2465B ... NVRAM DS1225Y problem or no problem?

[Joseph Bozzer]

What's visually happen at power, if the NVRAM is empty or dead?
The auto test work or no?

[jeanguypataterub]

What's happen at power, if the NVRAM is empty or dead?

I unplugged it and also took out the internal battery for 5 minutes. Also did a default reset. Didn't help

[LazyJack]

You will need to do a complete calibration.

[Joseph Bozzer]

complete calibration??

Pfffff ....  Everything worked well until today

When I saw that my NVRAM was from 1991 I decided to replace it
1) So I removed it and installed it in my programmer to read and save its content
2) I immediately noticed that there was a problem because the content seemed empty! ... a great aera with only "00" and rest with "FF"

3) I put it back in the 2465B and now at startup, the self-test turns on the "A WSP" led then stops
4) I done a hole to access at the pin7 to mesure the internal voltage : still 3v after 29 years!!

So what did you think happened?

[rolkinas]

The same happened on me couple years ago...
If battery is just on threshold of keeping data, putting such NVRAM to programmer to read content leads to failure for some reason. It was mentioned in eevblog somewhere. The most interesting thing is that writing data to old NVRAM, works just fine. Data in that old Dallas was OK after couple of months without being powered.
Luckily there is local person who service and calibrate Tek 2400 series scopes. so there was no problem for me with calibration. Now I have copy of my NVRAM in all my computers...

[BravoV]

Suggesting to have a cup of coffee, spare some time and read this before you do drastic change/repair -> https://www.eevblog.com/forum/testgear/tektronix-2465b-oscilloscope-teardown/

[james_s]

I had the same thing happen with 2 different DS1644's in TDS300 scopes. Somewhere in the process of desoldering them and trying to read on a programmer they got corrupted. Later with a DS1744 I found the first read attempt would fail, but if I immediately tried again it would read valid looking data. I'm not really sure why, maybe related to the timing of power being applied to the chip and trying to read it.

[Joseph Bozzer]

I will try .. but when I launch the reading I see the programmer baragraph working and his message "read complete" ... and the result is a great area of only "00" and the rest is "FF"

I think that the NVRAM is read and it just empty of calibration data
At power I have a error4 fail 11 .. I think that I have my confirmation

Do you know what's happen if I use a .BIN of an another 2465B?
(I'm not use this scope in professional domain but just in hobby ... extremest precision is not necessary for me)

[james_s]

It should work, but the calibration will not be quite right. How far off it will be is probably impossible to say, I have no idea how much variation there is from one 2465B to the next but it's probably a good place to start.

How difficult is it to calibrate these things? I have been unable to come up with a pulse generator that satisfies the TDS340 calibration routine, even the voltage calibration is extremely sensitive, I can sometimes get it to pass using an old precision voltage standard, I have no idea why it's so delicate, almost makes me wonder if they made it deliberately tricky in order to encourage people to send in the scopes for calibration or replace them when the Dallas battery dies.

[bob91343]

I just replaced the NVRAM on a 2440 and now the scope won't work.  I checked my work and it looks good but that doesn't mean a lot.

I bought the new parts from China so that may be the problem.  But shouldn't the scope function, if uncalibrated?

I get a nonsense display and no panel controls do anything.

[james_s]

If  you bought a Dallas NVRAM from China you bought a counterfeit part that is most likely non-functional. I bought 3 or 4 different ones to play with at one point and 3 of them were fakes that had various problems while the fourth was an old used part that still had data on it, the top had been sanded down and re-painted with a new date code.

Do you still have the original part? If so try putting that back in, hopefully you used a socket?

[bob91343]

I still have the original parts but one pin broke on one.  It's a major job.

What I really want to know, is it likely that this problem is caused by a bad IC?  I don't mind going through the exercise of switching back but wonder if it will fix it.

Before the procedure, the scope worked well but didn't save its configuration between startups.  I started to cut apart one unit but it's very hard epoxy and a lot of work.  I did buy some better cutters for my Dremel so maybe I can get in there and replace the batteries.  If it has a chance of fixing it.

[BravoV]

If you've spent some time reading that thread I was suggesting, alot of your troubles (not only this one) can be avoided, like loosing the cal, especially on weak NVRAM by using the scope EXER option to retrieve the cal values.

[bob91343]

Well I guess it's too late for that...

[james_s]

I still have the original parts but one pin broke on one.  It's a major job.

What I really want to know, is it likely that this problem is caused by a bad IC?  I don't mind going through the exercise of switching back but wonder if it will fix it.

Before the procedure, the scope worked well but didn't save its configuration between startups.  I started to cut apart one unit but it's very hard epoxy and a lot of work.  I did buy some better cutters for my Dremel so maybe I can get in there and replace the batteries.  If it has a chance of fixing it.

It's easy, I've fixed a handful of Dallas chips. Use a hot air pencil or heat gun to warm the epoxy and it softens right up. Takes a little patience, put the chip in a vice, heat it, dig a little, heat it, dig a little more, cut the battery out and solder wires to the pins. Careful not to tug too hard on the nickel strips or you can break the solder on the PCB. A magnet can be used to locate the battery, it's just under the top surface. Cut the plastic shell off with a knife before you attack the epoxy.

[Joseph Bozzer]

It's easy to add an external battery at NVRAM

1) I used this Dremel head
Before begin .. insert the NVRAM in a socket to protect his fragile pins !!





2) I started in front of the first "A" of text "DALLAS" ... this is where is located the pin 7 of IC
(Luckily the DS1218 is a old DIL8 case with his large pins which makes research easier)





3) Carefully I came down and widen the hole very near of case until i seen the copper appear..
Then i widened it enough to be able to solder a wire
(It's the good moment to measure the internal battery voltage)

Maybe a piece of advice .. do not widen it too much on the side of the "L" of dallas so as not to short-circuit, with the dremmel head, the pins 7 and 6 of DS1218 because I don't know if that could kill  it?..
on the "D" side there is no danger .. see schema below









Result :
Before this operation I couldn't read the NVRAM on my programer and after I could do it .. I was happy!


My final solution :
I decided to keep the original NVRAM equipped of an external battery
For the battery I chose this one on Amazon

[bob91343]

Wow this is great information!  I wish I had seen this before I ruined my scope.  Well then perhaps all isn't lost at that.

I will print the info and get to work on the unit.  I am hoping the problem wasn't caused by my workmanship, which isn't all that sterling (to be kind).

[Labrat101]

HI nice job .
I did 3 before reading this . but I drimmeled down from the top and with a little care I cut the top lead off the battery and them very carefully cut the entire battery out .
leaving the old battery in they start to leak . my battery was easy to spot as it was lifting the plastic up . so i just drilled round the bulge. (a magnet is good for locating)
recalibrated  and all back to normal.  I never removed the chip I did the whole operation with chip in place . This is the third time I 'v done it this way so far non of my patients
have died . Make sure there is Large coffee and a cigarette and a steady  hand