I dont like the audio jack gone

@Martynas you are trying to imagine a setup that is not really “appropiate” for an iGPU.

For something that is 2x 2K with 120Hz or even 165Hz I would suggest to use an (d/e)GPU, especially if you are doing GPU-heavy tasks.

For most other setups, a thunderbolt or even usb c dock should suffice, even with data and display information going through the same cable (especially with DP-alt).

If you want an extreme solution, take a look at the Anker 777 Thunderbolt Docking Station.

Edit:
So my suggestion: Use one thunerbolt port for displays + power and the other for data only. That way, you should have the maximum efficiency possible.

2 Likes

@Martynas These are Thunderbolt ports at 40Gb/s. Also capable of handling data and power at the same time. High refresh rate gaming at above 4K? You’ll need an eGPU or the dGPU module regardless so you that either gets it’s own dedicated TB cable or dGPU module. Everything else can fit on a single cable. 10Gb LAN plus audio jack plus 4 ports at 5Gb/s is still only 30Gb/s. Charging is a non-issue. Unless you want everything to have it’s own port on the laptop? I’m a little confused if a dock is acceptable to you or not.

I personally use an eGPU with a built-in USB hub that services all my needs. Granted, my USB hub actually requires a separate cable USB-A cable to avoid leeching Thunderbolt bandwidth from the eGPU but still…

I have 3 USB ports in that hub, one for mouse, one for keyboard, one for a USB to RJ45 Gigabit adapter. All get max bandwidth I think. A proper Thunderbolt dock will have even grater connectivity built in.

EDIT: Sadly, no, all my extra connectivity does not get max bandwidth, they must all share a USB 5Gb/s connection. That sucks but oh well. Realistically I’ll never saturate that bus with the attached peripherals anyways. Even if I attach an SATA SSD in the eGPU enclosure as well, I’ll still never saturate it.

I’d only connect one screen because I don’t need more and I only talked about theoretical limit because 2x4K@60Hz should be the same amount of traffic as 4K@120Hz. But these docks don’t say that this is what they do. I tried the dock that I talked about and it didn’t do 120Hz on one screen but maybe my driver was bad or something.

Anker dock that you linked didn’t also say that they do more than 60Hz.
They say that they can do 8K which I don’t need but if it’s 60Hz on 2K then what’s the point of trying it. Also if they use CPU a lot or introduce lag then it’s also a problem.

So I think that unless somebody here tested a dock that connects via USB and gets more than 60Hz we can’t say anything.

Of course you can connect LAN, Mouse, Keyboard and everything. But Mouse and Keyboard don’t really take much traffic as a monitor with high refresh rate. Also LAN lag isn’t that big of a deal if it’s only shared with mouse and keyboard.
But if you’d connect a 2K@120Hz screen then it should in theory take about the same traffic as 4K@30Hz. And if the USB dock supports up to 8k@30Hz then it could in theory keep up but based on the way that DisplayLink driver works I don’t think they’ll allow more than 60Hz in any configuration because CPU would immediately go to 100%.

@Martynas Oh, you mean like office work at 120hz? at 4K? Oh, that’s easy. All one cable then. Thunderbolt will just embed a DP signal. Image pulled from manufacturer website. It’ll do 4K @120hz.

Honestly you don’t need something that expensive though for what you describe. 3 Displays at 1440P@120Hz would be pushing the limits of Thunderbolt but just one display? That’s easy and can be handled by one cable along with peripherals.

Alright. Maybe it will work then. We’ll see what will be offered during release.
I also don’t think people even need 4K screens but we’ll see.

If anybody has one that they bought then it would be a good thing to know their opinion and to know whether they work in Linux. There was a driver for Linux but it didn’t work properly or at all (I don’t remember).

Please take a look here.

2 Likes

this is sort of irrelevant when discussing the lack of a headphone jack. audio bandwidth is negligible in most cases for a thunderbolt cable, unless one actually wants to plug a mixer with 40x 24bit inputs on at the same time, which isn’t exactly comparable to a headphone jack.

extra screens are completely off topic here, and once again, audio bandwidth is negligible in comparison.

3 Likes

The outcome is that you only need one port to cover all your monitor needs. Whatever fancy the setup you mostly only need one port. Which means that there is space for a built-in audio jack because not all ports are needed. (also I’m not a big fan of slim front-left and front-right design and it could’ve been thick and accomodate a jack and a little larger battery)

So you’re not a fan of removing the jack but you’re not a fan of the jack quality in FW13?

Wouldn’t this also be an issue if they would bake-in a 3.5mm jack into FW16 and somebody would not like the quality? Also do workstation docks offer better 3.5 jacks and you know that you wouldn’t get into the same issue? At least you can change the 3.5mm expansion card.
Also inFW16 you can have the jack in any side of the laptop.

