Hi In a previous video, you saw me have a little play around with a week on a 10 K laboratory resistance transfer standard that I've got hooked up to. my Keithley DMM 7510 multimeter here is just having to play around and we noticed some drift in the reading up here and I said I'd do a second video where it actually put the resistance standard inside my thermal chamber here and that's what I've done overnight. so let's take a look at what I got when I came in this morning. Now if you haven't watched the previous videos, I won't go through the data again so this won't might not make much sense.

So click here and watch that one first before we continue. Let's go. Ok, so basically nothing has changed in the setup here except I've put my weak on resistance standard inside my thermal chamber here I've set it to out 20 degrees. it shows 19 degrees up there, but I've actually got my thermometer here.

it's currently it's a well 19.8 nineteen Point six something like that. So it's been like that overnight and I've basically just been not continuously recording here although there was a where I disconnected the thing so I could put it through the port on the side here. and yes, as his standard industry practice for our thermal chambers. I've plugged it up with a rag Beauty works every time, no worries.

So yeah, and so there was a little glitch in there when I disconnected, but I've continued the data because I thought that might be valuable rather than just started again and look for stability because we've already got that. You know previous history I Wanted to carry on from the previous video so let's take a look at the data now. The data we got up previously showed a point double-o-double O 600 difference between what the resistor datasheet here tells us or what the resistor calibration sheet told us and what we actually measured on the meter head and my guess was that that point double-o-double O 600 k discrepancy was coming from the actual Keithley multimeter itself up here and we had a look at the specs for that and it seemed the temperature coefficient for this meter on its resistance range and that you know it seemed to be accountable. In fact, the datasheet value of 2.5 ppm per degree C So for a four degree C difference we're looking at actually seem to be larger, so it seemed you know it just as a rough order seem to be better than the spec, but the spec of this meter could certainly account for that.

hence why today I've actually got the resistance standard in the thermal chamber so that we can ensure that the resistance standard stays at a constant temperature and the only thing I'm going to change now is the temperature of this meter by changing my aircon here in the lab. So at the moment I'll show you what we did last night. this is where I started last night basically and see that big dip In there that was when I actually disconnected the contact. So it's good that it didn't actually give a reading outside of that I really like that and like mess up your data although it does show a little tiny red dot in there I'm not sure if you can see that but it that indicates that I Guess there was an error condition like it actually went open because we're using four terminal resistance measurement here.

Anyway, look the first thing you notice is you get we get a hell of a lot more noise on here. Look at that to what we got previously and that's not surprising. We've got this thermal chamber which uses a DC to DC converter in it to drive the Peltier device to, you know, keep it stable temperature and we've got unshielded shouldn't touch them Unshielded test leads on here, you know, so it's not the best setup so maybe we should actually turn the filter on for this one because we didn't need it before. but now that noise is getting massive.

but anyway, you could see it dropped up here when I left last night actually had the aircon on. This is when we ramped the aircon back up and we could actually see the difference. Yeah, we could actually I believed that we could actually see the oscillation of my aircon in there. You can see that like that.

And basically what I did is I turn the air con off I probably maybe I turned off well and I probably kept going anyway I turned it off before I went home. Okay, so there's no longer any aircon cycling okay, but we are getting look. we are getting some noise so you know that's like periodic noise on there, so that's rather interesting. So I don't think that's the Peltier device in the chamber.

Switching off and on. The thermal response of the resistance standard wouldn't be anything like that. You know it wouldn't be anything near that quick. So yeah, that's probably just you know, maybe switching crap from the DC to DC converter in the thermal chamber.

So when I went home last night it was I had the aircon on to ramp up like that and then it got to it was about 22 degrees or there abouts. then I went home, turn the air con off and you can see that we've got a negative temperature coefficient. As the temperature increased, we can see the change there and it's currently 26 and a half degrees in here. Let's just call it 26.

So we got that four degree differential that we had before which is handy and and once again, the temperature in the resistance standards been constant. although there might have been a little bit of thermal lag in there just to bring that up to temperature, but you know, still, it stayed stable all overnight. So really, the only change would be due to the meter here the Keithley meter. So we're because we in the previous video we suspected this meter had a negative temperature coefficient.

The datasheet doesn't tell you that, it just says plus/minus right? It doesn't tell you what the actual temperature coefficient is so, but we realize that it must have been negative. That was our theory anyway. So that was our hypothesis and so it's going negative like this. Exactly what we expected.

So let's get the difference there and there and see. You get the values and see if it's equal to that point. Double-o-double Oh 600k Discrepancy that we got last time and we can do that by going into our reading table here. So I'll read off that figure.

