TDS 684C NVRam Backup

[Armadillo]

Dear fellow members;

After reading some of the episodes of NVram problems and the desoldering difficulties, I decided to backup a copy of my TDS 684C Dallas NVRam for  safekeeping purpose.

The NVram Chip used in the TDS 684C is a DS1250Y which is 512K X 8 non-volatile static rams. This is in contrast to the DS1650Y chip as used in the TDS520 scope of same memory size and same pin counts.

However,  I can only find memory map details for the TDS520 scope. The NVram base address is at 0400 0000–040F FFFF. But I am unable to find the NVram memory map details for the TDS 684C scope.

Has anyone backup the nvram data from the TDS 684C before? Any info would be appreciated. Thanks.

I assumed the memory to be same as the TDS 520 and downloaded the 'nvram' data from my scope, which looks something like attached. However, I have no idea if the data is any correct. Any ideas? thanks.

[james_s]

IMHO the easiest and most reliable thing to do is desolder the chip and dump it with an EPROM burner. Obviously this is a good time to install a socket so you can easily change it later. AFAIK there's nothing critical in the DS1250, it's used for saving waveforms and user settings. The timekeeper holds the options and error log, it can be backed up in a similar manner. Cal data is held in eeprom on the acquisition board.

[Armadillo]

Thanks james_s, that is really a important piece of information there.

[Armadillo]

I am inclined to think that the GPIB read of the DS1250Y of the TDS 684C was OK. Attached interesting pattern comparison of feedback.loop TDS540 DS1650Y with my TDS 684C. You can see the peculiar "white 00" pattern similarity between the 2, though some of the bytes are not the same. Particularly not the same is the error logs starting page, but I don't expect the 2 to be the same anyway. When I find time, I will desolder the chip and read it with my GQ-4X to ascertain once and for all that the GPIB read can be reliable. I hope you can visualize the pattern attached.

Left [TDS 684 from Armadillo] ===========================================Right [TDS 540 from feedback.loop]

[nctnico]

Do a memory clear (this may be called secure erase or something like that). Most of the data stored in the NVRAM is reference waveforms. If you do a memory clear this data will be zeroed.
If you desolder the NVRAM then solder it into a socket with turned pins and put a socket with turned pins in the scope's CPU board. Soldering the NVRAM into a socket is essential because the tin on the pins will cause poor contact in both the EPROM reader and socket. I've done this a couple of times in various scopes but it seems I never took a picture.

[chihaxinh]

Hi All

Anyone have NVRAM data TDS520B  could share me ? I dump data but see allmost blank (FF) 

[Armadillo]

Do a memory clear (this may be called secure erase or something like that). Most of the data stored in the NVRAM is reference waveforms. If you do a memory clear this data will be zeroed.
If you desolder the NVRAM then solder it into a socket with turned pins and put a socket with turned pins in the scope's CPU board. Soldering the NVRAM into a socket is essential because the tin on the pins will cause poor contact in both the EPROM reader and socket. I've done this a couple of times in various scopes but it seems I never took a picture.

OMG, there are just reference waveforms.!

Thanks for the information. Regarding turned pins socket, I am afraid they are rather rigid and the outline dimension don't seems to align well with the existing holes. Furthermore like you say the need to solder-in socket, in good practice principles, I am really not in favour of adding extra capacitance to the already slowed speed of the chips by installing IC sockets.
However, for the ease of swopping the chips in future, saving another desoldering and risks lifting the pads, I am considering to use the cheaper flat pins socket instead because the legs are more flexible and will align well with the holes. In your opinion, are there any areas of concern here? thanks.

[Armadillo]

Finally found time to unsolder the NVram and the time-keeper chips;

So the comparison between GPIB read and the GQ4X read and the findings are as follows;

Findings;

1.0 The TDS684C NVram memory address is the same as the TDS520 map
2.0 From 0000 to 1FFFF [131,072 bytes], the content differs between the 2 methods [could be waveform storage space]
3.0 From 20000 - 7FFFF [393,216 bytes[, the contents are completely the same
4.0 The time keeper DS1486 bytes at 33 to 3A [8 bytes] will change between read indicating that the real time clock is still alive. [time register bytes]

Verdict;

IMHO: The GPIB read of the NVram is reliable. I would recommend this method to backup the NVram before any soldering works as the batteries of the NVrams may be so weak that it cannot withstand the stray current of the soldering iron tip. This concern is was also raised in the TEK forum.

