EEVblog® Electronics Community Forum
Electronics => Repair => Topic started by: cte7ds on March 03, 2016, 12:26:15 pm
-
Watching EEVblog #855: (https://www.youtube.com/watch?v=F2gX-R1k7MM (https://www.youtube.com/watch?v=F2gX-R1k7MM) - Ceramic Capacitor Piezoelectric Effect ) got me motivated to make a post about this issue I've been having with my Logitech G500 mouse;
When this mouse is idling on my desktop (powered on), a constant high pitched loud whine/noise can be clearly heard from within the mouse.
Armed with a plastic straw, a directional amplified electret microphone and an oscilloscope, I set out to find what the exact culprit is...
After couple hours of debugging, I'm pretty sure that the whine is localized around the SMD ceramic capacitors on the bottom layer of the PCB.
Worst problem areas being the areas between caps C7/C22 and C23/C31. It's surprisingly hard trying to pinpoint a high pitched noise like this :-BROKE
Not quite sure what my next step should be. I'm considering to de-solder all the ceramic caps and measure their values, so I can order better quality replacements. :-//
Logitech Customer Support couldn't help me at all, because the product serial number tag on the USB cord was worn out... why didn't they just put the serial number on the sticker at the bottom of the mouse like in every other mouse?
Here's what the whine looks like on spek. Recorded with a decent quality electret microphone, right next to the biggest source of whine I could find on the pcb:
(https://i.imgur.com/8nUdcrJ.jpg)
-
If it were me, I'd first try putting a blob of silicone on one of those caps. Seems like it would have a hard time resonating with a rubbery blob damping it.
-
If that doesn't work, just replace them. I've never heard of whine coming from a mouse, though. That's pretty interesting. :-/O
-
An old and well known issue.
Look for a cap near to the scroll wheel IR emitter, that was the classic culprit.
-
If it were me, I'd first try putting a blob of silicone on one of those caps. Seems like it would have a hard time resonating with a rubbery blob damping it.
It would need to be some type of silicone that can be easily removed, in the case it doesn't work.
I've never used any silicone compounds with PCB's, so, got any recommendations what to use?
10 years ago I remember trying to use hot snot to quiet down a coil whining graphics card and ended up damn near destroying the card in the process :P
If that doesn't work, just replace them. I've never heard of whine coming from a mouse, though. That's pretty interesting. :-/O
This is quite a common issue with many Logitech G-series mice (MX510, MX518, G5, G500). I have all of those mice, but the G500 has the loudest whine by far, enough so to get really annoying.
It has been documented somewhat, here: http://jdc.koitsu.org/logitech/ (http://jdc.koitsu.org/logitech/)
My guess is that Logitech cheaped out on the electronics and used cheap ceramics. The PCB itself has a real cheap look & feel and the whole "daughterboard soldered to pin headers"-arrangement is just awful for repairs (because it has to be soldered at exactly the correct vertical position on the header).
-
I have that exact model, but no whining caps. Mine, however, is plagued with the disconnect bug: every now and then it will disconnect from the PC and reconnect moments later.
-
Using ceramic caps is not 'cheaping out', it's quite normal.
The noise is merely an unfortunate side effect most people cannot hear.
I have that exact model, but no whining caps. Mine, however, is plagued with the disconnect bug: every now and then it will disconnect from the PC and reconnect moments later.
That would usually be a broken cable.
-
I have that exact model, but no whining caps. Mine, however, is plagued with the disconnect bug: every now and then it will disconnect from the PC and reconnect moments later.
OH! I always wondered why I sometimes get "USB reset" on my mouse for no apparent reason.... double whammy for me, it seems :scared:
Using ceramic caps is not 'cheaping out', it's quite normal.
The noise is merely an unfortunate side effect most people cannot hear.
I'll have to re-watch Daves video, but didn't he talk about different quality ceramics, which do not generate audible noise?
But yeah, I'm just assuming that the silent solution would have been more expensive to implement.
-
All ceramics are capable of generating noise. Whether one does or not is not a simple case of how much it costs or what type it is.
It's most likely that they weren't even aware they made noise. It's very unlikely anyone designing these can hear such a high frequency.
-
That would usually be a broken cable.
I thought so too, but it isn't. The problem is completely random and I cannot repeat it if I wanted. Besides, I've already tested the cable for continuity, bad connector, bad solder, etc. It's all perfect. This seems to the specific to the G500 and lots of other users have reported the same issue, but Logitech won't admit it is a faulty hardware.
