Topics

TDS3012B NVRAM

Ferenc Rózsa
 

Good afternoon for everybody,

I've got such a Tektronix TDS3012B scope and
run out the CR2032 coin cell in the NVRAM.
The type of the NVRAM is DALLAS DS1742W-120
So this is a 3.3V VCC, 2 KByte, JEDEC paralell programming NVRAM. Where can I buy it or which is the replacement type of this ?
This is 24 pin EDIP type.
Thank you very much for you help: Ferenc

 

Hi Ferenc,
You'll find what you're looking for in my message #157465 in thread " Clock error on TDS3034", including link to pictures of the Dallas Module.

Raymond

 

On Mon, Nov 18, 2019 at 11:45 PM, Raymond Domp Frank wrote:


You'll find what you're looking for in my message #157465 in thread " Clock
error on TDS3034", including link to pictures of the Dallas Module.
That's if you're prepared to open the module.

Raymond

David Kuhn
 

https://www.nacsemi.com/

When replacing the chip, install a VERY LOW PROFILE socket (pin carriers)
on the PCB. I bought my DS1742W-150 from NAC. Now the chips are probably
still 20+ years old. It may be a crap-shoot, but worked for me.

Some people dremel the chip and find the battery, unhook it, solder wires
and put a coin cell battery socket outside of the chip.



On Mon, Nov 18, 2019 at 2:58 PM Ferenc Rózsa <kapcsolattarto@...>
wrote:


Good afternoon for everybody,

I've got such a Tektronix TDS3012B scope and
run out the CR2032 coin cell in the NVRAM.
The type of the NVRAM is DALLAS DS1742W-120
So this is a 3.3V VCC, 2 KByte, JEDEC paralell programming NVRAM. Where
can I buy it or which is the replacement type of this ?
This is 24 pin EDIP type.
Thank you very much for you help: Ferenc



 

On Tue, Nov 19, 2019 at 05:00 PM, David Kuhn wrote:


When replacing the chip, install a VERY LOW PROFILE socket (pin carriers)
on the PCB.
That is not necessary. I've added a few pictures of the initial test-setup in my TDS3034 (now TDS3054) to the album "Clock error on TDS3034", before clean-up. It can be seen that there's ample space for a normal socket and putting a Li cell on top of the RAM

Yesterday, I wrote:

You'll find what you're looking for in my message #157465 in thread
"Clock error on TDS3034", including link to pictures of the Dallas Module.
Link to the photos:

https://groups.io/g/TekScopes/album?id=91131

Raymond

David Kuhn
 

Hello Raymond,

"That is not necessary. I've added a few pictures of the initial test-setup
in my TDS3034 (now TDS3054) to the album "Clock error on TDS3034", before
clean-up. It can be seen that there's ample space for a normal socket and
putting a Li cell on top of the RAM"

With my TDS3032 (no bloody A, or B, or C...), I tried a zero insertion
force socket, and another lower profile version, and neither would fit in
my scope as the I/O module would hit it. I ended up using lead carrier
pins (real old school form the 1980s there, maybe earlier). With lead
carriers, it is just slightly higher (maybe 1/8" - 3/16"), and even then,
on my scope, the chip "just" cleaned the GPIB/VGA/RS232 module. Maybe on
the TDS3034, the chip is in a different location?

Dave

On Tue, Nov 19, 2019 at 1:03 PM Raymond Domp Frank <@Raymond>
wrote:

On Tue, Nov 19, 2019 at 05:00 PM, David Kuhn wrote:


When replacing the chip, install a VERY LOW PROFILE socket (pin carriers)
on the PCB.
That is not necessary. I've added a few pictures of the initial test-setup
in my TDS3034 (now TDS3054) to the album "Clock error on TDS3034", before
clean-up. It can be seen that there's ample space for a normal socket and
putting a Li cell on top of the RAM

Yesterday, I wrote:

You'll find what you're looking for in my message #157465 in thread
"Clock error on TDS3034", including link to pictures of the Dallas
Module.
Link to the photos:

https://groups.io/g/TekScopes/album?id=91131

Raymond



David Kuhn
 

Ah, I just saw your pics. You are tight too. That must be a very low
profile socket. I think maybe even more so that my lead carrier pins. My
business depends on my original TDS3032. When this chip failed, I ordered
another TDS3032 from Alltest Instruments in NJ (~$1500). While I waited
for it to arrive, I did replace the DRAM in my original. I was way too
stupidly nervous at the time (I had to get back up and running) to dig the
chip out, so I bought a NOS ram. It was okay and worked good. I still use
my original as my primary (it's my baby - lol and seems to work slightly
better than my "new" one). If I have to do it again, I will definitely do
it the way you did!

