|
I like know is possible read Dallas nvram whitout desolderig the chip
the device work good bat the date of dallas is 1990
than for collaboration
mirco
from italy
|
Hi Mirco,
I recently managed to copy the nvram without desoldering it. Please search for my post under heading "NVRAM 2465 /2467 replacement" in this forum. If you don't understand anything let me know.
Ta
Das
from Australia
|
Ciao Das
Thank you for your reply.
I see you are expectingi hope you can help me.
My instrument works well but the date of the dallas chip is 1990my instrument I think that soon the data will be lost.
I ask you if i read the calibration data on the monitor with the Exer 02 function is possible to copy them in the Dallas device manually ?
There is a document that indicates where to position the data in the .bin file?
Many thanks for your collaboration
Best regards
Mirco from italy
|
Hi Micro,
Using Exer 02 to keep a copy of cald data is a good idea , the cal data is read from nvram location starting 1e00, so in theory can recreate, except there is probably a check sum of the data stored somewhere on the nvram, I am not sure where exactly where, there is no documentation on the checksum anywhere I have been able to find.
So if you are keen to copy the nvram, use the instructions, I previously mentioned in my post "NVRAM 2465 /2467 replacement". And you can get the data out. Desoldering the nvram to read is risky, because in these old nvram chips, the data gets corrupted easily by the desoldering process. After I had copied my nvram data out, I desoldered the nvram to replace it and found the contents in the old nvram corrupted by the desoldering.
cheers
Das
|
thanks Das for replay
I would like to use your method but I did not understand how to do it.
I ask you if you have a document that explains step by step the procedure.
after the first reading I have to copy the data on the pc?
with which syntax I add the 2 hex files
from which address to where?
I did not understand this passage
Thank for collaboration
best regards mirco
|
HI mirco,
Both ramlow.dat and ramhigh.dat are 8k in size, each of these programs copies half of the nvram.
Please copy ramlow.dat into the new nvram to be used for replacement, remove u2360 or u2260( eprom) from a5 board which contains the scope firmware andreplace the eprom with new nvram with ramlow.dat programmed into it .Turn the machine on, after 2 to 3minutes (copy should only take 100ms), short pin 1 and 2 on the 6802 cpu (u2140), to force Non maskable interrupt(NMI) for cpu.While shorting these two pins, also move jumper p503 to "diag" position to force cpu reset. The scope is then powered off. The last two steps prior to scope turn off may not be necessary, but are precautionary, in case with power down, brown out causes erratic read/write cpu behaviour.
The above procedure is repeated with ramhigh.dat. Remember to unshort pins 1 and 2 of cpu and put p503 back to "norm" position prior to restarting the scope, with the second program.
The old nvram contents (which contains the cal data) can be found in the new nvram starting at location 200hex. to 11FFhex inclusive. Both the ramlow.dat and ramhigh.dat programs copy the old nvram data into these locations.
For the next steps use a PC with a hex editor to combine the lower 1000hex bytes (obtained using ramlow.dat program) and upper 1000hex bytes data (using ramhigh.dat program) together. To combine the lower 1000 hex bytes with the upper 1000 hex bytes, simply append the upper 1000H bytes to lower 1000H bytes so the new combined data file will be 2000H bytes long. 2000 Hex bytes is the size of the nvram, Location 0 of the combined data needs to be copied to address 0 on nvram and last byte at position 1fff of the combined data needs to be copied to address location 1fff of the nvram.
hope it helps
Cheers
Das
|
Ciao Das thanks for the reply.
Maybe we can succeed in the adventure with your great help
Now the procedure is more clearer to me.
My cpu is SC67127p may be fine anyway ? is same 6802 ?
The jumper between pins 1 and 2 on Cpu I do it with 2 wires or a metal tip ?
I also ask if the procedure is invasive
if I destroy something or not.
thank for collaboration
ciao mirco
|
HI MIcro,
Yes i soldered temporary wires between pin 1 and 2 of cpu. The cpu is 6802 on all the tek 246x oscilloscopes despite the label "sc67127p". Dont be afraid the cpu assisted copy of data is much less likely to go wrong then if you desoldered the old nvram and then tried reading it, which often fails..
Cheers
Das
|
You all are going to a lot of work to save calibration
data that ceased to be relevant probably more than a
decade ago.
If the scope had something important, like firmware,
stored in the NVRAM I could see spending the effort,
but what you are talking about doing is surprisingly
similar to looking for ways to change a new fender
into a dented and rusted out duplicate of the old
fender on your car... so you can replace the old
fender...
-Chuck Harris
Claimer: I trade in discount calibrations, so my opinions
are clearly biased in favor of doing calibrations...
ADas via Groups.Io wrote:
toggle quoted message
Show quoted text
HI MIcro,
Yes i soldered temporary wires between pin 1 and 2 of cpu. The cpu is 6802 on all the tek 246x oscilloscopes despite the label "sc67127p". Dont be afraid the cpu assisted copy of data is much less likely to go wrong then if you desoldered the old nvram and then tried reading it, which often fails..