Luckily the scope still works and the old Dallas Chips still last a lifetime! [touch wood]

 

[Armadillo]

Part 2

[Armadillo]

Part 3..

[Jwalling]

FWIW, I have a DS1486 that went bad in a TDS700 scope (Read/write data failures on addresses from 0 to 3F) but the batteries (there are two inside the chip!) are still OK. The one closest to pin 1 measures 3.01VDC, the one closest to Pin 16 measures 3.13VDC.
The date code of the chip is the 30th week of 1998. I'm beginning to think these will easily make 30 years or more before they crap out.

[Armadillo]

I'm beginning to think these will easily make 30 years or more before they crap out.

That opens up another area of thoughts. Basing on this formulation, in another couple of years, these oscilloscopes are going to crap without the replacement chip. We know that ebay chip is fake crapped, maybe worse than year 1998. What are we going to do about it?

Any positive ideas how we can prolong the life of these scopes without everyone scrambling for the chip?

[Bushougoma]

IMHO: The GPIB read of the NVram is reliable. I would recommend this method to backup the NVram before any soldering works as the batteries of the NVrams may be so weak that it cannot withstand the stray current of the soldering iron tip. This concern is was also raised in the TEK forum.

Don't trust ANY dump from any Dallas SRAM on a GQ-4X it is very likely your programmer is the issue.

I had an issue where mine wouldn't read a DS-1220Y the dump was just repetitive hex. Reading a parallel SRAM is as simple as it gets and apparently they couldn't even get that right.

So I put it into my TL866 and it read and programmed the replacement properly.

Not wanting to let this go I sent the chip to the programmer manufacturer to investigate and they sent me a new version of the programmer software that did read the chip properly. So to test it I went to program a new Dallas chip I had (same part number) and it again didn't write or read properly I could read and write to the old chip though .

It can't even program 27C series EPROMs properly out of the box. If you do use it for EPROM programming click the "i" button to verify that all of the voltages are correct because the majority of the time they are not.

For example by default it sets the WVCC (Write VCC) on most EPROMs to 5.5 volts when most datasheets clearly state to raise the VCC to 6.25 volts when programming.

Another example many EPROMs have their read VPP "not set or invalid" by default don't know if that means the VPP pin is connected to ground or the pin is floating... never checked it should be connected to VCC when reading of course. This all can be modified in the devices.txt file which is where I spend most of my time getting it to work properly instead of actually using the programmer .

When I do have to use it which isn't that often thank goodness (when dealing with older 21 volt VPP EPROMs for example) I expect to have to tweak something to get it to work.

</RANT>

[ArcticGeek]

I would agree with the problems mentioned with the GQ-4X programmer.  I have had all kinds of issues programming and reading various parts on the GQ-4X programmer, so much so that I refuse to use it any more.  My most recent experience was with an AMD 29LV160 flash part - when reading the part I get different data depending on whether the speed was set to 0, +1, or +2.  The TL866 programmer read it fine every time.  I've also had numerous problems reading Dallas NVRAMs that are used in the Tek TDS600/700 scopes.

The contents of the DS1250 are not critical, as others have mentioned it only stores saved waveforms and saved user settings.  In fact, the scope will boot just fine if you completely remove the DS1250.  The original DS1650 that was used in these scopes is long obsolete, but the DS1250 works fine as a substitute. I've replaced dozens of these parts in various TDS600/700 series scopes.

The DS1486 is much more important - it actually holds the scope options as well as the date & time.  These parts went obsolete in the 2010 time frame as I recall.  You can buy these parts on Ebay, but they are fake.  I know this because some of these parts will have a datecode of 2014 or newer!  I've also verified that the some of the mechanical dimensions of these fake parts are also out of spec, another indication that they are not authentic.  For example, the pin length on the fake China parts are longer than what the Maxim spec indicates.  These parts might work okay, but I question for how long.  I do plan to start a new thread on an analysis of some of these fake parts, but just haven't had the time yet.

