What is the maximum PS/2 refresh rate?
- Thread starter Jeff7
- Start date
Just try the highest setting, if it doesnt work, try a lower one.
The refresh rate is simply how many times per second the computer samples position data from the mouse. The ps/2 port was designed to have the speed changed, using one thats high wont fry anything.
The refresh rate is simply how many times per second the computer samples position data from the mouse. The ps/2 port was designed to have the speed changed, using one thats high wont fry anything.
No such thing as a refresh rate here. PS/2 port is a packet driven serial bus. The mouse actively sends data packets when moves or clicks occur.
Maybe some mice are internally configurable to send more or fewer packets on mouse movements. More packets mean smoother mouse operation, but also much higher interrupt and bus load ... getting data off the PS/2 mouse port is a real pain in the butt, software wise, since that 8042 generic I/O controller that's being used for keyboard/mouse dates back to the late 70's.
regards, Peter
Maybe some mice are internally configurable to send more or fewer packets on mouse movements. More packets mean smoother mouse operation, but also much higher interrupt and bus load ... getting data off the PS/2 mouse port is a real pain in the butt, software wise, since that 8042 generic I/O controller that's being used for keyboard/mouse dates back to the late 70's.
regards, Peter
- Jan 4, 2001
- 41,596
- 19
- 81
Buuut....if there's no such thing as PS/2 refresh rate, why then does: a) the Logitech Mouseware have a refresh rate setting for the PS/2 port, and b) why is there a program specifically written to affect PS/2 refresh rate?
I set it down to 10 as a test, and yes, it does have an effect - the mouse got really jumpy; set to 200, it's pretty smooth. I just want to be absolutely sure that a rate of like 800 won't kill something on the mobo.
I set it down to 10 as a test, and yes, it does have an effect - the mouse got really jumpy; set to 200, it's pretty smooth. I just want to be absolutely sure that a rate of like 800 won't kill something on the mobo.
I just want to be absolutely sure that a rate of like 800 won't kill something on the mobo.
I suggest putting an active cooling solution on your mouse to prevent frying it. I've overclocked my mouse to 1600 and when I jerk it fast the icons fall right off the side of the monitor. Boosed my 3Dmark score by 340 points too.
The rate in the logitech control panel is a sampling rate. It specifies how often the OS looks data from the mouse.
I suggest putting an active cooling solution on your mouse to prevent frying it. I've overclocked my mouse to 1600 and when I jerk it fast the icons fall right off the side of the monitor. Boosed my 3Dmark score by 340 points too.
The rate in the logitech control panel is a sampling rate. It specifies how often the OS looks data from the mouse.
- Jan 4, 2001
- 41,596
- 19
- 81
<< I just want to be absolutely sure that a rate of like 800 won't kill something on the mobo.
I suggest putting an active cooling solution on your mouse to prevent frying it. I've overclocked my mouse to 1600 and when I jerk it fast the icons fall right off the side of the monitor. Boosed my 3Dmark score by 340 points too.
>>
Oh please stop, my sides hurt...
I'm thinking that sampling rate and refresh rate are the same thing. Ok then, what is the maximum sampling rate that is safe/possible?
<<
I'm thinking that sampling rate and refresh rate are the same thing. Ok then, what is the maximum sampling rate that is safe/possible?
>>Try it, the worst that could happen is the computer will lockup. I doubt it will work with 800. You won't see a difference over 200 anyways, I personally can't tell a difference between going to 200 from 100. You won't fry anything as your not overclocking anything, your simply changing how fast the computer polls the mouse.
Argh, does that driver actually POLL the mouse rather than wait for an IRQ12 to happen and then just fetch the data? Or maybe they actually do configure the mouse to send more packets representing tiny movements, or fewer ones representing larger movements, as I said above.
The polled mode driver would explain the huge CPU load caused by PS/2 mice as observed by some people on this forum.
regards, Peter
The polled mode driver would explain the huge CPU load caused by PS/2 mice as observed by some people on this forum.
regards, Peter
Set it and leave it @ 200 Reports/Second. All the PS2/rate changers I've seen can only go up to 200 anyway. If you crave something smoother, buy a USB mouse =)
USB mouse aint smoother, it is locked at a refresh much lower than 200. I think it is locked at 60 hz if i remember. And locked means cannot change.
Oh well, do we really have to start this thread over?
USB mouse handling IS smoother, simply because the USB root hub device delivers the
mouse movement packet data to system RAM through bus master operations and _then_
triggers an interrupt to the CPU - and the CPU has the data right away, and can move
the mouse cursor accordingly the same instant.
On PS/2, the keyboard controller (an ancient 8-bit ISA device) triggers an interrupt, but
delivers no data. The CPU must then go fetch the data first, before it can finally move the
mouse cursor.
