Hi check out what I scored. It's a HP 35608 Dynamic Signal Analyzer or DSA or sometimes known as an Fft analyzer in the industry and I can see myself There we go in the screen nice big CRT screen. This thing is an absolute beast and uh I've always wanted one of these for the home lab and uh, never had one because they are quite pricey especially if you get one new and working. but this one is faulty.

So um, hopefully we'll be able to uh repair this thing and get it back on track because I have been told that uh, it is fully working. apparently apart from the screen, the screen doesn't uh work. For some reason it went bust. Um, apart from that, it was working beforehand so you know I don't know.

Anyway, we'll find out. uh, hopefully we'll get a decent Uh repair video out of it. Could be be very simple or could be very complex or could simply be uh be Beyond Economical repair perhaps. So what is a dynamic Signal analyzer or Fft analyzer? Well, it's a pretty uh specific instrument used in the Uh sound industry, the vibration, uh industry, the acoustic and underwater uh sonar industry and for measuring uh mechanical and uh plant vibration and shock response and all that sort of uh, low frequency stuff which you get with mechanical.

any any sort of mechanical system or you know, acoustic, uh type, you know, underwater system of course I come from the Uh seismic Marine underwater background and these things uh DSA was as common as a general purpose oscilloscope in every lab and on every bench because they were just so incredibly useful. Uh for measuring the performance over frequency of very low frequency signals. Basically Uh DC up to 100 khz pretty much is the frequency range of these things. and they've got uh, usually they've got at least Uh dual Channel inputs and a signal uh source as well which can generate sign and noise and stuff.

and the Uh two wide dynamic range inputs. um, you know, they've got various gain, you know, really high gains in there, sort of low noise, uh stuff. Not absolutely. you know, industry leading low noise stuff, but good enough for sound and vibration and mechanical and shock response.

So you'll hook these things up to Uh accelerometers and things like that for doing uh drop testing, vibration testing, finding vibrational modes on circuit boards and things like that. I've done tons of testing like that with these Dynamic signal analyzers or Fftt analyzers and you can get all sorts of real, interesting Uh parameters. With this thing, you can get uh, crosscorrelation between signals. You can get coherence between signals, which is a very important concept.

I could do entire videos on those, but anyway, these are very specific Um instruments and well, hopefully we can uh repair this one and they are a bit of a beast. I Mean, check this thing out. Already taken the cover off it and it weighs an absolute ton. It is enormous and uh, well.

I don't even know how to get into the CRT in here. so um yeah, it's got some clips on here I don't know I've never actually taken one of these apart I've used them for like you know, more like 15 years or something. I've used these things I never had an opportunity to take one apart because they are. They were always calibrated.

they had the calibration stickers on them and if you took them apart in the lab, there would be hell to pay. So um yeah, this is great I finally get to uh, look inside. well no, sorry I have looked inside not one of these HP ones I've looked inside other brand ones say um, onos, sooki and uh, if I'm pronouncing that correctly and Rockland and um Stanford research, uh, make one I think it's the 785 which is uh uh, a pretty good um DSA and uh yeah, there's a whole bunch of Niche companies making these, but uh, pretty much um, you know h P Agilant are still the duck's guts at them and they still manufacture them although there are a lot of PC based ones these days cuz pretty much all it is is basically uh, you know, front end gain on here with you know, a wide dynamic range analog to digital converter and the source of course the signal audio signal source and then the rest is just you know, digitizing that and processing and getting the Fft result on that. So a lot of the more modern ones are done on PCS of course.

So you get like a little box which plugs into your USB and you know that's really all you need so you don't need the huge um, you know, processing built into the machines. although having them stand alone like this is rather convenient for lots of uses. So anyway, I might have to actually read the manual for this sucker to uh, figure out the best way to get into it. and uh, there's the power supply.

It's telling us our voltages and uh, you know, the first thing you do when you're at troubleshooting something like this of course is measure the power supply vol and we can tweak all the uh adjustment pots through there without having to take this out. But it is quite a beast and you can see the two are shielded. Obviously, they got uh two shielded uh, cars like this for one for each Channel and the front end stuff is shielded here. and then they've got the more you know, uh, digital type, uh stuff up the back here, all connected in a classic arrangement with ribbon cables like that.

and there's our processing board down there which will generate the signal for the Uh, do all the processing and uh Drive the screen and everything like that and there's no, there's actually no uh bus board down the bottom like that. All they do is sit in, sit in connectors on the bottom like that and they join them on the top instead of having your more traditional motherboard Arrangement down the bottom just saves an extra board really. And this uh Beast of course has long since been discontinued. They do, uh, sell and upgrade ad model to this under a different part number still.