[james_s]

I'd be curious to know more about the fakes. I got a DS1250 from ebay because the one in my scope has a bulge in it and it turned out to be a re-mark with clear evidence of the top being sanded and a bigger giveaway is that it arrived with data on it. It may be a genuine chip that was simply relabeled so it may be usable if I mill out the battery and replace it. I got a refund so one can get these fake/relabeled chips for free. I think the MO of many of the sellers is to just refund anyone who complains and bank on the fact that most people won't notice or won't bother. I'm really tempted to buy samples from a bunch of different sellers and xray them along with a real one.

It would be easy to make a DS1250 clone using a SRAM and battery backup management IC but I'm not sure about the ones that include a RTC. I wish the Chinese companies would just offer a compatible IC under their own brand rather than a counterfeit Dallas chip, the Dallas parts are so expensive that they could easily make a quality replacement that is not a fake.

[Armadillo]

Would someone provides the precise micro surgery template reference to left edge and Top Level of the IC, on how to expose the battery terminals for external connection. Photo of this precision surgery operation would be greatly appreciated. So instead of cutting, we can just precisely mill/drill 2 small holes for insertion connection.

I have just paid for a knowingly fake IC from ebay ready to perform this micro surgery operation in anticipation for the future and coming TDS apocalypse.

In probably 5 years time, the ebay will be flooded with TDS oscilloscopes, I imagine.

[james_s]

I'm going to try to xray a batch of stuff over the weekend if I can find the time, in which case this is one of the items on my list. I've done quite a bit of de-potting epoxy potted bricks but something small and delicate like these are a lot easier if you can see what's inside ahead of time.

[Bushougoma]

Would someone provides the precise micro surgery template reference to left edge and Top Level of the IC, on how to expose the battery terminals for external connection. Photo of this precision surgery operation would be greatly appreciated. So instead of cutting, we can just precisely mill/drill 2 small holes for insertion connection.

Just look at the epoxy under a bright light and you usually can see the outline of the cell. At least that was the case with the old Dallas SRAMs I have don't know about the RTCs.

[james_s]

I xrayed a DS1250 last night, still need to scan the films and then I can post the images. The battery is on the end under the Dallas logo laying flat above the board. It sits very close to the top surface, there is maybe 1mm of epoxy over it. Below the battery is a PCB with what looks like a standard surface mount SRAM and below the battery is what I assume is the power management IC.

[Armadillo]

I xrayed a DS1250 last night, still need to scan the films and then I can post the images. The battery is on the end under the Dallas logo laying flat above the board. It sits very close to the top surface, there is maybe 1mm of epoxy over it. Below the battery is a PCB with what looks like a standard surface mount SRAM and below the battery is what I assume is the power management IC.

Thanks James_s, that is a very complete electronic lab you have there.

Is the dS1486 same construction as the ds1250?

The x-ray is really something new. Maybe a descriptive introduction to us?

I only knew Mikeelectricstuff, Applied Science Ben, and now James-s have the machine. who else?

[Armadillo]

I would agree with the problems mentioned with the GQ-4X programmer.  I have had all kinds of issues programming and reading various parts on the GQ-4X programmer, so much so that I refuse to use it any more.  My most recent experience was with an AMD 29LV160 flash part - when reading the part I get different data depending on whether the speed was set to 0, +1, or +2.  The TL866 programmer read it fine every time.  I've also had numerous problems reading Dallas NVRAMs that are used in the Tek TDS600/700 scopes.

The contents of the DS1250 are not critical, as others have mentioned it only stores saved waveforms and saved user settings.  In fact, the scope will boot just fine if you completely remove the DS1250.  The original DS1650 that was used in these scopes is long obsolete, but the DS1250 works fine as a substitute. I've replaced dozens of these parts in various TDS600/700 series scopes.

The DS1486 is much more important - it actually holds the scope options as well as the date & time.  These parts went obsolete in the 2010 time frame as I recall.  You can buy these parts on Ebay, but they are fake.  I know this because some of these parts will have a datecode of 2014 or newer!  I've also verified that the some of the mechanical dimensions of these fake parts are also out of spec, another indication that they are not authentic.  For example, the pin length on the fake China parts are longer than what the Maxim spec indicates.  These parts might work okay, but I question for how long.  I do plan to start a new thread on an analysis of some of these fake parts, but just haven't had the time yet.

