What more i can do?

[222Lab_Test222]

IC16 removed, here are outputs of DO D2 and D5. Its the Non working board.

[CaptDon]

You could do a million more screen shots of data and address lines but it won't help to zero in on why the unit doesn't work since when it isn't working the busses are unpredictable. You almost need to write a simple loop program and burn it to an eprom and use it to examine what the busses are trying to do. Several hundred posts and it is no closer to working today than it was. Maybe you can find a local 'hands on' technician willing to help you have a look at this unit? This thing is so FUBAR'ed we can never repair it through posts even if we approach 36 pages of posts. You need the help of a good technician with their hands and eyes on the board.

[asis]

Hi,

222Lab_Test222
Please look at the schematic fragment from .pdf (p.14).
-
On the back side of the PCB, near the power connector, there are components D1, Q1.

If they match the schematic fragment, then there is a high probability that even if you short-circuit the battery to reset the CMOS, this will not be enough.
uPD 4990A can still be powered for a long time from the residual voltage from the +5V Bus.

You simply do not have enough patience.

Try to realize this and reset the CMOS not only by removing the battery and short-circuiting the contacts of the seat, but also by short-circuiting the +5V bus to GND.
(You can do this on the chip itself).
Naturally, with the equipment disconnected.

[Harry_22]

Look, all levels are pulled up to +5V. For example a 10k resistor is used for this.
Take a similar 10k resistor and connect NL2 rail to ground.
Watch the signal (without IC16). It should be around +2.5V.

We are looking for other signals on the Data bus if exist.


PS
If you are tired you can rest.
I don't require you to work hard.

[222Lab_Test222]

If you are tired you can rest.
I don't require you to work hard.

Why you think i am tired.
I need to sleep, eat and also give some time to family too.

[Harry_22]

Ok, good!

Where the trace?

[Harry_22]

@CaptDon,

it’s commendable that you’re worried, but the OP says he’s not tired and is ready to continue. I also don’t really care about the number of pages. Even the fact that we overload EEVblog with pictures is nothing for me.

What makes me despondent is that sooner or later the topic will be closed and our communication will end.

PS
By the way, how do you feel about valve guitar amplifiers?

[222Lab_Test222]

Ok, good!

Where the trace?

I don't have 10K resistance, I need to buy one.

[222Lab_Test222]

You simply do not have enough patience.

Try to realize this and reset the CMOS not only by removing the battery and short-circuiting the contacts of the seat, but also by short-circuiting the +5V bus to GND.
(You can do this on the chip itself).
Naturally, with the equipment disconnected.

I tried that too But unfortunately it didn't work, same beep.
I left the 5v and GND for 30 min while taking out CMOS battery and also sorting the battery terminals.
I also sorted the Pins of IC23 .

[222Lab_Test222]

You could do a million more screen shots of data and address lines but it won't help to zero in on why the unit doesn't work since when it isn't working the busses are unpredictable. You almost need to write a simple loop program and burn it to an eprom and use it to examine what the busses are trying to do. Several hundred posts and it is no closer to working today than it was. Maybe you can find a local 'hands on' technician willing to help you have a look at this unit? This thing is so FUBAR'ed we can never repair it through posts even if we approach 36 pages of posts. You need the help of a good technician with their hands and eyes on the board.

Lets see, to be honest i am learning in first hand and repair in second.
So i do not want to go to repair shop to repair.

[222Lab_Test222]

What makes me despondent is that sooner or later the topic will be closed and our communication will end.

Why will it be closed? 

[Harry_22]

Ok, good!

Where the trace?

I don't have 10K resistance, I need to buy one.

Take another one from 1k. Take several in series, parallel.

[m k]

On the working machine the data sequence is ODE9h, 00F8h, 0000h, 0000h...
On the not-working is 0D0Dh, 0000h, 0000h...

2 1 4 3 data would be JMP(rel16) F80D.

Faulty machine is doubling other side of 16 bit data.

[Harry_22]

So, the results of signal collection from the Address and Data buses of BIOS EPROM were processed.
We found out the following:

1. The Address signals on the working and non-working machines are the same for the first eight clock cycles.
2. Data signals on the lower EPROM differ from the first clock cycle.
3. In the case of mechanical removal of the lower EPROM, all data signals are set to logic level 1. That is, they are pulled up to the power bus by resistors.
4. It follows that it is EPROM chip that sets logical 0 at D5 line in the first clock cycle.
5. I am sure that today you will confirm that nothing prevents the chip from setting logical 0 on D2 bus.

Conclusion:
Data from EPROM IC16 is loaded with an error despite the fact that all incoming signals are correct.

I know that you checked IC16 on a working machine and everything was Ok.
It seems that the “Don’t believe your eyes” situation has arisen, but this is not so.
I assume that one or more address signals do not reach the chip body and are lost in the crib contacts. The fact that on a non-working machine you see a different breakpoint every time indicates the contact bounce. At high frequencies a fraction of an ohm is sufficient.

This is my current version. But I admit that I could be wrong.

PS
I usually detect bad contacts with a milliohm meter because I need to determine the exact point of failure.
Cleaning and reassembling is not the right approach if you are working with customer equipment.

