Replacing 80xx series ICs?

[DaveC]

It's in an old (mid-80's) industrial machine. The French mfr. is out of
business. The circuit I'm trying to resurrect takes a digital measuring
encoder as input and displays the measurement on 7-segment LED displays.
Display went wonky for a few weeks, then went blank, entirely. DC voltages
and ripple are all good. Tried reseating socketed components; no joy.

Lacking any documentation, I'm going to try some "shotgun" troubleshooting by
replacing the major components on the board (they're socketed).

Large components include:
D8085AHC
D8155HC
D8279C-2
uPB8212C
All parts NEC mfr.

When replacing with components available today, what are my limitations? (For
example, I know "H" means HMOS, and that is a "must".) What else should I be
careful re. the letter digits of the p/n's when selecting parts with the same
numerical digits of the p/n's?

I know that the 8279 is a programmable controller, so replacing that without
having the source code to program it with, is out of the question.

Just trying to help a client keep this otherwise good machine in service and
avoid "binning" it.

Thanks,

 

[William at MyBlueRoom]

http://www.csee.umbc.edu/~plusquel/310/slides/8086_IO3.html\

looks like the 8279 has an internal OTP EPROM.

Might be tough to find a replacement

Bill

www.myblueroom.com

 

[William at MyBlueRoom]

Opps my bad, looks like the 8279 is a programmable keyboard I/O chip.
So you should be able just to pop one in. (assuming that's what's wrong
with the device)

Bill

 

[DaveC]

Thus spake William at MyBlueRoom:

http://www.csee.umbc.edu/~plusquel/310/slides/8086_IO3.html\

looks like the 8279 has an internal OTP EPROM.

This is what I need help determining. If this is a "production programmed"
device (burned at time of manufacture), it's a deal-killer for my customer --
finding a programmed replacement is pretty much not going to happen.

But if this is a "dynamically programmed" device (ie, counters are loaded at
run time and reset by external devices), I presume that off-the-shelf it will
replace the suspect one in the machine.

Is the 8279 a one-time programmed device?

Thanks,

 

[DaveC]

Thus spake William at MyBlueRoom:

Opps my bad, looks like the 8279 is a programmable keyboard I/O chip. So you
should be able just to pop one in. (assuming that's what's wrong with the
device)

Oh, you answered my question while I was posting another message to ask you
this very question. (Such is the down-side of asynchronous communications!)

Thanks!

 

[Pooh Bear]

Thus spake William at MyBlueRoom:


This is what I need help determining. If this is a "production programmed"
device (burned at time of manufacture), it's a deal-killer for my customer --
finding a programmed replacement is pretty much not going to happen.

It appears not to be.

http://www.csee.umbc.edu/~plusquel/310/slides/8086_IO3.html

Graham

 

[Pieter]

It's in an old (mid-80's) industrial machine. The French mfr. is out of
business. The circuit I'm trying to resurrect takes a digital measuring
encoder as input and displays the measurement on 7-segment LED displays.
Display went wonky for a few weeks, then went blank, entirely. DC voltages
and ripple are all good. Tried reseating socketed components; no joy.

Lacking any documentation, I'm going to try some "shotgun" troubleshooting by
replacing the major components on the board (they're socketed).

Large components include:
D8085AHC
D8155HC
D8279C-2
uPB8212C
All parts NEC mfr.

When replacing with components available today, what are my limitations? (For
example, I know "H" means HMOS, and that is a "must".) What else should I be
careful re. the letter digits of the p/n's when selecting parts with the same
numerical digits of the p/n's?

I know that the 8279 is a programmable controller, so replacing that without
having the source code to program it with, is out of the question.

Just trying to help a client keep this otherwise good machine in service and
avoid "binning" it.

Thanks,

Hi,

All those parts are still available in the market.

Pieter

 

[nospam]

Is the 8279 a one-time programmed device?

No, it is just a peripheral with some registers to 'program' its function.