You guys are right on, the GQ-4X is a piece of crap, just shit out on me and hanged the software.   
Not going to waste my time on it.

[Bushougoma]

You guys are right on, the GQ-4X is a piece of crap, just shit out on me and hanged the software.   
Not going to waste my time on it.

I hated when I had to come to terms with that fact especially since the base model cost me more than double what the TL866 did. Guess you get what you pay for doesn't always apply.

[james_s]

Thanks James_s, that is a very complete electronic lab you have there.

Is the dS1486 same construction as the ds1250?

The x-ray is really something new. Maybe a descriptive introduction to us?

I only knew Mikeelectricstuff, Applied Science Ben, and now James-s have the machine. who else?

I don't have the machine in my lab, it would be very convenient but it's a rather large piece of equipment and I don't think the powers that be would appreciate me operating such a thing in a residence. I have pretty much full access to it after hours though and can xray various non-living stuff as long as I provide the film.

I can't say about the DS1486 without looking but I suspect it to be the same. If anyone has a dead one to send me I can find out.

[Armadillo]

Thanks James_s, that is a very complete electronic lab you have there.

Is the dS1486 same construction as the ds1250?

The x-ray is really something new. Maybe a descriptive introduction to us?

I only knew Mikeelectricstuff, Applied Science Ben, and now James-s have the machine. who else?

I don't have the machine in my lab, it would be very convenient but it's a rather large piece of equipment and I don't think the powers that be would appreciate me operating such a thing in a residence. I have pretty much full access to it after hours though and can xray various non-living stuff as long as I provide the film.

I can't say about the DS1486 without looking but I suspect it to be the same. If anyone has a dead one to send me I can find out.

Yeap!, it would be extraordinary to include a X-Ray machine in an electronic lab. Cause even the signal path Shahriar with his close to $million dollars equipment don't have one, not even a wire bonding machine.  I think the room need to be lined with lead to comply with country code. Ben is that scientist class, exception.
 

[KE5FX]

Yeap!, it would be extraordinary to include a X-Ray machine in an electronic lab. Cause even the signal path Shahriar with his close to $million dollars equipment don't have one, not even a wire bonding machine.  I think the room need to be lined with lead to comply with country code. Ben is that scientist class, exception.

Y'all need to check out Aurora's thread.  It hooked me.  Low-energy cabinet machines are perfectly safe and not all that expensive.  The trick is finding one that's being sold with its digital imaging hardware.

[james_s]

Ok here are the images. My scanner is terrible at scanning film but I was able to achieve reasonable results. This is a DS1250 and also a DS1386 I had on hand, I suspect the internal layout is very similar if not identical to the DS1486. The top down shots are taken at two different kVp settings, the higher the kVp, the more penetrating but the less contrast you get with lower density materials. You can clearly see the pads marked + and - where the battery attaches near the middle of the PCB.

[Armadillo]

Are we looking at these 2 pins? Is that 2 batteries, one piggyback onto each other.? Is this a fake or a genuine? thanks.

[james_s]

I think the - pin is the one below the + pin, the one in the upper-left I'm not certain, it may be for mechanical support. I plan to do some careful milling to remove the battery at which point I'll know exactly where the wires are. I think these pictures provide enough to do some reasonably educated surgery.

The DS1386 I know is genuine, the DS1250 has had the top sanded and re-marked but I believe it too is a genuine part based on the image. I think it's just an old pull that they re-marked and sold as a new one.

I just realized one top-down shot of each IC is flipped vertically, that happened when I scanned it. I believe the high-kvp shot is correct, looking down from the top with the battery under the Dallas logo based on the ground pin being in the lower-right but now I don't remember for sure. I really need to remember to keep track of this.

[Jwalling]

This is a DS1250 and also a DS1386 I had on hand, I suspect the internal layout is very similar if not identical to the DS1486.

A while back, I posted a picture of a DS1486 that I cut open. I can't seem to find it for some reason, so here's another picture.
This chip has TWO batteries, so it can't be like the chips you x-rayed.