2 Likes

we agree here. ideally i would like to have a headphone jack because “why not” but of course in the fw case the whole audio component seems a bit sub-par, but still very usable. for whoever buys the 16 this won’t be a problem, but i still hope for a better component for my 13.

1 Like

I can handle the removal of the audio jack purely for the reason it means I can choose where I want my wired cans connected. Depending on the headphones I’m using and what I’m using them with, I can want the cable to be on the left or the right of the device.

I know I’m an odd case - but it is all about the - user defined - options.

5 Likes

And welcome, friend :slight_smile:

On the flip side, my new hearing aid has no facility for a direct input cable. So 3.5s are no go and my options are either bluetooth (built in to hearing aid) or a £2,500 specialist radio system designed for the hearing aid. I don’t have that kind of money lying around.

1 Like

Quite frankly, the main concern of a 3.5 audio jack on an expansion card is what it’s going to do to battery life? Pretty much any of the expansion cards other than the USB-C pass-through cause significant power draw, both when the system is on and when the system is suspended; even passive cards such as the USB-A.

1 Like

Even passive cards ??

  • USB A seem to require a basic background voltage connect, so yes
  • USB C do not So No
  • The 3.5 Jack would seem to have a DAC inside, to covert Digital to Analogue. A similar chip would be used on the main system anyway which could not be removed so by removing the Jack Expansion card less power may be used, whilst connected it may be the same.

… but now with the overhead of a USB-C interface sitting in-between and the logic for allowing the thing to be hot-plugged. It looks to me like a straightforward implementation would likely use more power (when present) than a hard-wired-in jack. You’ll probably have to work hard to reduce that to (near) nothing. I hope they can and will do that.

I’m pretty sure they were caught of-guard by the USB-A using noticeable power. A wired USB-A on a motherboard does not use that kind of power when nothing is plugged into it. Similarly for their HDMI cards etc.

The modularity of doing everything over USB-C definitely comes at a (power) price. Had I known, I would have bought 4 USB-C expansion cards for when on the road. It really saves quite a bit of battery life when suspended.

I wish in some sense I understood better

the overhead of a USB-C interface

if the power is on the interface to monitor the plug and play then sure that is to the USB C port on all ports and nothing to do with the USB A or HDMI

If on the other hand if the power is required only after and continuously whilst the USB A is plugged in what is it for. If the USB A requires power via a USB why if it doesn’t require power when hard wired. It is really about the USB A? So it’s not really the USB A that uses power but the main board once it has to communicate with a USB A.

So with a jack socket that is attached to a DAC surely the power is used by the DAC and some little by the communictaion with it.

If that is the case why would having a DAC removable via an expansion card would use more power.

I would think it uses less power in the sense it can be removed and there is also no comms with it so the mainboard uses less too.

I just don’t seem to see the problem at all

1 Like

@amoun The issue is that the Thunderbolt port “sees” something plugged in all the time and thus does not suspend when the computer suspends, thus draining power when it shouldn’t. The latest firmware revision for the Displayport card tricks the Thunderbolt card into “seeing” an HID instead of Displayport when no display is attached, so the port is allowed to power down/go to a low-power state. Similar problems exist for all but the USBC card because that is simple pass-through and not conversion like the others, so the TB ports “see” nothing when those are plugged in.

1 Like

So the issue is only about power use when suspended and as such when power available is a premium and this would mean if the 3.5 jack is plugged in when the computer is being suspended it the port will use power to check on it, whereas if it is hard wired it is just ‘disconnected/dead/non existent’

So the problem only exists as removing the cards is too much trouble to do before a quick suspend, that makes sense…

So the OS has to be targeted to sense that if nothing is connected to the 3.5 jack it disconnects from it.

1 Like

@amoun I used suspend as an example but the power drain occurs regardless. Normally, when no display is connected to an Displayport or whatever port you choose, the controller for that port goes into a low-power state waiting to wake up when you plug something in. As far as the Thunderbolt port is concerned, something is plugged in, the expansion card. Thus it is kept in it’s high-power/active state at all times. I choose poorly in explaining that before, my bad. Since the USBC cards are dumb and contain no circuitry themselves, they alone are immune from this condition.

OK so any circuit plugged in requires the main board to communicate.

But going back to the 3.5, as this is not a suspend issue, if I understand correctly do you think the DAC or whatever in the audio card would require more power when it is plugged in that the DAC would use when it’s onboard.

Surely the circuit that detects a 3.5 is plugged in has to work wheter the DAC is in card or on board.

Also if the DAC is on board isn’t it being power all the time whereas with a n expansion card I can remove it.

I hope I’m not being too dumb etc. as I note I have been ‘told’ off for the way I question people’s posts and answers etc.

1 Like