

Anyone have an idea on how I can identify these PCI-E lines?

 

One of the JMB385 card reader controllers is on the daughter board. So I would think that the PCI-E bus wires need for a second fully functioning Mini PCI-E in the AAO are available in the harness connecting the daughter board to motherboard.

 

It could be, theoretically, although it is still almost not possible to figure out each line of the PCI-E bus from the daughter-motherboard connection. I think the only way that might work is to ask ACER for the pinouts information of that connector, or maybe if lucky, it could be found somewhere else.

 

Isn't the 2 card reader on 1 PCI-E connection? The JMicron Flash controller has 2 channel. 1 channel connects to the right card reader and the 2nd channel connects to the left one. So 1PCIE for WIFI, 1PCIE for the 2 card readers and 1PCIE for the network controller, and one spare one which is where the "3G" card is.

 

Pretty sure there are two JBM385 chips. I am assuming that is one PCIe per JBM385 chip.

 

Or there could be 1 JMicron JMB382 chip connecting to one lane of PCIe. I'm assuming Acer is cheap that they wouldn't waste more money on 2 chips instead of 1 more "efficient" and cheaper chip. I could be wrong though.

JMB382:
Host IF: PCIE
Card Reader: x2 Combo
Package: QFN64
"“x2 Combo” means 1 set of “4-in-1 Combo Memory Card Reader Mechanism” supporting SD/MMC/MS/MS Duo/MS-HG (1/4/8-bit) & xD and 1 set of “3-in-1 Combo Memory Card Reader Mechanism” supporting SD/ MMC/MS/MS Duo/MS-HG (1/4/8-bit)”. "