But the good thing about it? you can go to the Agilant website and download the full service manual complete with schematics and everything. Fantastic. So you know that makes makes servicing and repairing these things so much easier. Uh, this one doesn't really have a date code on it visible um anywhere.

but uh, based on some of the chips, I can see through the top there some of the digital logic. uh, it's dated around about uh, probably late 1990. That's a start anyway. I've taken off a whole bunch of screws on there, able to get that back panel off.

still haven't read the manual? I probably should read the manual to figure out the order that I need to take this apart because this uh CRT up here I suspect uh you know is not easy to get into because it looks like the shazzy wraps around over the top here. So H I don't know? there you go I Didn't expect to find another massive digital board under here which goes almost you know 3/4 of the way to the front There and that just contains a lot of Uh 74 series logic and uh uh, some unpopulated memory there by the looks of it and uh, it looks like we got a bottom. The bottom board down here of course is for the Uh Gpib and the Uh is that serial interface. This is really quite nice.

Look at the Um metal Emi tabs on there and it looks like this board maybe slides out. Ah, it's tough, but yep it does. Beauty I Don't know why I'm taking it out, but I just feel like taking it out. Why not? There we go.

I Didn't see that it's got a uh rotating, uh, pull ring there that you can get your finger in and pull the board out. didn't even notice. Neat and well, there you go. I mean I Thought the processor board was uh up in here, there's our huge, uh, classic Mc68000 processor and uh, all the miscellaneous stuff to go along with that.

So obviously that is the main processor board down in there. and uh, the board that was on top is the main memory board. obviously are only like half populated here and almost all of this sucker is made in the USA All the main boards and everything else is except for the CR unit up here. and here you go.

that's a Matsu shter distributed by Panasonic made in Japan There you go, It looks like you know completely. Um, self-contained module in there for the uh CRT so you know, uh, really? um I the fault is most likely to be in there. you know I suspect because uh, there's lots of analog type stuff there to go wrong. I mean you know, the digital processor and all that sort of stuff.

Unlikely to. Well, it's the least, uh, likely to be foldy. So yeah, probably something inside that module. I've just got to figure out how to get that damn thing out.

now. this power supply modu is, uh, rather interesting. I Love the modular way it just flips out like that with these ribbon cables. That's the most interesting thing about this bit of Gunk on there bit of dust from 20 years.

This thing is, uh, probably been in service Ender yeah, interestingly uses these um standard IDC ribbon cables to Output most of the power so that's rather fascinating. and I Originally thought that, uh, this was some sort of power connector going off. um, maybe down to the CRT module which requires 12v uh DC at 1.4 amp down there. but uh, no, that didn't quite make sense while they're using Idc's for everything else and just that one going out really.

and uh, and it looks like it goes down to the front panel. so that is, uh, most like the the front panel uh switch. So that's just a, uh, a soft switch. uh, presumably it doesn't look like a Mains rated cable or anything to, uh, switch the power supply off.

So what you've got here is most fascinating. This main ribbon cable coming out of the power supply just goes up under there and then pops out here like that you can see it wiggling there. and then that is the power for all those boards. I Mean you know, first glance, you would have you know sworn that uh, data going across there.

Well, there could be some data, maybe, but it's most likely that that's just all dedicated to power. And of course, we've got some signal. uh, going across the Uh from this looks like a controller board for the two channels over to the two analog channels. and this one here is actually a data cable which then just wraps over there and goes down the side of that case.

