Hi Welcome to Tear Down! Tuesday Another world exclusive. We've got the Agilant Truevault 34 61a Bench Multimeter. Oh yeah, Precision porn I Love it! We're going to tear this sucker down. See what's inside it? now? This is their, um, it was supposed to be released yesterday, but they jumped the gun on their website.

So yeah, everyone knows about it. Now it's their new and A2 Digigit uh multimeter that replaces the venerable old 341a. This is a 34 611a so it's complete replacement for that one. Presumably they're making that obsolete now.

Can't get the parts anymore. It's got the old vacuum fluorescent display. this one's got a sexy new color do Matrix Display: You can get histograms Trend charts, all sorts of stuff so very nice bit of Kit and it's cheaper than the existing model. Um um, this one is the entry level one which is the 346a is only 945 bucks.

bargain and this topof the range unit. Um, that I've got the 34 61a is only 1,095 us. So as far as bench multimeters go, it's pretty affordable. So this is a rather exciting addition to the Bench multimeter market and a lot of people want to know how this puppy performs and more importantly, what's inside.

We don't call it tear Down Tuesday for nothing. You know what? Say here on the E blog Don't Turn It on. Take it apart Now I Know I wouldn't normally do this, but let's take a look at the marketing uh, blur for this thing just for a a minute or two here. and uh, nice new graphical display.

It's got the histograms, the trend chart huge, um uh display like this, it can show all your stats and everything. Much better than the uh existing uh, green vacu vacuum fluorescent display on the 3401a meter and um, interestingly which we might find out inside it um offers or so it claims. Here's the true: Vault Let's have a look at it. Here's the true Vault multimeter what it means to you, blah blah blah.

Anyway, they reckon the uh noise and injected Uh current is much lower than brand a brand B and band Brand C I wonder what brands they are I don't know anyone want to has it a guess So it claims it's um, the best in terms of noise and injected current and also the best in terms of input uh, bias current as well. Once again, better than the three other it's too noisy to measure on Brand C Brand C must be really crap. Come on, let's find out what brand C is and uh yes, it doesn't have your traditional um RMS uh converter chip. It does digital um Acrms conversion.

so there we go so we won't find one of those. you know your your traditional Analog Devices um uh RMS converter chips in there at all curiously low. Why it didn't offer the Legacy um Rs232 if it's supposed to be a complete replacement I'm just uh surprised that they didn't offer that um built in. Anyway, it does now have Lan um and USB which the old 341 a didn't with its clunky old uh GP interface.

Now why they've put these stupid promo stickers on the front I Don't know. Just no. no. Speak your mind and you can win a prize.

Yeah, great, Thanks for that Anyway, Agent True Vault H Access Demos. There you go. I'll leave that on there for a bit of wank. Factor There we go.

but that's just no go away. She probably doesn't even work at Adelent give me a break. Ah man. Anyway, um, very nice.

uh. construction on the thing by the way. I'll have to do a full review of this thing of course, but uh, trust me, it feels really top quality. So let's rip this sucker apart and uh, looks like we got a couple of Torx screws on the back there and uh, that should slide out.

So essentially the big question is, uh, what voltage reference is going to be used inside this thing? Will it be a traditional, you know, LM 399 as is used in most uh meters like this. There we go. nice little captive uh, screws in there rather like that and uh, this hasn't got any on the bot. anything on the bottom? Yeah, one screw on the bottom holding on on and it should just slide off.

Shouldn't need to take the tiling bail off which is good, but uh yeah, it'll be interesting to see. This is their new Whizbang meter for the future. Here we go. So I expect um, although I don't expect them to have gilded the lily of course I mean it's only a replacement for the three double uh for 10A in terms of specs.

so I don't expect them to have this is really. no, it's not coming off, it's not coming off. so I don't expect them to Guild the Lily in that respect. So okay, this could be a pebcak error.

It's it's not the no, simply not coming out. Not sure what the deal is there. Aha, there you go. It was the tilting Bale It does have sort of a couple little hooks in there which then go into there so there we go.

apart from that should just slide out. Oh I'm seeing some big HRC fuses in there and Gpib card little Shield Hey, check it out. All right, this looks very nice. There's practically no wiring in that at all from the Main's Transformer None of your traditional uh rubbish in terms of Um wiring from your uh Jacks of course, like you know, a lot of them will come over here.

They'll wire directly from the Jacks as we've seen in other Um Bench multimeter tear. Downs They wire straight over here to the big main clunking um input switch over here, but this one hasn't Looks like they've custom Um designed and engineered some big solution for the Jacks There some big holder for the Jacks that looks really very nice. I'm impressed by that. You'll notice how theyve folded the metal over there on those input.

Jackson Uh, it wouldn't surprise me if they were um, well, I'd be surprised if they weren't proper low. EMF uh contacts. maybe a, uh, some form of torrium, copper or one of those other low EMF uh contact materials. So the first thing that Springs to mind is just this is so neat, tidy, and well engineered.

