NVRAM of oscilloscope TDSxxx serie C and D

[Tantratron]

The topic seems to have been discussed on different threads through time but here is my synthesis along with suggestion of way forward.

Unless wrong the design of these legacy TDSxxx serie C and serie D are quite solid, resilient and these oscilloscopes are still quite worth to be used for private or professional. Make a note that I'm radio electronics hobbiest for decades as well as running a single man consulting firm specialized on special RF testing like plasma lamp in France. One thread which attempts to discuss the quality of these scopes was initiated here https://www.eevblog.com/forum/repair/tds-500-and-tds700-series-a-b-c-or-d-what-subsystem-are-common/

The key problem with these TDSxxx concerns both NVRAM's which have internal batteries potted, guarantied 10 years, most of these scopes sold circa mid-90's. The good news in 2020, most of the scopes still work with these NVRAM's which means the DALLAS design has been super good (I have two TDS540C where one is now repaired, the issue was EEPROMs failure). However at some point, the batteries will fail soon after more than 25 years hence the scope will brick or never work again plus DALLAS chips are obsolete.

Some individuals or companies mainly in USA have tried to solve the issue by designing, manufacturing NVRAMs replacement on eBay https://www.eevblog.com/forum/testgear/i-noticed-nvram-replacements-for-the-tds784d-on-ebay/ and lately by member cuebus https://www.eevblog.com/forum/repair/dallas-ds1486-and-ds1250-replacement-designs/

One member of EEVblog who might post public gave me some critics on US replacement kit. This member was supposed initially to build for me under payment two units designed by cuebus but later cancelled arguing different technological issues. No idea if this correct or not but i'm wondering if it would not be good to have say a replacement solution in Europe (my company is based in France).

Do you think there is a market or a need of these NVRAMs replacement kit, what price and could we have commun order to reduce cost. Of course there is legal topic to know if the gerber file or the new design using non-obsolete components is doable on massive scale.

I'd appreciate frank and straight forward analysis on if the effort is worth or better to go after more recent high end oscilloscopes.

On a side note, how does the oscilloscope industry has solved say the programmed obsolescence or use of NVRAM's with potted batteries then no NVRAM replacement... what are the modern design of oscilloscope to avoid the known issue of these legacy TDSxxx.

Thank you, Albert

[madao]

1.) Problem:  Where can i buy  CR1632 battery with  2 pin terminal with shipping to germany(or EU ) ,  not for usury-price.  (Battery with usury-price is often to much old, long storage time)
   Mouser, Digikey, etc doesn't shipping Lithium-Batttery outside USA (or to EU ).    I can't sell  Kit with thinkering with other battery type.  I don't recomming: soldering on battery, lifetime would be much worst than before.  I have solder kit with CR123a, it would live longer than me. But for only me and friends. 
Did you know why, i don't want  sell  kit with thinkering of battery-type/soldering on battery.
Reason: I didn't feel like, owner cries problem in few years or about higher height and i lost again my time.


2.) DS1384G is obsolent, of coruse, you haven't a other choose.  And i have one DOA.  It is funny 12€-DOA , is not ?

Most cheap (& safe) solution: Crack  DS1486 and  replace battery (i have much  CR2430 with legs) . I would doing this way in future, because i can avoid stress with replica.

Now, i know:  why want seller 100$ for one set. Honestly,   replica is to much stress for me. (DOA,  battery, SRAM )

Greetings
matt

[Tantratron]

1.) Problem:  Where can i buy  CR1632 battery with  2 pin terminal with shipping to germany(or EU ) ,  not for usury-price.  (Battery with usury-price is often to much old, long storage time)
   Mouser, Digikey, etc doesn't shipping Lithium-Batttery outside USA (or to EU ).    I can't sell  Kit with thinkering with other battery type.  I don't recomming: soldering on battery, lifetime would be much worst than before.  I have solder kit with CR123a, it would live longer than me. But for only me and friends. 
Did you know why, i don't want  sell  kit with thinkering of battery-type/soldering on battery.
Reason: I didn't feel like, owner cries problem in few years or about higher height and i lost again my time.


2.) DS1384G is obsolent, of coruse, you haven't a other choose.  And i have one DOA.  It is funny 12€-DOA , is not ?

Most cheap (& safe) solution: Crack  DS1486 and  replace battery (i have much  CR2430 with legs) . I would doing this way in future, because i can avoid stress with replica.