Rather fascinating. And the other ribbon cable here is, um, obviously just the power for the main processor board down the bottom here. Fascinating. And if you can see down in there a whole bunch of test points on the PCB little nice little hooks I Really like those, um, just you know, soldered uh Tinder wire hooks on there.

got power supply and all sorts of stuff and there's lots of those including over here on the Uh analog Um boards as well. Fantastic test points which we can hook on to. and uh, here is our main d cable going over to the CRT that looks like the only thing going over I'm not sure if that carries power and the data for the screen as well, but yeah, um, the nice modular construction like that at least allows us to troubleshoot. I mean we can just, um, you know, probe the signals coming out of that IDC header down the bottom and see if we've actually got all of our required data going to the CRT module, so we should be able to narrow it down to that pretty quickly if you look down at the main Pro processor board down in there.

classic, uh, multi-layer uh routed construction for all this dip packaging very typical of the time. I've uh explained this in previous videos, but on the top layer here, you'll pretty much see all of the routing lines going in this. uh Direction like this, the vertical or horizontal Direction depends on how you want to look at the board and then the uh next layer inside you can see this is a multi-layer board. oops, and you can see the tracers going in there, the darker ones down in there, and they're generally going to go in this direction like this and you can see that they're just right angles like that.

So you know clearly. this thing has, you know, most almost certainly been Auto routed I'd say and uh, you know, someone's manually placed all the Uh chips, so there's some smarts going into that, but uh, then, they've pretty much probably let the auto router uh, do the job. This was pretty, uh, typical of the time because as I've explained in previous videos, you know all this stuff doesn't operate at high speed. You know we're talking.

Uh, what's the main Crystal down here, you know it's probably like 8 MHz or something like that, you know 4 MHz What does a uh 68,000 run out? like 8 maximum or something? So you know, like under 10? MHz Usually not very quick at all. So really signal Integrity is not a huge issue. Multi layer board. It's got all the bypass caps it's got.

you know, tons of bypass caps. It's got the ground planes, everything else, so not a problem whatsoever. So it's all digital circuitry. You can just let the auto router rip.

I Mean to manually route these by these sorts of boards by hand? I've done it. It's a pain in the ass, but this is one where it's very suited to an auto rou. The algorithms are really easy. When you've got one layer, you know, mostly going in One Direction another layer going in, another one.

you've got ground power, light planes, and all that sort of stuff. Auto routers Uh, just the auto routing algorithms. just eat that sort of stuff up. Really piece of cake.

You can see exactly the same thing on the memory board as well. On this top side here, almost all the traces running in this direction. Yeah, you know you got the occasional you know Wiggles going across. but generally you know the auto rout is told to give priority to you know signals in that direction.

So you're not going to find a trace going from here right down to there cuz it just ruins that entire layer. so the auto router is only going to know. Oh okay, I can only jump. You know, a few handful of pins, half a dozen at most in both directions.

Like that. Looks like we got uh, two traces between um, pads up there, which is, woohoo, you know, pretty big. I Mean you know, the big decision back in those days was oh well, does my PCB technology, uh, manufacturing technology and uh, uh, Trace which allow two traces between pads like that? There we go. You can see the two traces in there if it does, you know.

Brilliant. But back then you know that was a huge, huge decision. Not a big deal these days obviously. but uh.

anyway, on the bottom side of that, you can see that they all go vertical like that. So this layer, of course, there's some horizontal stuff on here because these are memories like, uh, maybe you know you would have told the auto router to, uh, you know, just route out. uh, the memory first or something like that. or give priority to that layer in that direction or something like that.

But as you can see, everything else, this or part of the circuitry or the miscellaneous stuff has been given priority in that vertical. Direction Here we go. I Just got these clips out here. still haven't read the manual? of course you know me.

um and bingo. There we go. We've popped out our front panel. we can see our CRT and it looks like yeah, I take out the screws here and this CRT module is just going to pull out.

That was really the only way it was ever going to happen. Check out the RFI shielding they've gone to the trouble to do on this thing. There's the there's the front panel. but look at this.

it's got. Look, Look at that gold wrap over that conductive wrap over the back of it. like that and all the nice little RFI uh, you know, tabs like that. It's just just beautiful.

They've really spared no expense there. They've really gilded the Lily and check that out. They've gone to the trouble to put uh, what looks like some sort of, you know, anti- glare filter or something like that on the Um over the front of the CRT. It's not a polarizing, uh filter because, uh, regard.

regardless of what angle you put it at, I can still see my LCD watch. There you go and check out those RFI tabs. Look at that, uh, basically connecting this, uh front panel board through onto the main Chazzy down here. Brilliant.

Actually, on second thought: I shouldn't take those out because that's taking the CRT out of the internal uh metal box. So really, what I want to do is there's I think there's a couple screws on the side and a couple on the top I think taking those out, this should just slide out. although that should pop out of there and that folks is a really nice modular design. Look at that.