I'm very very impressed. Look at the uh attention to detail that they've done on that uh, cover there for the Main's um, input there. they just you know, looks like does that slide that looks like it slides back somehow I don't know something like that and comes off. but uh, that is really neat.

They've really put a lot of thought into how this thing is going to be assembled I think most impressive. that's uh, really is quite quite nice. and yes, it does have a uh fan in it. Uh, it's not excessively, uh, loud.

but you know, if you didn't like it, well, you could get in there and disconnect it. or you could you know, uh, replace it with a a low noise substitute or something I guess. but M Very impressed and I'm impressed by the huge 10 amp. uh well, 11 amp HRC fuses down in there on the input because this thing's only Cat 2 rated.

Usually you don't find big HRC fuses and input protection on these bench multimeters. It's just not the application. Uh, they're designed for, you know, putting in a rack, designing into a system. you know you usually don't get the usual um, you know, surge type overload conditions that you would on a handheld meter which can used in the field and abused and used in all sorts of different situations.

So yeah, big thumbs up on the two HRC fuses there. Very nice because the thing does have curiously. Um, it does have a 10 amp input jack and a 3 amp input jack. So uh, I you know why they've had to go for those two different Uh values I don't know.

Haven't found the current shunt resistor yet? I Guess we'll uh, have to have a close look at that. But yeah, very nice. You can see the traces in the PCB there, so they're actually in the inner layers. um, rather than on the top or bottom.

I've checked the bottom of the board. they're not on there, so they're certainly um, in the inner layers of this board. This is, you know, probably a a six layer board or or something like that. Speaking of the board, they're up to revision four already.

Well, no surprises. You know they're going to tweak this thing before they release it cuz I've got one of the uh, very early units here and you would have noticed already that there's real, no huge processing GR on the main board of this thing, so it's all up under there on the front display board. I Like the nice board to board interconnecting there. Very nice, very well designed, very professional.

You have a battery backup battery there. No, it wouldn't be for the cow uh stuff. as in old, uh, multimeters. it just be for the uh, real time clock, That' be it.

but uh, that's obviously the processor up in there running. uh Windows CE which this thing does. and I Love how the front uh, input, front, rear input switch there has just popped up through a cut out in the PCB That is very nice, well engineered. It's not big, clunky, long arm.

Ah lovely. Now, as for these input: Jacks I'm assuming that this plastic clip here somehow retains that entire assembly. I'm not not 100% sure what's going on there. if you look down, they're soldered into the bottom like that.

and but that red clip look looks like it's got some sort of covering on there. nice little cut out there. They've done that for, uh, isolation reasons. Very nice now.

I'm trying to actually figure out how to get this whole thing apart. and uh, by the looks of these, Bnc's on the back here and the folded metal shazzy, this board has to slide out like this somehow. I Mean there's obviously a couple of screws holding that down in there. and there's that metal screen over there front and back.

I Can take those off, but then I'm not sure how the front actually comes out. It's uh. I don't know. it's a it's rather weird.

There's sort of molded plasticky bits all over the front and there's another separate thing in there I Don't know. Do I have to read the service guide recalibration required after removing? Shield Are you got to be kidding me? Really? Really? Just to remove the damn Shield recalibrate it. Ah. I think they're taking the Mickey there.

That's just ridiculous. But hey. I see a sensor there? We? there's the reference under there. There we go that looks for all the world like a Uh linear technology.

Um, second sourced LM 399. All right. I have resorted to the service manual and apparently I have to, uh, pry out this plastic black plastic clip first first. So H there we go.

Hey, Tada and then to release the Front plastic panel I've got to push this clip and then I don't know something's supposed to happen. Aha I Got it. These clips hold in these um, side studs here for the uh rack mount kit so you just got to push that in and taada Taada. There it is.

Oh look at that! Brilliant. That's actually very, very clever I Like that. So I've no doubt voided the calibration on this thing. even in the service manual.

it tells you you remove that metal shield in there. You got to re-calibrate it. Yeah, right. Um I doubt it.

I Think they're just, uh, scaremongering there. but uh. anyway. I have got the front panel out and the other clip in the back here and this board I think is just going to slide out.

or it should. It's already slid. A little bit should just slide out because there's a retaining hook in there which holds the board in place. There we go.

TDA There we go. The whole plastic assembly comes out. beautiful. Look at that and there's all the mains wiring under the cover.

There looks very professional. No problems there at all as you'd expect. I Rather, uh, like the way that thing's physically constructed. and as we looked at before, the input circuitry here 2 11 amp U fuses for both of the fuse ranges.

we got a massive poly switch there. by the looks of it, we' got our ourselves a spark gap. brilliant. and uh, over here there's our Diode Bridge We've got some mosfets in there, couple of three switching reays and there's no input wiring at all.

and uh, it really is all sort of. uh, you know, above and beyond your usual um, bench multimeter. It is, uh, very, very nice. I Don't mind that at all.

and it looks like we got ourselves a high voltage. Network Here you can tell by the resistors, each one individually, uh, compensated with a parallel cap. There they're in a high voltage string. We' got ourselves a 400 volt DC cap and what looks like um, some sort of uh, ceramic resistor Network high voltage High Precision Perhaps have to look at the number on that one.

