| Products > Test Equipment |
| I noticed nvram replacements for the TDS784D on ebay |
| << < (10/11) > >> |
| Tantratron:
--- Quote from: james_s on September 23, 2020, 04:00:52 am ---My recollection is that one of the two is necessary for the scope to boot, the other can be removed and it will still boot up. In both cases if you replace the missing part normal operation will be restored. Symptoms I've seen when similar Dallas parts get low batteries are the RTC stops advancing while the power is off and in the ones that are just NVRAM they will start randomly getting corrupted. --- End quote --- Since I have now a few TDSxxx/C/D where I did install 32 pins sockets for ease to install/remove the DS1486, the DS1250Y and replica, indeed the TDS can correctly boot and run with only DS1486. So wether there is a DS1250Y present or removed, the boot will only check presence of DS1486 to complete the initialization. I did try to actually install one DS1250Y inside the DS1486 socket then program via GPIB-USB a copy of the DS1486, we loose the RTC (clock feature) but the TDS boots then runs very OK with all its other functions. So unless someone needs to save waveforms, otherwise the DS1250Y is not really useful for the oscilloscope to function good. It all comes down then to find a replacement kit for the DS1486 to offer the time/day/year or just put instead a DS1245Y as it was suggested by @ArticGeek which can be purchased. |
| ArcticGeek:
I have been lurking and reading this forum but haven't been posting - mostly due to the fact that I don't spend that much time with my hobby in the summer - too busy with yard work and barbecue! I have built the DS1486 boards that cuebus designed and have had decent luck with them. You can find that post here https://www.eevblog.com/forum/repair/dallas-ds1486-and-ds1250-replacement-designs/msg2954864/#msg2954864 I have built a few failing Cuebus units due to my lack of soldering skills with a TSSOP package and fine pitch parts. I have also designed a small board that adapted a DS1486P part to a 32-pin DIP package. The DS1486P is the same part that is in the 34 pin PCM package. The DS1486P is still somewhat available, I've found some in China and I also found some in Canada. There are a few pros/cons of that approach: 1. The 34pin PCM package is a lot easier to solder than TSSOP packages. If you're skilled at soldering this isn't a big deal, but for me it's much easier. 2. The SRAM parts used inside the DS1486P are typically much lower power than anything you can buy. I've measured the power of both the Cuebus design and the purchased DS1486P power. The Cuebus design can vary wildly depending on which SRAM you buy even though specs are identical. For example, using an ISSI low power SRAM I measured the battery current to be 2.71uA on 1 unit and 2.47uA on another unit. This equates to a battery life of somewhere around 11 years if using qty 2 of the 120ma-hr lithium battery. Replacing that SRAM with a Renesas RL1P part reduced the battery current to 784nA and 735nA respectively - improving the battery life 3x to 35 years plus. Typical battery current of several DS1486P parts was 297nA, 382na, and 378nA. All of these measurements with taken with register 0x09=0x03 which means the oscillator is ON. (Running with oscillator off is MUCH lower, around 20nA). This works out to be about the same battery life, but you need to be careful which part you buy, and be VERY careful with cleanliness after you solder. 3. The DS1486P battery can be replaced easily without soldering. 4. The Cuebus build-your-own approach is a bit cheaper, depending on how much you pay for the DS1486P part. I can also confirm that the TDS series of scopes boot fine without the DS1250, its the DS1486 timekeeper that is critical for boot operation. If I can find the time I might post my design for the DS1486P adapter, but right now I've been busy at work and summer activities seem to consume a lot of my time. |
| Tantratron:
--- Quote from: ArcticGeek on November 20, 2019, 03:22:38 pm ---I actually purchase one of these just because I was very curious how he designed this. I only purchased the DS1486 replacement, and he was willing to sell that to me for $70. This is still very expensive, but I had to know what he did. The central part of the design consists of a DS1384 RTC controller; this part is identical to the DS1486 without an SRAM. The DS1384 sits underneath the battery, and an SRAM is on the bottom side of the board. Once you add the SRAM to the DS1384, you have something that is identical to a DS1486. I suspect this is exactly what is inside the encapsulated DS1486 - a DS1384, an SRAM, and a couple of coincell batteries. The DS1384 actually has inputs for TWO backup batteries if desired. I must admit it's a fairly clever design, although the layout seems less than optimal. The bad part of the design is that the DS1384 is obsolete just like the DS1486, so if you buy