I did learn that I have to hold the B TRIG in while booting it to initiate
that RAM chip. The screen came up in all kinds of pretty colors without
doing that - lol. Its been good now for about a year and went through
ISO17025 certification well.

Dave

On Tue, Nov 19, 2019 at 3:40 PM Daveyk021 . <daveyk021@...> wrote:

Hello Raymond,

"That is not necessary. I've added a few pictures of the initial
test-setup in my TDS3034 (now TDS3054) to the album "Clock error on
TDS3034", before clean-up. It can be seen that there's ample space for a
normal socket and putting a Li cell on top of the RAM"

With my TDS3032 (no bloody A, or B, or C...), I tried a zero insertion
force socket, and another lower profile version, and neither would fit in
my scope as the I/O module would hit it. I ended up using lead carrier
pins (real old school form the 1980s there, maybe earlier). With lead
carriers, it is just slightly higher (maybe 1/8" - 3/16"), and even then,
on my scope, the chip "just" cleaned the GPIB/VGA/RS232 module. Maybe on
the TDS3034, the chip is in a different location?

Dave

On Tue, Nov 19, 2019 at 1:03 PM Raymond Domp Frank <@Raymond>
wrote:

On Tue, Nov 19, 2019 at 05:00 PM, David Kuhn wrote:


When replacing the chip, install a VERY LOW PROFILE socket (pin
carriers)
on the PCB.
That is not necessary. I've added a few pictures of the initial
test-setup in my TDS3034 (now TDS3054) to the album "Clock error on
TDS3034", before clean-up. It can be seen that there's ample space for a
normal socket and putting a Li cell on top of the RAM

Yesterday, I wrote:

You'll find what you're looking for in my message #157465 in thread
"Clock error on TDS3034", including link to pictures of the Dallas
Module.
Link to the photos:

https://groups.io/g/TekScopes/album?id=91131

Raymond



 

On Tue, Nov 19, 2019 at 09:40 PM, David Kuhn wrote:


Maybe on
the TDS3034, the chip is in a different location?
Hi David,
Did you have a look at my pictures? https://groups.io/g/TekScopes/album?id=91131
The NVRAM in my TDS3054 (born as a TDS3034) sits near the side of the I/O module location, towards the right side of the 'scope. I don't know if the NVRAM location is different in the TDS30x2 but that seems to be the case.
Traditional (Textool) ZIF sockets would be too high but a standard turned pin socket easily fits, with the Li coin on its head.

Raymond

 

On Tue, Nov 19, 2019 at 09:53 PM, David Kuhn wrote:


You are tight too. That must be a very low
profile socket. I think maybe even more so that my lead carrier pins.
It's an ordinary turned-pin socket, maybe 4 mm (0.16") high and there was still room for another 4 mm of foam. It is however necessary to remove the I/O module plastic part before removing or inserting the NVRAM.

Are you sure holding B TRIG in while booting initializes the NVRAM? I'm still looking at a - 2945263 hours or so total up-time in my error log after replacing the battery...

Raymond

David Kuhn
 

" Are you sure holding B TRIG in while booting initializes the NVRAM? I'm
still looking at a - 2945263 hours or so total up-time in my error log
after replacing the battery..."

Hello Raymond,

On mine, it partially booted with its brains scrambled. I held in the
B-TRIG (saw that somewhere on here or on EEVBLOG) and it reset it. When I
use Tek Openchoice to get a screen shot, it starts out with the nice blue
background. When I run some other 3rd party program that take screen
shots, the background of the screen shot it clear from that point forward.
If I want the background to be blue again, I have to hold in the B-TRIG
when booting up. Then it resets it. I was not aware of a programming
command to change the background of the screen shot, but there must be
one. At least when the background is clear, it saves printer ink - lol
(but is not as pretty).

I wouldn't mind updating the BW of one of my TDS3032's but not sure I
should. I wonder what the Cal house would think if the TDS3032 booted as a
TDS3052? I get then ISO17025 calibrated each year. I know I have to
downgrade the software (where is the old code available at)? Then I see
the commands. Are those commands only through an internal programming
port, or can I give the commands via GPIB?

Dave

On Tue, Nov 19, 2019 at 4:18 PM Raymond Domp Frank <@Raymond>
wrote:

On Tue, Nov 19, 2019 at 09:53 PM, David Kuhn wrote:


You are tight too. That must be a very low
profile socket. I think maybe even more so that my lead carrier pins.
It's an ordinary turned-pin socket, maybe 4 mm (0.16") high and there was
still room for another 4 mm of foam. It is however necessary to remove the
I/O module plastic part before removing or inserting the NVRAM.

Are you sure holding B TRIG in while booting initializes the NVRAM? I'm
still looking at a - 2945263 hours or so total up-time in my error log
after replacing the battery...