By the way, we got more protection on the backside sockets here. Another spark G and another mauve and that's an IRC brand. It model looks like 4592 I did a quick Google and uh, nothing. uh, special.

It was like a .1% 100 volt ceramic Network 100 PPM Uh, nothing special. so I'm not sure if that that's the exact one or not I Don't know. Need more research. And if you're wondering about the current shunt, it's on the bottom side of the B board.

There there it is a Dale top quality brand four terminal shunt. Not sure why they've left off a fourth Reay there. Um yeah. I mean this is the fully optioned up unit.

So yeah, your guess is as good as mine. And yes, the soldering on this thing first class cannot fault it at all. And that unusual looking four Leed component, there is actually a Precision resistor. Um, yeah, it's 1% uh, but it's to be I believe a low uh Tempco resistor.

and they probably just, uh, software calibrate. So that's tucked in there right between the rear, lays near the rear sockets. and there's what everyone wants to see. Yes, LT Um, 1826 I Believe that's the agilant.

Uh, you know, internal part number, but it's just an LM 399. although you know, might be uh, you know, specially graded for them or something like that. but this is not a particularly, uh, high-end bench. multimedia.

Yes, it's 6 and 1/2 digits, but you know they, they do make better ones. Now what's interesting to note here is it seems to be stuck in a four pin machine socket there. that is really unusual. I'm not going to touch it to, uh, take it out I Don't want to stress the poor little fell, but uh, yeah.

I I Don't know why have they done that? Perhaps they're um, individually, uh, you know, uh, tweaked. perhaps for this? Um, up for this 61 model instead of the 6 Z So maybe they're individually bined and characterized and then, uh, put on the board as the last uh step so they don't actually go through the assembly. uh, Reflow uh, process. That would be another reason is that they're not, uh, thermally shocked at all.

And there's a 40K 0000 but it's marked 1% just like we saw in that uh, keithly, uh, tear down which these Precision uh, resistor these Precision film resistors here. Um, you know, usually you don't get them in a 1% grade. Anyway, that's going to be an incredibly low Tempco resistor there. So uh, that's where it's getting its stability.

and uh, of course all the magic is done in software and if we take a look around the reference here, uh, not too much happening at all. Adg 128 They're an 8 channel marks from analog. Devic has got at least three of those. Oh H C 4053 You got to have some 4,000 series camos.

We' got an OP uh 27 there Precision Opamp 8677, another Precision Opamp Adg 413 We've got ourselves that's a uh a Max um/ switch. We' got a couple of those surrounding that Precision resistor that we saw up 27 and um, that's an Interel 28110. that's a precision Jfet. and but, uh, apart from that, um oh LT Triple2 there.

oh LTC triple 2 was a 4 channel switch. but yeah, so just lots of uh switching happening around our main reference. There we got some IRC resistant networks. Again, they I haven't looked up the number.

they'd probably be low temp car of course, more switching happening. Oh 74 HC 4094 Good to see all like it. There we go and um, yeah, lots of you know nothing. Uh, blowing my socks off yet and we've got some ADC action happening here.

We've got an Analog Devices uh, Ad 9200. Now this is only a 10 bit 20 meg sample. um ADC and uh, even less over here. We got the Ad 9283 and that's only an 8bit ADC 50 meg samples per second.

So uh, obviously that then not using either of those for the main conversion and we got some 74 Lvc surrounding that. What else we got We've seen all that before, so that is pretty much it for that uh top part surrounding the reference there and it looks like all the sampling action happens inside a lattice LFX uh 2p Fpga This is one of those uh, nonvolatile um Fpga which is not your usual uh style. This one doesn't need an external um esqu prom to power it up and just bang instantly Powers up. And of course we've got our JTAG interface next to that.

so uh, you could get in there and have a play around if you really wanted to, but uh yeah, most people wouldn't then we've got some uh, you know, H boring uh, switch mode converters down there. Nothing special. Curiously though, we got a few unpopulated parts down here and you can see all the guard rings to stop the Uh leakage around all the individual pins. but the Um chips aren't populated at all.

I mean this one here is populated this uh, Adg 126, um 8 channel Marx and they're really stopping that leakage getting into individual pins on there. but uh yeah, I don't know. They left a few parts off this sucker. and curiously, on the flip side of the board, under that Fpga is an LM Texas Instruments Stellaris Mike R microcontroller LM 3s 1D 21.

