I'm here with Uh, Maz Acosta from the Ceo of Ness and he's going to take us way too long. Way too long. 1972. Exactly.

Yep. And he's going to take us personally through their production facility. Shorian, let's tell us about this: Uh, what's the history behind this? Uh, well. you probably don't want to hear all the history.

We'd be here for the next 24 hours. But a short brief synopsis. Okay, um. well, how about the the production area first? Okay, through a chain of events, we ended up, um, establishing our own facility.

And we did it because of problems we had with supply supply issues. that sort of thing. So this was an empty warehouse. Yep, uh.

We started with copper under the floor. It's uh, it's vinyl clad and then over the vinyl we we have a an anti-static solution. The idea is that it's uh, it's a, um, an anti-static environment. We are wearing anti-static straps.

We are wearing anti-static straps. Um, and everyone is Earth. Everyone that works on our boards as earth so we can freely pick up boards and not cause Vsd problems, right? Okay, we will start with the Uh. the process.

The Smt process the loading of components onto a board. They're um, we use Yamaha equipment because it's flexible equipment. It's quiet. Um, unlike the big Panocerts and Fuji machines, the turret machines that pound the floor and sort of like a mini earthquake.

The other beauty about it is that it's a very modular system so that if we need greater capacity, we split the line. We insert another little machine and we've got greater capacity, because we have enough uh, capacity at the far end of the line in our Hiller ovens that do the soldering. So essentially this one little line here is the equivalent of probably about 600 girls inserting components overseas, sort of sitting down at benches, inserting components onto a board and pushing him down with a paddle pop stick. Fantastic.

Which is what used to happen. What sort of investment does it take to set up a full line like this like four parts? It's in the millions. It's in the millions. Yes, it is.

But surprisingly our equipment is has been extremely reliable. Yep, Another good thing about Yamaha, but more to the point is that we once a month we, uh, we close the factory, everyone has an Rdo, and we maintain our own equipment. Um, all in-house So how old is all this gear? Um, it's at various sort of ages. The first lot of the equipment is Execom used to be an Awa um, company, right? So you actually bought it.

Secondly, we we bought some of that. A lot of that Uh Is has now gone, has been replaced. Um, but there's a couple of pieces that still remain along the way. We've bought um, uh, new machines like this one will.

Yup, just awaiting the process. So we have a board stacker at one end. So this stack of beer boards, yep, go into a into a feeder. First stage here and you can just see the process.

barely. Yeah, Um, it's silk screening the solder paste onto the bear board and that becomes your soldered joint. Similar to uh to a silk screening process, a painting screening process. Everything is automated.

so the board that board is ready to go into the next process. It's fully automated. Um, the computer knows where everything is so it's just awaiting the uh, the following process. Now each machine here is, um, uh, how did you put it? It's a full vision system.

So we have cameras that, um, that monitor the process. So boards are awaiting the first pick in place. Now this is what we call a chip shooter machine. This is one of our later machines.

It's rated at 17 000 components an hour to give you some idea and we're actually achieving about 15 000 components with optimization. Which is damn good is that inside it is. So we've got eight heads aside. Yep, you can see the um, the size of the components.

Well, you can barely see them. That's an O603. Okay, now that's the beauty about it. Um, number one is is that it's a full vision system when it starts.

When the process starts, you'll see that it's picking up components. It's going over that little, uh, see where? Yes, that light. Yep, that's checking to see if the component has fallen off or whatever because it actually picks up the components with suction. Or you know, the vacuum.

So if it should fling one off or it's it hasn't picked up properly. Process stops. Um light, red light comes on, operator comes along and sort of and sorts of thing. but it will not load a bad board.

Yep, that's the beauty of it. And yet does it have to go so you can fit all of your components on this one machine? Yeah, the view have to go through a second. No, we have multiple um depending on the boards. I mean, you know.

But what we do is we have multiple um. chip shooter pick and place machines. Um and they share the load and that way it just speeds up the process. You can see it can't automatically go into there, it will when it's ready.

Yeah, we have that split so that we can freely walk through here to get to your screens, our screens or whatever. Can we have a look at one? Um, we're able to pull one out. There We go. Yeah, Yes.

Yep. Excellent. Okay, you can see it's a nest of boards. You're right.

Yeah, yeah, it's a it's A. Boards are penalized. It doesn't just build one at a time and these. So how many different uh, products do you assemble here? Just your own products Because you do subcontract as well.

Yes, we do. Um, any ideas. Uh, many many hundreds. I won't say it's not thousands, but it's many many hundreds.

How do You juggle hundreds of different products on just two lines? Okay, the beauty about it is that the the types of sorry the equipment that we've got is highly flexible. and it's another good thing. Um, we have a, uh, a sort of. It's almost like a pit stop sort of process.

