Hi Just a quick follow-up video on this: Prema 6047 classic 1989 vintage or thereabouts uh 7 and 1/2 digigit uh Precision multimeter. It's like a Metrology uh gray one, you know, really expensive for its day, sort of top of the line and uh, if you haven't seen the previous video, the tear down one of that, and a bit of playing around I'll link that in down below. so check it out now. I did mention in the previous video that um, it was slightly.

It appeared to be slightly out of calibration. I would have expected it to be better than this, especially on the DC volts here. and uh, by the way, I figured out how to switch the Uh time here. It was easy.

It was this uh integration button here and I can switch that down and it actually I can switch the number of Uh digits there. so I can go to 1 second and then curiously. look, it's got a nice little uh countdown timer in here so counts down to when's the the next Uh measurement interval and it does the update. Really quite neat.

You can go up to 80 Seconds here. Anyway, in the previous video, I mentioned it seemed to be a little bit out and I suspected that might be due to the nonvolatile ram up the back here. and if that's lost, its uh, memory contents, then that is you. That has to be where the Uh calibration constants are stored for this thing.

CU these are electronic calibration con constants. There's no celebration pots in here that you tweak or anything like that. The only Um pots in here aren't pots are actually trimmer caps designed for the Uh frequency adjustment on the AC measurement range. so nothing actually to do with the calibration at all as such.

So the calibration constants must be kept in that non Dallas non, uh, presumably. Yeah, Dallas cuz they're the only ones who made it back in the day. Um, they will stay of the art back in the day. we'll have a look at it, but uh, they must be stored in there cuz there's nowhere else to store it.

We've just got an Eom It's a 65 CO2 processor in there. Absolute classic. Nothing else. Constants must be stored in there.

There's an internal backup battery in there which has a nominal life of 10 years. Uh. and then you basically have to throw the chip out cuz you can't replace the battery in the thing. Now this thing, of course, is like in the order of 24 25 years old now I've personally seen these Dallas non volatile Rams still working after 20 almost 20 years I think is the longest I've ever seen.

but you know we're really starting to stretch it now. they're only guaranteed for 10 years. So really, um I Pretty sure the calibration constants in that are gone. But anyway, what we're going to do is, uh, rip that out.

have a look at it and see if it is actually blank and see if it makes a difference actually taking it out. Now here's the schematic of the main processor here and you can see there is the 65 CO2 They've drawn it tiny because, uh, they've got just the parallel data buses coming in and out of it so you don't need to show uh much at all. and they've got the other uh, miscellaneous stuff that you find in any uh, microprocessor solution. Based around the 65 CO2 there's the Eom.

there's the Um main SRAM and here, uh, u12 here is uh, the battery backed Dallas realtime clock chip and basically what it is is it pin for pin compatible with a regular SRAM chip of the same size except that it's got a build-in uh, lithium, uh, primary battery in there. It's not rechargeable. it's primary. It's got a guaranteed 10year Factory life when you buy it.

I I Think you can still buy them these days I Don't know I haven't looked recently. but anyway, the somebody on uh in the comments previous video wanted to know, um, how easy was it to remove. Can you remove these things? Well, yes you can because once you power the thing off, okay, there's the VCC here. once you remove the power to VCC it's got internal control circuitry that detects that and then automat ically uh, right, protects the memory.

So as long as you guarantee that that power is off, um, and it's you know Split Second later or something like that as long as it's off, then you can pull that chip out no problem whatsoever. It doesn't matter whether the right P pins float or or anything like that. Now you'll notice that there is a calibration switch. This is on the back of the unit here, my one hasn't been fiddled with.

it's still got the cow sticker over that switch and as you can see uh in the calibration protection position which is up up, it's just pulling that uh pin. It's pulling the not right pin permanently high so nothing can be written to that chip so you can't override, accidentally, overwrite your calibration constants and other stuff which is stored in there, presumably on something like this. It's probably only the calibration constants maybe, uh, maybe they're storing like the last used mode or something like that, perhaps. but there's not much else in this thing that'll be stored and if you get in the back panel and flick that switch, of course, then it goes down to the some of the Gated circuitry down here which can enable the right switch.

and you can go into calibration mode via the front panel and you can, uh, set new calibration constants in there and calibrate this thing. So I'm going to rip that out and we should be able to read that as a regular. SRAM Let's see if there's anything in it. My guess is it'll be blank and I'll just get in there and check the rail.