So they've got an Fpg and a micro on there. So obviously the Um you know the Fpga is probably just doing the um, you know, the sampling and stuff like that. And then we've um, got the micro doing all the housekeeping and check this out. On the bottom side, there's two large surface mount resistors here and it looks like they've been, uh, flipped and solded upside down.

They're obviously designed to do that, but check that out. Isn't that weird? Just these white modelist there. and uh, as you can see, there's really, um, nothing else on the bottom there. There's the Stellaris art micro.

but it's just all you know, uh, passives bypass. You know, the odd uh, transistor or network here and there. but there's you know, there's nothing else over there. And that, of course brings us back to the ADC.

Where is it? It's not one of these two fast, uh, 10bit and 8bit ones and you can't just uh, Cascade those or, uh, anything like that to get anywhere near the performance required for a 6 and 1/2 digit meter. It's just not going to happen. So where is it? Well, you won't find it as an individual chip. It won't be an off-the-shelf chip in a multimeter like this.

It won't be a Delta Sigma You know, a 24-bit Delta Sigma converter or uh, anything like that, which you might, uh, typically find in, say, a handheld Uh multimeter these days. It basically goes back to the very old technique of the first digital multimeters of dual slope integrating. Adc's what is that you ask? ask Dave CAD time And here's the basic technique of a dual Slope integrating analog to digital converter and the technique has been used in Uh multimeters since Uh digital multimeters since the dawn of time. really.

And basically it's just an integrator. If you know your basic Uh building blocks, you'll know that an OP an with a capacitor in the feedback uh, loop like that and the input tied to ground, it is basically an integrator cuz when you feed a voltage on the input here, this reference Uh node here of course due to Ideal opamp action is going to be ground. So you're going to have a fixed noan current or a constant current flowing through this resistor here. So you're going to have a constant current flowing through your capacitor.

and when you get a constant current flowing through a capacitor, you get a linear ramp voltage like that just ramping up. So this Um dual Slope integrating technique is basically going to have a it starts off at zero assuming the capacitor's discharged, you apply say your input voltage or it could be a reference voltage for example and then your capacitor charges up like that and it takes x amount of time so that's the key. But when you get to a certain point you can then switch it backr back down to ground and it's going to discharge like that and that takes a certain amount of time and based on uh, very precise measurements of this time. And remember it's very easy and very cheap to make very precise time measurements with.

you know, a high stability Uh Crystal oscillator that sort of thing. Very cheap, very easy. So you can take as long as you want to measure these Uh ramps here and these ramp times and from these value Uh from these measurement times. Then and from a No and reference Uh charge up point you can apply a reference, you have a No and voltage takes a know amount of time.

Well you can work out by some very simple formulas what your input voltage is and because you have a fixed clock rate effectively, the longer you let this ramp up, the more accurate you can measure the time or the greater the resolution. You can measure the Uh ramp up and ramp ramp down time and therefore the greater resolution you can measure the input. That's why these Uh bench multimeters. because they use an integrating ADC Uh technique like this dual slope or multi slope as we'll get into Um then that's why they take longer to Uh get a 6 and 1/2 digit result than they will to get like a 3 and 1/2 or 4 and 1/2 digit result because they just change the Uh ramp time based on the current used to charge the capacitor, the longer your capacitor takes to charge up the Greater your resolution.

But therefore you're trading off Uh measurement resolution versus speed. But a modern high-end Uh bench multimeter like this Agilant or even the previous model or the previous model before that, or before that. Um, they don't use just basic U run-of-the-mill dual slope um um integration techniques. They use what's called multi slope integrating ADC and it works exactly the same way.

It's just that it has some extra switching in here and you'll notice that we had a lot of switching on the board before and we'll get back to that. But you can basically switch in multiple known resistors in here. I.E Multiple fixed currents going in and we're going to have positive and uh, negative references as well because the input can be positive and negative and you can do uh, various stuff with that, but it's basically exactly the same technique. It's just that what they do is they're optimizing this uh down ramp period here and then they're getting uh switching in multiple currents to try and improve that down ramp time.

I Know it looks like the same time period as before, but basically the whole idea of this is to be able to get a faster um faster measurement technique. I.E faster samples per second. You know this thing's capable of, you know, a th000 uh samples per second. So you need some techniques to try and get the Uh the ramp Tim down.

In this case, um, the ramp up is um going to be very similar although they can inject other currents in there to help ramp up and ramp down. But basically what they're doing is because you're injecting you can. This is a sum point and you can actually sum in along with the input. So it's not just the input ramping up anymore, you can get the input to ramp up.

Let let's say these switches here were off for example and we just were charging up with the input. Well, the ramp might go up like that and it might take a long amount of time. Well, you don't want that. Let's speed this sucker up.

We have a no one fixed voltage reference no and fixed resistor doesn't drift. Nice stable current. We can sum that in. so we're injecting more a know current in there.