[Armadillo]

Looks like a very hard material. The second battery is buried deep, will be hard to reach. 

[james_s]

Many epoxies will soften if heat is applied, so it may not be all that difficult. I have repaired quite a few potted HeNe laser power supplies, and even completely de-potted a couple of them. Just hit it with a heat gun and then dig the epoxy out before it softens. These Dallas chips are in a plastic housing that will stink so probaly best to cut the top off that first.

[Jwalling]

Looks like a very hard material. The second battery is buried deep, will be hard to reach. 

It is quite hard, as it's epoxy. Since I did that surgery with an X-acto knife, I've invested in a Dremel. 

[Mosaic]

Hi,
Would a TDS694C have the same NVRAM to replace?

[Jwalling]

Hi,
Would a TDS694C have the same NVRAM to replace?

Yes, it has the DS1486 as well.

[Mosaic]

Ok, My 694C was calibrated and is still in cal so I'd like to preserve the NVRAM.

I have a Willem PCB50 parallel port programmer, runs on XP, can anyone advise which settings to use for the DS1486 and DS1250Y?

http://www.mcumall.com/support/DualPoweredWillemUserGuide.htm#Supported%20Device%20List

[Armadillo]

So the chips DS1486 and DS1250Y finally came. It looks black like charcoal, every bits of it looks fake.

Anyway, I read the DS1486 using my TL866 set to DS1245Y but all read back as "FF". The time clock 14 bytes never change upon reread. Can anyone advise if this is normal or the fake needs to be rejected.

The DS1250Y all read as "00".

Thanks.

[james_s]

The DS1644 I read has the time bytes change every second so they're different with each read, I would expect the DS1486 to be the same. The DS1250 should all read FF IIRC, the one I bought came with data on it. Yours are probably re-marked pulls, if they were sold as new you should ask for a refund, you'll probably get it and be able to keep the chips. They may well be genuine internally but they're not new.

[Mosaic]

I ordered a pair from here:
http://www.ebay.com/usr/hkutsource?_trksid=p2057872.m2749.l2754

We'll see what they are when they arrive.
I hope the Willem PCB50 parallel port programmer can do them. I did a DS1225Y for a TEK2465B scope with it before.

EDIT: BTW since this exercise requires going inside the scope....have a look at this (my project) concerning the acq. chips cooling.
https://hackaday.io/project/12087-tektronix-tds-694c-trigger-cooling-workaround

[Armadillo]

I ordered a pair from here:
http://www.ebay.com/usr/hkutsource?_trksid=p2057872.m2749.l2754

We'll see what they are when they arrive.
I hope the Willem PCB50 parallel port programmer can do them. I did a DS1225Y for a TEK2465B scope with it before.

EDIT: BTW since this exercise requires going inside the scope....have a look at this (my project) concerning the acq. chips cooling.
https://hackaday.io/project/12087-tektronix-tds-694c-trigger-cooling-workaround

Same from UTsource.... :-(
Fake and defective.... what a nonsense.

[Armadillo]

The DS1644 I read has the time bytes change every second so they're different with each read, I would expect the DS1486 to be the same. The DS1250 should all read FF IIRC, the one I bought came with data on it. Yours are probably re-marked pulls, if they were sold as new you should ask for a refund, you'll probably get it and be able to keep the chips. They may well be genuine internally but they're not new.

ebay is directing me to paypal and paypal is directing me back to ebay.... both of them playing the Chinese taichi, pushing the problem around... 

[Mosaic]

What about  using an external Tadiran batt with a schottky diode to maintain the NVRAM Vcc at  3.6V so the 3.0V internal batt never comes into play?

[james_s]

That ought to work for the NVRAM-only variety, those with the RTC are not going to be as simple although it's certainly possible to construct something if you have the time and motivation.

I think the DS1644's I ordered are coming from utsource, I've gotten some good stuff from them before but I fully expect these to be re-marked pulls. Hopefully I can mill into them and replace the battery, I just want to have a good spare rather than risk damaging the original ones that are in the scopes. If they're not genuine I'll ask for a refund, and if they don't send me a refund I'll file a claim. Ebay's protection is heavily slanted in favor of the buyer, something I'm not real fond of but it's useful in cases like this.