Cheers
Das
|
On 2018-04-04 8:31 AM, Chuck Harris wrote: You all are going to a lot of work to save calibration
data that ceased to be relevant probably more than a
decade ago.
If the scope had something important, like firmware,
stored in the NVRAM I could see spending the effort,
but what you are talking about doing is surprisingly
similar to looking for ways to change a new fender
into a dented and rusted out duplicate of the old
fender on your car... so you can replace the old
fender... I think this is an excellent point, but are calibration procedures documented and possible for end users? --Toby (I haven't looked into that question yet, but I do own a TDS460A & others that are pre-digital)
-Chuck Harris
Claimer: I trade in discount calibrations, so my opinions
are clearly biased in favor of doing calibrations...
ADas via Groups.Io wrote:
HI MIcro,
Yes i soldered temporary wires between pin 1 and 2 of cpu. The cpu is 6802 on all the tek 246x oscilloscopes despite the label "sc67127p". Dont be afraid the cpu assisted copy of data is much less likely to go wrong then if you desoldered the old nvram and then tried reading it, which often fails..
Cheers
Das
|
The manual has the entire procedure.
-Chuck Harris
toby@... wrote:
toggle quoted message
Show quoted text
On 2018-04-04 8:31 AM, Chuck Harris wrote:
You all are going to a lot of work to save calibration
data that ceased to be relevant probably more than a
decade ago.
If the scope had something important, like firmware,
stored in the NVRAM I could see spending the effort,
but what you are talking about doing is surprisingly
similar to looking for ways to change a new fender
into a dented and rusted out duplicate of the old
fender on your car... so you can replace the old
fender... I think this is an excellent point, but are calibration procedures
documented and possible for end users?
--Toby
(I haven't looked into that question yet, but I do own a TDS460A &
others that are pre-digital)
-Chuck Harris
Claimer: I trade in discount calibrations, so my opinions
are clearly biased in favor of doing calibrations...
ADas via Groups.Io wrote:
HI MIcro,
Yes i soldered temporary wires between pin 1 and 2 of cpu. The cpu is 6802 on all the tek 246x oscilloscopes despite the label "sc67127p". Dont be afraid the cpu assisted copy of data is much less likely to go wrong then if you desoldered the old nvram and then tried reading it, which often fails..
Cheers
Das
|
hello
I did not understand this last post
from what I read on internet, if the calibration data were lost, not all tests would pass.
in my instrument all the tests are ok
and work good
thank for collaboration
best regards
mirco
|
I agree ideally for a scope of this vintage a recalibration with a fresh nvram installed is the only proper way to get the scope working to its absolute specs and keep it there.
But if the typical user doesn't have the timing generators, step generators etc for the recal, then a copy of the current cal data, is still worth something. If the scope is currently usable, then a copy of the nvram at least keeps it usable. An old NVRAM needs replacement anyway, even if the scope was recalibrated, as they are well and truly past their use by date, and data corruption an imminent threat.
The cal data copy procedure is pretty simple really, takes very little time, and the only equipment needed is a PC and ram reader.
With a copy of the cal data in a new NVRAM, when the opportunity comes up a recal can be performed.
|
I stand corrected.
-Chuck Harris
ADas via Groups.Io wrote:
toggle quoted message
Show quoted text
I agree ideally for a scope of this vintage a recalibration with a fresh nvram installed is the only proper way to get the scope working to its absolute specs and keep it there.
But if the typical user doesn't have the timing generators, step generators etc for the recal, then a copy of the current cal data, is still worth something. If the scope is currently usable, then a copy of the nvram at least keeps it usable. An old NVRAM needs replacement anyway, even if the scope was recalibrated, as they are well and truly past their use by date, and data corruption an imminent threat.
The cal data copy procedure is pretty simple really, takes very little time, and the only equipment needed is a PC and ram reader.
With a copy of the cal data in a new NVRAM, when the opportunity comes up a recal can be performed.
|
If you would quote some of the message you were responding
to, others could figure out what you were talking about.
-Chuck Harris
zoldanmirco via Groups.Io wrote:
toggle quoted message
Show quoted text
hello
I did not understand this last post
from what I read on internet, if the calibration data were lost, not all tests would pass.
in my instrument all the tests are ok
and work good
thank for collaboration
best regards
mirco
|
good morning
to the whole group
I live in Italy
I followed this discussion
and it's very interesting
I have a 2465b in which the dallas chip needs to be replaced (DS1225Y-200)
I'm afraid to unsolder it because the calibration data is lost
I have all the equipment to recalibrate but I want to try using the ADAS method
I need some info and I think you can help me
can I replace the dallas chip with one with a lower access time? (for example DS1225AD-150 or -70)
what is the name of the hexadecimal editor program to use to merge the two ramlow and ramhight files?
what programmer do you use to read / write the dallas chip?
is the TL866CS or TL866A minipro suitable for use?