Where to change your line doesn't take us all that long. We have setups like this where we pre-load yeah, we preload the reels onto the trolleys we basically wheel them in and we we can change over a line. It varies, you know, depending on, on, uh, on the product, but it could be half hour or something it used to take with old equipment. It used to take anything up to eight hours.

Wow, that gives you some ideas so it's very flexible now. it's very impressive. All right, nice. And we and we need to be in an operation like this.

and in a country like this. And you run two shifts here. we run two shifts, we have run three, or we do run through where we need to. We've run weekends or whatever.

but uh, the capacity is surprisingly high for such a small plant. Surprisingly high. Um, you know. Like I was saying, we we build ten thousand boards for a customer contract customer a month.

a month, you know. And that's just one customer anyway. So um, we're on to the next uh machine, which is another pick and place. This one is not a twin head machine.

It's only a single eight. Yep, Um, it's just a waiting. the board coming out of that process being passed on to this. this one, which we'll see in a moment you might want to.

and then eventually once it comes out of this machine, it goes on to a Um under the third machine. Which is okay. Here we go, There we go. Almost missed it.

There we go. Also gets pulled by that. That's excellent. Okay, on to the the next pick and place.

it's picking up again. Relatively simple sort of components and relatively light mass components. So it's it's relatively high speed. Why didn't you fit those on the other machine? Then we're trying to balance the.

um, the load. This, this is a slower machine than that. Oh, I got it. So yeah.

it's right. it's all about balance here. It's all balancing because some components are heavier, so they have to move slower. physically.

Exactly right. Exactly. The third machine is actually quite a slow machine, but it is a it. It's our odd place machine.

It can place things like microprocessors or whatever it it's capable of, uh, orientating the, uh, the component, turning the component around 90 degrees, trigonometry if it needs to, or whatever. It picks up things like, um, capacitors, for instance that have fairly relatively high mass. Oh yeah, okay. those machines at that speed would fling off.

It's also capable of taking microprocessors out of a off a tray feeder. Yep, What will happen here is that it will pick up. In fact, there it is. it's picked up a ship.

Oh, and and it shoots. Oh okay, that's neat. Yep, really neat. So these machines, although they look relatively simple, can load virtually any type of type of component, whether they come on reels yes, in trays, or even on sticks.

Occasionally, where we have no choice We, it's not our preferred option because the tubes are only relatively short and they don't hold and you've got to keep changing them. Yep, Yep. But if you're running out of components sometimes and that's all you can get, that's it. Um, machines are capable of having and you've got a huge component store over the other side there? Yep, Again, we run a Kanban system.

Yeah, right. Um, and again. A Through a walk through. Um, so they're the.

uh, they're the reels. and those components are the ones that go on to those, uh, those trays. And does the customer supply their own parts? Or do you supply a lot of them for them Or no. We, um, for all of our own manufacturer.

Obviously, we do our own purchasing for contract customers or our subcontract customers as we call them. We will supply probably on average 80 or 90 percent of components. The only things that we ask them to buy in because we've been caught before, is is any components that are specific only to their own product. Um, so they buy them basically and we store them.

Sure, another slight variation of Uh machine. But and and here's an example of this: There's our tube. Yeah, yeah, yeah, they're very short. You can only fit like 30 50 for two minutes.

You know somebody's got to come along. Exactly The beauty about it is that we can run these two lines When everything's optimized. We can run these two lines with probably an average of about three people that move around between the two lines. Three people, right? three, three start for two lines.

And Smt is the thing that's kept Australia still in the game. Yes, everyone thought we were crazy when we set up this plant in October in 95. outsourcing was the catch. cry back then and um, but you can't beat having like being able to walk downstairs and solve problems Our engineers exactly yes.

And and your yield you were saying is like now we're well into that. it's 99. We're well into this and that's first pass. That's not.

That's exactly right. Very impressive. Okay, so the process is that they, uh, they there is a small inspection station that we use on some boards. We don't have to use it everywhere.

Um, then of course you have your Heller ovens. Yes. and they use a similar profile for most boards. Or do you always have to keep changing the private? Uh, it's a good question.

We do change profiles depending on the type of board. Uh, I can't tell you specifically how often that is without getting some technical assistance. but that's all right. Anyway, so the board's coming through.

Yeah folks, this is the Ceo by the way. I mean, you know, look, he, he knows his own gear. He's an engineer. So yeah, I have to worry about finance as well.

Okay, um, so the boards go through the oven. they're essentially soldered. Then the the. the components are actually just sort of stuck down with the pick and place machines.

Just sort of sit the component onto the board. They've got a sort of a sticky back to them and so they they just remain in place. They only have to remain in place through the process until they go through the oven and the oven, then solders that solder paste. it's already been pre-pre-loaded Um, to the board.

And then they come through the, um. the process and they go into the trays and they pop out into the trays here. That's it. Yeah.

