N160JME-GQA. N160JME-GTK. N160JME-GT1. N160JME-GEK. N160JME-GE1. NE160WUM-NX1.
N160JME-GQ1. N160JME-GL2. B160UAN01.QNE160WUM-NX2. NV160WUM-NX2. NE160WUM-NX3
There’s 5 pages. i am not listing them all. Here are those.
I just search by physical size, connector position, connector type, and eDP signal lanes. Note that you will still need to double check the pinout, preferably by getting the datasheet, which require a Panelook account.
I dont have such a account, but there are services available online (for cheap) that I can access that can give me datasheets for panels I request, for $0.5 a pop.
If you want to get some “free stuff”, I am probably buying a dozen of the sheets for 1200p touch panels, and also see if I can get cables made.
I found 3 “shops” (or individuals) able to make a custom-to-order eDP/I-PEX mini-coax cable. So. Time to dig through more displays!
My previous post is actually fairly comprehensive. That’s everything on Panelook that will fit inside the Framework’s height limit (of 3.2).
There’s quite a few other interesting ideas, but a lot of them dont have the reverse edge connector, so it’s 5mm thick.
Unless you want forward edge connector which .. hm. don’t exist. Fair enough.
So it’s between “N160JCN-ELK”, “NV160WUM-T02”, “B160UAK01.2”, “B160UAK01.5 HW0A/HW2A/HW3A”. and “B160UAK 01.8”. 7 panels.
It’s a mix of I-PEX 20738/20525 40-pins, and STM 40-pins. But beggers can’t be choosers.
If you don’t care about touch, there is 100 models. Well, 93, since the 7 touch models are also workable.
I have strong feeling you can just drive a 4-lane display with 2-lane, but I can be wrong.
I plan on driving some 2-lane display with Framework, so.
I wouldnt, since I can find people that provide the whole custom cable. I dont have to get a weird crimper, get connectors, all that.
For you in not china it might be a bit of a pain. Which is why I am planning on getting 5 cables made, while I am here.
I havent found any touch that is OLED, but I am only looking at touch right now, so.
And coincidentally, all the touchscreen that fits in framework are 2-lane, 1200p.
It would be great if someone at Framework can help confirm the feasability of this madness. Firmware support, lanes and all that. Cause voltage, I can match, no problem.
It would also be a great idea to not test with my Framework 16, so as to not make a $1,000 mistake. Though, so long as there is no short-circuit, it should be fine?
if one of them works out definitely let us know, that’ll open a lot more option.
edp displays don’t have scalers and stuff that can deal with sending them a lower res signal and stuff so you have less flexibility. An even if you do have enough bandwidth on 2 lanes you also need a display using faster edp limiting options.
And I am not looking at touch at all XD. I get the feeling he have somewhat different criteria on displays but the way of connecting them overlaps a lot so it’s still interesting
Judging from my “just plug the 40pin oled into the 40pin lcd connector on the t480s” episode that sounds like a very good idea to avoid. There was (mostly fixable) damage to the mainboard and I later found out the screen is completely dead.
I was looking at Youtube of replacements from Framework and it looks like they come with two vertical strips with holes to screw them onto the lid. The third party displays don’t it looks like so those mounting strips would need to be recreated or cannibalized. It looks like the fit is a bit finicky and being a little off can make the snap on bezel pinch the display. I’m going to wait until I see someone actually make it work before I try anything myself.
If you open one up, you realize that the “two vertical strips with holes” is just two metal brackets glued onto the display. If you hold onto the two metal brackets from your old display, you can reuse those.
I think Framework use stretch-release adhesive, which make this pretty easy.
You are correct in that you need to line these up pretty well, but its just fancy double sided taping. Shouldnt be that hard.
Edit:
It’s almost shocking how little customization Framework did to the display. Of all the laptop I have opened up, this one might be the cleanest.
Traditional solution have like a crazy frame with a dozen screws holding the thing from the sides, or just glue the display to the bezel glass. The formet is very time-consuming, and very thick (> 5mm), while the latter is very pain to separate the display.
Alright we have datasheets. Let’s given them a read!
You know what, chart time.