 

[bg]

DaveC wrote in message

It's in an old (mid-80's) industrial machine. The French mfr. is out of
business. The circuit I'm trying to resurrect takes a digital measuring
encoder as input and displays the measurement on 7-segment LED displays.
Display went wonky for a few weeks, then went blank, entirely. DC voltages
and ripple are all good. Tried reseating socketed components; no joy.

Lacking any documentation, I'm going to try some "shotgun" troubleshooting by
replacing the major components on the board (they're socketed).

Large components include:
D8085AHC
D8155HC
D8279C-2
uPB8212C
All parts NEC mfr.

When replacing with components available today, what are my limitations? (For
example, I know "H" means HMOS, and that is a "must".) What else should I be
careful re. the letter digits of the p/n's when selecting parts with the same
numerical digits of the p/n's?

I know that the 8279 is a programmable controller, so replacing that without
having the source code to program it with, is out of the question.

Just trying to help a client keep this otherwise good machine in service and
avoid "binning" it.

Thanks,
--
DaveC
[email protected]
This is an invalid return address
Please reply in the news group

I do believe that the 8279 is not a programmed device. I could go look it
up, but that would take work!
The 8279 was often used for keyboard input and display output. The display
will be multiplexed. There
should be a BCD to seven segment converter chip that drives all of the
displays in parallel. Then each display will have a power transistor that
switches Vcc on for that display. The power transistors on signal is sync'd
to the BCD word so that the data appears on the correct display. Make
sense?

 

[Arfa Daily]

It appears not to be.

http://www.csee.umbc.edu/~plusquel/310/slides/8086_IO3.html

Graham

It seems to me that everyone is "going off on one", before the OP has
applied at least basic proper fault finding techniques. Before going in
shotgun and replacing parts that - trust me - seldom if ever go faulty, you
should first check that the 5v supply is absolutely good - voltage, ripple
AND noise, that the reset generator is working correctly, and that most
importantly, the system clock generator is running, and producing a correct
amplitude and frequency, signal. Also that any following clock distribution
buffers are working.

If none of those checks turn up the problem, and in at least 80% of cases on
this old simple logic, they will, then the next move is to download a
datasheet for the display controller, and figure why it's not driving the
displays. The required inputs are not many, and not complex, and require
only a 'scope to verify.

From years of experience repairing electronic equipment, I can tell you that
blanket replacement of components in the fond hope of hitting the right one,
is about the worst approach that you can take, and likely to result in more
problems than the probably simple one that you had in the first place.

Arfa

 

[linnix]

Hi,

All those parts are still available in the market.

Perhaps the local fea market. Look for old 8 bits micros. The 8085
is an old 8 bits micro, so there must be an EPROM somewhere.
Unfortunately, EPROM data are not designed to last for 20 years. It's
amazing to have lasted so long already. My first bet is the EPROM
losing contents. Unless you have a copy of the data somewhere, forget
it.

Sound like this is a frequency counter/display or something. A single
chip micro (AVR or PIC) can do the same.

 

[Wild Bill]

It sounds as though the instrument is a Digital Readout display
associated with machining or other production equipment.
I don't have much advice for troubleshooting other than what's already
been suggested by Arfa.
It's been too many years, and too little practice since my training in
these types of circuits. I have a completely dead Anilam DRO display,
and I've had difficulty understanding the circuits even though I have a
schematic.

If you proceed with IC replacement, check any datasheets you can find
to confirm that the operating voltages of replacement parts are the
same.

Newer used DRO systems are plentiful, depending upon your searching
abilities. If the linear encoder scales used with the DRO display unit
are the commly used 5 volt TTL squarewave types, you could possibly
just replace the display unit with one that's compatible.

Finding a suitable replacement will probably be a little tougher if the
DRO display also has outputs for interfacing with NC or CNC machine
controls, but numerous displays have various optional modules to
convert inputs to outputs.

There are numerous surplus suppliers of older ICs in the U.S, and more
semiconductor sellers are gradually appearing on eBay.

Cheers
WB
...............