Now, i know:  why want seller 100$ for one set. Honestly,   replica is to much stress for me. (DOA,  battery, SRAM )

Greetings
matt

You must be joking... where can you NOT buy 2-pin CR1632 in Germany, in Europe, in France where for example the RENATA battery reference from Cuebus BOM is available in France https://www.conrad.fr/p/pile-bouton-cr-1632-lithium-renata-125-mah-3-v-1-pcs-1009401 or https://www.conrad.com/p/renata-cr1632fh1-lf-button-cell-cr1632-lithium-125-mah-3-v-1-pcs-1009401

RENATA is Swiss company so maybe DIGIKEY or MOUSER decides to not sell these batteries to Europe (why not idea) but hey... just buy in Europe instead of criticizing the great NVRAM design of Cuebus https://www.eevblog.com/forum/repair/dallas-ds1486-and-ds1250-replacement-designs/

Now you say one DS1384G to be DOA (Dead On Arrival)... how about a soldering issue, these things happen sometimes

[madao]

i and joke?  Thanks.

I thought, i have explain (over mail and here) enough , why i dont wan't buying batteries with usury-price. 2.5€ is usury.

DOA, no solder issues , high battery consumption (~0.5mA ),  you can put car battery on this modul.
You have gerber file and  you did found  shop with battery. you can solder kit.   

[Tantratron]

i and joke?  Thanks.

I thought, i have explain (over mail and here) enough , why i dont wan't buying batteries with usury-price. 2.5€ is usury.

DOA, no solder issues , high battery consumption (~0.5mA ),  you can put car battery on this modul.
You have gerber file and  you did found  shop with battery. you can solder kit.   

Hello @madao,

To be honest, I still do not understand your critics against the solution of @cuebus or the eBay replica kit done by KellyJo plus I notice lots of contradiction with different mails you have written since last december.

In my opinion, there are three fundamental topics when going after a NVSRAM replica kit: the battery quality, the consumption of the chipset when the TDS is turned off and the obsolescence of some chip.

Regarding the eBay kit of kellyjo, you wrote last december yes,  DS1486-replica, i know it, it works by friends good, but battery is a bit small.. Indeed both DS replica only have one battery as shown on his eBay item sell but you confirm it works.

Later Cuebus came with a new design, this time two batteries for the DS1486 but you wrote last july I can't buy CR1632 battery with solder joint (2 pin horizontal) in europe. see "shipping restricted" on digikey or mouser, because  many parcel service see lithium same as  weapon-plutonium. sad, welcome to overregulation-world

I just made a kind post few days ago mentioning that we can buy two-pins CR1632 in Europe but now you say cuebus circuit is draining 0.5mA and would need a car battery !!!

Frankly I do not follow your logic here unless you can explain exactly what chip consumes 0.5mA from the BOM attached, how did you check this issue in your lab. What is certain, you argue an issue about CR1632 and clearly this argument is incorrect. Of course one need to choose quality CR1632 but this should not kill or criticize amazing work done by kellyJo and Cuebus.

Hello All,

Another thing I've been thinking but no idea if anybody has done or inquired such approach. How about using a small processor but enough CPU power along with flash memory (no ore battery needed) to emulate the work of DS1486 and the DS1650 kit ?

Last question, at least my two TDS540C with time code 1997 (see attached pictures) still work very good, it seems other TDSxxx serie C and serie D made in the 90's still work with no DALLAS battery failure. Do you know what is so special with the legacy DALLAS design and internal potted batteries to be able running after more than 25 years whereas it was warrantied 10 years ?

Albert

[TheMG]

On a side note, how does the oscilloscope industry has solved say the programmed obsolescence or use of NVRAM's with potted batteries then no NVRAM replacement... what are the modern design of oscilloscope to avoid the known issue of these legacy TDSxxx.

These days, pretty much anything that requires non-volatile memory uses flash EEPROM, which has gotten much cheaper and more durable, that's what has pretty much made the Dallas modules obsolete.

For RTC purposes, RTC ICs with external coin cell batteries, or supercapacitor backup.

[cuebus]

Given that the batteries have about 130mAh capacity, and since I'm not sure which one is being referred to, 0.5mA drain would be about 260 hours for the DS1250 and 520 hours for the DS1486. That's about 11 and 22 days respectively. I've had these in my scope now for 8 months with no voltage loss so clearly the design isn't an issue...
I think it's important to get RAM with a low standby power. From memory the spec to be looking for is <1uA. A lot of datasheets will quote a min/max standby current draw and the max can sometimes be quite high, but it's usually because the datasheet is covering the entire usable temp range. Look for a graph that shows standby current draw at 25C. Low power models generally fit that 1uA value. The original module used a Sony low power chip that I can't remember the model off of the top of my head but it too had various current draws depending on temperature for the standby power.