I Know it's dead, but you know, um, this is what you should do just to be absolutely sure that, uh, there's not any residual power left on that rails. It's not. It's dead. So that chip will be thoroughly right protected with the internal circuitry.

So I can now safely simply remove that and whack it in my um, uh, eom reader to check it. So I'm going to be really mean to this thing and I'm going to power it up with no chip in there. There it is. it's gone.

So um, I we'll see if it's just got the calibration constants and it was reading out garbage before and it was happy to read it out then. well, you know everything's hunky dory, so let's so it should power up exactly the same as it powered up before controller one. It goes through takes a few seconds, but uh oh, no error. eight There you go.

No, it obviously write some uh data to that. No, no it. it writes some data to that SRAM and tries to read it back and there's nothing there. But hey, we could replace that with a standard uh SRAM If I've got one and there's the sucker, look at the date code 50th week 1987 and it's supposed to have a nominal life of 10 years.

So um, yeah, you can see the outline of the battery in there. There we go. You can see it if you get the right angle there. There's the Lithium coin cell battery like a CR you know, 20 uh50 or something probably stacked up and solded directly on there.

And there's the chip poking out the bottom and and fully potted of course. So you can't really get these things and replace the battery. So if the battery is gone, of course, it's an SRAM. So if it has gone, there's no way I can recover the calibration, uh, constant out of this thing.

Not a chance. And by the way, no, there is no way that you can measure that battery voltage on the pins. It's purely internal, so all you can do is read out the contents and uh, presumably if the contents aren't uh, blank, then well, it's probably got the original contents, but hey, they could have been corrupted if the battery is marginal or something like that. So I'm going to use my little Mini Pro Teal 86 uh programmer you've seen in a previous video.

really cheap. These things. Everyone should have at least one of these. I Mean it's like 30 bucks or something.

It's incredibly cheap. Now if we have a look at the uh software for this sucker, please excuse the crude screen capture here. but uh, look it already. It supports the Dsr2 1220.

We've got the 1220 Y, but that's going to be close enough I Don't know what test there means, but anyway, we're going to select that Otherwise, we could have just selected any generic um SRAM for the same size now I'm a bit concerned about the test part of it. There could be like test mode as in testing a chip and look at this test range. Yeah, I don't like the sound of that. So uh, we don't want to test it, we only want to read the contents out.

That's all we want to do Now let's say we selected one of the other ones here and of course we could go and do what standard SRAM if we knew the size I haven't looked up the data sheet yet. but no, these are test ranges too. So maybe if we go into uh, standard SRAM down here, let's choose a 616. No, it's got that test range as well.

So I don't want to test a chip I want to actually read a it out. So here we go. if we I was in SRAM uh Dam before I've now gone up into ROM flash nonvolatile Ram there and Dallas Ds220 RW So there you go if we select that Bingo we're not in the test menu anymore. So yeah, if we accidentally hit that, oops, we could have screwed our chip.

If it was good, we certainly don't want to do that. So now we can use the existing tools here to actually uh blank check and then uh ver well no want a ver read from chip? That's what we want so it shouldn't enable that right line assuming there's no fault in the programmer and we should be able to read that contents out. Anyway, there's the data sheet for the chip. Yep, it's the uh, it's the bog standard one back in the day, the 16k nonvolatile.

SRAM There you go. 10 years minimum data retention in the absence of external power. Um, and well, it's You know if this thing was powered up for long periods of time, then it wouldn't use that. Uh, it wouldn't use the battery in there.

But of course you're basically when you're looking at 10 years, you're talking about the shelf life of the Lithium battery lithium primary battery in there anyway. So even if you had the power off, it's still probably only going to last 10 years or so. But as I said, I've seen one's last almost double that. But definitely not guaranteed.

All right here we go. Moment of Truth: Let's read it in yeah, the whole range. Uh, by default, it tells us exactly the orientation we want. Fantastic.

It's all in there. I've made sure it's around the right way and here we go. read: Tada Done! I It's I Think it's silly that you have to cancel, but anyway, it is done. And look, we have.

We have data. Woohoo! Look at that. There is 80. There's lots of 80s in there, but there is all this other data now just to make sure that's not just a random gibberish.

What we want to do up here is we want to go up and we want to rify that so it will compare what's in the chip to what it just read in. So yeah, Verify. Successful. Yes.

and we can just run that. Just run that a couple of times just to make sure everything's hunky dory and let me actually physically take the chip out and put it back in and just verify it again. Yep, so there you go. We have actually successfully read the contents of that and it is not empty.