 

[Zak]

Perhaps the local fea market. Look for old 8 bits micros. The 8085
is an old 8 bits micro, so there must be an EPROM somewhere.
Unfortunately, EPROM data are not designed to last for 20 years. It's
amazing to have lasted so long already. My first bet is the EPROM
losing contents. Unless you have a copy of the data somewhere, forget
it.

Reducing the supply voltage a bit (was it? or increasing?) might still
recover the data.


Thomas

 

[Spehro Pefhany]

Perhaps the local fea market. Look for old 8 bits micros. The 8085
is an old 8 bits micro, so there must be an EPROM somewhere.
Unfortunately, EPROM data are not designed to last for 20 years.

Do you have any data which supports this claim?

Most manufacturers claim 100 years or more at normal temperatures.

It's amazing to have lasted so long already.

Best regards,
Spehro Pefhany

 

[Spehro Pefhany]

Reducing the supply voltage a bit (was it? or increasing?) might still
recover the data.

Yes.


Best regards,
Spehro Pefhany

 

[Michael A. Terrell]

It's in an old (mid-80's) industrial machine. The French mfr. is out of
business. The circuit I'm trying to resurrect takes a digital measuring
encoder as input and displays the measurement on 7-segment LED displays.
Display went wonky for a few weeks, then went blank, entirely. DC voltages
and ripple are all good. Tried reseating socketed components; no joy.

Lacking any documentation, I'm going to try some "shotgun" troubleshooting by
replacing the major components on the board (they're socketed).

Large components include:
D8085AHC
D8155HC
D8279C-2
uPB8212C
All parts NEC mfr.

When replacing with components available today, what are my limitations? (For
example, I know "H" means HMOS, and that is a "must".) What else should I be
careful re. the letter digits of the p/n's when selecting parts with the same
numerical digits of the p/n's?

I know that the 8279 is a programmable controller, so replacing that without
having the source code to program it with, is out of the question.

Just trying to help a client keep this otherwise good machine in service and
avoid "binning" it.

Thanks,

All of those parts were originally from INTEL. NEC was a second
source. What type of RAM does it use?

--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

 

[Erik Baigar]

Yes.

....also had good experience with carefully cooling
the EPROMS during readout...

Erik.

 

[Tzortzakakis Dimitrios]

Perhaps the local fea market. Look for old 8 bits micros. The 8085
is an old 8 bits micro, so there must be an EPROM somewhere.
Unfortunately, EPROM data are not designed to last for 20 years. It's
amazing to have lasted so long already. My first bet is the EPROM
losing contents. Unless you have a copy of the data somewhere, forget
it.

When you say EPROM, you mean those chips with the little transparent window
that could be erased with special UV light?If yes, could this just happen
with long exposure (many decades) to normal sunlight (even if dim in the
bowels of the machine).I strongly doubt though, that you could find nowadays
labs with the UV light and programm devices to programm (again) those chips,
now that flash chips are so common and cheap (EEPROM).
Are there any chips with the characteristical transparent window?

 

[linnix]

...

Most manufacturers claim 100 years or more at normal temperatures.

People who made these claims in the 80s are mostly retired or dead.

Can't find any typical EPROM (2716 or 2732) used in the 80s. A newer
one like AT28C16 says data retentions of 10 years. If you want to bet
$1 per year, we can find an independent engineer for verification.

 

[Pieter]

Perhaps the local fea market. Look for old 8 bits micros. The 8085
is an old 8 bits micro, so there must be an EPROM somewhere.
Unfortunately, EPROM data are not designed to last for 20 years. It's
amazing to have lasted so long already. My first bet is the EPROM
losing contents. Unless you have a copy of the data somewhere, forget
it.

Sound like this is a frequency counter/display or something. A single
chip micro (AVR or PIC) can do the same.

There are still lots of devices in older stocks available. And for
many devices replacement parts have been created by others.

I just did a quick search and found hundreds of the D8085HC of NEC at
my contacts. It is expensive to get 1 part due to the shipment costs,
but it is possible. I havent searched the other partsm, but I am
certain they are available too.

Pieter