5150 Clone (MTM Scientific) Has Strange Beep

[ajacocks]

All,

I just finished assembling MTM Scientific's 5150 motherboard kit, and it POSTs successfully, with 64k of RAM onboard, and an IBM EGA card installed. However, I get a weird behavior on the speaker; I get a long, drawn-out, garbled-sounding beep, instead of the expected single tone.

Does anyone have a suggestion as to where to start debugging?

Thanks!
- Alex

 

[modem7]

and it POSTs successfully,

Please clarify. Which of the following best describes the symptom:
1. POST executes, and then motherboard boots to DOS (or BASIC). All works except for the speaker problem.
2. POST executes, and then motherboard 'hangs' coincident with the long garbled beep.

However, I get a weird behavior on the speaker; I get a long, drawn-out, garbled-sounding beep, instead of the expected single tone.
Does anyone have a suggestion as to where to start debugging?

I'm sure you realise that the probability is high that there has been a construction problem (as different from a faulty component).
Presuming that the symptom is point #1 above, then I suggest:

1. Check/resolder the solder joints on the following components (involved in generation of the speaker signal). I assume that the MTM motherboard uses the same component designations as the IBM 5150 motherboard.

U34
U63
U95
RN3 (resistor network 3)
R10
C9
P3 (connector for speaker)

2. Verify that RN3 is in the correct orientation.

 

[ajacocks]

modem7,

Many thanks for the suggestions. To clarify, the boot of the system proceeds to the point that it attempts to boot from a floppy disk, which fails, because I do not have one connected. The long garbled beep seems to not affect POST.

I acquired (from Retro Innovation, thanks Jim!) a 27c64 and 2364 adapter, and programmed the Supersoft/Landmark BIOS, and got the following results:
- U34 8253 Timer Channel 2 failed, but only on 2nd pass
- Keyboard controller failed
- Keyboard scan lines passed

Since, according to the Supersoft manual, timer channel 2 has to do with sound generation, that probably explains my beep problem. I'll trace all the joints, as you suggested.

The keyboard, I'll have to check, as well. I, unfortunately, don't have a real PC/XT keyboard, but I breadboarded an XT/AT keyboard adapter, from the wiki, here. I'm not sure that it's completely working, but the keyboard LEDs do flash, during POST, and the pressing of any of the lock keys (caps, scroll and num) causes the appropriate LED to light, or go out, on the keyboard. I have one concern about assembly, when it comes to the keyboard. The two caps next to the keyboard connector are 3-holed, for 3-pin caps, and MTM supplied standard 2-pin MLCCs. The instructions for the standard decoupling caps, elsewhere on the board, specified that the caps should be installed to the center pin, and either one of the two outer pins. The keyboard caps, however, were not specified, and the only picture that I could find seemed to indicate that they were installed on the outer 2 pins, which is what I went with.

Thanks!
- Alex

 

[modem7]

I acquired (from Retro Innovation, thanks Jim!) a 27c64 and 2364 adapter, and programmed the Supersoft/Landmark BIOS, and got the following results:
- U34 8253 Timer Channel 2 failed, but only on 2nd pass

Puzzling as to why it did not fail on the first pass, but I have seen these diagnostics do some funny things.

For diagnostic purposes, the channel 2 output of U34 (8253) is directly fed to a dedicated input on U36 (8255A), so that code can monitor the output. So, I one of the things that the Timer Channel 2 test will be doing is verifying that the channel 2 output of U34 is switching as expected. Failure of that test suggests that you should concentrate your effort around U34.

Ensure that all chips are correctly orientated.

- Keyboard controller failed
- Keyboard scan lines passed

Unlike on AT-class motherboards, XT-class motherboards do not have a keyboard controller chip, so I am unsure of what exactly that the PC/XT version of that test does.

Since, according to the Supersoft manual, timer channel 2 has to do with sound generation,

In the 5150, it is also used for generating data for the cassette port.

The keyboard, I'll have to check, as well. I, unfortunately, don't have a real PC/XT keyboard, but I breadboarded an XT/AT keyboard adapter, from the wiki, here. I'm not sure that it's completely working, but the keyboard LEDs do flash, during POST,

In most cases, that is a result of the keyboard receiving +5V and consequently doing a self test. The LEDs momentarily turning on is part of the keyboard's self test.

and the pressing of any of the lock keys (caps, scroll and num) causes the appropriate LED to light, or go out, on the keyboard.

That's a good sign. Code in the AT2XTKB unit controls that.

I have one concern about assembly, when it comes to the keyboard. The two caps next to the keyboard connector are 3-holed, for 3-pin caps, and MTM supplied standard 2-pin MLCCs. The instructions for the standard decoupling caps, elsewhere on the board, specified that the caps should be installed to the center pin, and either one of the two outer pins. The keyboard caps, however, were not specified, and the only picture that I could find seemed to indicate that they were installed on the outer 2 pins, which is what I went with.

You need to converse with MTM about that.

 

[Chuck(G)]

The point of the 3-hole caps was that it was impossible to install them backwards if the outer leads were the same. So, IIRC, the center lead is +; the outer leads are -. A brief chapter in component packaging; sometimes called "3 lead keyed polarity" capacitors.