Point 10 Point Double-o-double Oh say Four zero and then we'll get this value at the end here and right at the end there getting basically put Nine Point Nine Nine Nine Nine Eight You know, Eight One, something like that eight - So last time we had that discrepancy as I said of the meter - what we got from the resistor Cal sheet here and we had a point: Double-o-double Oh Six Zero Zero basically difference there. So we had that discrepancy which our hypothesis was that it was caused by the temperature coefficient of the Keithley multimeter. And what did we get when we only change the temperature of the Keithley multimeter today? A difference? A Delta point Double-o-double Oh Five Nine, let's call it six Zero. We see these all make practical e the exact difference there.

and we're not even using a really, you know, decent, you know, like really controlled proper setup here. Yet we were able to convey ceccolini difference between these two. Neat beauty, huh? So what we're going to try and do now is I'll turn the air con back on. Okay, Oh, of course our temperature chamber is going to stay exactly the same temperature.

It's not. You know it can easily ka, you know, cater for external whites, low external temperature changes caused by the aircon. No worries, see our resistance standard. It's going to stay at our constant temperature which is currently.

Come on. you can do it. You know, twenty odd degrees it you know it's going to be cycle. You know, maybe half a degree around that or something like that.

so there's not much in that and I'll put turn the air con back on. it is currently twenty six degrees and I'll I think I'll turn some filtering on just so that we get a filtered signal there because that's too much noise Anyway, because we now have confirmed that this thing has a negative temperature coefficient. If we drop the temperature back down to 22, then we expect this to rise up. Let's see if we get it.

Actually I could have sworn I heard the filter turned off, but I actually have the filter turned on with the in power line cycles. Oh geez. Anyway, so it looks like we're not going to filter that any better and I don't want to go touching the setup and you know, shielding the leads and doing everything else so you know. Let's just go back to the graph and just simply turn the aircon on anyway we can see.

Even though we've got a lot of noise there now, we can see, we should be able to see that change. So I expect that now to start ramping back up. I'll give it an hour and or two and we'll come back. And I think as we saw in the previous video, I just started, it's been like five, ten minutes or something.

just started turning on the air corners. Of course you'd expect it to go immediately up in value because the temperatures coming down right the air. Khan's I've measured the temperature. you know at the vent it's pumping out like you know, 13 14 degrees C air right? So you'd expect it to go immediately up, but it doesn't.

It looks like it, just like it's not the correct term, but it's sort of like under shoots like that and then it's going to go back now. I Suspect that's because of the control loop for the reference inside the Keithley meter actually trying to. You know either the thermal response of the you know the the heater in there for the reference or you know However, they've got that control loop set up so that's probably a response from that and you know it would be nice if we had a nice noise rate. You know you could do it better if we actually put had the meter in a thermal chamber everything else and we're just measuring a fixed resistor with no noise or whatever.

You know you'd be able to see that, but you can actually see that dip go down and now it'll You can actually see it. It's starting to head back up and I believe Over the next hour or so it'll shoot right back up. So that's rather interesting that we could actually see that it's another five minutes later or something and you can see it is actually heading up there. And of course we've already hit.

you know, near our 22 degrees there. So because the ambient air is just circulating around but it hasn't immediately shot back up because there is going to be some thermal lag for the for the air to actually get in through. you know, the vent holes and this things hot inside and and it's got its own thermal characteristics inside. so it's going to take some time for this thing to actually ramp back up.

Okay, it's been an hour or so and yep, you can see that we're getting basically the same ramp we got here. not quite as big or instant because well, where rub you know we've got that extra noise on there now. and I guess the resistor wasn't adding on because it is a quarter of the value that we had before. So now we haven't got that so.

but it's slowly ramping up as you can see and it'll most likely get back to that average value that we saw there. Absolutely no doubt. And yes, I have been monitoring the temperature in side this thing and inside the thermal chamber and it has been, you know, stuck around 20 plus minus 0.5 So the thermal chamber was doing the business and keeping that even though the air kind of ramped down by four degrees because it's 22 degrees in here now the resistance standard stayed at that temperature. So what we're seeing there is just the Keithley meter and here you go, you can see that we've actually flattened out and we haven't quite come back to the peak there.

If we have a look we've come back to is that it's that straight there we go, we've cut well, almost almost pretty darn close to it. But of course as I said, there would have been some settling there of the resistor in the thermal chamber as well. So you know, maybe that and that sort of exact in quote marks figure we got a point double O double O 600. That difference we measured it might be you know it's gonna be slightly out, perhaps.

but anyway, you can see That it has recovered there, so no problems whatsoever. Look at that. So this Keithley DMM 75 one zero on the resistance range has a negative temperature coefficient and it matters. You can actually see it in the data which is really really interesting.

but look at that. look at that cycling there. There is definitely definitely cycling in that well have to get a a time period on that You can see that's actually 200 seconds or there abouts per division. So it's nice that actually tells you that.

