• Please review our updated Terms and Rules here

Keyboard not work anymore on motherboard AP8548.

A

andromeda92

Experienced Member
Joined
May 6, 2020
Messages
263
Location
FRANCE
Hi,
I have an AP8548 socket 7 motherboard (80486DX2-S CPU at 66Mhz), the keyboard no longer works.
I can boot but the keyboard does not work.
I changed the 5 pin DIN socket, the fuse, there is indeed 5v in the DIN socket, I tested with several other keyboards, but the keyboard still does not work.
The battery is ok.
May be the DM7406N buffer circuit ?
It is possible that it is the Via VT82C42N controller, how can I test if the problem comes from the VT82C42N micro controller.
Thanks for your help.
 
andromeda92 said:
May be the DM7406N buffer circuit ?
It is possible that it is the Via VT82C42N controller, how can I test if the problem comes from the VT82C42N micro controller.
Click to expand...
It sounds like you have already worked out the keyboard related circuity on the motherboard. Using a VT82C42N, that circuitry is expected to be simple, like that shown at [here] for the IBM AT (IBM 5170).

By convention, the computer's power-on self test (POST) will have asked the keyboard controller chip (e.g. VT82C42N) to do a self-test and report the result; a negative result expected to be reported by the POST to the user as an on-screen error. In the IBM AT, the 8042/8742's self-test includes a loopback test from P26 to TEST0, and P27 to TEST1, and so most of the circuitry (including PCB traces) is tested. But not all.

You do not state whether or not you are seeing an on-screen keyboard related error.

Assuming that you are not seeing an on-screen keyboard related error, then with the keyboard related circuity on your motherboard worked out, diagnosis should be able to be done using a logic probe or multimeter.
 
I have no errors related to the keyboard, it's just that the keyboard doesn't work, the boot is done normally.

For the test I just turn on the motherboard and a graphics card as well as the keyboard, I don't have a floppy drive or hard drive, I don't need it.
I have the message Press the F1 key to resume or DEL to access the setup, which is normal because as I said I do not have a hard drive and floppy drive.

I don't know if the problem become from VT82C42N.
 
In summary:
* Keyboard operation was working, but not now.
* No keyboard related error on-screen.
* You ruled out the keyboard by trying other keyboards.
* You ruled out a bad DIN connector on the motherboard.
* You measured +5V at the DIN connector on the motherboard.

Presumably, those other keyboards are also AT-class keyboards (because XT-class keyboards will not work).

We know you have a multimeter because you measured +5V. Do you also have a logic probe? That is better than a multimeter for measuring logic levels.

Are you able to draw a circuit diagram of the keyboard related circuitry (like [here])?
Hand drawn is fine.
If so, add the logic level (high or low) (or the voltage) that you measure at each point on the circuit, and then from there, we can provide diagnostic instruction.
 
My keyboard is an AT keyboard.
I will get a logic probe by Tuesday, but I don't have a schematic of my motherboard, the connection schematic of VT82C42N is standard on the datasheet.
The same that your link. P26 and P27.
 
If i disconnect keyboard i have message:
Keyboard error or no keyboard present

if i connect keyboard, i don't have this error.
so the keyboard is detected byt VT82C42N but it not work, i can't enter to bios for example.
 
andromeda92 said:
... but I don't have a schematic of my motherboard, the connection schematic of VT82C42N is standard on the datasheet.
The same that your link. P26 and P27.
Click to expand...
I see that you are referring to figure 2 in the VT82C42N data sheet, figure 2 shown at [here]. Figure 2 is a suggestion only, to quote the data sheet, "Design Example". The engineer is free to change the circuit (within certain constraints). Pull-up resistors for the 7406 and 7407 are not even shown.

You may discover that the maker of your motherboard board has done something different. They may have even imitated the IBM AT motherboard. If the problem cause is between the VT82C42 and DIN connector, then to assist you remotely, knowledge of the actual implemention is probably required.

andromeda92 said:
If i disconnect keyboard i have message:
Keyboard error or no keyboard present
if i connect keyboard, i don't have this error.
Click to expand...
That is interesting. The POST will be sending (via port 60h) an FFh (reset command) to the AT-class keyboard, and the AT-class keyboard is reponding with an AAh (BAT passed).
It suggests that the keyboard circuitry on the motherboard is working.

andromeda92 said:
... so the keyboard is detected byt VT82C42N but it not work, i can't enter to bios for example.
Click to expand...
And F1 is not working when you are prompted to press F1.

The keyboard was working with the motherboard before, and you have tried other AT-class keyboards.

Odd.
 
F1 not work.

Yes, the keyboard was working with the motherboard before.

I have tried other AT-class keyboards and is a same, it not work.

Maybe without doing it on purpose I had to remove the keyboard then reconnect it without turning off the pc.

