I had the first generation Anker TB2 doc and that was the first time I ever saw them produce a product that was a dud.
They sent me a replacement and I had the same problems. Then found out it was a design problem not a manufacturing one. They could send me 10 more and they would still not wake up all the ports properly. I don't even recall now what series of ritual steps I took to wake it up. I think unplugging it completely from everything. Which is actually more cable wear than not having a dock.
Please take this gently, but I did not really walk away from this post having gained anything. I don't have any problems with the post in isolation but I just don't feel it really offers enough weight to be here if that makes sense.
I read a few of your other posts and your writing style is direct and pleasant in a way I appreciate. Thank you for actually writing things.
My experience with non-Apple Thunderbolt products has been pretty dicey across the board. My ASUS motherboard has a port and theoretically works with it but across several firmware, BIOS, and driver upgrades, I’ve never actually got it to successfully recognize my TS4 as a Thunderbolt device, only USB.
I use two different docks. Dock 1 is a cheap chinese dock with some ports that worked (it has many ports). Dock 2 is hp g2 (an expensive dock). Occasionally the screen shuts off and resumes a second later in the dock 1. And resuming from sleep, the laptop doesn’t detect the keyboard. I figured it was until the laptop is fully up, I should not plug in the usb. On dock 2, the laptop goes to sleep and won’t turn back on from the keyboard until I press the power button.
Thunderbolt/USB-C protocols are basically at the edge of what's electrically possible. If you look at Ethernet equipment that's roughly in the same speed range and notice how much it costs, it's clear there are heavy tradeoffs with the consumer stuff.
I use a bunch of Lenovo Thunderbolt docks, with both Macbooks and Thinkpads. The older models go for as little as $20 used nowadays. I had to update the firmware on the really old ones, but after that they work flawlessly and let me connect 2x4k60 monitors.
I think there's a solid case for someone to make a monitor with a working, built-in dock. I mean, many monitors already have USB-outs for keyboards and 2nd displays and such.
I think monitors are a sweet spot - they tend to stick around longer than computers, and docks really don't need to change a lot over time, at least now with thunderbolt out there. Fewer cables, too.
I like the idea of standalone docks, and I purchased a few, but none reliably worked for me.
I just looked up who runs the Thunderbolt specification, and was surprised to see that it's owned by Intel. But Thunderbolt 4 was donated to the USB Implementers Forum so maybe that's why things are improving.
My Lenovo Thunderbolt 3 docks wakes up reliably, the problem is that it keeps randomly nodding off for a second or so causing the monitor to lose signal and have to sleep/wake itself.
What I would have done is tested it with an Android phone. My Samsung S26 Ultra allows me to connect to an external display via a USB C hub. So the test setup would be:
1. Plug the phone into hub.
2. Check that you get display output, the mouse and keyboard are working whilst the screen is not turned off and the phone is not locked.
3. Lock the screen and wait for the screen to turn off, thereby mimicking suspend.
4. Try to make the phone resume by using the attached keyboard and/or mouse.
If all goes well, the monitor should turn on after unlocking the phone or using the keyboard / mouse.
# Other things to try
Again try and reproduce the issue with a Samsung tablet that supports display out via USB C.
Connect the keyboard via Bluetooth and try to make the laptop resume. I found that connecting a Keychron keyboard via bluetooth allowed me to resume a Macbook Pro. I was kinda surprised that this worked.
Does the same issue occur on Linux as well?
I guess you could attach a USB packet type of debugger to ensure that data is being sent to the laptop by attaching the debugger to the cable that connects to the host machine. If you see USB traffic when the monitor goes to sleep and you are pressing keys in the keyboard and the dock still has power that means the dock is working fine so you can rule that out.
I wish I could have submitted this entire thread as an exhibit to every commenter in the "Mac Pro Discontinued" thread from a few months ago telling me "jUsT uSe tHuNDeRbOLT!"
I hate to say but this thread gives me comfort that: a) it’s not just me, b) it’s not my particularly unfortunate combination of hardware, and c) I made the right call to permanently retire my Caldigit TS4 and rely on my Lenovo ultra-wide monitor’s ports.
Learning this lesson (on the unreliability of systems using a very expensive docks) cost me over $1200 in chopping and changing hardware in the last few years.
MY CalDigit TS4 is not what I would have blamed for this issue, given that every time it happens for me, opening the lid of my Mac shows that it's asking, again, if this CalDigit accessory should be allowed to connect. This used to happen after every single reboot, but now it's rare, though I haven't changed anything about the setup in a year. I attributed this to MacOS problems, for that reason.
The VGA->DVI->HDMI chain retained the following simplicity of: 1. Send color of pixel 2. At the end of the line + blank send a pulse on HSYNC pin 3. At the end of the screen send a signal on the VSYNC pic. That's it (almost, there's EDID). If you can implement this, chances are you can display what's sent down the cable accurately.