And we're looking at basically one cycle. Um you know around about I just over two divisions. There may a bit more like it. I Think it actually varies.

It's gonna vary a bit, but that's you know. Let's say you know 400 to 500 seconds per cycle. So it might be tempting to think that this is the cycle of the aircon because it does seem to match the time period of the aircon cycling off and on. but it's not I've actually switched the aircon off at the moment.

so there is no that the you know cycle of plus minus 1/2 degree air. actually you know doing anything. So what I'm gonna do now is just last thing is just turn the thermal chamber off and I won't open. I Will just turn it off and see what happens And there we have it.

I Switched off completely and disconnected the mains power cord from our thermal chamber. I haven't opened it so the temperature hasn't changed so it looks like, well, you know you could say that that's a temperature cycle change with inside the thermal chamber, but I don't necessarily hear it actually. you know, changing that quick or though you know it could be. But as I showed in the previous video inside the thermal resistance chamber, it has a high, you know, like it's gonna have a very big thermal lag, So I'd be very surprised if it could actually change that quickly.

although it possibly could, so we might be able to see a smaller term change ie. that one quarter which we're looking at before. I'm not going to go any further in this and muck around with it, but suffice it to say that that's switching does certainly come from the thermal chamber. So there you go.

I Hope you found that video rather interesting. It started out as a quick thing I wasn't even going to make a video and I just wanted to, you know, plug that reference in overnight and I see what happened? It's turned into a two-part video getting a you know, a fair bit of data here, but it's interesting to note that that Keithley meter we've proven actually has a negative temperature coefficient substantial. one or one. that's You know, it's still within spec.

Everything we're seeing here is still within spec of the instrument. But because we're measuring right down in the noise, we can actually see the changes in the temperature coefficient, so it's rather interesting. We can see the thermal lag in there as well, you know? So if the Keithley instrument and the reference heater and other things inside, we can see that little undershoot there. and it's you know thermal response of the reference? That's all.

It's all rather interesting stuff, and this was pretty much a hack together test as I went along. I didn't you know really? design it from the start to make you know to do a proper test or still able to get some real numbers out of that that actually matched up. Fascinating stuff. So if you like that, please give it a big thumbs up.

It's another 20 minutes of waffle and anyway, I think some people might really find that rather fascinating. And if you've got you know, a decent meter with this sort of resolution, you can, you know, do and play around with that test like this. Of course you know, like all you need to do is put like a resistor short you know, straight in the input terminals for example, or a battery voltage. you know, just hook a battery straight up to it.

you know, so you don't get noise and everything with the cables and test system and all that sort of jazz so you know and you can eliminate noise and you can actually start seeing drift in your actual meter so you know. it's It's handy to know when you're doing precision measurements like this. you can't just magically take the meter and you know it's within spec. You know, plus -5 degrees outside that? okay, but within plus -5 degrees of the temperature was calibrated that it's spot-on Well know the thing, Drifter all the time.

So there you go. I Hope you enjoyed it. Catch you next time. Just a very quick follow-up I left this running over Christmas time.

it's now Boxing Day I Came back in the inside the chamber I switched it off twenty five and a half degrees and inside the lab here haven't been to the lab for like you know, a day and a half. two days or whatever. So if 25 odd degrees and this is the response I Got over that time so the aircon is definitely off, everything's switched off and there we go, but it's fascinating. You can see that's reached a new low down here.

where it's You know it hasn't really increased in temperature much since here, but yeah, like it's You know it's been tracking there because that's been the temperature here. In like the building and you know, the corridor. It seeps in through the roof, the walls, everything else. So I've had some slight temperature variations here, but looks like it's reaching a new low there.

and this brings up interesting things in, you know, long-term trend plot in Wicked Me, you know, plot. This thing for two weeks in a temperature chamber and stable and everything else and we might see some drift. And you know things like that, It might you know, settle down and you know all that sort of jazz when you're talking about the the precision, the resolution we're looking at here and all the little subtle things, matter and long-term aging and and drift and other. You know they could even be some sort of, you know, quantum effects in the reference or whether you might like weird, you know, a physical phenomenon at play in you know it could be fascinating stuff if you leave it for weeks and months.

And you know some people leave them for years to see the long-term drift in things that can be absolutely fascinating. So there you go I wonder where it's going to go? But yeah I think I'm gonna call it quits now. you.