I mean that is fantastic I Really love that we can now take that apart and uh, work on that on its own and of course we can. Uh, this cable looks long enough or if it wasn't, we just extend the cable to go in there and we should be able to probe and operate uh this while it's all connected and the power supply is uh, still connected to the thing and we should be able to hook up the uh keyboard. You know, the keyboard just plugs down into there and easy to access. I like it and a quick glance through the service manual for this Beast It doesn't show any schematics at all for the CR T unit.

so I don't know. maybe they didn't have the rights to do it or they just deem that that wasn't necessar. they only um, did the schematics for the HP part of it which is all the rest of the circuitry up to the input here now of course I still haven't uh measured anything at all on this thing and uh I probably should at least do some uh uh voltage checks on the power supply as I said. but I can't help myself of course I've got to pop this uh CRT system open.

although I did switch it on right I I switched the thing on and some of these screws are hard I did switch it on and uh, it did send it like it was accepting the button presses um and it was beeping as I remember and I could hear the input relays switching Auto ranging and stuff like that. So just as I remember using this so it's um, that seems to uh, tie in with what I was told that uh, this thing did work apart from the CRT So I'm assuming that the power supply is correct I mean it's got. you know it does have uh 5, Vols and 12 Vols and plus - 15 Vols and 18 volts and stuff like that on it. but uh, I suspect that's not going to be the reason why the Crt's fail perhaps although I could be wrong.

But and of course you could argue that before you even measure power supply voltages, you look for obvious signs of damage and smell as well. Use all your sensors and uh, so it's worth just popping the lid off a for curiosity and B just to see if there's anything obvious you know Is there any blowing caps? you know? are they bulging it? Look like they're leaking? Is something? I don't know. uh, broken or something? Somebody dropped the unit I don't know. Who knows.

but uh, definitely worth taking a look inside. Let's crack this thing open now of course. uh, dangerous as it warned you on the back. Dangerous voltages are on in here, so even when it's powered off, they can still be charged up.

So just be very careful. You got to know what you're doing and uh aha. I see the problem straight away there There we go. This was worth looking.

Um Can Anyone see the issue? Look at that the neck board here and touch that the neck board has come off. Look at that. it's just fallen off. There's the uh socket for the tube.

It looks like the pins are all okay. Jeez. I Hope that's the only fault that would be fantastic. Um, troubleshoot the Yeah couple of little bent pins on there.

maybe I I won't try and bend them back. I'll definitely just try and stick the socket back on and uh I don't know. Vibration? Maybe um has caused the board to fall off. The board.

looks in good. Nick Um, anything else on the on the board. there is a fuse. There's a fuse down in there.

It looks intact, but I will measure that. So there it is down in there, it's socketed. Very nice. A Little M205 glass Jobby and uh caps look good.

Diod. nothing seems broken. although those Doos look funny because they've got like a purple, uh silk screen on them. It almost looks like they're burnt, but uh, um, they're not.

So uh, and of course, uh, there's a reason that they have, uh, the protection around here. nasty voltages straight out out of there. You don't want that. but apart from that, sorry, it's hard to get the camera right in there.

but it does look pretty good. and of course, well, it's definitely not going to work with that Bloody neck board hanging off like that. So I'm going to uh, plug that back in and then just fire this thing up and see if we get lucky. Oh, that'd be awesome if that's the only issue.

I Can't be that lucky though. Surely Murphy's got to get me somewhere and that pushed back on there very nicely. No, no issues there at all. So hopefully, um, we'll power this up and we'll see what happens Here We go.

let's give it a go. Let's give it a bow. actually what? I've done, which is a little bit more convenient, just slide the CRT back in there and uh, hook the power switch back up and uh, just have the front panel in position. So here we go.

A now come on. can't be that easy. Easy Murphy Will get me. it's powering up.

Powering up these things can often take a bit of time to. hey, look at that, look at that. We're in. We're in.

but look at how adjustment. oops. Um, our rotation is a little bit out. um no, it's flickering on camera there.

That's just the power on Testing Progress Copyright 1988 CET Packard Corporation But it looks like apart from a trace rotation issue, we have a working DSA searching for application. blah blah blah blah hey, that looks good. There we go. Calibration in progress: It's always doing itself Cal I Can hear the little read relays clicking.