Excellent. And do you do only single-sided load? Or do you do boards that have no, We two-sided But they've got to go through the machine a second time. So we load damn good questions. So yeah, so we load this the surface and it's an interesting sort of a thing in that.

Um, the components that are loaded on the bottom have to be glued down. So we have a gluing machine that we kind of, uh, did we bypassed? We bypassed that? We actually only because we don't do that much. double-sided sort of work. We only have a glue machine on one line.

Got it? So yeah, we bypassed it on this short slot. Yep, it's okay. So the boards are then in trays. They are in trays.

There we go. They're all ready to, uh, they're all assembled. And they've got some big thermal mass stuff on there too. Not just, uh, little.

no large thermal mass. Now some boards that come through the Smt process are 100 ready for testing, but probably most won't. There's still some degree of either hand insert or whatever or got it or even automated insert. but just these huge connectors on there.

Exactly. Things like that. Yeah, yeah, yeah. now things have changed over time.

When we first implemented this plant in October 95, we had perhaps 20 or 30 girls doing that hand loading. Such was the sort of the state of the nation on what was available in surface amount or might have been a cost sort of, uh, issue or whatever these days. you know, a couple of girls. it's mostly Smt, but you do have a wave soldier.

We do have a wave soldering machine and old school. the looks Simple. Yep, nothing sort of all like great, but this was the first of its kind in the country. really.

Do tell in that nitrogen nitrogen based. When we first installed this machine, nitrogen was just becoming the thing after the uh, the device of Cfc cleaning which we've already spoken about and uh and contamination issues that can occur which is a big issue on high reliability boards like you need for alarm panels exactly. You need environmental issues Yeah, contamination in flux where it can cause leakage across tracks and cause all sorts of issues in in rainy or humid weather. So you can formally coat any boards and products.

No, we we actually do, but as a rule we don't It's very little of our uh of our production needs to actually go outside um and conformal coding we learned the hard way is as not as simple as it seems and one layer is not sufficient either. Not even two or three layers, right? That's a whole other subject. Okay, um, this process. suffice it to say that this process here as simple as it might look, it's just one machine.

Prior to this process, a supplier, an overseas supplier of our boards when we were having our boards loaded overseas near centers, broke right contamination in the solder bath um, cause failure rates up around the 30 sort of level. that's you can't stay in business with worse. Worse still is is you can't find those sorts of issues in an air-conditioned environment because you know the humidity levels are low. It's only when you get outside.

It can't happen with this machine because we can use Um fluxes that are only about five percent as active as they would normally have to be, right? Uh, in fact, this this line is actually capable of fluxless soldering. We don't quite push it that far. Okay, um, fairly close. Fairly close right? Um, it's um, and partly because or mostly because sorry, because it's nitrogen-based We have a huge nitrogen tank outside.

it floods the chamber with nitrogen, gives us an inert atmosphere, takes oxygen out of the atmosphere. Oxygen is the killer. Oxygen is what causes the oxidization. With the, With the elevated temperatures, we the boards have to undergo a preheat cycle.

This particular machine is actually fairly advanced in that it's capable of reading a barcode on the board and you can changing this profile. It's changing its profiles Exactly. Very nice. Uh, meaning that if you have large ish boards with large thermal mass or whatever or different types of components to very small boards.

Um, it can change not only temperature profiles which are stored in memory and are displayed. you can see those yep on this on the screen. It also tracks your, um, the progress. Yep, that's going through the preheat cycle there that's going over the or it's about to hit the wave of the bath and we have.

If you, Yeah, we can. If you want to stand up on that one up there. I've done this before. Well, there's our molten bark in there.

The interesting thing is that we have multiple Um waves. Yep, oh, multiple waves. Okay, and you'll notice that as the board comes up that the wave sort of suddenly rises and here we go. It's coming in and the multiple waves prevent things like shadowing.

Yes, as the board passes through at a certain speed, there's a a a shadow where solder doesn't get on the back side. especially from um, from boards that that have, uh, this, sorry from components that are say perhaps through hole with uh, that with leads that are either that either poke through a hell of a long way or they might be a can for argument's sake. you know, with these little tabs to poke through and uh, it gives us much better yields. it prevents things like short circuits.

and yeah, that sort of thing. And there's our fully uh, wave soldered because uh, there's no components on the bottom. No, that's right. Yeah, that's the interesting thing.

and precisely that. where we have components on the board and they're glued and they're glued. Yep, they'll go through the wave solar probe still go through. So far, it's a fairly, I would say fairly typical except for this particular machine.

Very few people have. you know a wave solar machine is as good as this, but we really don't need it that often because most of our soldering is done in the Smt process. This this is probably just magic. Flying probe was in its absolute infancy when we bought this machine.

