VOGONS


First post, by LunarG

User metadata
Rank Oldbie
Rank
Oldbie

I was wondering if anybody knows if it is possible to just piggyback a keyboard controller chip, for the sake of diagnostics.
I have a board that doesn't receive any input from the keyboard port, but visual inspection doesn't show any damaged traces or anything indicating damage to the board.
This leads me to think that it is probably a faulty keyboard controller chip. I have a replacement chip, so I was wondering if anyone knows if it's possible/safe to just piggyback the new controller chip to see if that fixes the problem?
I'd rather not have to do the whole soldering job to replace the chip if that isn't going to fix the problem, which is why I'd like to be able to test it before desoldering the old chip.

WinXP : PIII 1.4GHz, 512MB RAM, 73GB SCSI HDD, Matrox Parhelia, SB Audigy 2.
Win98se : K6-3+ 500MHz, 256MB RAM, 80GB HDD, Matrox Millennium G400 MAX, Voodoo 2, SW1000XG.
DOS6.22 : Intel DX4, 64MB RAM, 1.6GB HDD, Diamond Stealth64 DRAM, GUS 1MB, SB16.

Reply 1 of 9, by kaputnik

User metadata
Rank Oldbie
Rank
Oldbie

Can't give a better reply to your original question than I'd guess you'd have to fully replace the chip to be completely sure. Lifting only the VCC pin(s) might be enough, wouldn't bet my right hand on it though.

However, assuming that's an AT or PS/2 port, I've seen some examples of controllers that won't play nicely with modern dual protocol USB/PS2 keyboards, the kind that detects what type of port they're connected to, and choose protocol accordingly. They'll only work with older pure PS/2 keyboards. Have you tried with one of those?

Reply 2 of 9, by LunarG

User metadata
Rank Oldbie
Rank
Oldbie

In this case it is an actual old AT keyboard connected to an AT port on a 486 Opti chipset board.
The keyboard is "known good" as it is the one I use on a pretty much daily basis for my 486 DX4 100 build.
I guess I'll just have to get a new soldering iron (my current one is a bit crap) and get on with replacing the keyboard controller.
Currently using a board from A-Trend, but I'd quite like to see how my QDI board compares to it, and I can't do that when it won't boot due to keyboard error.

WinXP : PIII 1.4GHz, 512MB RAM, 73GB SCSI HDD, Matrox Parhelia, SB Audigy 2.
Win98se : K6-3+ 500MHz, 256MB RAM, 80GB HDD, Matrox Millennium G400 MAX, Voodoo 2, SW1000XG.
DOS6.22 : Intel DX4, 64MB RAM, 1.6GB HDD, Diamond Stealth64 DRAM, GUS 1MB, SB16.

Reply 3 of 9, by Tiido

User metadata
Rank l33t
Rank
l33t

Piggybacking is likely going to destroy one of the chips sooner or later, you cannot connect two outputs together without harming both. Lifting VCC pins isn't gonna do any good either, then the chip will power itself up via ESD diodes on the signal pins and that will damage whatever drives those IO pins aswell (chipset, ISA bus drivers or some other parts).
Only way out is to desolder the original chip, socket it and try new one in there. The controllers should be interchangeable for the most part.

T-04YBSC, a new YMF71x based sound card & Official VOGONS thread about it
Newly made 4MB 60ns 30pin SIMMs ~
mida sa loed ? nagunii aru ei saa 😜

Reply 4 of 9, by LunarG

User metadata
Rank Oldbie
Rank
Oldbie
Tiido wrote:

Piggybacking is likely going to destroy one of the chips sooner or later, you cannot connect two outputs together without harming both. Lifting VCC pins isn't gonna do any good either, then the chip will power itself up via ESD diodes on the signal pins and that will damage whatever drives those IO pins aswell (chipset, ISA bus drivers or some other parts).
Only way out is to desolder the original chip, socket it and try new one in there. The controllers should be interchangeable for the most part.

"Sooner or later" sure, but I'm not talking as a repair, only for testing for the time it takes to switch the PC on, and see if it will accept input from the keyboard. The idea is that if it does accept input with the piggybacked controller, then I'd just go for the whole repair job.
But I guess I better not risk it. I'll just wait until I can afford a new soldering station, and then just replace the chip and see if that fixes it. Too bad it's located in such a position that adding a socket would prevent ISA and VLB cards from being inserted completely.