There might be fewer packets delivered, depending on what your mouse hardware is like,
but the latency from hand movement to mouse cursor movement is much reduced on USB.
That's what makes it feel smoother.
Besides, has anyone figured yet that the entire discussion beyond this is rather pointless?
What's the point in sampling the mouse 800 times a second when the mouse cursor
won't be updated any more often than 75 to 85 times per second, depending on your display's
refresh rate?
regards, Peter
USB mouse handling IS smoother, simply because the USB root hub device delivers the
mouse movement packet data to system RAM through bus master operations and _then_
triggers an interrupt to the CPU - and the CPU has the data right away, and can move
the mouse cursor accordingly the same instant.
On PS/2, the keyboard controller (an ancient 8-bit ISA device) triggers an interrupt, but
delivers no data. The CPU must then go fetch the data first, before it can finally move the
mouse cursor.
There might be fewer packets delivered, depending on what your mouse hardware is like,
but the latency from hand movement to mouse cursor movement is much reduced on USB.
That's what makes it feel smoother.
Besides, has anyone figured yet that the entire discussion beyond this is rather pointless?
What's the point in sampling the mouse 800 times a second when the mouse cursor
won't be updated any more often than 75 to 85 times per second, depending on your display's
refresh rate?
regards, Peter
- Jan 4, 2001
- 41,596
- 19
- 81
"Besides, has anyone figured yet that the entire discussion beyond this is rather pointless?
What's the point in sampling the mouse 800 times a second when the mouse cursor
won't be updated any more often than 75 to 85 times per second, depending on your display's
refresh rate?"
That's not really a good arguement though. Otherwise, the entire issue of changing the sampling rate would be pretty pointless in the industry. Same would apply for videocards - why have them capable of over 200fps when a monitor can't refresh at that? The fact that the mouse samples that fast means that it's ready to send data regardless of what the monitor is showing. Plus, synchronizing with the monitor refresh requires additional overhead - I disabled vsync in my games because it lowered performance, sometimes even to half the framerate of the refresh rate.
What's the point in sampling the mouse 800 times a second when the mouse cursor
won't be updated any more often than 75 to 85 times per second, depending on your display's
refresh rate?"
That's not really a good arguement though. Otherwise, the entire issue of changing the sampling rate would be pretty pointless in the industry. Same would apply for videocards - why have them capable of over 200fps when a monitor can't refresh at that? The fact that the mouse samples that fast means that it's ready to send data regardless of what the monitor is showing. Plus, synchronizing with the monitor refresh requires additional overhead - I disabled vsync in my games because it lowered performance, sometimes even to half the framerate of the refresh rate.
It is a good argument. Why? Simply because mouse cursors are only moved during the VSync beam return phase, so that the cursor itself is always being displayed properly.
Besides, the mouse itself is sampling movement all the time, collecting data in its own internal buffer. All serial, PS/2, USB mice have ever worked like that.
Same for the graphics card - even with VSync method off, you'll only see 75 to 85 of your oh-so-brilliant 250 frames per second simply because the monitor doesn't update the display any more often. The cathode beam passes by each pixel 75 to 85 times per second, any inbetween changes to this pixel will not be seen. All you'll get with VSync off is tearing artefacts from showing half the previous picture along with half the next one.
The reason why that "mouse refresh" thing exists is marketing. Got to have something that justifies asking $90 for a mouse when others ask $5, don't you?
You folks who believe in this, without looking at the technical background, is what keeps the hype afloat. Think about it.
regards, Peter
Besides, the mouse itself is sampling movement all the time, collecting data in its own internal buffer. All serial, PS/2, USB mice have ever worked like that.
Same for the graphics card - even with VSync method off, you'll only see 75 to 85 of your oh-so-brilliant 250 frames per second simply because the monitor doesn't update the display any more often. The cathode beam passes by each pixel 75 to 85 times per second, any inbetween changes to this pixel will not be seen. All you'll get with VSync off is tearing artefacts from showing half the previous picture along with half the next one.
The reason why that "mouse refresh" thing exists is marketing. Got to have something that justifies asking $90 for a mouse when others ask $5, don't you?
You folks who believe in this, without looking at the technical background, is what keeps the hype afloat. Think about it.
regards, Peter
- Jan 4, 2001
- 41,596
- 19
- 81
I think I'm going to try and get my mouse working with the USB ports and see how it goes - I say try, because last time I used the USB port for the mouse, it never powered up at all.
Peter, where do you learn all this technical info? Got any links I can check out, or do you just sort of pick it up as you go?
Peter, where do you learn all this technical info? Got any links I can check out, or do you just sort of pick it up as you go?
I sort of pick it up as I go to work every day ... I'm a BIOS engineer ya know ... but much of all that marketing hype is easily disguised if you throw a brain cell or two at it
regards, Peter
regards, Peter
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 24K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 20K
-
-
-
Facebook
Twitter