Oh soon after we set up the plant, it would have been uh, late 90s? Probably no less? 95? Probably 97? 97? Okay, somewhere around there 98 I could be mistaken, but I won't be too far off now. Every board has to be tested. At least we believe it does every single board one way or the other. We test every ball.

We don't rely just on on batch testing. now. everyone's always struggled with testing, from the bed of nails test to the uh, you know, to dedicated pieces of test equipment when we design a board. Uh, in R D, we spend almost as much time sort of building and designing.

Yeah, you have to. We saw this particular machine many, many years ago at a show yep, and it was being used to test computer boards at the time. There'd be an operator stand here, would open this lid, place the board, the board would read the fiducial marks and error correct, and then he'd wait for a couple of minutes for the board to get tested. It sounded like it had some merit.

We hoped that this process might be good enough to automate so that you didn't have an operator here standing there. Yeah, yeah, yeah, just comes in the boards. Just come in. We were wary about the longevity of the probes.

We've got six probes, and you know you can imagine these things working you know how many hours a day? So we tested one of these machines for three months. They were the company we, you know were, uh, that we were dealing with, were gracious enough to allow it to test and also to compare with another machine. This other machine looked way more impressive. It was a magnetically operated machine.

It sort of stood about this high that was. I mean, your heart said that's the one you wanted to buy. But ultimately, this one was way more reliable. So as basic as it might seem, it's been an incredible machine.

And what life do you get out of those probes? Probes? Uh, it's in the millions. It's in the millions of probes. Um, so we don't have to replace them all that often. Way, way better than we expected.

Now, what's the beauty about this machine? Number One: it's flexible. We take the Gerber files Um, that we generate in R D Um, when we're uh, we're designing the Pcb layout. Now, those Gerber files go off to a supplier to supply us their Pcb. We use conversion software to convert that information into something that the machine can understand and it's ready for testing.

It's fully flexible. We don't have to build the bed of nails. We don't have to. We don't have to write the software.

We don't have to. so it takes, it can take weeks or a month out of the process. It can save the money for it because those gold-plated contacts costs a lot of money. It takes and it's the labor for two.

We used to have a generated tester and that was one of these testers that used to suck the board down onto a bed of names. And the interesting thing there is is that if you ask anyone in the country that uses and typically everyone does, every all all manufacturers use those types of machines at least they did before this and you ask them what would they do if the board fails, they'd say well, you'd put it through again. Of course you sort of think to yourself, well, how do you know then that you know it was an actual failure or not? Yeah, but that that's just the way it was, You know, And I guess you took a little bit of a risk because the thinking was that it was more likely to be a little piece of flux or some sort of oxidization. Just a bad connection.

Yep, you know you put it through again if it passed. Everything was fine, but with this thing, it's a little bit more intelligent than that. What happens is that as these probes come down, they're doing it. They're doing a test.

They can do everything from uh, you know, measuring uh components, a resistors, um, a resistance, or a reverse polarity short circuit. All sorts of things. So your dozen three get a numbers profile? Yes, but it's a static test. It's not underpowered, but does it have the capability to power up and do not? this particular machine.

But we'll we'll get to that a little bit further down the track. But it's capable of doing anywhere between about an 80 and a 95 test of the board. and it can find things that you would normally not find. Um, and of course, this is a machine.

When we do our first off, we load we. We've loaded a new set of boards onto the machine. We've run it through the whole process. We get to this point and only at this point do we say good board all the right components because this would find find if it was otherwise.

Then we press the Go button. We don't build 10 000 bad boys. What it's what it's capable of doing is is it's nearing the end of this. Now there is one, two, three.

There's six boards there on that nester boards that it's doing. So okay. this particular one is the winner. It's going through to the past.

Fail sorter. Yep. the pass boards go that way that way. if it had failed.

now. Okay, we've got to switch back here quickly, because okay, now you're measuring the fiducial mark. Oh yes. Yep.

Error correcting, forward alignment. Yep. And then away it goes and away goes. So it would have popped out here if it, exactly, and we would have got a readout saying r22 out of tolerance or whatever.

Yeah, and does it physically mark the board? No, no, no, it just just puts it up here. It puts it over here and then then it matches the uh, um, are the boards? Are each board barcoded? Uh, yes, Right. So it knows so we can use the barcode. And then we get our first pass yields from this machine.

So we know without doubt you know what sort of yields we're getting. And we monitor those yields. And that's how slowly we've been able to get them well into the 99 region. So you're you're getting 99.9 something percent depends on the board.

Some boards may not be quite as good as that, but you know, But that's out of this process. We were probably five years ahead of any other company that we know with this sort of setup anywhere you know. And we're talking about companies Much much larger than our production facility. Um, so much so that down the track.

There were so few companies that even knew anything about this process of these types of machines that down the track. We, um, we managed to pick up a second machine at a bargain basement basement price and I had an option. Nice. Yep, so we have two.