So instead of a ramp that goes slow like that, we can speed up the ramp like that and get a get the same result in in less time because we're injecting a KN current so we can just you know, subtract that out of the equations later. Piece of cake. So they do a similar thing for the Uh down ramp and then we get some overshoot and it's correcting like that. And by the time you start adding in all these different techniques, it's um, ultimately going to end up more like a Delta Sigma uh conversion type technique than uh, your traditional integrating technique.

But we're not done yet. What if we could do a trick to basically eliminate this down ramp here? Well, it's not that hard because you don't have to wait for that capacitor to discharge. What you can do is just tap off like that into another ADC and that gives you the residual voltage across that capacitor and you can just read that pretty instantly using a very fast uh flash ad DC and you can basically eliminate that downramp period because you know exactly what the residual voltage on that capacitor is. And because you've calibrated the whole system and everything else.

Bam, you can speed it up even further. So that's probably the technique that Agilant are using in here Pro Plus some other uh uh tricks as well to get a really accurate and Fast 6 and 1/2 digigit converter. And if we go back to our board here, we can see what's going on now. Um, it says in the data sheet that this new meter uses a new multi slope.

Um, it's a Agilant patented thing. the multi Slope 4 conversion technology. The older 3401a used the Multislope three converter, so they've obviously improved that for this one. but it's not absolutely brand new cuz it's been out for a while because the Um Agant L41, a system multimeter also uses.

apparently this according to its data sheet this new multi slope for conversion. So clearly what they're doing is they've integrated all of this multi Slope 4 conversion into this lattice. um, Fpga here in the previous 3401a multimeter. Um, it was done in some uh, custom, uh, you know, system as6.

Uh, Basically, so they've replaced that. you know, too expensive. don't want to do a custom Asic for that? So we can. They've decided.

oh, we'll just whack some of that in the Fpga here. So that's what they're doing now. of course. as I said, with all that residual measurement, that probably explains these fast 8 and 10bit converters in here.

why they've got two of them I'm not, uh, exactly. uh, sure. but that could explain what they're doing there. they're getting maybe a residual measurement to eliminate their uh, downramp phase of The Moldy slope? uh, conversion? perhaps? Oh, you know that is best guess.

Anyway, I don't know for sure we'd have to get some sort of, um, you know, app node or something like that of how the Uh M slope for uh conversion technique actually works. So basically all that explains also, all of these switching, um, uh, chips we saw around here. all those Adg ones and stuff like that. Lots and lots of switching.

There's that Precision resistor down in there, of course. Um, you know that could be say. for example, the uh reference input resistor for example there. because all these resistors, of course they're going to be.

um, you know they don't need to be, you know, Point triple 1% or anything. they just need to be really low. Tempco Ultra Stable resistors. That's it.

They can be 1% doesn't matter. So that's why that one down there, as we saw before, actually has 1 1% on it. Its value doesn't matter, you can actually calibrate that out uh, later in software. So basically that's all that uh circuitry around there is.

plus the Um Ad plus the Fpga to control it all is handling that. uh, multislope for conversion. and there's the main uh Crystal there that uh, dries all the time and it actually is marked 10 16.25 MHz So there you go. That would be a reasonably low, reasonably, uh, stable.

um, uh Crystal oscillator there to uh, generate all that uh, timing for that. So there you go: the Multislope 4 conversion. If anyone has any links to any um, you know info from Agilant that uh, explains the multi slope floor uh for conversion, then uh, please link it in and I'll add it to the notes. So I think that explains most of the Uh circuitry on the on the top side here, but on the bottom side why they need this um, Arm processor here? You know it's no slouch.

um I'm not entirely sure whether that's part of the multi-slope 4 conversion or maybe it could be uh, maybe something to do with a new um, uh digital RMS uh conversion? uh digital? true RMS uh conversion or something like that. perhaps? I don't know. Uh, we'd really need the schematic and some decent uh app Notes on this whole thing to figure it out. Now there's one thing I can't find in here and that's the Uh integration uh capacitor here because um, it needs to be a super ultra stable uh cap of course and I can't see anything obvious uh on the top side, you know I'd expect it possibly to you know, be around.

you know one of the well, the Um integration uh Op amp uh down here which is going to be one of these Precision Op amps that we uh saw around the reference in there. but uh yeah I can't see any obvious uh High stability cap and that's on the bottom side of all that circuitry around there, you can see the guard rings. of course there's lots of Uh rings around there I think that is the main reference resistor there. Um, but yeah, I don't uh, it must.

I Guess the integration cap must be some little uh High stability uh, surface mount one? Perhaps it just doesn't stand out. Now of course, this recalibration required after removing Shield uh, warning here on both the board and in the Uh service manual as well. You know they're probably spinning a bit of a yarn there. It's uh, where're we're obviously going to be? you know, really down in the noise.

