Electronics advice needed. Discharging an IC?

Mugsy

Posts: 772   +203
Long story short, I need to clear the Winbond FlashRAM (pdf pg 7) chip in my tablet to reset the Bios.

I found information online that suggests I can do this by grounding one of the pins, but before I attempt this, I could use some advice. What's the best way to "ground" an IC to discharge it?

I was thinking of simply using needlenose tweezers to link the required pin to Pin-4 (labeled "Ground" in the schematic.)

Would it be better to "ground" it to something else? Should I use low-gauge wire instead of steel tweezers for this (does it matter)? Does the tablet need to be On to do this? I have a multi-meter and was wondering if I could use it to perform this?

Any information is appreciated. TIA
 
The picture says it all -- any wire from the pin on the chip to any mounting screw on the case
 
Thx for the reply. Any advice on the best method? Could a multimeter be used for this?
 
Guess so - - be sure the MM is on a setting to measure current or AMPS and hold your connection for 4-5 seconds.
 
Thx Jo. Does it matter if I use the red probe or black? (I don't want to fry the chip.)

I'm wondering b/c normally I'd touch the red probe to the positive (power "out") lead, but now I'm wondering if the red probe is "hot"?
 
ALL POWER should be off. The leads from the MM don't matter here, given you are just shorting to ground.
 
I believe the tablet must be on to clear the Bios (according to the few sources I've found), but I will (of course) try it with everything off first.

I was hoping I'd be able to see any charge remaining on the chip drain in real time using the multimeter.
 
After years of building shortwave radios, this ALL SOUNDS BOGUS to me - - sorry.

  1. Capacitors hold charge, I've never see an IC do that.
  2. to drain a charge, you need power off - - and the MM will not do much other than just flip a tad and drop to zero immediately IF AT ALL.
  3. The BIOS doesn't hold charge but is powered by the onboard battery.
  4. There is a technique to reset a hardware password via a short but I'm not certain this is it
Best wish and good luck
 
Thx for the reply. I'm trying to avoid the "long" story, but in a nutshell: I inadvertently disabled the USB ports in my tablet's Bios. And while I can get INTO the Bios, I can't reset it w/o a keyboard (which requires a USB connection.)

I found the linked screenshot above on resetting the Winbond Cmos FlashRAM chip manually, but simply grounding pin 2 as shown wasn't enough. Neither did disconnecting the battery. So I assumed the IC was somehow holding a residual charge.

I heard back from Winbond and they told me the procedure to reset the chip. Basically, it must be switched to "Reset Mode" with "Write Enabled" in order to clear it. Here are the steps but it's mostly Greek to me (the pinout labels are in the Datasheet in my first post):

===================
First of all, you have to convert from /hold pin to /reset pin. you have to write the status register-3 for that.
Please read the SR3 first with 15h. I think it should be 60h or something else. If it’s 60h, you have to change it to E0h with 11h command.

Here is the write status register-3 sequence.
06h (Write Enable)
11h (Write SR3)
E0h (SR3 value)
Wait for BUSY bit clear or ~15ms after /CS high.
15h (Read SR3).
===================

I don't know what this means I'm supposed to do? Apply charges to individual pins to "set" them? Which pins and for how long?
 
Hmm; that is so terse, I can't see how to use that info either.

I feel your pain...
 
Good, so it's not just me. I'll contact them again for clearer instructions.

Thx.
 
The instructions they gave you are meant to be done by software, something the average user won't be able to accomplish unless they are familiar with programming and hardware(which I am.) From looking at the data sheet, it seems that you should be able to do a reset by powering up your system, then doing a momentary reset using a wire connecting to ground and then touched to the /RESET pin. Using a multimeter may work but it has resistance and the internal pullup in the Windbond chip means that the meter would form a voltage divider with that pullup and it may not pull the pin down to a vaild low value. It is better to use a wire which will make sure the pin is brought to a low level. If your board has the 8 pin SOIC type chip(shown on page 6 of the PDF document), then using the 'dot' on the one corner of the chip, you should be able to touch pin 4(labeled GND) to pin 7(labeled /HOLD or /RESET). Make sure that you are using the GND pin and not the VCC pin. You should be able to measure the pin with your multimeter and read the GND as 0 volts and the VCC pin should be just below 2 volts. When you are sure that you have the ground pin(or a valid ground), then carefully short the /RESET pin.
If you don't have the 8 pin SOIC chip, then see if you have the other SOIC chip and use the correct pins. If you have a surface mount type chip, there may be traces leading to the correct pins that you can use. Otherwise, if there is no way to contact the /RESET pin, it would have to be done per the instructions.
 
you should be able to do a reset by powering up your system, then doing a momentary reset using a wire connecting to ground and then touched to the /RESET pin.

Great information. That is what I was planning to try next. When you say "momentary", how long? Just a second? Longer?

I have the 8pin Winbond chip. I uploaded a photo of it here.
 
Follow-up:

Using my multimeter's black probe (plugged into "COM") and the meter switched to DC mV, I tried:

1) Computer off, touching Pin 7 ("/Hold, /Reset") for a couple of seconds. Removed probe. Switched on. No change (I was hoping with the computer off, if this worked, I could add my own Reset button.)

2) Computer on, same thing. No change on restart.

Did the same thing with Pins 3 & 4. No change.

It's possible using the multimeter in this way did not provide sufficient ground (meter reading remained at zero volts), but that seems unlikely. Does it sound like I did something wrong? Anything else I might try? TIA
 
If you insist on the MM, put it on a current (amps) setting

Thanks. The only reason I prefer to use it is because I'm worried about static shock from other sources (plus the pointed tip of the probe makes it easier to avoid the other pins.)

Still using Negative "COM"?
 
Using a multimeter may work but it has resistance and the internal pullup in the Windbond chip means that the meter would form a voltage divider with that pullup and it may not pull the pin down to a vaild low value. It is better to use a wire which will make sure the pin is brought to a low level.

Hmm. No go. :(

I tried both the multimeter probe (with meter tuned to uA) as well as a pair of steel needlenose tweezers. I tried grounding Pin-2 to Pin-4 as well as the probe to pin 2), tried with Power off and Power on (interestingly, the tablet refused to power on while holding Pin-2 & Pin-4 together using the tweezers until I released them.

I couldn't touch Pin-7 to Pin-4 using the tweezers (too far) but I tried holding the probe on it prior, during & after booting, and nothing had any effect. I think shorting the Winbond chip isn't going to work. :(

I have one last idea: I noticed there are some tiny barrel capacitors on the motherboard. I previously had a friend desolder the power connectors to see if that would clear the Cmos but it didn't work, but I'm wondering now if that's because of the capacitors? So I'm going to try disconnecting the power leads from the MoBo again and discharging the capacitors by repeatedly attempting to power on the tablet until they are discharged before reconnecting power.

How do I go about checking the capacitors to see if they still contain a charge before reconnecting the power? TIA
 
Final update for anyone who was following this thread.

I had to give up. I tried every trick in the book. No combination of shorting pins on the Winbond chip worked. Grounding pins before, during and after... didn't matter. Nothing worked. :(

I decided to try again to disconnect the batteries in hopes of clearing the Cmos, this time attempting to turn the tablet On 4-Times in hopes of draining any capacitors or residual charge, then letting it sit for an hour before reconnecting the power, but despite all that, the old Bios settings were STILL there.

That made me wonder if it was reading the settings from the built-in SSD, so I disconnected it yet the Bios settings are STILL loading.

So that's it. I can't think of anything else. I will NEVER buy another tablet without a reset button (or w/o a touch-screen bios.) Anyone need an 11.6" lighted serving tray. :(
 
Back