DIY 5k Monitor - success :-)

[PaulD-UK]

Part of the reason that the R1811 was designed with USB2 ports is that the bandwidth available with USB-C cables is limited and it’s at the limit to run multiple USB3 ports and transfer 5K video.

Now computer GPUs handle DSC - Display stream compression, so it’s possible.
The SA1 is newer than the R1811, so it is possible it manages to do it - Apple does in the Studio Display, but the older LG 5K Ultrafine can’t, it only manages 4K over USB-C, because it has USB3 ports.

That’s why the SA1 USB-B port may be an input.

Thunderbolt 3/4 cables have more bandwidth for video so possibly may be a better choice.

 

[PaulD-UK]

@SubDriver There is also this post from @ecma to consider…

https://forums.macrumors.com/threads/diy-5k-monitor-success.2253100/post-33625639

He’s using Windows with his SA1, so it might be that his USB-C setup doesn’t have DSC, but the 8 bit issue needs clarification.

 

[SubDriver]

@SubDriver There is also this post from @ecma to consider…

https://forums.macrumors.com/threads/diy-5k-monitor-success.2253100/post-33625639

He’s using Windows with his SA1, so it might be that his USB-C setup doesn’t have DSC, but the 8 bit issue needs clarification.

Agree! @PewQ also had some issues and was able to get 10 bit working over USB-C, but only YCbCr 4:2:2 (not 10 bit RGB) for some reason (post 1994). He was using a MacBook, but didn't specify a model.

 

[PaulD-UK]

It’s something Better Display might help in troubleshooting, hopefully someone will try. The free version would be OK.
These things come down to the quality of the firmware.

 

[ColeHarris]

@erihp, @SubDriver, @beginswithaJ

Thank you all for your advice, really appreciate it. Unfortunately, I've been unable to fix any of the stuck pixels using the suggested techniques. I almost wondered at first whether the issue was caused by a faulty GPU or display cable given the severity of the issue, but at this point I'm content with just writing it off and finding a replacement. Based on my research it seems unlikely that I could find a 2019 or newer iMac for less than $400 anyways, which is roughly the total cost of just replacing the display on this one (possible less if I can part out the components from my current unit).

@PaulD-UK

For anything later that the (D1)) panel I would recommend using a R1811 controller board.
Because the results are stunning...

Thanks for the info. Do you (or anyone else in this thread) happen to know whether there are any meaningful differences between the LM270QQ1(SD)(E1/E2) and the *(SD)(F1/F2) with respect to reliability/performance? I'm a little paranoid now about display defects after experiencing and reading about the various issues these displays are prone to (primarily pink-fringing and image-retention/ghosting). I just really don't want to spend a bunch of money on a new(ish) panel only for it to develop these issues again in a few years, as much of my work is fairly color-critical.

One last question and I'll stop derailing this thread. Can anyone who's connected the built-in speakers tell me whether there was a noticeable reduction in the sound quality post-conversion? I was pleasantly surprised by the quality of the sound coming from my late-2015 unit, but I saw some people earlier in this thread suggesting that theirs sounded perceptibly worse afterwards, if that's the case I probably wouldn't bother. I just wasn't sure if Apple was doing some DSP magic the get those results or not...

 

[yurazazik]

I closely inspected the eDP cable and noticed some oddities, at least from my perspective. Is it supposed to be like this?

 

[PaulD-UK]

@yurazazik Yes that looks completely normal.
It's soldering the earth strap attached to the GND wires to the outer shell of the plug.

 

[yurazazik]

@yurazazik Yes that looks completely normal.
It's soldering the earth strap attached to the GND wires to the outer shell of the plug.

Then I only have to test my led panel, but I can’t test it in the iMac because I tore the eDP cable during disassembly and am now waiting for the replacement to arrive.

 

[PaulD-UK]

@ColeHarris I think (speculate) that when Apple worked on the design of the iMac Pro (in 2015-17) they also gave the screen design and manufacturing a workover to cure any problems.
So screens C1 panels onwards are all better than earlier ones, although C1 panels produced late 2017-19 may be the best for the 2017 iMac.

The built in speakers certainly benefitted from Apple's DSP magic, but only the 2020 iMac has the T2 audio processor which is completely digital.
The AliExpress audio connector/crossover that is used by some in this thread has its problems, which I have analysed, but with a tweaked crossover the result can be very acceptable.

Edit: @5lives Also analysed the performance of the speakers, in post #1833, and how to improve the sound.

 

[traxap]

Yep... hoping @traxap or @PewQ or @kevinherring or someone else who has used the SA1 can confirm. I have tried asking StoneTaskin but I am having trouble explaining to them...

Even if it ends up that my idea doesn't work for the SA1 board, I would basically be in the same place I would be with the R1811 board since I have no desire to even try with its USB 2.0 ports.

