Overclock.net › Forums › Components › Mice › Why is an ADNS-9500/ ADNS-9800 bad and a PMW3310 good? They both have acceleration?!
New Posts  All Forums:Forum Nav:

Why is an ADNS-9500/ ADNS-9800 bad and a PMW3310 good? They both have acceleration?! - Page 10

post #91 of 99
Quote:
Originally Posted by qsxcv View Post

i feel like a logitech shill posting this, but here goes

summary of procedure:
1. empty local csgo server, sv_cheats 1, bind F3 "setang 0 0", cl_showpos 1
2. line up mouse against a hard surface like the edge of a keyboard. i used a mini projector with rubber feet.
3. press f3 to reset angle
4. swipe right quickly. i used a medium speed, around 1-1.5m/s. the length of my swipes correspond to roughly 360 +-15degrees motion in-game.
5. swipe back slowly until you reach the hard surface again. i moved back at a speed of roughly 5cm/s.
6. record second angle from cl_showpos 1, which shows the horizontal displacement.

here's my data. ten measurements for each mice.

g100s teensy mod (am010) Warning: Spoiler! (Click to show)
1.52
2.92
1.40
1.80
1.69
-0.77
0.57
-1.74
1.80
1.77

roccat kpm (3310) Warning: Spoiler! (Click to show)
-4.08
-4.14
-3.47
-1.68
-2.68
-2.43
-0.96
-3.22
0.18
-1.53

g303 (3366) Warning: Spoiler! (Click to show)
0.14
0.21
0.79
1.07
-0.50
1.64
0.93
0.50
1.29
1.57

evga x5l (9800) Warning: Spoiler! (Click to show)
7.08
7.08
5.79
5.79
7.51
6.94
6.22
5.86
4.86
7.01


bbbbbut they told me 3310 is a flawless optical sensor rolleyes.gif

This is such a great idea, thank you for this. The difference between a G500 and a G400 in this test is insane.

~ 20cm of movement with a peak speed of ~1.4m/s on an old black Taito 5mm. Both 800CPI @ 1000Hz. (short movement because I wanted to compare against my tiny I-2 pad)

G500 Warning: Spoiler! (Click to show)
-5.28
-4.8
-9.85
-4.70
-3.54
-6.6
-4
-8
-9.35
-3.85
-7.79
-5.02
-5.39
-7.74
-8


G400 Warning: Spoiler! (Click to show)
0.21
1.32
0.61
-0.45
-0.16
0.24
0.92
0.87
-0.08
-0.11
1.98
-0.79
0.03
0.9
-0.55


I would have loved to include my G100s but the sensor has developed a strange bug where the cursor will just wander around by itself. RIP.
post #92 of 99
for your g100s, this could be obvious, but maybe the lens is dirty?
main
(15 items)
 
old
(14 items)
 
 
CPUMotherboardGraphicsRAM
4770k maximus vii impact nvidia gtx 970 crucial ballistix tactical 16gb 
Hard DriveCoolingOSMonitor
crucial mx100 noctua nh-c14 windows 7 ultimate sony cpd-g520 
KeyboardPowerCaseMouse
kbp v80 matias quiet silverstone sx500-lg ncase m1 v3 logitech g100s with mcu replaced by teensy2.0 
Mouse PadAudioAudio
allsop raindrop xl chord mojo hifiman re-600 
CPUMotherboardGraphicsRAM
i7 920 evga x58 sli le galaxy gtx 460 crucial something 3x1gb 
Hard DriveCoolingOSMonitor
intel 330 180gb scythe kotetsu windows 8.1 pro sony cpd-g520 
KeyboardPowerCaseMouse
logitech k120 silverstone st75f-gs nxzt h440 evga torq x5 
Mouse PadAudio
allsop raindrop mobo 
  hide details  
Reply
main
(15 items)
 
old
(14 items)
 
 
CPUMotherboardGraphicsRAM
4770k maximus vii impact nvidia gtx 970 crucial ballistix tactical 16gb 
Hard DriveCoolingOSMonitor
crucial mx100 noctua nh-c14 windows 7 ultimate sony cpd-g520 
KeyboardPowerCaseMouse
kbp v80 matias quiet silverstone sx500-lg ncase m1 v3 logitech g100s with mcu replaced by teensy2.0 
Mouse PadAudioAudio
allsop raindrop xl chord mojo hifiman re-600 
CPUMotherboardGraphicsRAM
i7 920 evga x58 sli le galaxy gtx 460 crucial something 3x1gb 
Hard DriveCoolingOSMonitor
intel 330 180gb scythe kotetsu windows 8.1 pro sony cpd-g520 
KeyboardPowerCaseMouse
logitech k120 silverstone st75f-gs nxzt h440 evga torq x5 
Mouse PadAudio
allsop raindrop mobo 
  hide details  
Reply
post #93 of 99
don't we need to normalize the cm/360 to even begin to compare the results?