Well, I'm actually rather surprised now. Of course, we can't make heads or tails out of that what stuff is stored in there, but it looks like it has several sort of blocks of data stored in there. So that would be all of the calibration stuff and maybe some other uh mode things or something like that. But yeah.

wow. Okay, after what 24 years or something that battery looks like, it's still good. Well, it's still got something. We don't know if it's actually corrupted the data, but usually these things will go blank.

So look what? I've dug out of the old junk bin a616 2K By8 SRAM 16k bit SRAM should be exactly the same pin out. Let's plug it in see if the thing boots up without that error message cuz if I'm right, it is actually right in to the SRAM when it boots up. And if it can't read it back somehow, then it knows You know that error8 is saying it's an SRAM fault. All right, here we go: I Got the SRAM installed.

It's around the right way. Yep, let's Power It Up Controller One Two Error Eight No, it's exactly the same. There you go. So maybe it's not reading the uh check sum that he expects or something like that.

So maybe our original Dallas non volatile Ram is actually good. So I don't know, We have to read the Uh manual I Guess see if we can get something on error8 Bingo There you go. Error 8 Error: during self test it Compares As I suspected, compares the check sum with the Uh one in the nonvolatile ram that Dallas chip. So obviously for this thing to power up previously, that Dallas chip must still have the correct data in there.

Otherwise, it, uh, would give you an error eight every time you boot up. So hey, the Dallas chip has to be good. All right. So I'll put the Dallas chip back in and we'll power this thing up again.

and uh, it's probably going to come good. Exactly like it did last time. So yep, yep, everything's fine. So I because it passes, that check sum is stored in that Dallas SRAM If that battery was flat, then that Che check sum would be incorrect.

and uh, it. You'd get that error8 message every time you power up so it must still be good. Wow, after all these years. Unbelievable.

1987 That one is. Uh yeah. 1987 50th Week 87. Unbelievable.

Still going, but of course, if you really cared about keeping this thing going, then the first thing you would do. read the contents out of that. uh, save it and then try and find a replacement chip that you can reprogram whack back in there. And yes, I have read in the contents of the ROM as well.

that's a Uh 27 C256, just your classic type and uh yes, I've got that uh contents that I will upload it uh into some repository on the net. for those who need it in the future. always do that with this old sort of gear. Now there's one interesting thing to note about this meter is that even if this uh SRAM here does fail, then it's not a problem because uh, you don't lose the calibration of the instrument cuz Prema have actually programmed the original calibration values at the factory into the Eom here.

So uh, the manual actually tells you how to recover if your uh battery fails in here. you can't get, uh, your last calibration values of course unless they rewrite the ROM for you. Uh, but you can get the original Factory calibration values. There's a mode you just flick the switch on the back, you power it up, and then it automatically copies the contents over to your new SRM Fantastic! And for those of you curious to know if it's possible to turn the 6047 into the 6040 I 8 I.E Get that 8 and a half digit resolution? that extra one digits.

Well, it might be possible to hack the thing in the ROM to actually do that. Um, that could be a distinct possibility because, um, the only major difference between the 6047 and the 6048 is the LTC 1000 voltage reference. There's some circuitry difference in there, but you might it should in theory, probably have the same software and just be uh, limited to that 72 digits in the uh firmware for the 6047. So um, I've got the ROM dump here and uh, as you can see: um I've searched for premer I've searched for 604 7 I've searched for 47 I've searched for 48 and none of it's in there.

Basically, it's all just gibberish. I can't uh, find anything so you would have to disassemble it. Uh, disassemble the 65 CO2 binary file and to be able to, um, try and find that if there's a bit in there that gets flipped or something like that? Um, so I Was hoping that yeah, it would be as easy as like there'd be a string in there saying 6047. you might change that to 6048 and then the firmware just checks that when it boots up, but that doesn't look to be the case.

Not that easy. So yeah, not not impossible. but I thought I'd give it a shot Anyway, and if you're wondering why my room sounds very very echoey here, it's because I am um, actually recording this at home and we're renovating the house and I do actually have my uh, sound absorbing wall panels in here, but still, there is no furniture in this room so they don't cover the walls 100% So uh, really, we're getting some pretty bad echo in here, so there you go anyway. Um, hope you enjoyed that quick video.

Catch you next time.