@PaulD-UK and SubDriver, in my test, when I tried one of the USB 3 ports and connected a 4 port hub the OSX immediately disabled it due to not enough power. In my opinion it shows that the OSX Sonoma, in my case, recognized the USB port on the SA1 board which was connected via USB C to my mac pro 7,1. I use an 8K USB C to USB C cable.

 

[SubDriver]

@PaulD-UK and SubDriver, in my test, when I tried one of the USB 3 ports and connected a 4 port hub the OSX immediately disabled it due to not enough power. In my opinion it shows that the OSX Sonoma, in my case, recognized the USB port on the SA1 board which was connected via USB C to my mac pro 7,1.

Thanks! I am leaning towards getting the SA1 and giving my concept of using the USB-A ports as data links to a HDD and SSD a try. If it doesn't work, I simply fall back to using a separate dedicated cable coming from the SATA/USB converter board to the host computer.

 

[traxap]

Thanks! I am leaning towards getting the SA1 and giving my concept of using the USB-A ports as data links to a HDD and SSD a try. If it doesn't work, I simply fall back to using a separate dedicated cable coming from the SATA/USB converter board to the host computer.

Later I will try connecting a ssd and an USB drive to see if the SA1 board acts as an external device and accepts OSX's USB drivers to communicate directly with external devices connected to the board - I let you know.

 

[PaulD-UK]

@traxap @SubDriver
I've found a review of a 4K/160Hz monitor than is driven by a JRY-W9RQUHD-UA1 board with the MediaTek MT9801 chip.
The specification page for the monitor describes the USB ports as:

• 2*USB3.0-A
• 1*USB3.0-B (uplink)

Which means the USB-C circuit is probably separate.

There is also a Reddit thread that indicates Asus are using MediaTek chips in a 27" monitor (1440p).
However that means they may also be using the MT9801 in their new ProArt 5K monitor???

This pic comes from a YT review of another monitor, which talks about its USB 3.1 10Gbps hub chip, which is also USB-C compliant...:

 

[SubDriver]

@traxap @SubDriver
I've found a review of a 4K/160Hz monitor than is driven by a JRY-W9RQUHD-UA1 board with the MediaTek MT9801 chip.
The specification page for the monitor describes the USB ports as:

• 2*USB3.0-A
• 1*USB3.0-B (uplink)

Which means the USB-C circuit is probably separate.

There is also a Reddit thread that indicates Asus are using MediaTek chips in a 27" monitor (1440p).
However that means they may also be using the MT9801 in their new ProArt 5K monitor???

This pic comes from a YT review of another monitor, which talks about its USB 3.1 10Gbps hub chip, which is also USB-C compliant...:

View attachment 2462741

View attachment 2462745

@PaulD-UK - I hope we find out the USB-C is not separate. From doing some more digging, I think all these USB-A and USB-B ports were initially intended to function as a KVM switch to support using only one keyboard and mouse with different host computers connected to the various inputs. When I reviewed the monitor you mentioned, which is also manufactured by TCL, I noticed what is in the picture below.

Therefore, I think it is very likely the USB functions of the boards are not separate since there has to be data flowing from the mouse/keyboard to whatever computer input is selected, including a computer connected to the USB-C port!

USB 2.0 is more than sufficient for such a basic task (keyborad and mouse) from a bandwidth perspective. And I expect it would be cheaper than USB 3.0. So I am still holding out hope we find out the USB 3.0 ports will indeed support the higher data rate it is capable of!

🤞🤞🤞

 

[DaveF]

I'm guessing this is what I need.

ZYAMY 10PCS 6P Jumper Line 6 Pins 2.54mm Pitch Female to Female Jumper Cable Connector Multicolor Jumper Wire for Breadboard 20CM

Feature: Brand: Jumper Wire Jumper Cable Length: 20cm Connector: 6P Female-Female 6 pin header Use: Can be used for PCB project, pc motherboard etc Pin Spacing: 2.54mm

www.amazon.com

No the R1811 has a 2mm pitch socket. That cable would probably fit the DZ-LP0818 CC board.

It looks like a 6 pin JST PH 2mm pitch plug to a JST XH 2.54mm pitch plug.
But it probably needs higher gauge wire than PH cables normally use, as the backlight LEDs are quite high current.
Maybe it will be better to get it made, or DIY….

I agree about the R1811 being fine without it. Unless you’re going to use the screen in sunlight.

Jumpers I bought are too big for both boards. But now I’m pretty sure both connectors on the driver and current boards are 2mm pitch, so 6 pin JST PH 2mm socket-socket connector should work.

The downside, I can’t find such cable Amazon anywhere, short of spending $30+ to have the kits and gear to crimp custom cables.

EDIT: StoneTaskin replied, said there was a fulfillment error, and will send me the missing cable.

 

[traxap]

@PaulD-UK - I hope we find out the USB-C is not separate. From doing some more digging, I think all these USB-A and USB-B ports were initially intended to function as a KVM switch to support using only one keyboard and mouse with different host computers connected to the various inputs. When I reviewed the monitor you mentioned, which is also manufactured by TCL, I noticed what is in the picture below.