I would also like to try using the RAMTRON FM1608
is the TL866 compatible with fram 1608?
thanks for the answers and your help
Simone
|
If you are going to recalibrate, there is no need
to save, merge, or hexedit any data in the Dallas chip...
no need at all. Following the calibration routines
saved in the ROM of the A5 board takes care of everything.
I have no idea what you mean about "trying the ADAS method"
of recalibration...
There is only one method, tektronix's method, and it is
embodied in the A5 card's ROMS, and the service manual.
Any other method is almost certainly wrong.
The NVRAM part you want is: DS1225AD-200 IND+ .
It is readily available from distributors like DigiKey, or
Mouser in the USA. Do not buy your NVRAMS from ebay!
The Chinese buy scrap PC boards in mind blowing quantities
from US recyclers. Every scrap yard has a population of
Chinese scavengers that heat the boards over trash fires
burning in oil drums, and they shake (smack) the boards so
that the parts fall off into tubs.
They sell these ugly solder splatted scrap parts to other
Chinese scavengers that dip the parts from the tubs into
hot peanut oil to melt and reflow the solder, making the
leads look nice and shiny.
Other Chinese scavengers remove gold plating in acid baths,
weld new leads on those that were trimmed too short, and
then replate the leads with something shiny and cheap.
The NVRAMS they remove from these scrap PC boards get
"refurbished" by cleaning up the pins, sanding off the
manufacturer's markings, and stamping them with a popular,
but fake, part number and data code.
You can buy rubber stamps with the markings for a Dallas NVRAM
from China.. There are companies over there that will create a
new stamp for you from a scan, of an IC top.
All of these "refurbished" old/used NVRAMS with fake markings
get sold in bulk in markets that most folks will never see, and
resold on ebay.
The RAMTRON part is readily available in SMD form. You
have to install it on a carrier PCB to convert it to a
DIP pin arrangement. The carrier PCB's are readily available
on ebay. Buy your FRAMS from a reliable source!
You likely can't program the FRAM correctly with your EPROM
programmer. The FRAM uses the falling edge of the CS signal
to capture the address lines... EPROMS don't. You may see
the first and last addresses contain wrong data.
I have installed FRAMS in many of my customer's scopes, and
one of my own. I haven't heard any complaints so far...
-Chuck Harris
simone minelli via Groups.Io wrote:
toggle quoted message
Show quoted text
good morning
to the whole group
I live in Italy
I followed this discussion
and it's very interesting
I have a 2465b in which the dallas chip needs to be replaced (DS1225Y-200)
I'm afraid to unsolder it because the calibration data is lost
I have all the equipment to recalibrate but I want to try using the ADAS method
I need some info and I think you can help me
can I replace the dallas chip with one with a lower access time? (for example DS1225AD-150 or -70)
what is the name of the hexadecimal editor program to use to merge the two ramlow and ramhight files?
what programmer do you use to read / write the dallas chip?
is the TL866CS or TL866A minipro suitable for use?
I would also like to try using the RAMTRON FM1608
is the TL866 compatible with fram 1608?
thanks for the answers and your help
Simone
|
hi
chuck
thank you for your help and advice
you are right that it is recommended to recalibrate after so many years
but my curiosity is to be able to read the data of the old calibration
when I say "try the ADAS method" I refer to the ADAS member who wrote the program in order to read the calibration data without unsoldering the dallas
I read on other forums that you can replace the dallas with one with faster access time
you can not do the opposite, that is, if the oscilloscope has a 70 ns nvram, you can not put a 200ns nvram
what do you think about it?
I will certainly buy the nvram on digikey or mouser (I know very well the Chinese filth)
what programmer do you use for NVRAM dallas ds1225ad ?
what programmer do you use for FRAM RAMTRON FM1608?
Thanks for the info
Simone
|
simone, from what I understand you will loose the info if you use a grounded tip soldering iron to remove it. ungrounded tip should be fine. as for programmer I am in the same boat with a 2430a. if you find the solution let me know as well. I have the chips coming and the sockets zif as well on the way. I need the program from mine as well.
|
Hi all,
here's how I replaced the Dallas chip in my 2465B's.
I modified my soldering station a long time ago:
Originally the chassis of the control-unit was grounded, and so was the tip.
I disconnected the ground-wire to the tip and put an 1 MOhm in series to ground.
This way the tip always has a zero volt potential, and if I touch something still a bit live (like a capacitor) nothing really catastrophic will happen.
I modified my vacuum de-soldering station the same way.
Using a narrow nozzle on the de-soldering station I removed the Dallas from my 2465B's.
I could clear the tin from all pins but one, the ground pin.
For this I needed the soldering iron to keep the pin hot while pulling the Dallas out of the board, because that pin is connected to the ground plane inside the board. (too much cooling)
Using my modified soldering station I put in a new socket in the SMD A5 board.
I used a TL866 MiniPro to read the old and program the new Dallas chips I bought at Mouser's.
Running like a charm since.
Un saludo,
Leo
I used this to
|
Gif Sim
satbeginner