Excellent, Terrific. How do you program your boards firmware wise? Oh, we use uh we we flash program a lot of boards. Yep, um, we also do a bit via a header cable. You would actually plug a header cable on would you? and it depends on.

Okay, right? So you wouldn't do it at the flying probe. You don't have a better nails that. Oh, some products we have better nails, right? Okay, we put the board on the boards of flash. The better nails jig right there pins on the board, the ics are pre-programmed in gang programmers and the light before they get on before they get on.

Or you can order them from the manufacturer already flashed. Yep. Okay and the way things are going, I guess. um, more and more.

Where I mean our latest, um, range of engineering sorry engineered products, but such that you can just download the program anyway and you can feel programmed flash upgrades or whatever through reports or whatever. Or the cloud. Or the cloud. Exactly.

So we're about to go and have a look at, uh, another one of um, Steve's favorite machines. All right. An Aoi machine. Automatic inspection machine.

We previous to this, we had a group of girls that had sit down and inspect the machine for anything for you know, things like joints or whatever and even and even little cosmetic things. Um, some of our customers are extremely finicky about things like, uh, alignment? You know, a pin header. Yeah, they didn't like pin headers. sitting at sort of 20 degrees or whatever.

the perpendicular you know? Well, that indicates a poor manufacturer. Yeah, for some it's important. So um, so to take as much labor out of the process because again, that's what keeps us in the game in this country. Um, we invested in in this machine, which was the latest of its kind at the time.

It has six high definition cameras and is actually capable of even inspecting solder joints to a standard. Top of that, you can train the machine. Um, you can see over here on this monitor this sort of magnification that you can get and you can train the machine to look for things. It could be anything from, um, you know, a the alignment of a uh of a header or whatever to such things.

As for argument's sake, we may have a chip that, um, a garden variety chip that we may have several different brands of. It might be a simple sort of, you know, Cmos chip or whatever, a gate or whatever. and we have Motorola and national or whatever. and they'll all have different markings and different exactly.

But in this specific board, we know that it really has to be a Motorola. For whatever reason, the engineers have specified that so we can check that you know someone hasn't done the wrong thing and placed the right part with the wrong brand onto a board by actually having it look for that. There are many other examples of what this machine can do, so you would bring the boards on their stats via a trolley over to this line. Well again, you can see that we've got an inline sort of a process here that it's it's actually coupled up with our uh, flying probe tester.

Oh right. Okay, so some boards we find there are certain things that Aoi is even better than probe at finding and some things that probe is better than Aoi got it. The thing is that we have both, so each board will get both. Whoops.

Sorry. if you want to pan around to the screen at some point and there we go. That's a nice even light too that it lights up. very nice.

Beautiful. Yep, I can stop a beautiful image and these six different cameras on there. Can it test six different things at once? Or it's just for a 3d view? Yeah right. it's actually a 3d view, right? I'll just show you what it does with the cameras.

So first up the board goes through and does a uh, fiducial correction yep to align everything to the all the components, etc. to the Pcb. Yep. and then we go through.

and then you look at a typical component such as this Ic. Here, the the um, this is in the top down camera view. it's looking at the component marking here to identify it's the correct component. It's also looking along here looking at the individual solder joints.

You can see where the arrow is Yes. And it's also doing a check of the lead banks to ensure there's no shorts between the leads. Oh yes. okay.

now if for example, we want to switch up and we want to get a good closer view, we can switch up onto the high we can inspect with the high mag camera. Okay, and that's the sort of issue. That's actually a separate camera with a with a high magnification here coming straight down on the top. We can do that then also too.

We've got angular cameras coming in from the sides. Yeah, so we can come in from this side. Oh yeah, Look, you can see the leads yep at a different angle. Yeah, come in from that side.

We can also come in from this side side and and come in from the back side. nice And then once again. then we go back to the normal thing so we can use any of those camera views to inspect what we we want to inspect. Is it testing all of those angles at once or does it go and capture them and then goes in and processes it.

It generally does all the down views and then comes around and picks a certain camera and goes through and does the the side views. Got it across the board with that particular camera. Um when it gets when there's component checking that needs to be or inspecting with a high high mag camera, it'll then flick to the high mag camera and go around and do the high mag views. And how long does the board typically take? Oh, that's a that's a tough question.

You know what depends on the ball? Well, this sport here, for example. Well, there we go. Um, all right. 80.

finish 82 seconds, right? Okay, 82 seconds for that. That's the nest of six boards. That's it. That's actually six board panels and does this? Is this Able to identify the barcode on the board as well and also tag the borders? Fail.

Yes, it does. Oh yeah, there we go. So there's there's the individual barcode, barcode, individual serial number for this panel, right? And then what will happen is the machine will store all the information that's found within its hard drive. Uh, the board will then come out.

