Jump to content

US$189 AKiTiO Thunder2 PCIe Box (16Gbps-TB2)


Recommended Posts

By the way, I did continuity test from the screw hole (black probe) to the C2 lower part (red probe), and multimeter gave a continuous beep, so that is ground and means good news for us? :)

Yes, all test points near the DC jack look good. Nothing fused as yet. Only odd one is the -0.36V you are getting at the PCIe slot.

Can you attach the yellow/black cable, put the black probe on the screwhole (GND) and then test the 12V, GND and 3.3V PCIe slot pins (on the top side of mechanical key, pins 1 to 11)? :: ST-Service ( .

Link to comment
Share on other sites

My Gigabyte 750Ti works with a MacBook Pro Iris on Win8.1 with 6-pin power cable plugged (but doesn't fit inside the enclosure). 750Ti fans spin at full speed before starting the computer, which is quite annoying. The differences are that you use a NUC and your GPU is only PCIe powered.

By the way, what version of the Geforce drivers are you using for your Gigabyte 750 Ti?

Thanks for doing all the voltage tests on the Thunder2 PCIe Box. It would be great to know if it does indeed have a 25W limit.

Link to comment
Share on other sites

Yes, all test points near the DC jack look good. Nothing fused as yet. Only odd one is the -0.36V you are getting at the PCIe slot.

Can you attach the yellow/black cable, put the black probe on the screwhole (GND) and then test the 12V, GND and 3.3V PCIe slot pins (on the top side of mechanical key, pins 1 to 11)? :: ST-Service ( .

I did some additional tests yesterday and if I remember correctly, I wasn't able to get 0.36V again from the 2nd PCIe pin, the first 1-3 pins showed very small voltage (0.05V) when the black probe was on the screwhole (GND). Other pins were 0V (up to 11 pin). The same result, when TB card and its little cable removed. Only when thunderbolt cable was attached to the MacBook Pro, PCIe 2pin voltage was 12V and 750Ti's fans started to spin.

When the system was shut down, I also did continuity tests from the screw hole to the PCIe GND pins, and it gave a beep sound. A very short beep sounds occasionally when touching 2nd PCie pin and from the both sides of C2, switching probes place (black/red) resulted a short beep sound again.

Link to comment
Share on other sites

Accualy I was so happy after reading this, but when I entered Akitio site to purchase it, I just noticed it says BOOTCAMP incompatible, so ye that's why I finnaly did not make a purchase - can you tell me please how did you manage to make it works (mac + windows) on your MPB?

I have latest 13 inch MBP retina, 600W PSU and 580 GTX waiting for right solution - I'm not so much into tech etc. - all I need is something simple that works with no harm and something that's not as expensive as 999$ Sonnetech solutions...

Link to comment
Share on other sites

[ATTACH=CONFIG]12297[/ATTACH]

I did some additional tests yesterday and if I remember correctly, I wasn't able to get 0.36V again from the 2nd PCIe pin, the first 1-3 pins showed very small voltage (0.05V) when the black probe was on the screwhole (GND). Other pins were 0V (up to 11 pin). The same result, when TB card and its little cable removed. Only when thunderbolt cable was attached to the MacBook Pro, PCIe 2pin voltage was 12V and 750Ti's fans started to spin.

When the system was shut down, I also did continuity tests from the screw hole to the PCIe GND pins, and it gave a beep sound. A very short beep sounds occasionally when touching 2nd PCie pin and from the both sides of C2, switching probes place (black/red) resulted a short beep sound again.

This is bad news. It means we can estimate the PCIe slot can only reliably receive 25W, not the 75W per PCI SIG spec when using the DC jack. That is:

- there is no direct link from the DC jack to the 12V used by the PCIe slot

- power is routed across the yellow/black cables, via the edge connector to the PCIe slot board.

- the yellow/black cable is not of sufficient gauge to safely and reliable pass the 75W for the slot + say at least 10W for the TB board

- the edge connector also unlikely capable of passing 75W via it to the PCIe slot

- even if the yellow/black cable could pass 75W, it's likely then that the TB board applies a fuse to restrict input power to 25W

So the only safe workaround to get 75W to the shot would be to short 12V directly to the PCIe slot. One way of doing that is using a powered PCIe riser. Inconvenient as then are using the AKiTiO Thunder2 PCIe Box as a board rather than enclosure. So then:

Where can 12V and GND be soldered onto the PCIe board as a ATX PSU compatible molex plug so a PCIe riser isn't used?

Firstly, soldering onto any part of the AKiTiO internals isn't for everyone. You'd pretty much be blowing your warranty away. So as a service for those who aren't fazed by warranty loss, pls proceed to:

- check if the 12V PCIe slot pins are all tied together. If not, use only the 12V pins that are registering 12V on them for continuity tests.