If they just sold the things as used pulls for the same prices I'd still buy them, I don't know why they bother faking them. I'm also not sure why no Chinese company seems to have come up with a line of self-branded compatible devices, the technology within these Dallas chips is all commodity stuff, good quality replacements could be made for a small fraction the price of a new one. If I didn't have so many projects in my queue already I'd do it myself.

[Mosaic]

So the chips DS1486 and DS1250Y finally came. It looks black like charcoal, every bits of it looks fake.

Anyway, I read the DS1486 using my TL866 set to DS1245Y but all read back as "FF". The time clock 14 bytes never change upon reread. Can anyone advise if this is normal or the fake needs to be rejected.

The DS1250Y all read as "00".

Thanks.

It could be the DS1486 is genuine and has not been initialized to start the clock.
have a look at page 5 here:
http://www.dataman.com/media/datasheet/Dallas/DS1486_DS1486P.pdf

[james_s]

If you're lucky, it's a genuine part that has been re-marked with a current date code.

[Mosaic]

What is the date code on the parts?

EDIT:
I see the DS1250Y is still active, in production:
https://www.maximintegrated.com/en/products/digital/memory-products/DS1250Y.html

The suggested replacement for the DS1486 is the DS1556:
https://www.maximintegrated.com/en/products/digital/real-time-clocks/DS1486.html

[Jwalling]

What is the date code on the parts?

EDIT:
I see the DS1250Y is still active, in production:
https://www.maximintegrated.com/en/products/digital/memory-products/DS1250Y.html

The suggested replacement for the DS1486 is the DS1556:
https://www.maximintegrated.com/en/products/digital/real-time-clocks/DS1486.html

The DS1586 is not a drop-in replacement. The clock registers for the DS1586 are at the top of memory space while the DS1486 registers are at the bottom. I hate Maxim with a passion.

[Armadillo]

Question: What if instead of the DS1486, we plug in DS1245Y the memory non-volatile static ram instead without the clock, date, watchdog etc.... will the scope still stay to function. Anyone tried that?
Well it is definitely much cheaper than the DS1486 which is destined to be extinct.

[Armadillo]

Sorry Guys, I checked my DS1486 backup from the original scope IC and noticed that the first 50 bytes are actually "FF". The bytes that will varies automatically actually start at byte 51.
Now how do you tell TL866 to start programming from bytes 51 onwards?

Edit: So, that's the reason why TL866 does not have a device named DS1486.

[Mosaic]

What is the date code on the parts?

EDIT:
I see the DS1250Y is still active, in production:
https://www.maximintegrated.com/en/products/digital/memory-products/DS1250Y.html

The suggested replacement for the DS1486 is the DS1556:
https://www.maximintegrated.com/en/products/digital/real-time-clocks/DS1486.html

The DS1586 is not a drop-in replacement. The clock registers for the DS1586 are at the top of memory space while the DS1486 registers are at the bottom. I hate Maxim with a passion.

No, that is not so clock data are at the top n both cases. The DS1556 has an extra 2 bytes of data used for the clock, for the full century, I think= Y2K ready.
https://www.maximintegrated.com/en/products/digital/real-time-clocks/DS1486.html
https://www.maximintegrated.com/en/products/digital/real-time-clocks/DS1556.html/tb_tab0

[Armadillo]

Jwalling is correct, not direct drop in replacement.

[Mosaic]

Can u be more specific as to why it isn't a drop in replacement and if the fix is doable?"

[Armadillo]

Can u be more specific as to why it isn't a drop in replacement and if the fix is doable?"

Just compare the address spaces of the two [page 8]

[deepskyridge]

Sorry Guys, I checked my DS1486 backup from the original scope IC and noticed that the first 50 bytes are actually "FF". The bytes that will varies automatically actually start at byte 51.
Now how do you tell TL866 to start programming from bytes 51 onwards?

Edit: So, that's the reason why TL866 does not have a device named DS1486.

Did anyone figure out if you could use a DS1486 in a TL866A. I have a TDS744A i need to backup.