Obviously, removing just that uh shield on there isn't going to do Jack Uh, really, they're probably more concerned with you know, people, uh, touching this thing and you know, causing leakage and you know stuff like that so you know I it's It's not like oh, your calibration is going to be suddenly you know, half a percent out or something like that just because you remove your Shield not going to happen. So as long as I don't uh touch that, breathe on that or uh, you know, sneeze on it or I don't know, uh, spill my drink on it or something like that, we should be just fine. And if you're curious to know more details about you know the Dual slope and The Moldy slope and the residual uh techniques and stuff like that, then uh, there is actually a decent wiki page for this thing. uh, surprisingly.

or maybe unsurprisingly, but you know it's a you know, it's not a mainstream uh thing, but uh, somebody's gone to the effort to put uh, a lot of the detail on there and I think there's links to uh, you know, HP application notes and stuff like that as well. And if we have a look at the main board, one obvious thing you need to do with that multimedia like this is um, isolate everything. You need to isolate all of the Uh outputs here. You know your Landan and your USB and your Um various trigger type stuff and everything else from your Um multimeter inputs because you don't want to have a common ground reference and you can pretty much see the isolation here they've got here.

We go. Let's go in closer, but you can see the isolation here. There's basically nothing all the way around there. It's isolating that you notice the two separate grounds there and they're isolating right along.

They've got a slot for the Um that's for the case to go into, and they've got another Uh slot along here. But you can basically see the separation between these and this circuitry around here powered from its own tap on the Transformer so completely isolated and that goes to the front panel processor of course. And then we have our opto uh, isolation uh, data path between the ADC stuff and the main processor up on the front panel there, and an isolation slot under there. So there you go.

It's all nicely isolated. Oh, and uh, if you're curious to know why they've got both a 3 amp input terminal and a 10 amp input terminal and why they didn't combine in one. Well, I Believe it's backward compatibility with the 3401a, which only had the 3 amp input. They actually, uh, tell you in the manual that the Um, if you want the best accuracy on this meter to use the 10 amp input instead of the 3 amp input, you get slightly better accuracy.

Now let's take a look at our front panel processor board here and there's a Um St Uh Spear 320 uh processor that's clearly running the Windows C operating system that's based on the Arm uh 926 uh processor core so you know quite a really Powerful Beast there. There's the memory coupled into it and we've got some power supply stuff around here and uh, the backup battery that would be for the real time uh clock? I'm not sure what that one is under there we'll have to take. We'll have to lift that sticker off and have a look uh JTAG interface Of course. Well, looks like a JTAG interface and this is interesting.

They've got a Serial Port down in here and there. It is a 10 pin header with a Max uh 3232 on it. So um, that's an Rs232 interface and um, if you remember before I said there was an Rs232 interface option missing on this thing that the 341a had uh, previously. So um, you might think oh, maybe that goes to you know that's another, uh, future option, perhaps? Well, I don't know.

It doesn't make sense to put it on the front panel like this and then uh, run a cable right AC ribbon cable right across to the back panel of the machine and then have it. uh, you know, actually. uh, go out there so it's probably more for internal um, debugging or uh, you know, uh, system. uh Diagnostics and system development.

Um, stuff. Although they have left it, uh, they have left it actually populated in the production. uh version. So yeah, maybe it has something to do with the production, uh, testing or uh, you know, a setup or something like that.

And when we reassemble this thing, we may have to, uh, actually, um, check to see what uh data is coming out of that thing. Probably you know all of the boot, um, you know code. It's very common in these Uh products with Um operating systems to you know have like a system monitor, um, output or something like that from the Um from the main processor and there it is. the Spear 320.

There's a memory surrounding it, 24 MHz processor and what else have we got there? That's rather interesting. It's a TPS 65701 that's a Uh battery management uh controller usually Um designed for, you know, lithium ion, uh battery, uh management and stuff like that. So not sure what's that doing on here. They're obviously not using it for charging, but they're uh, you know, probably using it for local uh rail generation.

Rail management as well, you know, Dropout and stuff like that. And we are getting our different rails out here 3.3 1.8 1.2 And there's our Smsc 8710 that's our Ethernet uh controller that goes down to the rear panel and we've got more DC to DC converter stuff happening down in that corner, large tracks, all the uh Power for this thing cuz it's going going to take a fair bit of uh juice to run this thing. and uh, let's have a look a few miscellaneous parts around here. That's probably our real time clock.

Certainly, there's our watch crystal there, so that's just ping our real that'll be a real time clock chip. don't even have to look up the number. Sure it is. and I thought at first that might be a JTAG um interface, but look, they've got uh cow secure on there for that pin and override.

so uh, obviously that's um, you know for uh uh, system, uh uh. programming calibration Factory type stuff and I might just pop that sticker off there and see what's under there. Well there you go. That's interesting.

It's an LPC 932 900 series old school I Believe it's a Legacy um, 8051 Microcontroller, 8051 processor. What's that sucker doing? You can see the USB connector on the front panel here. there's a PCB Mount vertical USB B socket there and that's just connected directly via some uh, um, input protection and filtering there into the Spear processor. And once again, they've done very well with these uh, plastic brackets and stuff that hold everything in place.