Therefore, I think it is very likely the USB functions of the boards are not separate since there has to be data flowing from the mouse/keyboard to whatever computer input is selected, including a computer connected to the USB-C port!

USB 2.0 is more than sufficient for such a basic task (keyborad and mouse) from a bandwidth perspective. And I expect it would be cheaper than USB 3.0. So I am still holding out hope we find out the USB 3.0 ports will indeed support the higher data rate it is capable of!

🤞🤞🤞

View attachment 2462793​

PaulD-UK and SubDriver, the SA1 board recognized several USB Flash Drives and a mechanical SSD, unfortunately the write speed is very slow, see pics below. I'll look around to find more samples and let you know. I suspect OSX does not have the necessary USB drivers to drive the SA1's USB connectors. I did a search for USB drivers for the SA1 board and I did not find any. However, the board is able to update its firmware via its own built in menu found under the "Other" selection. I ran it but I could not tell if an update was received.

 

[SubDriver]

Thanks @traxap. That is definitely not what I was hoping for. USB 2.0 speeds from USB 3.0 connections☹️.

What MacOS version you are using?

I see you clearly have 5K resolution based on the screen grabs. Can you tell if it is 8 bit or 10 bit color? And if the color scheme is RGB or YCbCr?

Thanks for all the info thus far for the SA1 board. Very helpful.

 

[PaulD-UK]

OK. Got a YT hit for the JRY-W9RQUHD-SA1 board.

The board comes from a Chinese 4K/160Hz monitor - the Jichuang G272UP.
From the translation of the video the USB-C performance is less than the DP 10 bit performance - only 8 bits, at the 4K the monitor runs at.
Also the USB ports are 'One USB 3 up and two USB 3 down.'
But there do appear to be PCB traces between the USB-C port, and the square USB 3 controller above the USB-A ports.

Edit: But it is possible only the USB2 signals are transferred from the input USB-C port to the controller.
The main USB (3) bandwidth may all go to the MT9801 controller for the video/audio signal?

Obviously this board has 5K firmware to use with an iMac screen, but the limitations are part of the design capabilities.

 

[DaveF]

Do the iMac speakers sound like they originally did with this controller board?

I’m deciding if I want to buy this kit and try getting the internal speakers working. Or I can use cheap external speakers I have. Using internal saves some limited desk space, eliminates some cables. But they take up limited cavity space and might make mounting the display drivers boards harder.

Treble Bass Speaker Crossover Divider for 5K 27" imac and our controller board | eBay

2 X Crossover Audio Board. We want you to be happy with our products.And your satisfactory is very important to us.

www.ebay.com

 

[PaulD-UK]

@DaveF "Do the iMac speakers sound like they originally did with this controller board?"
No, especially if you use those crossover units, which can only give a 'good enough' result.
My analysis of the problem.
What you are buying is a set of plugs to avoid soldering... 😵‍💫

"CAN the iMac speakers sound like they originally did...?"
Almost, but the crossovers need redesigning.

What I ended up with was a 10microFarad non-polarised polypropylene audio-grade capacitor in series with the tweeter, and a 0.15 milliHenry inductor in series with the low frequency driver.
However I used an additional set of iMac Pro speakers in series with the original speakers, to give an 8 ohms impedance load to the R1811 amp. With a crossover for each speaker, so I can run both, or just one set of speakers...
That isn't necessary, but is helps fill out the lower frequencies.

"CAN the iMac speakers sound good enough?"
Yes. There's some advice, and more advice from @5lives

I've spent quite a time working out what's needed so ask if you have specific problems.

 

[traxap]

Thanks @traxap. That is definitely not what I was hoping for. USB 2.0 speeds from USB 3.0 connections☹️.

What MacOS version you are using?

I see you clearly have 5K resolution based on the screen grabs. Can you tell if it is 8 bit or 10 bit color? And if the color scheme is RGB or YCbCr?

Thanks for all the info thus far for the SA1 board. Very helpful.

Hi SubDriver - please see the attached print screen. Mac Pro 7,1 on Sonoma. ARGB with 10bit per color.

 

[PaulD-UK]

@traxap I think that's only showing what resolution and bit depth the computer GPU is rendering the graphics frame buffer at.
What happens to the signal chain when the monitor processes it is not shown in System Settings.
You need to use Better Display>Display and Colour mode to be able to view the whole signal chain.
This is my R1811.

 

[traxap]

@traxap I think that's only showing what resolution and bit depth the computer GPU is rendering the graphics frame buffer at.
What happens to the signal chain when the monitor processes it is not shown in System Settings.
You need to use Better Display>Display and Colour mode to be able to view the whole signal chain.
This is my R1811.

View attachment 2463153
View attachment 2463154

 

[traxap]

Thank you PaulD-UK for your input. I will download the Better Display app and see what I get, and I'll post back.

 

[traxap]

UPDATE: ... it shows 10 bit at 60Hz for the SA1 controller.