- from the 12V pins do continuity tests against pads on the PCIe board to find suitable tap off pads. There are some silkscreen outlines headers on the upper left corner and top of the picture you posted (above) which would be a good place to start. Once you find the 12V points, check nearby locations for GND pins.

Link to comment
Share on other sites

Accualy I was so happy after reading this, but when I entered Akitio site to purchase it, I just noticed it says BOOTCAMP incompatible, so ye that's why I finnaly did not make a purchase - can you tell me please how did you manage to make it works (mac + windows) on your MPB?

I have latest 13 inch MBP retina, 600W PSU and 580 GTX waiting for right solution - I'm not so much into tech etc. - all I need is something simple that works with no harm and something that's not as expensive as 999$ Sonnetech solutions...

I refer to Tech Inferno Fan's answer http://forum.techinferno.com/diy-e-gpu-projects/7205-us%24225-akitio-thunder2-pcie-box-2.html#post98895.

Some GPUs do work with a standard Windows 8.1 Boot Camp installation. Unfortunately don't know about GTX 580. For OSX you need modified kexts, but I have not yet succeeded with that (external screen is black, however GTX 780 shows up under system report).

Link to comment
Share on other sites

I refer to Tech Inferno Fan's answer http://forum.techinferno.com/diy-e-gpu-projects/7205-us%24225-akitio-thunder2-pcie-box-2.html#post98895.

Some GPUs do work with a standard Windows 8.1 Boot Camp installation. Unfortunately don't know about GTX 580. For OSX you need modified kexts, but I have not yet succeeded with that (external screen is black, however GTX 780 shows up under system report).

Thanks for your answer.

To be honest, all I need is ONLY eGUP - makein' externet monitor working properly isn't a prioryty for me - I was planing to play o MBP screen.

All I need right now is to be sure, that after I purchase these solution there wont be a need to HACK A LOT (with im not good at), and ye I realize that my gtx 580 may not work, but my budget got some extra $ for new GPU (gtx is curently inside my desktop).

Link to comment
Share on other sites

Thanks for your answer.

To be honest, all I need is ONLY eGUP - makein' externet monitor working properly isn't a prioryty for me - I was planing to play o MBP screen.

All I need right now is to be sure, that after I purchase these solution there wont be a need to HACK A LOT (with im not good at), and ye I realize that my gtx 580 may not work, but my budget got some extra $ for new GPU (gtx is curently inside my desktop).

Only a few GPUs are just plug and play with internal retina screen (Optimus), most of them don't work even if using external screen. Internal screen also has performance loss and I think that your card is too old anyway. I chose EVGA, because people here have had success with Sonnet enclosures earlier. It would be very helpful if people who own MacBook Pro mid 2014 or late 2013 (13" or 15") and AKiTiO's box, would post their results.

Link to comment
Share on other sites

Where can 12V and GND be soldered onto the PCIe board as a ATX PSU compatible molex plug so a PCIe riser isn't used?

Firstly, soldering onto any part of the AKiTiO internals isn't for everyone. You'd pretty much be blowing your warranty away. So as a service for those who aren't fazed by warranty loss, pls proceed to:

- check if the 12V PCIe slot pins are all tied together. If not, use only the 12V pins that are registering 12V on them for continuity tests.

- from the 12V pins do continuity tests against pads on the PCIe board to find suitable tap off pads. There are some silkscreen outlines headers on the upper left corner and top of the picture you posted (above) which would be a good place to start. Once you find the 12V points, check nearby locations for GND pins.

According to the PCI spec pins 2,3 on side A and pins 1,2,3 on side B provide +12V to the PCIe connector. Pin 4 on side A and pin 4 on side B are ground pins. So, basically all +12V pins (2,3 side A; 1,2,3 side B) should be wired to the yellow cable on the molex connector and the ground pins (4 side A, 4 side B) should be wired to the black cable on the molex connector.

That means we should identify where are these pins on the back side of the board. I highlighted the region where these pins should be.

post-19093-14494998199101_thumb.jpg

@Tech Inferno Fan Is this correct?

Also, maybe there is another option. It is a little far fetched but I'd like to discuss it here.

I'm thinking to create a female-female PCIe 4x to x16 powered riser. In order to do that I want to combine a male-female x16 to x16 powered PCIe riser and a male-female x4 to x4 PCIe riser so that the final adapter has only female-female connectors.

post-19093-14494998199358_thumb.jpg

post-19093-14494998200025_thumb.jpg

Will such an adapter work ?

On the AKiTiO board there are a lot of other components which I have no idea what is their purpose. Maybe there is a clock there, which provides the frequency for the whole board. If so, using a female-female adapter will not be possible as there is no clock.

What are you thoughts about this?

Link to comment
Share on other sites

Also, maybe there is another option. It is a little far fetched but I'd like to discuss it here.