Likewise with audio, you just send a voltage level down the cable, if you want to be fancy, you use SPDIF, in which case you send a 16-24 bit number per channel.
That's it, no handshake, no negotiation.
I have a TB dock for my Macbook. Sometimes it doesn't register. Sometimes it reboots for some reason, giving me 10 seconds of black screen. There's a 10% chance on every connect or reboot that it will just plainly not work.
And it's way better than it used to be, but still pretty bad.
The thing is, the old protocols all have their limitations.
Analog audio - unless it's balanced - is highly vulnerable to all sorts of external interference. S/PDIF is normally two channels of PCM audio (that's decently easy), but it can carry up to 8 channels of variously compressed audio and not every sink or source supports all possible codecs (thanks to software patent licensing BS). In either case, 48 kHz sample rate and 20 bits precision are the fundamental limits.
For video... analog maps well to CRT displays but reconstructing signals from analog inputs in a TFT/LED screen is inherently a lossy process, and running anything more than 1080p is a gamble and requires good, short cables. And digital video at higher data rates becomes a nasty mess and practically devolves into RF black wizardry at the kind of stuff you can pull off with modern digital video signalling - 4K @ 240 Hz refresh rate, this is utter insanity if you just think how much data is being shovelled around.
32 comments
[ 0.23 ms ] story [ 55.8 ms ] threadThey sent me a replacement and I had the same problems. Then found out it was a design problem not a manufacturing one. They could send me 10 more and they would still not wake up all the ports properly. I don't even recall now what series of ritual steps I took to wake it up. I think unplugging it completely from everything. Which is actually more cable wear than not having a dock.
I read a few of your other posts and your writing style is direct and pleasant in a way I appreciate. Thank you for actually writing things.
any benefits in coding and reading hackernews
My chances of getting new monitors are slim in this climate but it's a new avenue of investigation at least.
I think monitors are a sweet spot - they tend to stick around longer than computers, and docks really don't need to change a lot over time, at least now with thunderbolt out there. Fewer cables, too.
I like the idea of standalone docks, and I purchased a few, but none reliably worked for me.
1. Plug the phone into hub.
2. Check that you get display output, the mouse and keyboard are working whilst the screen is not turned off and the phone is not locked.
3. Lock the screen and wait for the screen to turn off, thereby mimicking suspend.
4. Try to make the phone resume by using the attached keyboard and/or mouse.
If all goes well, the monitor should turn on after unlocking the phone or using the keyboard / mouse.
# Other things to try
Again try and reproduce the issue with a Samsung tablet that supports display out via USB C.
Connect the keyboard via Bluetooth and try to make the laptop resume. I found that connecting a Keychron keyboard via bluetooth allowed me to resume a Macbook Pro. I was kinda surprised that this worked.
Does the same issue occur on Linux as well?
I guess you could attach a USB packet type of debugger to ensure that data is being sent to the laptop by attaching the debugger to the cable that connects to the host machine. If you see USB traffic when the monitor goes to sleep and you are pressing keys in the keyboard and the dock still has power that means the dock is working fine so you can rule that out.
Learning this lesson (on the unreliability of systems using a very expensive docks) cost me over $1200 in chopping and changing hardware in the last few years.
I take modern USB-C Flash Drive, plug it into my Mac. Nothing. Take it out, plug it in again, and it mounts just fine.
Same with SD Cards.
I have no idea why this is, and I'm not sure if it's always been this way, but I can't remember a time when I didn't have to do this.
The VGA->DVI->HDMI chain retained the following simplicity of: 1. Send color of pixel 2. At the end of the line + blank send a pulse on HSYNC pin 3. At the end of the screen send a signal on the VSYNC pic. That's it (almost, there's EDID). If you can implement this, chances are you can display what's sent down the cable accurately. Likewise with audio, you just send a voltage level down the cable, if you want to be fancy, you use SPDIF, in which case you send a 16-24 bit number per channel.
That's it, no handshake, no negotiation.
I have a TB dock for my Macbook. Sometimes it doesn't register. Sometimes it reboots for some reason, giving me 10 seconds of black screen. There's a 10% chance on every connect or reboot that it will just plainly not work.
And it's way better than it used to be, but still pretty bad.
The thing is, the old protocols all have their limitations.
Analog audio - unless it's balanced - is highly vulnerable to all sorts of external interference. S/PDIF is normally two channels of PCM audio (that's decently easy), but it can carry up to 8 channels of variously compressed audio and not every sink or source supports all possible codecs (thanks to software patent licensing BS). In either case, 48 kHz sample rate and 20 bits precision are the fundamental limits.
For video... analog maps well to CRT displays but reconstructing signals from analog inputs in a TFT/LED screen is inherently a lossy process, and running anything more than 1080p is a gamble and requires good, short cables. And digital video at higher data rates becomes a nasty mess and practically devolves into RF black wizardry at the kind of stuff you can pull off with modern digital video signalling - 4K @ 240 Hz refresh rate, this is utter insanity if you just think how much data is being shovelled around.