Or better, express the results in the following form:

{physical displacement error +/- instrumental error} = [Delta Angle] * [cm/turn] +/- [sensitivity * m_yaw * cm/turn]

Unit is in cm of error per 360 degrees. Statistical error you already know how to calculate.
post #94 of 99
i always normalize the cm/360... just didn't write it down, but the sensitivities for all my measurements were equivalent to 1.3 sens at 1000dpi

but imo this method is so rough that it's not meaningful to do anything other than draw those rough conclusions
main
(15 items)
 
old
(14 items)
 
 
CPUMotherboardGraphicsRAM
4770k maximus vii impact nvidia gtx 970 crucial ballistix tactical 16gb 
Hard DriveCoolingOSMonitor
crucial mx100 noctua nh-c14 windows 7 ultimate sony cpd-g520 
KeyboardPowerCaseMouse
kbp v80 matias quiet silverstone sx500-lg ncase m1 v3 logitech g100s with mcu replaced by teensy2.0 
Mouse PadAudioAudio
allsop raindrop xl chord mojo hifiman re-600 
CPUMotherboardGraphicsRAM
i7 920 evga x58 sli le galaxy gtx 460 crucial something 3x1gb 
Hard DriveCoolingOSMonitor
intel 330 180gb scythe kotetsu windows 8.1 pro sony cpd-g520 
KeyboardPowerCaseMouse
logitech k120 silverstone st75f-gs nxzt h440 evga torq x5 
Mouse PadAudio
allsop raindrop mobo 
  hide details  
Reply
main
(15 items)
 
old
(14 items)
 
 
CPUMotherboardGraphicsRAM
4770k maximus vii impact nvidia gtx 970 crucial ballistix tactical 16gb 
Hard DriveCoolingOSMonitor
crucial mx100 noctua nh-c14 windows 7 ultimate sony cpd-g520 
KeyboardPowerCaseMouse
kbp v80 matias quiet silverstone sx500-lg ncase m1 v3 logitech g100s with mcu replaced by teensy2.0 
Mouse PadAudioAudio
allsop raindrop xl chord mojo hifiman re-600 
CPUMotherboardGraphicsRAM
i7 920 evga x58 sli le galaxy gtx 460 crucial something 3x1gb 
Hard DriveCoolingOSMonitor
intel 330 180gb scythe kotetsu windows 8.1 pro sony cpd-g520 
KeyboardPowerCaseMouse
logitech k120 silverstone st75f-gs nxzt h440 evga torq x5 
Mouse PadAudio
allsop raindrop mobo 
  hide details  
Reply
post #95 of 99
Quote:
Originally Posted by qsxcv View Post

so something with my modded g100s is degrading performance. maybe the sensor itself, maybe i got the lens or led dirty, etc... or maybe it does depend on the microcontroller/firmware somehow.
guess i'll have to make a new modded g100s to figure out tongue.gif

update:
put my modded g100s pcb to another shell
now the variance is as low as my other am010 mice :O

so it wasn't some magical difference in firmware

the difference is that with the original shell, i swapped the stock mouse feet with 4 small circular feet. the larger stock mouse feet of the shell i'm using now allow for a more even glide
my guess is that with the smaller mouse feet, whenever i did fast swipes, i press the sensor into the pad a bit more, leading to higher sensitivity and larger errors in this test.

tl;dr: size does matter
main
(15 items)
 
old
(14 items)
 
 
CPUMotherboardGraphicsRAM
4770k maximus vii impact nvidia gtx 970 crucial ballistix tactical 16gb 
Hard DriveCoolingOSMonitor
crucial mx100 noctua nh-c14 windows 7 ultimate sony cpd-g520 
KeyboardPowerCaseMouse
kbp v80 matias quiet silverstone sx500-lg ncase m1 v3 logitech g100s with mcu replaced by teensy2.0 
Mouse PadAudioAudio
allsop raindrop xl chord mojo hifiman re-600 
CPUMotherboardGraphicsRAM
i7 920 evga x58 sli le galaxy gtx 460 crucial something 3x1gb 
Hard DriveCoolingOSMonitor
intel 330 180gb scythe kotetsu windows 8.1 pro sony cpd-g520 
KeyboardPowerCaseMouse
logitech k120 silverstone st75f-gs nxzt h440 evga torq x5 
Mouse PadAudio
allsop raindrop mobo 
  hide details  
Reply
main
(15 items)
 
old
(14 items)
 