I'm thinking to create a female-female PCIe 4x to x16 powered riser. In order to do that I want to combine a male-female x16 to x16 powered PCIe riser and a male-female x4 to x4 PCIe riser so that the final adapter has only female-female connectors.

[ATTACH=CONFIG]12312[/ATTACH]

[ATTACH=CONFIG]12313[/ATTACH]

Will such an adapter work ?

On the AKiTiO board there are a lot of other components which I have no idea what is their purpose. Maybe there is a clock there, which provides the frequency for the whole board. If so, using a female-female adapter will not be possible as there is no clock.

What are you thoughts about this?

There you are assuming the edge connector used to plug the TB board into the PCIe board is using a x4 type PCIe connector. Certainly if it is then your proposed solution of making a powered x4 female to x16 female would alleviate having to do any soldering. You'd plug the TB board on the x4 end and the eGPU into the x16 end and away you go. No chassis space used up by a PCIe riser board and no warranty-voiding soldering either.

Someone with the AKiTiO board would need to verify the edge connector used to install the TB board has a x4 compatible electrical and physical connector.

Link to comment
Share on other sites

There you are assuming the edge connector used to plug the TB board into the PCIe board is using a x4 type PCIe connector.

The only verification I did is counting the pins, there were 21 pins and as per PCIe spec, 21 pins equals a x4 connector. I understand that is not enough to prove it is a x4 PCIe connector. It proves that only it is a compatible physical connector with PCIe x4 but no an electrical too.

Hope it is compatible electrically too :)

Link to comment
Share on other sites

The only verification I did is counting the pins, there were 21 pins and as per PCIe spec, 21 pins equals a x4 connector. I understand that is not enough to prove it is a x4 PCIe connector. It proves that only it is a compatible physical connector with PCIe x4 but no an electrical too.

Hope it is compatible electrically too :)

Interesting idea. There still can be a power limiting fuse somewhere after the "x4" edge slot however. Assuming that it is also electrically x4, is there any risk to break some components when feeding too much amperes to the slot Tech Inferno Fan? One drawback would be that the TB card is loose and more cables, but maybe they fit inside the enclosure.

Link to comment
Share on other sites

Interesting idea. There still can be a power limiting fuse somewhere after the "x4" edge slot however. Assuming that it is also electrically x4, is there any risk to break some components when feeding too much amperes to the slot Tech Inferno Fan? One drawback would be that the TB card is loose and more cables, but maybe they fit inside the enclosure.

Arclite would be mating the powered x16 female end to a x4 female end to attach directly to the TB board. It would eradicate the whole AKiTiO PCIe board altogether and feed 12V directly to the slot. Any fuses on the TB board would be circumvented.

Link to comment
Share on other sites

Arclite would be mating the powered x16 female end to a x4 female end to attach directly to the TB board. It would eradicate the whole AKiTiO PCIe board altogether and feed 12V directly to the slot. Any fuses on the TB board would be circumvented.

Ok, I misunderstood it. 4-pin yellow/black cable has to be in place even if using a normal powered x16 PCIe riser and without barrel plug. Otherwise GPU is never detected in device manager. Continuity tests also showed that there is a very short signal going through to the 2nd PCIe pin from near the tests points of DC jack(when testing without TB card), but not straight 12V line. Therefore I think that PCIe board has a bigger role and we cannot bypass it. But I may be wrong. I could play a little more with the multimeter if I know what to do... :D

Link to comment
Share on other sites

Ok, I misunderstood it. 4-pin yellow/black cable has to be in place even if using a normal powered x16 PCIe riser and without barrel plug. Otherwise GPU is never detected in device manager. Continuity tests also showed that there is a very short signal going through to the 2nd PCIe pin from near the tests points of DC jack(when testing without TB card), but not straight 12V line. Therefore I think that PCIe board has a bigger role and we cannot bypass it. But I may be wrong. I could play a little more with the multimeter if I know what to do... :D

Yes, please do play more with the multimeter. We are looking for 12V solder points on the AKiTiO PCIe board that have continuity with the PCIe slot 12V pins. If/when found, look for a nearby GND point. The 12V-GND pair then being a location where a molex plug could be soldered to the pads to bypass the need for the PCIe riser in order to supply 75W slot power.

Link to comment
Share on other sites

Why don't we collect some funds and send a Akitio your way nando? :)

Seems like the troubleshooting business would go a bit faster that way, hehe.

Thank you for the suggestion. I'm definitely interested in a TB2 eGPU but am underwhelmed by the Haswell notebook offerings with TB2 ports. Well, 15" rMBP is nice as is the MSI unit but both overpriced imho. REF: http://forum.techinferno.com/diy-e-gpu-projects/4109-egpu-candidate-system-list.html#post57511 .