Does VT82C42 have any connection with serial ports? because I had tried to connect a mouse with a ps/2 serial adapter (db9).
 
andromeda92 said:
Does VT82C42 have any connection with serial ports? because I had tried to connect a mouse with a ps/2 serial adapter (db9).
Click to expand...
The VT82C42 does not have RS232 serial port functionality.

When your logic probe turns up, it can be used to check (under our guidance) that logic levels/activity are as expected at various points. For example, maybe the 'output buffer full' signal (i.e. keyboard byte received) is not being generated, and the POST's keyboard check is unaffected by that because the check simply looks for a received byte a certain period after issuing the reset command.

For now, I suggest that you perform a 'CMOS reset' of your motherboard, however that is done on your particular motherboard (e.g. via jumper for that, battery removed for xxx minutes, etc.)
 
Yes, I will receive the logic probe this Tuesday.

I did a cmos reset two days ago, I cut off the power then remove the battery, and leave all overnight until the next day.

I suppose than P24 from VT82C42N is output buffer.

datasheet for VT82C42N is here

datasheet for DM7406 is here

too, it should also be known if the DM7406N is not out of service, it must be possible to know.
 
Last edited:
andromeda92 said:
I suppose than P24 from VT82C42N is output buffer.
Click to expand...
Yes, like as shown in the IBM AT diagram at [here]. On the subject, the VT82C42 data sheet contains (in part), "The VT82C42 also raises P24 or P25 to indicate a output buffer full". Other parts of the data sheet indicate that P25 is related to the mouse component of the VT82C42.

andromeda92 said:
datasheet for DM7406 is here too, it should also be known if the DM7406N is not out of service, it must be possible to know.
Click to expand...
By "DM7406N", you are being specific, suggesting that you have established that your motherboard is in fact using a 7406 chip in the CLOCK path.

If your DM7406N was faulty, then there would be no clock reaching the AT-class keyboard to clock in the FFh command put onto the DATA line. Accordingly, there would be no AAh response from the keyboard, and you would see the error of, 'Keyboard error or no keyboard present'.
 
Note too the P17 pin shown at [here]. In the IBM AT (IBM 5170), if the keylock switch is in the 'locked' position, the IBM POST will display, '302-System Unit Keylock is Locked'.

On the subject of P17, the VT82C42 data sheet contains, "When the VT82C42 receives data in the normal mode (pin 25 on DIP40 or pin 28 on PLCC44 parts connected to VCC) and the status of P17 is low, then the controller will not raise the P24, nor activate its internal OBF flag. It looks like the controller will consume the income data itself."

Is the P17 pin of your VT82C42 low or high?
 
while waiting to receive my logic probe, is it possible to simulate the DEL key to access the setup bios, by shortcut for example two pins of the keyboard controller or another circuit?

if not, is it also possible that there is a false contact at the pin of the dx2 processor?
 
andromeda92 said:
while waiting to receive my logic probe, is it possible to simulate the DEL key to access the setup bios, by shortcut for example two pins of the keyboard controller or another circuit?
Click to expand...
No.

andromeda92 said:
if not, is it also possible that there is a false contact at the pin of the dx2 processor?
Click to expand...
I would not expect a keyboard symptom if there was a CPU related issue. To get to the point where your motherboard is getting to, the CPU has been used extensively by the POST.

modem7 said:
Is the P17 pin of your VT82C42 low or high?
Click to expand...
You can use your multimeter for that. See [here]. If P17 is 0.8V or less, that is a TTL low. If P17 is 2V or more, that is a TTL high.
 
the measurement with the voltmeter is made between the VCC 5V and pin P17 ?

EDIT:
I made the measurements with the multimeter:
If I measure with the voltmeter between P40 (vcc) and P17, I have 0V, if I measure between P20 (GND) and P17, I have 5V.
 
Last edited:
Where can the problem come from, a resistor or a capacitor?

its possible to force P17 to LOW ?
 
Last edited:
Today, you get your logic probe. Verify the following:

NORMAL LOGIC STATE

1. VT82C42's P26 pin (CLOCK) should be sitting at LOW.
2. VT82C42's P27 pin (DATA) should be sitting at HIGH.
3. VT82C42's TEST0 pin should be sitting at HIGH.
4. VT82C42's TEST1 pin should be sitting at HIGH.
5. The DATA and CLOCK pins of the motherboard's DIN connector should be sitting at HIGH.

ACTIVITY

The 'pulse' LED on your logic probe shows activity, and so you are expected to observe:

1. During the POST, for a short period, activity on the P26, P27, TEST0, and TEST1, pins of the VT82C42.
2. The same activity appearing on the DATA and CLOCK pins of the motherboard's DIN connector.

When the POST finishes, then whenever you press a key (DEL, F1, etc.):

1. Activity (generated by keyboard) on both the DATA and CLOCK pins of the motherboard's DIN connector.
2. The same activity appearing on the TEST0 and TEST1 pins of the VT82C42.


temp_VT82C42N90457148978hgjiwe58.png
 
You must log in or register to reply here.
Back
Top