It'll go to the lady here. She'll scan the barcode and whatever the machine's found to be of an issue. It'll appear up here on this screen here and she'll She'll work her way through those errors. Awesome.

What? The machine? What? What Necessarily the machine may find? Uh, is an error made? Not necessarily A machine has what's called false false calls? Yep, we basically can't. There's so many variables, so many variations that, um, you do. You do tend to get full scores, so it does auto identify the board when it comes in based on the barcode and load the profile testing profile for that board. No, we've already.

We've already manually. We've already pulled up. We do, you know, hundreds and thousands at a time? Yeah, so so we you know. All the conveyors are adjusted, the magazines are set, the the programs are loaded to suit that particular board.

It doesn't matter where you pick up a board. Yeah, it'll be clean. Yeah, there'll be no flux. There's no flux.

There's no cleaning. Beautiful. No cleaning involved or necessary. Um, and you can see the V-grooves You find that v-grooves are the best solution for most of you.

That's another interesting thing. Yeah, um. well. it.

Obviously we gain efficiency by having standard sort of widths and we try and design to standard sort of sizes or whatever for productivity reasons. But the V-grooving Years ago when uh, with other contractors were loading our boards in the very early stages, some of them were actually just snapping. The V-grooves are designed to be able to stuff the we don't do that and when the machine comes along and actually puts force on the board, they can walk if you don't design it well and well. What's worse is that we've come up with a whole set of design rules.

Um, for instance, things like we cut: we don't face components within five millimeters of an edge because of the stresses that you can get because of the separation of the boards. Now I show you something that, um, that's important. It looks pretty serious. There's our defaultization machine.

Well, what did that? What did that machine cost us? Um, we've got a couple of them. I'm sure that they cost us something like seven grand a piece. Seven thousand dollars for a deep analyzation machine? That's probably going back a ways. Maybe they're a lot cheaper these days? I don't know.

I think they're about five and a half. Six hours are they each play? It's worth five hundred dollars. Well, five hundred dollars a blade? Yep. But the point is is that we do this to eliminate the stress that would otherwise be caught.

Um, yeah, they can. Yeah, no. If you actually snapped it by hand, there's a lot of stress. A lot of stress you can actually.

Yeah, you can damage people. and believe it or not, even this can still cause stress within a few millimeters of the edge. And that's why we have our design rules. Yep, absolutely.

But it it's it's the processes that give you a good board. In the end, you know they give you the sort of quality. It's really the design is one thing and the you know the the manufacturer per se or the testing or whatever is another. But it's all each.

It's a sum of all of these individual parts because if the testing and the inspection's a feel-good thing. Yeah, that's right, but you you can't stick it in. but it's like building a house. Exactly.

Your foundations are no good, the house is going to fall down. If you get all of these things right, then you're almost guaranteed a good board. Yeah, absolutely. Or even worse, there's these components you can't test for, like capacitors, for instance, Exactly.

They can have long-term uh, failure rates. Yeah, you you can get a you can get a fine crack in the component, but you cannot see even with the fracture exactly and it'll pass. Have you had any major, uh, field production? like in field issues like that, like you know where, Like, I just snuck through all your tests hand on heart, hand on heart. I can say that since we set up our own plant here and it's under our own control, we haven't had a single solitary um, I'd say I call it major issue, right? Sure, you might have some little minor things.

Never anything approaching the the sort of disasters that we used used to be commonplace. I mean, we used to have almost a disaster or a month, but you know, I mean not right. Or in a major one, Perhaps a year. And how many boards would be involved in a major one? Ah, sometimes.

like I say, probably the most extreme was that 30 percent failure rate. But like out in the field, would you have ten? You would you would? No. Well, fortunately, we've never had that. We had an issue many years ago with a fracture in a in a capacitor.

Yep, Um, and that caused, uh, I'm talking quite a lot of years ago. It was probably it might have amounted to five or six thousand boards out in the field. Fortunately, the nature of the uh of the failure wasn't that critical, that it sort of it. It came to a sort of a recall sort of situation.

The capacitor was actually, uh, it was a filter cap. So basically, unless you were in a sort of a bad environment, it really, you'd never know exactly because you usually over engineer the filter caps. And another time we had a leakage sort of problem, but that was really more a component failure. Since we've we've been under our own control.

we haven't had any of those issues out in the field, and that's the last 20 years now. October 95. Firmly, we used to spend many minutes. Um, if it's a very complicated board, it might be five or six minutes.

Um, testing a a board under powered sort of conditions. So you know. And in the early stages, a typical control panel would be set on a jig of some sort and you'd exercise all of the inputs. Yeah, that's right, and look at outputs and you know it was a long-winded sort of process.