Thanks
Gary

[Polarue]

I have looked at the datasheet for the ds1486-120 nvram in my TDS640A and it occurred to me that it might just be possible to add a common rechargable lithion ion battery to the Vcc of this chip in order to maintain the voltage of the Vcc to about 3.5v, when the main power of the scope is off.  When the scope is powered, a simple blocking diode from the Vcc to the new battery would regulate the voltage down from the nominal Vcc of 5.0v to 4.0v by placing two diodes in series and then one additional diode returning to the same Vcc leg of the chip(32) to yield about 3.3-3.5v when the scope is powered down but safely charging while powered up.

The above scenario is made possible as the integrated LTC cell in the chip only switches on if the Vcc drops to about 3.0v. 

Forget the whole "replacing the chip idea" and just bypass the LTC cell entirely. That battery will never again be relied upon if the Vcc never drops to 3.0v.

The new external cell will not allow the voltage to drop if the unit is switched on  a few times a year to top it up with charge and it is cheap and easy to replace every 5 years.

Am I missing something here like too much current drain on the board if the Vcc is powered by the lithium external when the scope is off.  If so, a small circuit could remedy this yes?  Is this a great idea for these TDS scopes or not?

[james_s]

That's not going to work, the other parts on the board will draw a lot more power than you think and quickly drain the battery. Also those Dallas chips shut down and switch over to internal battery when Vcc drops below something like 4.5V.

[Polarue]

Are you guessing, James?  Where did you get that info. regarding the chip?  See below.

"As VCC falls below approximately 3.0V, a power switching circuit turns the internal lithium energy source on to maintain the
clock and timer data functionality. It is also required to ensure that during this time (battery-backup
mode), that the voltage present at INTA and INTB (INTB) never exceeds VBAT. During power-up, when
VCC rises above VBAT, the power-switching circuit connects external VCC and disconnects the internal
lithium energy source. Normal operation can resume after VCC exceeds 4.5V for a period of 200ms."

The above is from the datasheet for the ds1486-120, so maybe you misdirected yourself, unless I misunderstood the above?

Secondly, have you measured the current draw as the battery can be much larger than the one in the chip! 

Third, could not a circuit be built to block the current to the other parts of the board with a transistor driven and triggered by the board's own 5v (tying the emitter and base of npn together) and then it would block the 3.5v from the external battery to the rest of the board when the unit is off? Of course this would mean lifting Vcc pin. 

If the details and proper component selection are worked out, this would be a heck of a lot easier than replacing the chip and further, the chips are costing more than I paid for my scope these days.

[james_s]

I was thinking of the related DS1220 datasheet, but I realized looking at it now that the voltage threshold I was thinking of was for the write protect, not for the backup battery.

"The DS1220AB provides full functional capability for V CC greater than 4.75 volts and write protects by
4.5V. The DS1220AD provides full functional capability for V CC greater than 4.5 volts and write protects
by 4.25V."


Yes you could hack up the board to isolate the Vcc pin to the Dallas chip but why would you want to? It's easy to cut the battery out of a Dallas chip and attach an external 3V battery. That then makes use of the built in changeover circuit and requires no modification to the original circuit. On top of that it's easier to modify the Dallas chip than the PCB in the scope. It's a superior solution all around.

[Polarue]

Easy?  I have not read that anyone has actually succeeded in doing what you describe to the dallas chip.  I have seen the xrays and also a pic of a partial removal of some epoxy etc, but never a completed mod while the chip is on the board.  There is risk, if you remove the chip from heat, although some have done it. High probability of damaging the data.

From my experience, there is great risk in carving up the working chip in the scope.  I would rather make up a circuit with battery and add it with minimal disturbance to the board and the chip.  Just lifting the pin 32 and connecting the  battery circuit to the pin and its pad seems way less risky, along with the fact that no programming is required either.  This way the chip is not at risk to the same extent, YMMV.

[james_s]

Not in a scope, but I have replaced the battery in a Dallas chip before and it was quite easy, in my case it was one that was used on a vintage PC motherboard.

IIRC the Dallas chips in the TDS600 series don't store the calibration data anyway, just the options, and you can enable any of those you need via GPIB. The TDS300 series is another story.