Raymond



 

On Tue, Nov 19, 2019 at 11:31 PM, David Kuhn wrote:


On mine, it partially booted with its brains scrambled.
Hi David,
My 3034 had no problem starting up after I replaced the battery.

As I understand it, TDS3032's and TDS3052's are the same 'scopes, hardware wise, as in fact are all TDS30x2's as one group and TDS30x4's as another.
I have modified my three 30xy's. Calibration, including frequency- and step-behavior were fine afterwards and the new x was kept after upgrading the firmware. I used a PG506, an SG504, a home-built 200 ps step generator and Leo Bodnar's 40 ps step generator to verify.
Tek most likely used automated calibration. Possibly, all TDS30xy's were born as TDS305y's and were downgraded where desired. That could explain why after upgrading, calibration is still OK, including frequency- and pulse-behavior. Possibly, factory calibration used a "backdoor" and didn't even need to go through the x=5 - then - downgrade procedure.
I'm pretty confident that a spruced-up 3032 cannot by behavior be distinguished from the original version. The firmware is the same for all x'es and y's
I guess your cal. lab. and what the client orders determines whether they accept modified 'scopes. I'm convinced that the behavior during calibration and adjustment is identical.
Send me a personal email for instructions and data for temporarily downgrading and modifying. GPIB is fine for interfacing, very simple procedure. 30xyB's can use the network interface. Done all that. For downgrading FW from 3.41 and re-upgrading, you'll need a working internal floppy drive.

Raymond

David Kuhn
 

" Calibration, including frequency- and step-behavior were fine afterwards
and the new x was kept after"

I am curious where the calibration value are stored? It's not in that
battery ram. I asked before and some said there wasn't flash memory
onboard. So I thought the hardware was just held that good (yaa right)
that changes to a calibration table wasn't necessary. When I get data
sheets, for my equipment, I see where calibration values for the Agilent
34401A Meter have been tweaked/adjusted, but never anything for the TDS3032
scopes.

I see TDS3012 scopes on EBAY for cheap (<$800) that have been upgraded to
TDS3052s. I am tempted to get a four channel version (although I do not
100% need it for certifications that I do) and upgrade one. Of course then
there is the expensive I/O module, but I'm up for that.

Dave

On Tue, Nov 19, 2019 at 8:03 PM Raymond Domp Frank <@Raymond>
wrote:

On Tue, Nov 19, 2019 at 11:31 PM, David Kuhn wrote:


On mine, it partially booted with its brains scrambled.
Hi David,
My 3034 had no problem starting up after I replaced the battery.

As I understand it, TDS3032's and TDS3052's are the same 'scopes, hardware
wise, as in fact are all TDS30x2's as one group and TDS30x4's as another.
I have modified my three 30xy's. Calibration, including frequency- and
step-behavior were fine afterwards and the new x was kept after upgrading
the firmware. I used a PG506, an SG504, a home-built 200 ps step generator
and Leo Bodnar's 40 ps step generator to verify.
Tek most likely used automated calibration. Possibly, all TDS30xy's were
born as TDS305y's and were downgraded where desired. That could explain why
after upgrading, calibration is still OK, including frequency- and
pulse-behavior. Possibly, factory calibration used a "backdoor" and didn't
even need to go through the x=5 - then - downgrade procedure.
I'm pretty confident that a spruced-up 3032 cannot by behavior be
distinguished from the original version. The firmware is the same for all
x'es and y's
I guess your cal. lab. and what the client orders determines whether they
accept modified 'scopes. I'm convinced that the behavior during calibration
and adjustment is identical.
Send me a personal email for instructions and data for temporarily
downgrading and modifying. GPIB is fine for interfacing, very simple
procedure. 30xyB's can use the network interface. Done all that. For
downgrading FW from 3.41 and re-upgrading, you'll need a working internal
floppy drive.

Raymond



 

On Wed, Nov 20, 2019 at 04:55 PM, David Kuhn wrote:

Please find my response inline

I am curious where the calibration value are stored?

Calibration is stored in flash memory, as is the firmware (upgradeable via floppy disk).

It's not in that
battery ram. I asked before and some said there wasn't flash memory
onboard.

There definitely *is* flash memory onboard, see above.

So I thought the hardware was just held that good (yaa right)
that changes to a calibration table wasn't necessary.
The calibration procedure (select calibration mode and follow prompts, using calibration equipment) is in the Service Manual.

When I get data
sheets, for my equipment, I see where calibration values for the Agilent
34401A Meter have been tweaked/adjusted, but never anything for the TDS3032
scopes.

I see TDS3012 scopes on EBAY for cheap (<$800) that have been upgraded to
TDS3052s. I am tempted to get a four channel version (although I do not
100% need it for certifications that I do) and upgrade one. Of course then
there is the expensive I/O module, but I'm up for that.
Raymond