And how much about how much cost would that add to the product? How much would that like? Would it be ten percent of the product? Well, you only have to look at sort of with your overheads to cover your overheads. I mean, you know every minute probably cost you a dollar in overheads. You know it's It's one of those sort of situations these days because of our process control, because of these other sort of automated test machines. uh, our processes in place.

We only need the sort of testing that we need at Um. Underpower is the sort of tests that might be used for instance, to test amplifier noise, uh, voltage settings, or that sort of thing. Or perhaps, uh, for an Rf sort of thing where we still have to we don't have to do this and and or, um, say, frequency or whatever tune for frequencies. So typically our tests might be anywhere from 10 20 seconds up to maybe half a minute or whatever.

we test for things like even a simple product like a Pir detector. People can't believe we actually even built a Pir in this country. I, I know it's stunning. Well, we do, and it pays you to do that financially.

or you just like to have control over a bit of both. A bit of both. I mean, um, our our products are renowned as being as bulletproof as they they can be. Um, people expect to put up a detector these days and not have not touched it for 10 15 years.

people expect a battery in a in a wireless detector to last for eight years and that sort of thing. So we build our products. We engineer our products. We build our products to last.

Um, we replace them. Now, if if a product goes 4k under warranty or whatever we bin them, we don't bother repairing it. So you know in the long run we get a payback that way. right? Um, this little tester is is a home brew machine.

We we tool these. Um, I don't know what material it is to be honest with you. But yeah, that looks like some sort of Dell room material. Yeah, something like that.

We have it all tooled and we we, uh, we have one of the engineers sort of design the. um, uh, the We. In fact we have a dedicated engineer. Yeah, that is off battery life.

Yep, Yep. Um, the product goes in, the door closes, and we even check for the balance between the two halves of a dual element sensor. Oh nice. So you know how a Pir sensor has nice dual elements.

We simulate the temperature of a moving target at a range. Oh, there we go. There's our board. There's our bit of nails.

Yep, flip the dip switches. Okay, this is our latest luxe Pir, which I'd love to show you later because this Pir actually has a a white light that turns on automatically at night when you're moving. So in the day time you just get your walk test light right at night time. Um, when you turn off that last light and you're trying to move from your kitchen to your bedroom without walking into walls, Right on comes your nightlight with movement with movement.

and it's just enough to give you sort of enough light just to maneuver without wasting. It's obvious it sounds obvious, but you know, wireless product. Believe it or not, it's not that simple. Oh okay, we get ideas because it's battery powered, so your pulse with modulating the lead at some ridiculously low current.

We can't divulge all of our secrets. Suffice it to say that yeah, you're driving it very low. We're getting eight years battery life. We have two batteries, one for the Pir and one for the the sensor light so that you know if if it should sort of go flat or whatever.

it won't impede the performance of the Pir detector. and we estimate eight years with the light coming on once per day or something. Oh no. Multiple times in fact, I have had out.

This detector is now from the early prototypes I've I think I've had mine in at home for about two years or whatever and I constantly monitor the battery voltage. So far, it still looks like a brand new battery. So you know. I mean, I know it's got years.

and what is it? Got a Cr123 Lithium battery? Or uh, I wish I could tell you that because I'm not okay with the batteries. We probably don't even have them down. No, you probably don't. We will.

Ah, yep, that's right there we go. Yeah, it is a Cr123. I know you're stuck. I know you're just now.

Yeah, yeah. okay. but I mean that's that's that's impressive. These and you do hundreds of products.

I just hundreds and hundreds of products. and we do things like, uh, in that little room, we have ultrasonic welders so you know those, um, uh, the pendants for media alarms. You know, the elderly you use. they have to be waterproof.

Yes, people wear them in the showers and that sort of thing. So we actually, we pioneered a process where we're using plastic moldings. Obviously, we use a rubber button, but the rubber button itself is not either glued in yep, or press fitted into the plastics. It's actually an integral part of the molding, so it's actually welded into the plastic molding.

Exactly. Yeah, nice. It's part of the process. Yep, today, that's not really that much a big deal, but when we first started doing that, probably 10 years ago, it was.

it was New Zealand company. Can we have a look at the ultrasonic bottles? Yeah, they're not actually operating at this point. but right, you can have a look and there's my board there. Okay, they're making the Uh D16x.

Another little thing that we do just to try and be as efficient as we can. We have a Carton Live system right to try and keep as many sort of boxes and things out of the, um, out of production off the floor and try and keep it as dust free as possible as well. because, uh, cardboard creates dust and everything. Yeah, so we feed components um, boards and things in by the Carton Live direct from our store.

A component store? Uh, they're picked up and they're sort of unpacked generally here and then. and I I haven't heard of that before. Carton Live is that that one is set up obviously to do our radio keys because even our standard key fobs are fully waterproof, not whole, and have been since 19. Oh, probably the early 90s.