WinXP : PIII 1.4GHz, 512MB RAM, 73GB SCSI HDD, Matrox Parhelia, SB Audigy 2.
Win98se : K6-3+ 500MHz, 256MB RAM, 80GB HDD, Matrox Millennium G400 MAX, Voodoo 2, SW1000XG.
DOS6.22 : Intel DX4, 64MB RAM, 1.6GB HDD, Diamond Stealth64 DRAM, GUS 1MB, SB16.

Reply 5 of 9, by quicknick

User metadata
Rank Oldbie
Rank
Oldbie

Good decision, I'd say the damage described by Tiido is likely to happen as soon as you power the board. AFAIK, only chips that have a float pin can be piggybacked the way you intend.

Are you sure there are no damaged traces running to your KB connector? Does the board have any damage from a leaky battery?

Reply 6 of 9, by LunarG

User metadata
Rank Oldbie
Rank
Oldbie

There is no visible damage. It uses a CR2032, so no battery leak or anything. I seem to remember using my multimeter to check that there was signal from the AT connector to the keyboard controller.
I will do that again tomorrow. Got all my spare motherboards out, but it is past midnight here now.
Thanks for the advice so far though.

WinXP : PIII 1.4GHz, 512MB RAM, 73GB SCSI HDD, Matrox Parhelia, SB Audigy 2.
Win98se : K6-3+ 500MHz, 256MB RAM, 80GB HDD, Matrox Millennium G400 MAX, Voodoo 2, SW1000XG.
DOS6.22 : Intel DX4, 64MB RAM, 1.6GB HDD, Diamond Stealth64 DRAM, GUS 1MB, SB16.

Reply 7 of 9, by Tiido

User metadata
Rank l33t
Rank
l33t

That sooner or later depends on chip manufacturing specs, and is measured in seconds in worst case and minutes in best case. It isn't hours or days, you'll see both chips get very hot very quickly and one will give before the other with a great likelyhood both will (and permanently at that). If you're unlucky you'll fry some bus transceiver or worse some chipset component aswell. The parts can only source or sink limited amount of mA per pin and two outputs together form essentially a short with maximum dissipation at any difference in voltages and currents between the two sides.

T-04YBSC, a new YMF71x based sound card & Official VOGONS thread about it
Newly made 4MB 60ns 30pin SIMMs ~
mida sa loed ? nagunii aru ei saa 😜

Reply 8 of 9, by LunarG

User metadata
Rank Oldbie
Rank
Oldbie
Tiido wrote:

That sooner or later depends on chip manufacturing specs, and is measured in seconds in worst case and minutes in best case. It isn't hours or days, you'll see both chips get very hot very quickly and one will give before the other with a great likelyhood both will (and permanently at that). If you're unlucky you'll fry some bus transceiver or worse some chipset component aswell. The parts can only source or sink limited amount of mA per pin and two outputs together form essentially a short with maximum dissipation at any difference in voltages and currents between the two sides.

Yeah, cheers. I've already ruled this testing option out. I'll just do another close inspection, and use my multimeter to check the traces from the chip to the keyboard port. If they show as being alright, I'll just replace the controller. It's been about two decades *shudder* since I did electronics at school, and I can't remember if we ever even discussed the possibility of piggybacking anything back then.
I know that piggybacking works with memory chips, but I wasn't willing to gamble with something more complex, which is why I was asking advice in the first place. 😀
Vogons is truly a fountain of knowledge.

WinXP : PIII 1.4GHz, 512MB RAM, 73GB SCSI HDD, Matrox Parhelia, SB Audigy 2.
Win98se : K6-3+ 500MHz, 256MB RAM, 80GB HDD, Matrox Millennium G400 MAX, Voodoo 2, SW1000XG.
DOS6.22 : Intel DX4, 64MB RAM, 1.6GB HDD, Diamond Stealth64 DRAM, GUS 1MB, SB16.

Reply 9 of 9, by LunarG

User metadata
Rank Oldbie
Rank
Oldbie

Right. Multimeter says that there is perfectly fine connection between AT connector and keyboard, so I'm guessing there's not much doubt that the controller must have failed. The board looks pristine, so I find it unlikely that there's a break in any signals anywhere. Right, that will have to go into the repair queue 😜

WinXP : PIII 1.4GHz, 512MB RAM, 73GB SCSI HDD, Matrox Parhelia, SB Audigy 2.
Win98se : K6-3+ 500MHz, 256MB RAM, 80GB HDD, Matrox Millennium G400 MAX, Voodoo 2, SW1000XG.
DOS6.22 : Intel DX4, 64MB RAM, 1.6GB HDD, Diamond Stealth64 DRAM, GUS 1MB, SB16.