geekhack
geekhack Projects => Making Stuff Together! => Topic started by: PancakeMSTR on Sat, 03 March 2018, 15:17:53
-
So I bought a KBD75 in the past month or so, and it's been a truly fantastic keyboard. I really love it.
It's been working great lately but as of the past few days I've been having some issues I don't know the best way to deal with.
Primarily, I seem to be shocking the case a lot. Every time I do, it shuts the keyboard off and I have to unplug and replug it back in.
This isn't that big of a deal, but I'm wondering if there is something I can do internally in the case so that the static shock grounds to the usb shield, or something. Whatever it may be, some method of discharging the static shock in such a way that it doesn't shut the keyboard off.
The thing that is prompting me to post this is because I think I killed some LEDs by shocking the board. That or I fried the LED controller.
As of like, a few minutes ago, not all the LEDs will turn on. I haven't identified the exact number but I seem to be only getting a few in the upper left corner of the PCB.
Like I said, My guess is, by shocking the keyboard, I've fried either a or many LEDs or the LED controller itself.
So I have two main questions
- Is there anything I can do to make it so that static shocks to the keyboard case are discharged in a non-harmfal way? E.g., some way of connecting the case to the USB shield?
- I seem to have killed some LEDs, presumably by shocking the keyboard. How do I fix this problem? Could it be that I need to just reflash the controller, or could it be that LEDs or the LED controller is actually dead and needs to be replaced?
Thank you, I appreciate any suggestions. Please let me know if I need to clarify anything.
PS: I checked each of the LEDs for functionality by applying power to them individually. As far as I can tell, they are all completely operational. For some reason, though, only the one closest to the USB plug (upper right looking at the bottom of the PCB) turns on when I plug the keyboard into the computer.
So I don't really know what to do. Replace the LED driver or something? IDK.
-
I bet it's your electric personality.
Sent from my SM-G930V using Tapatalk
-
In my experience, this is a problem with aluminum keyboards most people never mention or expect.
Keep in mind, I'm not an electrician or even great at electronics so some expert may point out that that this is a very bad idea.
While I have yet to do it on my 75, on the Filco I connected the usb shielding to the case so it acts as a ground. I can still get zapped, but it least it goes down the cable to the computer case and grounds rather than jumping to the internal electronics.
Remember, it follows the path of least resistance to ground, if your case is not grounded, it has high resistance and if that spark is strong enough it can jump from the case (or your finger) to the PCB trying to find a ground. The bad side of this is if there is lightening strike you have a slightly higher risk of getting zapped that way, but while I have seen several system hit by lightening and power spikes, none has ever left the computer tower and rarely gets past the power supply if you have an adequate power strip.
If you do this, you will need to sand through the anodizing as that acts as an insulator, and while the spark can easily make that jump, it creates a high resistance point, so if there is a path with less, it will choose that instead.
As for your current pcb, are all the leds dead, or just some. If you apply power to one of the burnt ones (from a working one), does it come on (it may be dim but should come on)? If they fail to come on with rerouted power, it's likely just the led that has been damaged. If they do come on then you have larger problems. A burnt trace, damaged controller or scrambled firmware. Odds of it being firmware though is pretty slim in my opinion simply because what are the odds it scrambled only the code for those leds while leaving everything else intact.
-
In my experience, this is a problem with aluminum keyboards most people never mention or expect.
Keep in mind, I'm not an electrician or even great at electronics so some expert may point out that that this is a very bad idea.
While I have yet to do it on my 75, on the Filco I connected the usb shielding to the case so it acts as a ground. I can still get zapped, but it least it goes down the cable to the computer case and grounds rather than jumping to the internal electronics.
Remember, it follows the path of least resistance to ground, if your case is not grounded, it has high resistance and if that spark is strong enough it can jump from the case (or your finger) to the PCB trying to find a ground. The bad side of this is if there is lightening strike you have a slightly higher risk of getting zapped that way, but while I have seen several system hit by lightening and power spikes, none has ever left the computer tower and rarely gets past the power supply if you have an adequate power strip.
If you do this, you will need to sand through the anodizing as that acts as an insulator, and while the spark can easily make that jump, it creates a high resistance point, so if there is a path with less, it will choose that instead.
As for your current pcb, are all the leds dead, or just some. If you apply power to one of the burnt ones (from a working one), does it come on (it may be dim but should come on)? If they fail to come on with rerouted power, it's likely just the led that has been damaged. If they do come on then you have larger problems. A burnt trace, damaged controller or scrambled firmware. Odds of it being firmware though is pretty slim in my opinion simply because what are the odds it scrambled only the code for those leds while leaving everything else intact.
I've clearly got bigger problems then, because all of the LEDs work under power applied to them individually. Firmware issue seems hard to believe too. Any ideas?
Probably will find some way of connecting the USB shield to the case, something like you said.
-
I've clearly got bigger problems then, because all of the LEDs work under power applied to them individually. Firmware issue seems hard to believe too. Any ideas?
Probably will find some way of connecting the USB shield to the case, something like you said.
I would be looking for a bad mosfet, but I couldn't even tell you if that uses one or if they are used for RGB. If it's not a mosfet or if there is one they it's probably the controller.