I didn't know that I've just got a crappy one from Mcm which is not not waterproof at all. Oh, we can't say anything about our opposition. They're all good people. Yeah, they are.

Okay, so yeah, um. that sort of thing. This one's doing our medi pendants. Yep, I can tell by the shape.

Yep, um. House is one of those ultrasonic welding systems. I'm not sure what they go these days. There's various types, but probably probably these particular machines are probably.

oh, I don't know. Eight to ten thousand new? Okay, and uh. and of course, then you've got to set up your little. You've got to do the custom jigs and everything else.

Yeah, Yep. nice. The Horns object. This is an Object Eden 350v.

Tell us about this one now. Well, you know, 3d printers these days are really not such a big deal. But the thing is that this is our second generation of product. I mean, we've had a 3d printer for so long, right? that back in the days we bought one, we were using external resources or an external company to produce our prototype plastics for things like we designed.

And there's the sensor housing. uh, sensor housing or whatever. Yeah, um. and you can do great things like this chain nail straight out of the machine.

Get it out of the machine, straight out of the machine. and this is like a six digit price machine, right? Well, um, they've come down dramatically. Like I say, uh, from a million dollars I think. We bought our first one at about a quarter of a million dollars and this is our second generation.

has greater capabilities. It's a better machine all around, but it's still an object. and uh, we probably paid under 200k for this one. But and we've had this for years.

so that's how long we've been doing this sort of stuff. And we do it because you know, I mean, look, we, we just do. We do everything in-house We've got full capabilities. So from concept, uh, to industrial design, hardware, software, we do everything.

Um, to out the door. Um, the importance of this is that we generate our Cad files. Our Industrial Design Cad files. Those files go off to a tool maker to produce the tooling for the housing in parallel.

with that, We could be running, um, several different Um versions if you like, and we can be testing everything from clipping actions to you know, we can be putting real boards in there, doing walk tests in a Pi in the case of the Pir, or whatever. We can clean it all up, paint it all up, and for all the world. It's like the finished product. So we can be doing our brochures and everything.

So we're doing things in parallel rather than in cereal. Which is the way things used to be done. Sure, Yeah, so time to market could be three months. You know, a three-month difference.

And when you're hundreds of products and you're always developing new ones, exactly. Yeah, it's important. And how? Uh, with your purse sensors? Who does your uh, Fresnel lenses? Fresnel? Uh, little things again? Um, our lenses. We have stuck with the company called Fresnel.

Oh, they were. They are the authority in lenses. Now a lot of people don't understand this and this is why. I mean you look at our Pir detectors for argument's sake, and there are of a certain size and and form and fit.

And they cost perhaps a couple of dollars extra. It costs us something like about four five, six times. In some cases, the price for a Fresnel lens compared to the lenses that we can buy out of Southeast Asia, right? Difference being that we get something like about 30 percent greater captivity. You know of, uh, energy captivity? You know in each being in each, Yeah, and that translates.

How does that translate? It translates. If you're in, say, uh. In in the field of whether it's Tv or whether it's two-way radio communication or whatever, everyone will tell you that the antenna is everything you know. The greater signal you can get in, the less gain that you then can use down the track, lower noise, it translates to lower noise and greater and, and, uh, greater immunity to other sort of airborne interference sort of signals and Pirs.

The biggest single problem with Pirs in the early stages was definitely it wasn't so much environmental. A lot of people didn't actually realize this. It wasn't It wasn't that it was detecting, um, full sort of, uh, heat signatures, moving or whatever. it was in fact, Rf.

Oh Rf for a number of reasons. One because the sensors were in the early days, they weren't as they are today. They're Rf hardened and whatever, but predominantly because you're playing with signals that are micro volts. Yes, and you're trying to amplify those signals.

It takes bugger oil, especially when you're hanging these things off a line, off a wire, going back to your control panel. The amount of Rf that you pick up at, and and it doesn't matter how well you try and filter these things, right, so there's no point trying to fill it. No, because you may, you may be very successful at filtering out certain frequencies, but you know it's going to be tuned. That bit of wire is going to be tuned with your electronics.

At a particular wavelength, something is going to get through. and these days, let's face it. I mean, you've got. you've got so much airborne stuff, it's crazy, let alone the other sort of things.

You've got electric motors, that sort, of, you know, that switch on and off, and all of these sorts of things. And we proved the hard way. When back in the installation days, you know that, um, that was the biggest single factor that so we tried to develop products that would be inherently, um, more immune if you like. And the lenses was one of the key sort of parts of gaining that reliability out of the Pir detector, so you wouldn't have to gain it up as much you want to susceptible to the Rf.

And yep, it all makes sense when you're well. thank you very much for the tour! Nice tremendous founder and Ceo of Yes! Thank you for coming. Thank you for your time. Thanks mate! You.