-
I had a mouse with these issues, random disconnects and reconnects. The problem was a broken wire inside the cable, in the area right before the cable exits the mouse, where there's an S which is there to hold the cable in place.
The cable coming out of the mouse was usually straight, but every once in a while i would move the mouse close enough to the monitor stand in front of the mouse that the cable would bend sideways enough to get that wire inside the S loose contact and then gain contact again as soon as I'd move the mouse again (being surprised that the arrow froze on the screen)
-
If it were me, I'd first try putting a blob of silicone on one of those caps. Seems like it would have a hard time resonating with a rubbery blob damping it.
It would need to be some type of silicone that can be easily removed, in the case it doesn't work.
I've never used any silicone compounds with PCB's, so, got any recommendations what to use?
"Neutral-cure" is the recommended type. Actually, any kind of sound insulation might work, like foam tape. Maybe even plain electrical tape.
-
Big news: I found a simple solution to effectively suppress the whine
Ended up purchasing another G500 in great condition for $20 shipped to run experiments on:
- De-soldered the two biggest MLCC's (C18 - 10uF, C5 - 2.6uF) and replaced them with parts salvaged off a laptop motherboard
- Added a 100uF cap to the 5V USB input
Neither of those did anything to the whine.
But then I realized that it was the actual Laser-module that was making the noise...
So I probed the modules exposed pins to find its 5V supply and I soldered the 100uF cap directly to those pins.
And whine? Well It's Gone. :-+ :box:
Feels amazing to finally have this issue fixed! I'll attach some pics for anyone who wants to replicate this mod.
-
Hello !
I was following your topic since the beginning and I'm happy that you find a solution :-+ 8)
Before, on my G500 I have put many glue in the motherboard from my Bosh GluePen and some foam inside my mouse, and now there's less coilwhine on idle time.
I'll try this solution tomorrow and will reply, thank you !
-
All ceramics are capable of generating noise. Whether one does or not is not a simple case of how much it costs or what type it is.
It's most likely that they weren't even aware they made noise. It's very unlikely anyone designing these can hear such a high frequency.
On the flip side, there is at least one product out there which can emit (quite loud) diagnostic beeps, but doesn't actually contain any speaker or dedicated piezo beeper. Guess how that was accomplished... ;)
-
All ceramics are capable of generating noise. Whether one does or not is not a simple case of how much it costs or what type it is.
It's most likely that they weren't even aware they made noise. It's very unlikely anyone designing these can hear such a high frequency.
On the flip side, there is at least one product out there which can emit (quite loud) diagnostic beeps, but doesn't actually contain any speaker or dedicated piezo beeper. Guess how that was accomplished... ;)
That's an interesting application. I would like to know more.
-
Using ceramic caps is not 'cheaping out', it's quite normal.
The noise is merely an unfortunate side effect most people cannot hear.
I have that exact model, but no whining caps. Mine, however, is plagued with the disconnect bug: every now and then it will disconnect from the PC and reconnect moments later.
That would usually be a broken cable.
or more exactly half-broken, depends on the tension in the cable, sometimes it is broken and sometimes it is not.
-
Big news: I found a simple solution to effectively suppress the whine
Ended up purchasing another G500 in great condition for $20 shipped to run experiments on:
- De-soldered the two biggest MLCC's (C18 - 10uF, C5 - 2.6uF) and replaced them with parts salvaged off a laptop motherboard
- Added a 100uF cap to the 5V USB input
Neither of those did anything to the whine.
But then I realized that it was the actual Laser-module that was making the noise...
So I probed the modules exposed pins to find its 5V supply and I soldered the 100uF cap directly to those pins.
And whine? Well It's Gone. :-+ :box:
Feels amazing to finally have this issue fixed! I'll attach some pics for anyone who wants to replicate this mod.
Hello,
As I told yesterday, I've done the mod too ;)
So I have tried many caps from 10uF to 220uF (10-22-47-100-220): the larger the cap is, the quieter the mouse is.
But with 220uF: the mouse's power stage doesn't start so my computer doesn't reconize my G500.
With 100uF: it works fine and the coil whine is about 80% canceled (I have good ears ;D ).