any of those parts you will have to buy off of Alibaba, Ebay, or a broker somewhere so the quality is questionable. However, the DS1384 does not have an internal battery like a DS1486 does so it is arguably less risky to buy these parts. But I worry that the mileage on these parts may vary wildly. --- End quote --- Regarding the eBay sold design of the DS1250Y, does anybody knows what Nonvolatile Controller with Lithium Battery Monitor chip is used ? When you look on some of his add on eBay with remote batteries, we see it is 8 pins chip (DS1250Y) and confirm 44 pins (DS1384) instead of DS1384G (48 pins) for the DS1486 board (see attached pictures). On a side note, does anybody knows if the DS1384G has two separate inputs (two batteries) or only one battery since it is used on DS1486P PowerCap hence one battery ? |
| cuebus:
A lot of good info here ArcticGeek. I looked at the DS1486P as well but I think at the time I designed this I couldn't actually find vendors selling them. I'm sure supply comes and goes for these kinds of surplus/obsolete parts. As you say, the SRAM selected by Dallas/Maxim is very low current draw. I imagine they had the resources to either purchase ultra-low current devices or select them from a standard batch. I tested the ISSI part you mention and also got 2.5-3uA draw. This part is labeled low power however it is anything but. If you use cypress, alliance, or the renesas memory you mentioned, you will get .7-1uA current draw with RTC on. With 2 batteries that equates to a lifetime of about 19.5 years worst case. There's been both successes and failures of people building these and I must admit that the reason for these failures seems to be complex and varied in nature- including whether a device is programmed on an external programmer or within the scope itself after initialization. I'm no longer pursuing any troubleshooting on this project so I haven't been able to figure out any exact causes but Tantratron has 3 of these (that I programmed externally and then tested in my own TDS540C) and has had trouble programming them in circuit in his scope via tekfwtool. He's been doing some low level debugging so I suggest anyone who is having issues reach out to him for more information. FWIW- my own replica I built last february is still going strong. |
| Tantratron:
--- Quote from: cuebus on July 23, 2021, 04:49:29 pm ---There's been both successes and failures of people building these and I must admit that the reason for these failures seems to be complex and varied in nature- including whether a device is programmed on an external programmer or within the scope itself after initialization. I'm no longer pursuing any troubleshooting on this project so I haven't been able to figure out any exact causes but Tantratron has 3 of these (that I programmed externally and then tested in my own TDS540C) and has had trouble programming them in circuit in his scope via tekfwtool. He's been doing some low level debugging so I suggest anyone who is having issues reach out to him for more information. FWIW- my own replica I built last february is still going strong. --- End quote --- First I really want to thank @cuebus (Robin) for his research, reverse engineering and design of his kit which really contributes to maintain these legendaries TDSxxx/C/D oscilloscopes. If there is enough interest, we (I) could create a synthesis new thread to cover both this thread and this other one https://www.eevblog.com/forum/repair/dallas-ds1486-and-ds1250-replacement-designs/ for deeper investigation. It seems clear that cuebus design works so it is just a topic to secure its use, its programmation because clearly we have reports of some replicas which just work fine and other ones which failed. Why no idea for the moment except I do confirm that one replica still works fine in my lab since june 24th in my TDS794D. However the other replicas are partially failed, one did work fine during 10 days in my TDS784D then started to brick my oscilloscope. The other one never worked including my TDS540C and each symptoms are different. I did connect one Console Port and might purchase a SALEAE logic 16 Ch analyzer which I need to repair another TDS784D major boot failure, see this thread https://www.eevblog.com/forum/repair/major-boot-failure-tds784d/ What I can tell you for the moment, during the boot there will be only test check on DS1486. For example, the boot firmware will try writing a sequence of 12345678 in the DS1486 then read. With one replica, I do get a read of 56785678 so it declares a failure but this very same DS1486 replica does generate some RTC including the clock. The other failed replica will pass the 12345678 test but will fail writing the RTC registers, no clock test output as a result as well as every 1Kb page full of FFs. Again the 3rd replica has been running strong since last june 24th... |
| Navigation |
| Message Index |
| Next page |
| Previous page |