did i kill my 'puter

O.K. Not clear from what was described. So the BIOS keeps the memory size in CMOS and prompts when it changes?

Hmmm ... not on the CP/M machines which I used from time to time. They used RS-232 connected terminals, and did not have a way to tell whether the keyboard was present. Perhaps some of the single-board CP/M machines with a built-in terminal, however.

And -- my Altair 680b did not even *need* a terminal. It could be jumpered to accept program loading and starting from the front panel switches. (The only way to load and run a program when I first got it. :-)

And the Sun workstations, which did have keyboards would simply detect the missing keyboard, and switch over to use the first serial port as a terminal -- expecting a terminal to be connected to it. It was a common way to set up servers, which did not need a graphics screen or keyboard -- just a terminal to get them started and/or to diagnostics. Leave out the keyboard or the graphics card (called a "framebuffer") to switch to a serial terminal as the console.

Enjoy, DoN.

When Sun workstations are booting, or running memory tests, the monitor ROM display the following characters, all in the same location:

'/', '-', '', '|'

which looks like a spinning propeller and gives a clue that things are still happening.

I remember an intersting memory test documented in a Motorola programming manual for the MC6809 CPU. It included a checksum on itself, and could be relocated anywhere in memory. It used a very good set of test patterns, and would spot any problem in the memory -- except speed. The accesses to sequential addresses were sufficiently separated so it would miss those -- until you relocated it to *run* in the slow memory. You could leave it in ROM, ready to run whenever you wanted it, and when it relocated itself, it corrected the checksums to be correct for the new address in which it was living. (It helped that the 6809 made totally relocatable programs easy to write. :-)

Enjoy, DoN.

Do subsequent boots all come up quickly or are they still slow?

Interesting cunundrum and one Ill pay attention to.

Gunner

The methodology of the left has always been:

1. Lie
2. Repeat the lie as many times as possible
3. Have as many people repeat the lie as often as possible
4. Eventually, the uninformed believe the lie
5. The lie will then be made into some form oflaw
6. Then everyone must conform to the lie

I do it with a green Scotchbright. Seems to work fine if not over done.

Gunner

The methodology of the left has always been:

1. Lie
2. Repeat the lie as many times as possible
3. Have as many people repeat the lie as often as possible
4. Eventually, the uninformed believe the lie
5. The lie will then be made into some form oflaw
6. Then everyone must conform to the lie

This stuff wasn't a cleaner, per se - but a "contact conditioner". Can't remember the name of the stuff - but it sure saved a lot of trouble on those old boards with about 35 or 40 sockets on them.

Back then we sold computers with a 3 year warranty. Without the "dope" an AMI motherboard didn't stand a chance. With it, they were pretty well bulletproof.

Yup. On MANY computers.

I'm talking CP/m 86 - virtually a PC. The first computer we had at the dealership back in the '80s.

Cramolin?

That can cause static damage, since it's nonconductive. Not all damage kills an IC right away.

There is a small EEPROM on the memory board that IDs the size. That code is copied & stored for comparison. The EEPROM is an I2C interfaced chip, and small. 24C01, 24C02, 24C04, 24C08 are typical I2C EEPROMS used on DDR modules. They are tiny eight pin surface mount ICs.

Cramolin worked well in the time of the AT&T 7300/3B1/Unix-PC (68010 based desktop unix box), and it was later discontinued for ozone later problems IIRC, and replace with DeOxit by the same company (both in aerosol cans). Also, I believe the high-end audiophile vendors were selling Cramolin's ingrediences in tiny bottles for really high prices under the name "Tweak". :-) The aerosol cans were not that expensive, and I bought several.

That product was not available when I was having the problems with the SWTP computers. I should have spent the extra to get the gold-plated molex pins and sockets. :-)

Enjoy, DoN.

Could be - I honestly don't remember. I just remember it was gaudawfull expensive, a clear thick gell like K-Y Jelly that we thinned with ethanol - and it was worth every cent. I think I've still got a bit hanging around somewhere.

Whatever you use, use it DAMP, and ground yourself. Gives everything a fighting chance.

Gunner on Mon, 19 Nov 2012 10:32:35 -0800 typed in rec.crafts.metalworking the following:

Sometimes, "Well this stick will work in this socket, but not in that one, and that one won't work in either configuration, But if we put this on in this socket, then the other one will work just fine."

-- pyotr Go not to the Net for answers, for it will tell you Yes and no. And you are a bloody fool, only an ignorant cretin would even ask the question, forty two, 47, the second door, and how many blonde lawyers does it take to change a lightbulb.

Shrug..maybe so. But Ive been doing it for over 15 yrs with no issues noted.

But thanks for the input. Maybe Ill switch over to pencil erasers or something?

Gnner

The methodology of the left has always been:

1. Lie
2. Repeat the lie as many times as possible
3. Have as many people repeat the lie as often as possible
4. Eventually, the uninformed believe the lie
5. The lie will then be made into some form oflaw
6. Then everyone must conform to the lie

That's even worse. Erasers are abrasive and non conductive, so they remove the plating and generate static. The approved process was a little contact sprayed into the socket, and reinsert the device or board. We used to add a thin film of GC 'Tunerlube' to keep air away from the mating surfaces.

Sure. Go into Setup to see all the things the BIOS detects and saves.

The lack of relative jumps was a major annoyance of the 8080 microprocessor. By the time the 6809 came out my homebrew computer was too far along (and too obsolete) to change, but I did borrow the video circuit from the 6809-based Radio Shack Color Computer.

I've designed a dynamic memory controller IC and was on the engineering team building industrial memory chip testers for National Semiconductor et.al., so I do know something about computer memory.

The testers could program UVPROMS much faster than their specs. We wrote the pattern in very fast passes repeatedly until all cells read back correctly, then hit it again a few times to be sure.

Dynamic RAM is supposed to be refreshed at a rapid rate to keep the charges from leaking out, but in practice it can retain data for many seconds at room temperature.

jsw

That delay isn't part of the normal boot sequence. The boot time varies too much depending on how many tuners and hard drives are connected for the memory to have much effect. Anyway 7 Media Center wakes the computer 5 minutes before a recording is scheduled to begin, and when I boot it manually I allow 15 minutes to aim the antenna for the strongest signal and fiddle with all the settings and connections of my complex and not completely reliable OTA TV installation. Doubling the RAM didn't change the power consumption enough to register on a KAW. jsw

The 1G sticks I removed have 34C021s on them. jsw