Here's are some picture:
>I have to bend the IR Led of the wheel :
(http://morpheus.stigmata.free.fr/elec/G500-IR_Led.jpg)
>Then I have put the capacitor like this :
(http://morpheus.stigmata.free.fr/elec/G500-100uF.jpg)
You can see that I have changed my click button because last year I was having the double-click bug.
These switchs are from a low cost logitech mouse that I've never used, and these are more confortable to click and less noisy than the (bad quality) Omron.
That's why I put my capacitor like that: I can desolder the 3 pins of the switch later if needed.
Since this, I have add a little CMS 10uF in parallel which is a little better, I think about 85% less noisy.
With all the foam inside my mouse, it's barely audible now 8)
Thanks :D
-
Or you could just replace the culprit ceramic and not have to cram a large cap in your device and risk pissing off your USB ports.
-
Or you could just replace the culprit ceramic and not have to cram a large cap in your device and risk pissing off your USB ports.
But replace which one :-//
I think cte7ds has done a great job to find a solution.
And I have just checked the consumption, idle=0.07A (with the very low coil whine) and in movement+mouse click actived at 2000dpi 0.08A, so far from 0.5A USB power rating.
But you are right, it's too hard to check the inrush start current, so I'm thinking to add a little 0.1A fuse (or a little more) just in case, I have to find one CMS for example.
-
A 100mA fuse won't do anything useful.
-
A 100mA fuse won't do anything useful.
Have you got some suggestion ? :-\
-
A 100mA fuse won't do anything useful.
Have you got some suggestion ? :-\
Yes: Don't pile massive amounts of capacitance into a USB device not designed for it.
-
I did the mod! With 150uf cap the mice is silent and works great! It's perfect! Thank you!
-
I did the mod! With 150uf cap the mice is silent and works great! It's perfect! Thank you!
Hello,
Thanks for having shared your mod improvement !
So I have add a 47µF to my 100µF and I can confirm it is better !
Thanks :-+
-
cte7ds
Thank you very much for the idea on the capacitor replacement! I've dug out my G500 and a G500s (which is a replacement Logitech have sent me due to the said whine...) and started playing around with one of them. After some initial probing with a cheap pocket oscilloscope I've learned there's noise on the sensor power line with almost 200 mV peak-to-peak amplitude at the frequency of around 10 kHz. The datasheet for the sensor used states that the laser is operated in pulsed mode so this is the most likely reason for this. Of course, the readings are to be taken with a grain of salt given the tool used, but it tends to perform generally well at these frequencies.
Since I was reluctant to desolder the whole daughterboard assembly from the mouse I couldn't really trace out the entire supply circuit to learn what was fundamentally wrong with it, so I just soldered a couple of wires to the sensor pads and started trying on some capacitors I've had laying around. Surely enough, a 100 uF part did the job, eliminating the electrical and acoustic noise completely, but shoving such a massive capacity into my mouse did not feel right at all so I went in even further.
The datasheet for the Avago S9500 sensor calls in the typical application circuit for a 0.1 uF ceramic and 4.7 uF electrolytic filtering caps in parallel on the power line for the 5V scenario. It also states the maximum noise of 100 mV on the supply line for that matter, so the sensor appears to be already operating out of its specs, emitting audible noise amongst other things. Logitech have installed two small ceramics in parallel, they are labeled C30 and C31 and located on the underside of the mainboard very near the sensor. I couldn't measure their value without removing them so I did just that with my assumption being that perhaps they did not put enough capacity there. Having learned this isn't the case as they turned out to be 2.2 uF each, I decided to try and play by the book by replacing them with a single 0.1 uF 50V 0805 which happened to be the only size I had in my parts bin. I have also learned that the noise did not come from the capacitors as the mouse would still whine without them installed.
A single 0.1 uF did not seem to make any noticeable difference so I added a 16V 10 uF electrolytic one — again, just because that was the only one I've had laying around. This had the immediate effect of reducing the audible noise to the point that it could only be heard with my ear very close to the sensor. Measurements with my oscilloscope have confirmed that the noise has reduced to about 60 mV which was still noticeable but still much better than in the stock configuration. I have also tried a size A 4.7 uF tantalum I've had laying around, but they did not seem to produce any significant effect and wouldn't fit on the underside of the board anyway as there's less than 2mm of height to play with. Theoretically, one could be installed directly on top of the sensor but I did not feel like experimenting any further, so I've just soldered the 10 uF electrolytic job to the pads on the sensor, putting it a little diagonally so that it wouldn't interfere with the mouse wheel assembly.