Broadwell should hopefully give us some more affordable TB2 options to mate with the recently affordable TB2 AKITIO eGPU enclosure (Thunder2 PCIe Box). I'd like to see HP re-engineer their current ULV 14" ZBook 14 to Broadwell i7-quad spec along with a TB2 port. That would do very nicely :)

So while I'd quite like a AKiTiO enclosure it would be only useful for electronic probing. I have nothing to plug it into :/

Looking ahead, I've got questions about modding it to make it useful for both 75W (eg: GTX750. no pcie-power) and other high performance eGPU use (up to say 300W, 6P+8P). goalque and Arclite are onto it. Our discussion might be useful to others with the same goal in mind?

Link to comment
Share on other sites

Yes, please do play more with the multimeter. We are looking for 12V solder points on the AKiTiO PCIe board that have continuity with the PCIe slot 12V pins. If/when found, look for a nearby GND point. The 12V-GND pair then being a location where a molex plug could be soldered to the pads to bypass the need for the PCIe riser in order to supply 75W slot power.

Now very promising news! B1 connector (rear side J10) on the right corner of "x4":

post-28870-14494998203738_thumb.jpg

8 small holes through the board, 7 of them gave continuous beep sound with PCIe 2nd pin, one is different but couldn't get GND continuity to it. The nearest GND is marked on the picture, that point is very small, but I was able to get continuous beep when paired with a screw hole.

EDIT: Voltage metering also proved that this is the right place to feed extra power. All those 7 points gave +12V. Is this a standard size power connector?

  • Thumbs Up 2
Link to comment
Share on other sites

Now very promising news! B1 connector (rear side J10) on the right corner of "x4":

[ATTACH=CONFIG]12324[/ATTACH]

8 small holes through the board, 7 of them gave continuous beep sound with PCIe 2nd pin, one is different but couldn't get GND continuity to it. The nearest GND is marked on the picture, that point is very small, but I was able to get continuous beep when paired with a screw hole.

Some welcomed news.

Did you want to check the pins on just the right side of mechanical key of the "x4 connector" to see if you have matches for 12V, GND and 3.3V using this as a reference: :: ST-Service ( ? The vicinity of those 12V points you found certainly looking like it. This would then give inclination that it is *electrically* a x4 slot. This test could then be used by those unconcerned about warranty to solder a molex lead straight onto the 12V/GND pins there.

Can you then visually inspect if the pin pitch matches that of say a pci-e video card you have there by pretend sliding it in. Of course, not being a open-ended x4 slot will mean it won't go in. That would then give some assurance it is *physically* a x4 slot. This would then give reason for those concerned about warranty to researching for the the same x4 female to x16 female board as exists on the existing AKiTiO PCIe board.

Maybe AKiTiO could even sell that board as a part for direct 12V/GND soldering at a reasonable cost?

Link to comment
Share on other sites

A powered PCIe riser has pins 1,2,3 on side B and pins 2,3 on side A connected to 12V. I illustrated in the picture below which side is A and which is B.

Also pin 4 on side A and pin 4 on side B are grounds. Can you do a continuity test for all these pins too?

Thank you very much :)

post-19093-14494998204328_thumb.jpg

Link to comment
Share on other sites

Some welcomed news.

Did you want to check the pins on just the right side of mechanical key of the "x4 connector" to see if you have matches for 12V, GND and 3.3V using this as a reference: :: ST-Service ( ? The vicinity of those 12V points you found certainly looking like it. This would then give inclination that it is *electrically* a x4 slot. This test could then be used by those unconcerned about warranty to solder a molex lead straight onto the 12V/GND pins there.

Can you then visually inspect if the pin pitch matches that of say a pci-e video card you have there by pretend sliding it in. Of course, not being a open-ended x4 slot will mean it won't go in. That would then give some assurance it is *physically* a x4 slot. This would then give reason for those concerned about warranty to researching for the the same x4 female to x16 female board as exists on the existing AKiTiO PCIe board.

Maybe AKiTiO could even sell that board as a part for direct 12V/GND soldering at a reasonable cost?

I checked the first 11 golden fingers of the TB card (upper side) and 12V, GND and 3.3V all matches the x16 electrical specification (Side A). Visually it is also like a x4 connector :) Should I check the rest GND fingers from 11 till the end to be sure?

EDIT: Already checked from the both sides and they all match! :)

A powered PCIe riser has pins 1,2,3 on side B and pins 2,3 on side A connected to 12V. I illustrated in the picture below which side is A and which is B.

Also pin 4 on side A and pin 4 on side B are grounds. Can you do a continuity test for all these pins too?

Thank you very much :)

Yep. Both sides (A & B) 12V golden finger of PCIe riser had a correct continuity to the test point (one of the 7 holes).

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use. We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.