[Jwalling]

Not in a scope, but I have replaced the battery in a Dallas chip before and it was quite easy, in my case it was one that was used on a vintage PC motherboard.

IIRC the Dallas chips in the TDS600 series don't store the calibration data anyway, just the options, and you can enable any of those you need via GPIB. The TDS300 series is another story.

SPC constants are stored in the DS1486 as well.

[Polarue]

Did you do it while it was still on the board?  Do you have any pictures to share? Exactly what tools did you use so as to not damage the rest of the chip etc.  This would be most appreciated. I do not want to reinvent the wheel.   

[rodd]

Sorry Guys, I checked my DS1486 backup from the original scope IC and noticed that the first 50 bytes are actually "FF". The bytes that will varies automatically actually start at byte 51.
Now how do you tell TL866 to start programming from bytes 51 onwards?

Edit: So, that's the reason why TL866 does not have a device named DS1486.

Dear Armadillo,
I desoldered a DS1486 from a TDS784D scope and read its contents with a TL866A.
It seems I have not lost all data, but I can't say anything about the integrity of it.
However, when I try to write the data to other DS1486 (known to be fake), the programmer writes one line of code and one line of zeroes.
Have you managed to find a way of using TL866A to write data back to DS1486?
Thank you

[madao]

I can't recommend  reading with TL866A.
Have you zero byte  from address 0 up to ~200? This is wrong !

I have always problem with reading of DS1486 and TL866a.  I have  a idea,  VCC lifting...but i use rather tektool.

(no success with  tektool? GPIB communication ok? )

Greetings
matt

[james_s]

I ended up making an adapter board for an Arduino mega to work with Dallas chips. I hacked together some firmware to get in and out of test mode, set the clock and a few other functions but I never got around to getting it usable enough to post. I can share the PCB layout in case someone else wants to take the project and develop it further.

[whatboy]

I have 3 TDS684 A/B/C, wanted to replace the DALLAS chips, bought them on ebay, and have a programmer, a LAB TOOL-48UXP, the original ones, can be read, and verify as good read, but with the Timer chip (DS1486) besides the first 4 bytes, there seem to be 8 to 16 bytes which are wrong data, don't know if the programmer is the bad one, or the chips... but with the DS1250Y mostly all data is wrong when write to the ebay chips, but it can read and verify correctly!, again, do not know if the programmer is the bad one or the chips, thinking on buying a TL866II programmer, see if a 60 dollar programmer can beat a 1200 dollar programmer! bummer!!!

[CaptDon]

There is a guy on Ebay selling a redesigned replacement set of
chips for around $100 for the pair and he can preload data for you
including options you may not have had. I bought a set for my
TDS644A and TDS644B. He even supplies the machined pin sockets
for you. 100% recommended!!! Download the data using the
nvram tool on eev blog!!! My chips lost their data when unsoldering
but I already had a good copy of the data!! B.T.W., if you look at
the hex data dump there is something in ascii that reads "invalid"
blah blah something or other.... It is normal, don't panic!!

[Per Hansson]

I ended up making an adapter board for an Arduino mega to work with Dallas chips. I hacked together some firmware to get in and out of test mode, set the clock and a few other functions but I never got around to getting it usable enough to post. I can share the PCB layout in case someone else wants to take the project and develop it further.

That would be nice if you could share, I guess it is for the DS1486?
There exists this one I have built for BQ3287 which is direct cross compatible with Dallas DS12887 quite common in old PC's.
I modified the code so it can also output HEX dumps, useful for checking things, if anyone wants a copy let me know.
https://fritzing.org/projects/ibm-rtc-on-arduino/

Finally found time to unsolder the NVram and the time-keeper chips;
So the comparison between GPIB read and the Mcumall GQ-4X read and the findings are as follows;
4.0 The time keeper DS1486 bytes at 33 to 3A [8 bytes] will change between read indicating that the real time clock is still alive. [time register bytes]

This was useful info, I found on the Tek forum that this can be done by selecting DS1245 when reading a DS1486 as they are pin compatible, did not know that before.
https://forum.tek.com/viewtopic.php?t=138969
I will play around and see if that would allow even a DS1286 to be read, but with device type DS1230Y I think.