You probably can't hear that, but trust me. I Can hear those a mile away and we have folks a Workin DSA Oh well, presumably woohoo Fantastic look at that oop, sorry, gone out of focus there. a bit beautiful. Bob's your uncle now of course.

that is, uh, very much a physical uh rotation uh issue in this thing. It's not like, you know I can just, uh, tweak the uh pots on the top and and fix this sucker. Something has, uh, physically happened to the uh rotation in this. CRT So I think we're going to have to, uh, take it apart again and take a closer look.

but hey, at least it's working I just turned the brightness up a bit. Doesn't seem to be any particular, uh, burnning in the Uh screen at all. Um, so you know it looks really good. So let's have a look at what we got here.

I got the lid back off and uh, proba. you know this isn't real. This issue really isn't surprising considering that the uh, neck board, uh, fell off. You know, neck boards don't just, uh, fall off on their own.

It's got to vibrate loose or be, uh, shocked, loose or a combination of both. I Don't see any physical signs of trauma on the uh main case of the unit. so I don't think it's been dropped. It's possibly just uh, vibrated, uh, loose over time and eventually just, uh, you know, popped off with the springiness of the uh socket there.

perhaps? but uh, let's have a look in here and oh, that. Cable's a bit annoying, but uh, hello, what? hey hello, hello. that assembly. There looks like some sort of magnet assembly is I Take it, that's not normal.

I Noticed that it was out here. It was protruding out here and sunken down in there. so something has happened there. I'm not sure if some other something else is broken off.

I Mean, there's nothing in there. so unless uh, so, there's nothing you know rattling around in there. so there's nothing loose, just a plastic ring assembly with uh, uh, possibly uh, magnets in there. So that's a you know, that's not like the focusing coil or the uh, uh, deflection uh coil or anything like that.

You can see those in there, but uh yeah, that's just hanging around flapping around in the breeze there. So and well. so let's have a poke in there. that's not really anything to do with the rotation as I expected.

So um, it's the physical, uh, assembly of the coils in there. we're going to have to get in there and, uh, manually rotate those suckers. Looks like we're going to have to get in there and uh, rotate our coil assembly and fix this sucker up. What we have to do is undo this clip in here and rotate the entire uh deflection assembly around the shaft there.

so that will fix, um, our rotation issue. So um, really, you know that's obviously, uh, come, uh, loose or something like that during the vibration or whatever event, um, caused the neck board to fall off. So really easy fix. By the way, when you're tightening these uh things back up, don't do them up too tight, you don't want to crack the neck of the tube down in there.

and if I hold my tongue at the right angle and just, uh, poke that board in there. Bingo. We have ourselves a nicely aligned screen. Oh what? Oh what? You know, it looks reasonably good anyway.

and because we've got some dead spots on the CR T around the corners here, then uh, we can pull in. uh, the horizontal width. Back in with the uh, horizontal width control which is, uh, that puppy down in there so we can adjust that after it's all uh, back in, not a problem through the uh adjustment controls on the top of the case. now I was able to uh, uh, find out where this magnet assembly goes back in.

There's a little uh key slot up the top and down the bottom. so I was able to uh, push that back in and then I put some silicon back in there to uh, hold it back in place. obviously it's uh, uh, you know, broken loose or come loose or something and that also, um, had effect on the uh deflection uh coil assembly around here which of course is what caused that um, uh, rotational issu which we've uh, fixed up but uh, I don't exactly know what those I mean I'm not big on my uh, CRT um, you know, systems, uh, knowledge, but those um, permanent magnets there four of them around there I presume that their permanent magnets probably um, helping with some sort of focusing I mean I don't think you know they're not like, uh, convergence magnets. uh, for example, because you'd only get those in a color CRT and of course this is a mono CRT so I might have to do a bit of research on that one and uhuh, see what they are.