Regarding battery shelf life - it's not likely to be an issue- some panasonic batteries (BR1632) are rated for 10+ years with virtually no capacity loss. In fact, the BR1632 battery chemistry may be the better choice here because of its lower self-drain, even though it has 10mAh less capacity. I still wouldn't imagine a CR1632 that's 4-5 years old say, would have any issues, as long as it's coming from a good proper battery manufacturer like Renata or Panasonic.
My only issue so far with fulfillment has been Mouser's packing of the cells - since the tabs sticking out make it very easy to short them out, Mouser should really do a better job than just placing them in some styrofoam- I've had some arrive with leads shorted against an adjacent cell/bent/etc.
The Panasonic cells with shrink wrapping around them are probably a better solution as well just for that issue.

[Tantratron]

On a side note, how does the oscilloscope industry has solved say the programmed obsolescence or use of NVRAM's with potted batteries then no NVRAM replacement... what are the modern design of oscilloscope to avoid the known issue of these legacy TDSxxx.

These days, pretty much anything that requires non-volatile memory uses flash EEPROM, which has gotten much cheaper and more durable, that's what has pretty much made the Dallas modules obsolete.

For RTC purposes, RTC ICs with external coin cell batteries, or supercapacitor backup.

Do you know after what year or from what model tektronix decided to use flash memory and external battery to avoid the DALLAS or internal potted battery memory critical storage on their oscilloscopes ?

[Tantratron]

What would be the PCB electrical test point one could DMM check the voltage remaining value or health status of both NVRAM's internal bateries say on the TDS540C main board ?

[andy2000]

What would be the PCB electrical test point one could DMM check the voltage remaining value or health status of both NVRAM's internal bateries say on the TDS540C main board ?

To avoid drain, there are no external connections to the battery.

[Tantratron]

What would be the PCB electrical test point one could DMM check the voltage remaining value or health status of both NVRAM's internal bateries say on the TDS540C main board ?

To avoid drain, there are no external connections to the battery.

The reason i've asked was to understand how the eBay seller here https://www.ebay.fr/itm/GENUINE-Dallas-DS1486-DS1250Y-NVRAM-pair-for-Tektronix-TDS-600-700-series/303131785473? DMM checked or confirms the actual voltage remaining in one of the two DALLAS chipset.

I'm copy-pasting his description from eBay below

For sale in this auction are a pair of GENUINE Dallas Semicondutor:

- 1 pc. DS1486-120 RTC/NVRAM unused, factory new, date code 2006, factory disconnected battery.
- 1 pc. DS1250Y-70IND NVRAM, used but in good condition and battery level above or still close to 3V, date code 2005 or more recent. Items are not desoldered or damaged, as they have been fitted into sockets.

They are suited as a replacement NVRAM for many TDS600/700 scopes. This is probably your very last chance to grab a genuine and brand new DS1486 and save your precious instrument or device for the next decade to come.

This auction is pretty unique as long as it will last, as especially the DS1486 is no longer being manufactured and all stocks have been depleted, worldwide. As you may know already, all "new" DS1486 sold on Ebay are Asian counterfeits, and, aside from the fact counterfeiting is illegal, they have their fair set of compatibility and quality issues as described in many forums. You can easily recognize them at the cheap thermoplastic dull-edged package and the different looking printed label with repeated date code. The components offered here have known distributor origins and have in addition been inspected under microscope and compared to reference genuine parts.

[TheMG]

That's an excellent question. As far as I know, it is not possible to measure the internal battery voltage without some slightly invasive methods, such a drilling a tiny hole in the right location to allow a multimeter probe to access the battery connection.

Maybe the seller just tested them and assumed if the NVRAM holds its memory, that the battery must be good?

Regardless, measuring the voltage of those lithium cells doesn't give you a good indication of remaining life. Under very light loads they tend to stay pretty close to 3V until nearly depleted and then the voltage rapidly decreases.

But if they are indeed genuine and manufactured in 2005 they most likely have quite a bit of life remaining. I've got a couple with date codes in 1990 (30 years ago!) that still hold their memory just fine.

[cuebus]

As andy2000 said, there is no way to tell the battery condition without hacking apart the module. There are no connections of the batteries to any of the outside pins- all battery + pins go to the controller chip.
I can only surmise that this seller is guessing that the cell voltage is at 3V but there really is no way to check, aside from drilling a small hole down to the battery and putting a probe in and measuring between the top of the battery and the ground pin. A risky move though, considering how thin 1632 cells are.