Possible solution for you:
Check the clamping force of the collets with a some chip leg.
Read the working IC16 EPROM with programmer, compare it and make a copy of new 27C2001 for test.

[CaptDon]

I was given a butchered up Peavey Classic 100 series (there are several similar models with the same name) that uses a pair of 6L6GC tubes as the 50 watt power amp. I restored it back to factory new condition. The combo amp is a twin 12" using the Eminence (code 67 12820PA) speakers. All of the preamp / predrive stages are solid state. Overall the amp indeed has a smoother more pleasing sound than my all solid state twin 12. My friend recently played his Martin acoustic through the Peavey and he said it had a 'nicer sound' than his Fishman. The tube amp although mostly solid state is more pleasing to the ear possibly because it sounds like it has a bit more pronounced low mids without sounding muddy. The amp sounds fast and articulate but not shrill or glassy on the high end. Perhaps the push/pull tube finals add a bit of even harmonic distortion giving all individual notes a fuller and richer sound? The all solid state twin twelve seems glassy and brassy but when reducing the mid and treble controls it goes quickly to a muddy sound. The all solid state amp just doesn't seem as rich or full, overall very lifeless compared to the tube amp. All of my P.A. gear for front of house and monitors are solid state using SMPS power supplies and class G multi-voltage output laddering. Many of the amps are Carver PM-2.0T models. They are Clair Brothers Touring modified with additional mods by me.

[asis]

Hi,

I dare to voice my version.
-
If you look at uPD4990.pdf (p15), you can see uPD4175BC = (IC22 74LS174), inputs (4, 5, 12, 13), which are directly connected to the junior tetrad  DATA BUS D0 ~ D3.

The command procedure is described in detail in the document.

It cannot be ruled out that at the initialization stage (POST) some process occurs related to entering into the shift register and checking the contents.
It is quite possible that uPD4990 are faulty.
-
I suggest studying the document, so as not to fall into euphoria.

[222Lab_Test222]

Ok, good!

Where the trace?

Here,

[222Lab_Test222]

Hi,

I dare to voice my version.
-
If you look at uPD4990.pdf (p15), you can see uPD4175BC = (IC22 74LS174), inputs (4, 5, 12, 13), which are directly connected to the junior tetrad  DATA BUS D0 ~ D3.

The command procedure is described in detail in the document.

It cannot be ruled out that at the initialization stage (POST) some process occurs related to entering into the shift register and checking the contents.
It is quite possible that uPD4990 are faulty.
-
I suggest studying the document, so as not to fall into euphoria.

I am buying new RTC, lets see changes when new RTC are swapped.

[222Lab_Test222]

Asis,

I am looking your currents comments and you seems to be interested in RTC IC23.
And we have done couple of measurements and observations.
If you need further anything let me know.
By the way you can read past posts, #Reply 109 and #Reply 119-123. It can give some answer you want.

[222Lab_Test222]

Possible solution for you:
Check the clamping force of the collets with a some chip leg.
Read the working IC16 EPROM with programmer, compare it and make a copy of new 27C2001 for test.

The main things is those Motherboards are so weird. Even in the working boards sometimes the results are not same. I am trying to get on one more motherboard which is good and is not repair by me in the past for further easy comparisons.

The things about BIOS is i ask my friend for programmer, he told me to that he will give BUT still i haven't got it. Looks like i will buy now. And in japan any delivery takes a week time not sure why is that.
And EPROM of non working board works in another board, so i doubt we can find anything on those files BUT i will send as fast as i can.

[222Lab_Test222]

Data from EPROM IC16 is loaded with an error despite the fact that all incoming signals are correct.

Shall i check for EPROM IC15, for further observations. The same things we did with the IC16?

[222Lab_Test222]

You almost need to write a simple loop program and burn it to an eprom and use it to examine what the busses are trying to do.

What type of program you suggest?

[amyk]

Reading this long thread is very satisfying, like a mystery novel.
If I had more time I'd try to help out, but for now I'll just leave this one observation:

The scope traces suggest that the beeping stops as soon as Reset is released. This in turn suggests that this is happening well before BIOS starts executing. However, the OP has reported that the beeping continues on a good motherboard if the EPROMs are not present. This would suggest that BIOS code is switching off the beeping. I don't get it. :-?

If the EPROMS are not present the data buses may float high and feed the CPU FF bytes, and FF FF is not a valid instruction, causing it to triple-fault and reset repeatedly.

[m k]

Data from EPROM IC16 is loaded with an error despite the fact that all incoming signals are correct.

Shall i check for EPROM IC15, for further observations. The same things we did with the IC16?

For doubled data an unchanging A0 is a good candidate.
But that requires that chips are one after the other.

Swap ROM chips and do address and data line scopes again.
Result for doubled data is E9E9 F8F8 ...

[222Lab_Test222]

Data from EPROM IC16 is loaded with an error despite the fact that all incoming signals are correct.

Shall i check for EPROM IC15, for further observations. The same things we did with the IC16?

For doubled data an unchanging A0 is a good candidate.
But that requires that chips are one after the other.

Swap ROM chips and do address and data line scopes again.
Result for doubled data is E9E9 F8F8 ...

ROM Chips means, Swap EPROM High and Low and watch signals on both?