 
CPUMotherboardGraphicsRAM
4770k maximus vii impact nvidia gtx 970 crucial ballistix tactical 16gb 
Hard DriveCoolingOSMonitor
crucial mx100 noctua nh-c14 windows 7 ultimate sony cpd-g520 
KeyboardPowerCaseMouse
kbp v80 matias quiet silverstone sx500-lg ncase m1 v3 logitech g100s with mcu replaced by teensy2.0 
Mouse PadAudioAudio
allsop raindrop xl chord mojo hifiman re-600 
CPUMotherboardGraphicsRAM
i7 920 evga x58 sli le galaxy gtx 460 crucial something 3x1gb 
Hard DriveCoolingOSMonitor
intel 330 180gb scythe kotetsu windows 8.1 pro sony cpd-g520 
KeyboardPowerCaseMouse
logitech k120 silverstone st75f-gs nxzt h440 evga torq x5 
Mouse PadAudio
allsop raindrop mobo 
  hide details  
Reply
post #96 of 99
Quote:
Originally Posted by qsxcv View Post


the difference is that with the original shell, i swapped the stock mouse feet with 4 small circular feet. the larger stock mouse feet of the shell i'm using now allow for a more even glide
my guess is that with the smaller mouse feet, whenever i did fast swipes, i press the sensor into the pad a bit more, leading to higher sensitivity and larger errors in this test.

tl;dr: size does matter

 

It should also put perspective on certain feet designs, tough the "Z-Axis Keepers" were imo only problematic on the G100S due to their material, i didnt like them on the G402 as well but not to a degree that i need them gone, the ones currently used are actually pretty rounded.

post #97 of 99
My G100s measures the following with different feet configs:

original -- exactly 1000 CPI
original + 1 layer G9 -- exactly 900 CPI
original + 2 layer G9 -- exactly 800 CPI
post #98 of 99
Quote:
Originally Posted by daniel0731ex View Post

My G100s measures the following with different feet configs:

original -- exactly 1000 CPI
original + 1 layer G9 -- exactly 900 CPI
original + 2 layer G9 -- exactly 800 CPI

with 1mm feet (maybe ~1.1mm including the double-sided tape?) i get ~910 on black cloth.
post #99 of 99
yea that's more or less what's expected if you compare against the graph in the 3090 datasheet.

actually i think the issue isnt that the circular feet were too small, but that they were ~0.2mm too thick and didnt allow the "z-axis keepers" to do their job
main
(15 items)
 
old
(14 items)
 
 
CPUMotherboardGraphicsRAM
4770k maximus vii impact nvidia gtx 970 crucial ballistix tactical 16gb 
Hard DriveCoolingOSMonitor
crucial mx100 noctua nh-c14 windows 7 ultimate sony cpd-g520 
KeyboardPowerCaseMouse
kbp v80 matias quiet silverstone sx500-lg ncase m1 v3 logitech g100s with mcu replaced by teensy2.0 
Mouse PadAudioAudio
allsop raindrop xl chord mojo hifiman re-600 
CPUMotherboardGraphicsRAM
i7 920 evga x58 sli le galaxy gtx 460 crucial something 3x1gb 
Hard DriveCoolingOSMonitor
intel 330 180gb scythe kotetsu windows 8.1 pro sony cpd-g520 
KeyboardPowerCaseMouse
logitech k120 silverstone st75f-gs nxzt h440 evga torq x5 
Mouse PadAudio
allsop raindrop mobo 
  hide details  
Reply
main
(15 items)
 
old
(14 items)
 
 
CPUMotherboardGraphicsRAM
4770k maximus vii impact nvidia gtx 970 crucial ballistix tactical 16gb 
Hard DriveCoolingOSMonitor
crucial mx100 noctua nh-c14 windows 7 ultimate sony cpd-g520 
KeyboardPowerCaseMouse
kbp v80 matias quiet silverstone sx500-lg ncase m1 v3 logitech g100s with mcu replaced by teensy2.0 
Mouse PadAudioAudio
allsop raindrop xl chord mojo hifiman re-600 
CPUMotherboardGraphicsRAM
i7 920 evga x58 sli le galaxy gtx 460 crucial something 3x1gb 
Hard DriveCoolingOSMonitor
intel 330 180gb scythe kotetsu windows 8.1 pro sony cpd-g520 
KeyboardPowerCaseMouse
logitech k120 silverstone st75f-gs nxzt h440 evga torq x5 
Mouse PadAudio
allsop raindrop mobo 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Mice
  • Why is an ADNS-9500/ ADNS-9800 bad and a PMW3310 good? They both have acceleration?!
Overclock.net › Forums › Components › Mice › Why is an ADNS-9500/ ADNS-9800 bad and a PMW3310 good? They both have acceleration?!