I Mean, look at that. That's just beautiful, really is. And on the main board here, they haven't skimped on the brand of the cap there. Nichon.

No worries at all. And you know how I was raving about these input Jacks here. Apparently they're not new at all. they aren't look to be the same ones that were used on the 3401a or at least that's what I can see in some internal photos of that one.

I hadn't actually taken the previous model uh, apart before. So there you go. Nothing new in terms of uh, the input uh, Jacks really? Or the switch looks to be exactly, uh, the same as before, but looks like they've added, uh, some fuses. They've changed the input, uh, stuff around a bit.

and of course, from the 3401 A, they've just, uh, you know, completely changed the entire architecture. the um, if you look at pre photos of the previous one, they used custom, uh, a couple of custom uh, Asex, and uh, custom, uh, hybrids and things like that. but uh, you know now it's just, uh, pretty much all. um, you know it's all off the shelf I don't really see anything Uh custom in here at all.

And of course, the multi slope bar for conversion. They've ditched the Asic and they've gone for the Fpga and uh, possibly the Um processor. I'm still not entirely sure whether or not that's to do with the Multislope 4 conversion or is just uh, doing housekeeping or the RMS uh, uh conversion or um, something else. but yeah, so it's um, it's similar in terms of uh and in fact, I think some of the mounting uh posts and things and the shazzy uh looks to be very similar, but they have changed the Uh.

Transformer um Arrangement that they uh, had up here. This is all sort of changed and rearranged and improved and stuff like that. But uh, so yeah, they've done quite a substantial redesign of the 34 01a and I'm uh, I'm really, uh, quite impressed. You know this Uh instills a lot of confidence in me.

One thing to note though: I mentioned that the um, the Agilant L411a System multimeter also uses the new Multislope 4 Uh conversion and presumably it's all the same. You know they they're just going to reuse all the same circuitry cuz the multis slope 4 isn't just the you know, the Uh smarts inside the Fpga it's sort of the whole Uh technique and the parts chosen. But that system multimeter is actually capable of a lot more speed than this one. So uh, possibly they're holding back on this one.

The actual Multislope 4 conversion technique might be capable of much, much higher conversion rates, but they just don't enable it in this one because, well, you know, just keep backward compatibility I guess with the 3401a. So yeah, you know they could have gilded the Lily on this thing, but I think their driving factor in the design of this thing was absolute complete compatibility with the industry standard they had the 3401a. I mean that one's been around for, you know, more than 20 years and uh, you know it's designed into a lot of systems and stuff like that People are used to it, so you know they really want a direct replacement. so they didn't really go all out to uh, you know, shock the market with um, you know, stunningly better uh, performance and stuff like that.

They've just replaced an old favorite. almost forgot about the Gpib board. tiny little thing. um, they've implemented that in a zyink Spartan Fpga.

So yeah, nothing special there at all, but that is an optional extra connected uh via a permanent ribbon cable, which uh, then goes down onto the Uh header there on the main board. TDA It works Beauty Well, it seems to be initializing and verifying Hardware the relays are clicking over and I'm sure it's not out of calibration just because I removed that damn. Shield Anyway, woohoo, there we go. Let's even whack it on.

Oh shall we? and uh, there's no leakage on there. and there's my uh reference res my 1K reference resistor there way way way within spec. It's a bit Dicky there based on the uh, just on the two wire uh, measurement there. but yeah, well, within spec.

All right, let's do some logic sniffing of the serial bus inside this thing. So I've hooked up some headers there, hooked up, uh, the scope to it, and uh, the first thing we need to do is, um, figure out the board rate and which pins are what now? uh, that was easy to figure out which pins were what because I um, just actually. uh while I had the board out I just buzzed it out and it was easy. Uh, pin one was ground.

in fact, pins 1, 2, N and 10 uh ground and uh, pin five is the transmit output. um and pin three is the RX input. So um, what I've done is I've captured uh, it on the Ryol here I've actually powered it down I uh, I captured some data when uh I pressed the power off button and uh, this is what we got here and uh, it's a whole bunch of stuff before it. uh, powered down and here's a benefit of the deep memory on the Uh Ryol here.

and uh, there we go. Let's have a look at uh, the time base there looks like we've got individual bit, uh, changes there. We're at 5 micros uh, per division. so we're at uh, you know around about that uh, 9 microsc? um period there.

So that corresponds to 115 K board. So that must be our board rate. Now we could of course use the uh scope to uh, decode that data, use a Serial decoder in there or logic analyzer or some other. um, you know, uh, serial decoda.

but really, um, that's just pushing brown stuff up a hill with a pointy stick. This thing is designed I'm sure is designed to give out um, you know asy cereal data so you know I don't think it's going to be uh binary I think it's going to be like a monitor debug? um uh thing. So we should get genuine asky out of this. So the easiest way hook it up to a PC got a terminal program running here um, it's all Rs232 wat level of course.

um so you can't just hook it up to like a bus pirate or something like that. You do actually have to hook it up. So I'm using an old notebook here with a uh serial Port built in and uh I've got just the uh transmit of the um uh Agilant scope. go into the receive of the Uh pin on the Uh D9 on the serial port and I got it set to 115 board.