But if any anyone knows exactly uh, what their uh purpose is in there, then uh, please let us know. Doesn't seem to make a huge difference. Uh, regardless of the rotational, um, rotational position of them or you know whether not, they're skewed at some angle or something like that. um, does I can't see a visible uh difference on the image and I've uh, just quickly reassembled it just so I could have a decent uh, play around with it and you can see the the dark fringing around the corners of the CRT still.

so I still have to, uh, deal with that I'm not sure what's gone wrong there, but anyway, I wanted to test the rest of the functionality and uh, as you can see, this thing has um, extensive internal uh self test I mean there's a TMS uh, 320 DSP processor in this thing. There's Gat Arays that uh, uh, DSP processor is probably I don't think I saw it on the board under the horizontal board, under the CRT so it's probably on the big uh vertical board we saw in there doing uh, doing the DSP goodness. Of course you know you do all that in one processor these days. Uh, you know basically to uh, Drive, the CRT and everything else and and handle the gooey and the keyboard and the coms and and to do the DSP too.

But anyway, it's got a ton of uh, internal self test this thing I love it all these Gatoray, all the ADC everything. everything passes Beautiful. Look at that so and we can set it to our Loop test as well. So uh, that's pretty good.

It's got a whole bunch of stuff and uh, you know it. it looks as, uh, claimed, you know, um, pretty much 100% functional although I yet to stick a seetal in it, but geez, you know it's looking pretty good apart from the uh, dark fringes on the CRT there and unfortunately, with that horizontal width uh, that here is about as close as I can get to. if I keep going, it goes back the other way. So there you go.

that's about as good as I can get. unfortunately. Um, with that, so you know I I can still I can start seeing the negative uh sign down there on my number down in the bottom corner on the vertical axis so it's not too bad. You can still, uh, read the menus of course, but would have been ideal if I could squeeze that in a bit more then I didn't have to worry about the uh dead spaces there.

And by the way, if you are adjusting uh, the these things use one of these um, non conductive, non-magnetic plastic adjustment alignment tools, especially for that particular control which is not just a Uh pot, but it's actually a Um slug tuned inductor because if you use a screwdriver and stick it in there, it's I'm not even touching that you're going to affect it because that's going inside the Uh coil itself. All right. So let's just do a quick test to see if it works. I've got the source hooked up to Uh Channel One here.

If go into Source let's just set random uh, noise and uh, the levels. Okay, just give us some sort of level because if you get nothing and you press start like that, of course you know we're down in the noise floor down here, you know - 85 uh Dbv. So um, let's plug it in and well, let's start that again. I Set it to 100 averages.

so as you can see, we're getting ourselves a pretty Flat Line there which is what we expect after average in all that random noise. So it's working over the full frequency range which is basically our DC to Daylight or what daylight is on a Uh DSA Dynamic signal analyzer 100 khz. Woohoo! And if we set the Uh Trace type here to instead of Uh frequency, we can set it frequency domain. We can set it to the time domain and there's our random noise.

We can go in and we can do a fixed sine wave if we want. and uh, there is our fixed sine wave Beauty set it to 1 khz. There you go and uh, we uh, we've got continuous trigger. That's why it's not uh, getting a stable display.

but if we set the channel one trigger, not a problem. The source and the channel one input works perfectly and the same thing on Channel 2. Beautiful! I Well, you know, and on a basic check, this is a fully working DSA Awesome! So there you go I Hope you found that at least somewhat interesting I Was hoping that it would be a decent troubleshooting uh video? at least have to get out the moldy meter for goodness sake, uh, let alone the scope. but it turns out it was a pretty darn, uh, trivial and obvious um fix.

and I need to do more with this and uh, but it looks in, uh, in pretty good shape. So I should be able to do some really cool videos with this. uh DSA And if you do want me to do something, um, please leave it in the comments. but uh, yeah, this is a really nice bit of Kit I mean it's absolutely huge I don't know where I'm actually going to put the damn thing.

it's probably not going to really sit up on my rack there, so I don't know. might just have to keep it under the bench and whip it out when I need it. It's not an everyday use tool at at least not the stuff I work on these days. it used to be at my, uh, former job I Used to use these things almost every day.

um, sometimes and uh, they are really nice instruments. little bit obscure, but anyway, could do some interesting stuff with it. So yeah, Murphy got me on a troubleshooting thing every time. I Always hope when I get something that doesn't work I Always hope that you know it'll take me days to fix and hours and troubleshooting and it'll be some obscure fault or something.

but nah, too easy. So I don't know what happened to it. they probably you know whether it doesn't look like it's been dropped or something I don't know Anyway, um, it's not in bad shape so hope you liked it. Catch you next time and.

.