Name
Depth
Shape/Style
Color (NTSC)
DP standard
Interface Brand
Mating Housing
Lamp Voltage
Config
Drive Voltage
Drive Power
Luminance
Supply (V)
Touch Supply
eDP voltage
HPD voltage
VRX-Vpp
RRX
Aux Vpp
RAUX
EN level
PWM level
PWM freq
PWM duty
Notes
NE160QDM-NZ6
2.15/2.3
PCBA Flat, <3.2
0.98
4 lane eDP 1.4 (8.1Gbps)
I-PEX 20682-040E
11S8P
05_12_21
9.1W
500
3.3V
/
2
2.25-3.6
1.2
120
0.29-1.38
120
2.5-5
1.44-3.6
200-2000
5-199
Framework stock
N160JCN-ELK
3.05±.15
PCBA Flat, <3.2
0.47
2 Lane eDP 1.2
I-PEX 20525-040E
20736-040T-01S
28.5± 1.5
05_12_21
4.3W
300
3.3V
3.3
2.25-2.75
2.2-5.5
2.2-5.5
190-2000
1-100
Lenovo Slim 5 16 AHP9
N160JCN-ELL
3.0 max
PCBA Flat, <3.2
N/A
30 pin
NV160WUM-T02
2.85/3
PCBA Flat, <3.2
0.45
2 Lane eDP 1.2
MSAK24025P40G
I-PEX 20453 240T
11S6P
05_12_21
4.1w
300
3.3V
3.3
2
2.25-3.6
100-1.32
120
0.29-1.38
120
2.5-5
2.5-5
200-2000
5-100
B160UAK01.2
3.05/3.2
PCBA Flat, <3.2
0.46
2 Lane eDP 1.2
I-PEX 20738-040E-02
I-PEX 20736
05_12_21
4.6W
300
3.3V
3.0-3.6
2.25-3.6
1.32
.27-.8
2.2-5.5
2.2-5.5
200-2000
1*-100
lenovo
B160UAK01.5 HW0A
2.85/3.0
PCBA Flat, <3.2
0.47
2 Lane eDP 1.2
MSAK24025P40
I-PEX 20453
05_12_21
3.65W
300
3.3V
-0.3-3.6
/
2.25-3.6
1.32
.27-.8
2.5-5.5
2.5-5.5
200-10000
5-100
B160UAK01.5 HW3A
2.85/3
PCBA Flat, <3.2
0.47
2 Lane eDP 1.2
MSAK24025P40
I-PEX 20453
05_12_21
3.65W
300
3.3V
3.0-3.6
2.25-3.6
1.32
.27-.8
2.5-5.5
2.5-5.5
200-10000
5-100
B160UAK01.5 HW2A
2.85/3
PCBA Flat, <3.2
0.47
2 Lane eDP 1.2
MSAK24025P40
I-PEX 20453
05_12_21
3.65W
300
3.3V
3.0-3.6
2.25-3.6
1.32
.27-.8
2.5-5.5
2.5-5.5
200-10000
5-100
B160UAK01.8
3.05/3.2
PCBA Flat, <3.2
0.47
2 Lane eDP 1.2
I-PEX 20738-040E-02
I-PEX 20736
05_12_21
3.6W
3.3V
-0.3-3.6
2.25-3.6
100-1.32
.27-.8
2.5-5
2.5-5
200-2000
1*-100
Ooo data format? @Adrian_Joachim we might be out of luck. Damn. But keep reading.
All 4 I’ve read so far use RGB 888, in that order, from top left top right, then top to bottom.
Seem to be just VESA spec. V1.3
Would not rule stuff out just because of that, level-shifters can be quite small and are electronically pretty simple.
Don’t think we would need to worry about that too hard, for one those pwm inputs tend to be quite tollerant and we can probably mess with it on the ec side if nessecary.
Don’t think that should be a problem if it is edp.
Since I tend to look for higher resolution/refresh rate displays lanes do seem to matter, especially since the displays usually don’t seem to support link rates fast enough to drive them at fewer lanes than they have. 2 lanes of HBR3(8.1Gbit) could theoretically drive 1440p120hz but if you find that resolution in a 4 lane display it’ll likely only support HBR2 or something. Faster link rates are more expensive to implement and tend to use more power.
We have B160UAK01.5, but I am concerned about the reduced Aux Vpp (differential pair voltage). The N160JCN-ELK also have a reduced HPD voltageu the im. Assuming the Framework use 3.3V, 2.75 is not acceptable.
Seems like the B160UAK01.5 is the more likely workable candidate. Consider everything is isolated via capacitors.
They also give you impedance no they don’t.
Yeah datasheet says 2.5V max, weird. 2.5V logic levels ±10%. IIRC HPD is output-only? So it shouldn’t be a problem for 3.3V hosts. At worst, can add a 1K series resistor, and a spot for an optional 2K pulldown.
This is un-true. HPD is pull-down by the display, I think.
HPD also can be not pulled down by the display if the display is initializing, etc. Though some of this “data” is from external DP, so not eDP. So it might actually be different.
That’s exactly what I mean - what I remember is, HPD’s driven by the display, as opposed to host driving HPD, which afaict they do not. That in turn means you don’t need 3.3V being accepted by 2.5V logic (needs components to lower the voltage), only to have 2.5V logic be accepted by 3.3V (which it will, 0.7*3.3 is 2.3). If a host happens to have a HPD pullup, that could complicate things a little bit, but nothing that a series resistor couldn’t solve.
An idea I’ve had for alignment is to screw the metal strips into the display lid, put adhesive strips on them. Then align the new display with the bezel, fix the alignment in-place with some electrical table (something that will not leave residue on the screen when removed). Put the screen with the bezel face down on the table, and then put the display lid with metal/adhesive strips on top. Theoretically, magnets will then enforce good alignment and fix the display to metal strips in the right position…