Um, all your usual uh stuff you know One Stop bit no parody and let's power this thing on and see if we get anything Here we go way there we go. You boot. Whoa. whoa, who? What's this? What's this? No, there we go.

came back. There's a whole bunch of debug stuff there. Yeah, as you know, that was almost guaranteed that we'd get uh, something like that. So there you go.

What I'll do is I'll probably uh, save that, capture it to a text file so that I can, uh link it in. um, it'll be on the I'll upload it to like the Forum thread or something like that for those who want to decode that. oh, failed local serial 2 in it got oh, it's still booting by the way, still booting blah blah blah Oh, there's tons of stuff in here we have to scroll back through. It's the scope is uh, meter scope.

The meter is, uh, still booting I've got apparently I've got, um, sort of not released firmware in this one so it does take longer to boot. There we go and we're done. We're in. we're in.

like Flyn There we go. it's now working and uh, that looks like it's not going to give us anything more out. that's just all the boot stuff so you know we can Scroll of course. uh, right back through that and uh, see all sorts of System Info stuff in guard buffer blah blah blah Cal Port blah blah blah.

But let's go right to the very top and uh uh, no, there should be something before that. so I don't think it's captured at all. Maybe my buffer is not big enough or something like that cuz I saw like a uboot or something. Uh, booted up right at the start.

We saw a glimpse of that. So yeah, I'll have to just try that again and reset my buffer size or something. By the way, that's what you get when you switch the thing off. You just get a couple of uh errors there.

Wind: SE E 600 Platform com blah blah blah kernel blah blah blah whatever that means. Serial 2 Break error All right. I've tried another terminal program here, which then I'm able to. This is termite.

I'm able to, uh, save that so I'll be able to dump all that uh data. and if we go back and uh, have a squiz at the uh top here, we can see there it is. Uh, uboot adelent p510 I guess that's the platform I think we've seen that in other implementations of Um Windows CE Agilant products. There we go.

Serial number: that is not my serial number. so uh, I'm not sure what's going on there. it's not outputting the correct serial number I guess they don't hold that in there and uh chip AA Board Rev 4 internal RTC No, No Ethernet found H Okay, well I guess it's not connected. Nand all that sort of stuff.

debug Pro Starting Kernel: no RTC again on the 320 chip. No, it's a separate Uh Windows CE Boot Loader There it is common Library blah blah. Boot loader for the Agilant P500 board. Yep.

adaptation performed by Agant in 2008. So uh, this is going back quite a bit. System ready preparing for download. Okay, so it's load in the image so it's got a it's just a boot loader and then it's loading the image from uh, somewhere else and Windows CE Kernel forarm there it is thumb enabled and Uh May 7 2012 setting up for cold reboot blah blah blah blah blah.

All those fans who want to, uh, afficionados who want to go in there and uh, check that stuff out by all means, do it. now. let's see if we can get anything out of this sucker. and uh I don't know what do we do? uh help games no no response question mark, no response, help, no response.

uh okay. um yeah. let me see if I've got maybe I need uh Line Feed or something like that. Okay I've appended carriage return now and still nothing.

um and go into the settings here and we can do like a pendant Line Feed and uh, stuff like that I Would have expected to get some response from this thing, but uh, no pin Carriage uh return and Line Feed help? No no idea. Debug no H no and I verified that the signal is uh, being transmitted with the scope there, so it's getting into it. but uh, it just doesn't seem to be uh responding in any way. So yeah, I got no idea I'll just leave it at that.

So there you go that's inside the new Agilant 34 61a multimeter and update for the over 20y old uh 3401a classic old instrument and they've uh, renewed it for? Well I was going to say renewed it for the '90s No, they've renewed it for 2,000 No, they've renewed it for the 10. What's this decade called? anyway I Have no idea. Yeah, H the 2010s? whatever. Anyway, they Have um yeah, they claim this is the new meter for the next decade.

Woohoo The true. Vault I Really like it? Um, design and construction is, uh, first class. so uh, and it's um, pretty darn cheap as well. I'm going to have to do a full review on this thing of course, which will, uh, be coming.

but uh, hope you enjoyed that tear down. and if you want to discuss it I'm sure there's no shortage of people who will on the Eev blog Forum being in the Ficado that they are, so that's the best place to discuss it. And if you like tear down Tuesday Please give it a big thumbs up and thank you to Agilant for uh, sending me this before it was released. Unfortunately They jumped the gun and well, it was released.

H they put on their website a couple of days ago. Fail. Anyway, catch you next time.