Unfortunately you've pretty much exceeded what I can help with at this point.
-
In my experience, this is a problem with aluminum keyboards most people never mention or expect.
Keep in mind, I'm not an electrician or even great at electronics so some expert may point out that that this is a very bad idea.
While I have yet to do it on my 75, on the Filco I connected the usb shielding to the case so it acts as a ground. I can still get zapped, but it least it goes down the cable to the computer case and grounds rather than jumping to the internal electronics.
Remember, it follows the path of least resistance to ground, if your case is not grounded, it has high resistance and if that spark is strong enough it can jump from the case (or your finger) to the PCB trying to find a ground. The bad side of this is if there is lightening strike you have a slightly higher risk of getting zapped that way, but while I have seen several system hit by lightening and power spikes, none has ever left the computer tower and rarely gets past the power supply if you have an adequate power strip.
If you do this, you will need to sand through the anodizing as that acts as an insulator, and while the spark can easily make that jump, it creates a high resistance point, so if there is a path with less, it will choose that instead.
As for your current pcb, are all the leds dead, or just some. If you apply power to one of the burnt ones (from a working one), does it come on (it may be dim but should come on)? If they fail to come on with rerouted power, it's likely just the led that has been damaged. If they do come on then you have larger problems. A burnt trace, damaged controller or scrambled firmware. Odds of it being firmware though is pretty slim in my opinion simply because what are the odds it scrambled only the code for those leds while leaving everything else intact.
Can you show us how to connected the shielding to the case?
-
I've clearly got bigger problems then, because all of the LEDs work under power applied to them individually. Firmware issue seems hard to believe too. Any ideas?
Probably will find some way of connecting the USB shield to the case, something like you said.
I would be looking for a bad mosfet, but I couldn't even tell you if that uses one or if they are used for RGB. If it's not a mosfet or if there is one they it's probably the controller.
Unfortunately you've pretty much exceeded what I can help with at this point.
Is there any way if I can tell if the LED mosfet is working right?
-
Can you show us how to connected the shielding to the case?
Ideally, you solder a wire to the housing and then use a screw to attach the wire to the case, but I just tape it onto the case and use some foam between it and the pcb to keep it pressed it against it.
Is there any way if I can tell if the LED mosfet is working right?
Trace the leds back to it, and see if there is power any coming out from it.
-
Replacing the SMD LEDs on the KBD75 PCB isn't too bad. You'll just need some solder braid and flux for each of the four contacts, soak up the solder n they come off easy peasy. I replaced mine with Adafruit's Neopixel WS2812B. They used to be on Ebay for 10 bucks for 100 of them... But looks like that listing is over. I contacted KBDFans a while back and they sent me here (https://www.aliexpress.com/item/10-1000pcs-SK6812-4pins-5050-SMD-similar-with-WS2812B-Individually-Addressable-Digital-RGB-Full-Color-LED/32527626500.html?spm=2114.search0104.3.9.VCmmOg&ws_ab_test=searchweb0_0,searchweb201602_0_10152_10065_10151_10068_10344_10342_10343_10340_10341_10307_10137_10060_10155_10154_10056_10055_10054_10059_100031_10099_9912_10338_10339_10103_10102_440_10052_10053_10142_10107_10050_10051_10325_10326_10084_10083_10080_10082_10081_10110_10111_10112_10113_10114_10179_143_10312_10313_10314_10078_10079_10073-9912,searchweb201603_0,ppcSwitch_0&btsid=0380c381-ca1e-4711-9471-7eb6d7f1fb82&algo_expid=0b9c4121-1225-49f4-96e0-58b88dac7d35-1&algo_pvid=0b9c4121-1225-49f4-96e0-58b88dac7d35).
If there's a break in the chain, it's usually just one LED that needs to be replaced. Which ever is at the end of the line.
Hope this is helpful somehow, I am still very sleepy.
-
Here is a picture of the PCB: https://imgur.com/a/RC8HZ
Like I said, all of the LEDs work if I apply power directly to them, at the right contacts. It's only when I plug the keyboard in that only the upper right (in the photo) works, and the rest don't turn on.
Someone could do me a huge, huge favor and use MS paint or something to draw on the picture of my PCB where I could check continuity or something to see if things are working, or to help me find out where they aren't.
Thank you.
-
If the Leds are wired in series, which is a possibility, Ygor may be on the right path, figure out which one is last in line that is working, and where it goes from there. It's probably a a bad trace or solder joint one the next LED in the line.
Yes, this means following the traces (I recommend a magnifying glass), but you also need to check the solder joints.
-
If the Leds are wired in series, which is a possibility, Ygor may be on the right path, figure out which one is last in line that is working, and where it goes from there. It's probably a a bad trace or solder joint one the next LED in the line.
Yes, this means following the traces (I recommend a magnifying glass), but you also need to check the solder joints.
I mean, the thing is they were all working, and then they weren't. If it were a bad solder joint, wouldn't I have had problems getting all the LEDs to light up when I first put it together? Why would it have suddenly stopped working when, previously, it was doing just fine?
Anyway, I'll...check. I'm not sure what traces I need to be looking for/following.