The reassembled G500 was much quieter than the unmodified G500s I used as the control: it was immediately obvious when an unmodified mouse was plugged into the computer. Surprisingly enough, after a few hours of working any remaining noise there was had ended abruptly — I couldn't hear it now even with my ear directly at the mouse. I've no idea to why this happened, but I certainly know this doesn't upset me one bit!
Having done all this, I'm now having mixed feelings about Logitech. They could have easily avoided all this with countless warranty replacements by just following the sensor manufacturer's recommendations on the circuit. Does make me wonder why they did not do it in the first place, but oh well. At least it can be fixed!
-
I have bought a G213 Keyboard last week: coil whine too when Leds are ON, increasing sound until full white then decreasing etc..
So I have sent back it to the store.
Logitech development quality is not what it was before :-[
-
Hello,
I am having an issue with my G500 mouse.
I was first experiencing an issue with my left click button not working at all and found the microswitches to be the culprit. (Bought D2F-01 instead of D2FC) :)
After replacing both of the switches, I have found that my left click acts as a right click now.
At this point, I have noticed that C33 appears to have exploded or was damaged by myself.
What can I use to replace C33? I have access to a plethora of SMD capacitors at work but I am afraid of removing C32 to find out what I need for fear of damaging it and having nothing.
In the attached picture, C33 is the capacitor closest to the switch SW2 which is the right click.
Thank you for taking the time to read this.
-
Thanks for sharing the research and fixes!
I disassembled the mouse and held three different Nichicon caps to the solder pads, to test.
Just a 10uf cap hugely reduced the movement sound, and it didn't do a "wow" amount of reduction for the idle sound, but it was still quite noticeable.
Both the 100uf and 150uf pretty much eliminated the movement sound, and greatly reduced the idle sound.
The 150uf did reduce the idle sound a small amount more than the 100uf but not enough to warrant using it, so I just went with what others did and used the 100uf.
I probably could've been happy with the 10uf, but I didn't want to chance not having silence once it was all together. The 10uf definitely would've fit between the boards.
I'm very happy with the 100uf and can only hear it, slightly, if I put my ear up to the bottom of it--I don't hear it at all, otherwise!
I'm an extremely amateur soldering guy, so you'll balk at my soldering job on this project, and I balk at mine, too. It was so tough to squeeze in there with barely-tinned, thin wires (26awg?). When I started putting it back together, one wire snapped off and I had to re-solder to the pad. The attached pictures were before I re-soldered; the re-soldering job had a little more solder on it, so I hope it lasts for the long haul.
The capacitor I bought from digikey, a Nichicon (https://www.digikey.com/product-detail/en/nichicon/UKL1C101KPDANA/493-13396-ND/2598379) (hoping better quality would let me use a 10uf), was too thick to fit between the boards, so I cut off a tiny wire from a (25mm?) donor, small computer fan. The capacitor is resting in the area circled.
EDIT April 16, 2019:
I forgot to mention that throughout my testing/deciding between the three caps, which was a good hour or so, occasionally my mouse "disconnected" from the computer. This never happened with the 10uf cap, but rarely with the 100uf and 150uf caps. I would place the cap on/off the pads many times within a minute to listen. I would guess that a total of no more than 15 disconnects happened, so it was very infrequent. So, multiple influxes (as someone mentioned) of power from me constantly tapping the larger caps did cause a disturbance in the manner which I was testing. But, it's never disconnected even once since I soldered on the 100uf.
-
Hello Moduvator! I noticed that you mentioned the Avago S9500 made by logitech and I'm wondering if you were referring to a copy of it's specific datasheet or the original ADNS-9500. I've been looking for the S9500 data, can't find it anywhere.
-
Hello Moduvator! I noticed that you mentioned the Avago S9500 made by logitech and I'm wondering if you were referring to a copy of it's specific datasheet or the original ADNS-9500. I've been looking for the S9500 data, can't find it anywhere.
Doesn't appear to be too difficult to find but that's the one I used: http://www1.futureelectronics.com/doc/AVAGO%20TECHNOLOGIES/ADNS-9500.pdf (http://www1.futureelectronics.com/doc/AVAGO%20TECHNOLOGIES/ADNS-9500.pdf)