Teardown Tuesday
What's inside Rigol's low cost DSA815-TG spectrum analyser?
Photos: http://www.flickr.com/photos/eevblog/sets/72157632115219155/
Forum Topic: http://www.eevblog.com/forum/product-reviews-photos-and-discussion/eevblog-391-rigol-dsa815-spectrum-anayser-teardown/'>http://www.eevblog.com/forum/product-reviews-photos-and-discussion/eevblog-391-rigol-dsa815-spectrum-anayser-teardown/
Datasheets on chips used:
XC6SLX25 http://www.xilinx.com/support/documentation/data_sheets/ds160.pdf
A3P030 http://www.actel.com/documents/PA3_DS.pdf
ADSP-BF526 http://www.analog.com/en/processors-dsp/blackfin/adsp-bf526/products/product.html
LTC3728LX http://cds.linear.com/docs/Datasheet/3728lxff.pdf
ADF4106 http://www.analog.com/static/imported-files/data_sheets/ADF4106.pdf
ADE-2+ Mixer: http://www.minicircuits.com/pdfs/ADE-2.pdf
Block Diagram Breakdown: http://www.eevblog.com/forum/blog-specific/eevblog-391-rigol-dsa815-spectrum-anayser-teardown/?action=dlattach;attach=33922;image'>http://www.eevblog.com/forum/blog-specific/eevblog-391-rigol-dsa815-spectrum-anayser-teardown/?action=dlattach;attach=33922;image
EEVblog Main Web Site:
http://www.eevblog.com
EEVblog Amazon Store:
http://astore.amazon.com/eevblogstore-20
Donations:
http://www.eevblog.com/donations/
Projects:
http://www.eevblog.com/projects/
Electronics Info Wiki:
http://www.eevblog.com/wiki/

Hi welcome to tear down Tuesday with another Spectrum analyzer. The Ryol DSA 815 TG with the optional tracking generator. Thought we'd crack it open. Take a look inside because Spectrum analyzers are usually a bit more interesting than other bits of test gear like your runof-the-mill scope or your multimeter or whatever bit more engineering poured in these things.

So you know what we say here on the EV blog. Don't turn it on, take it apart. and here it is. You've seen it before in previous videos and it's a real nice solid design.

Real solid bit of Kit this one I Really like it. It weighs like 4.2 kilos. There's a lot of heft in this thing, which uh, indicates a lot of uh shielding. Maybe some uh, diecast, uh stuff.

Of course for the RF front end perhaps um, but it should be uh, as easy to crack open as uh, other Ryal gear is. couple of screws here, couple of screws up the top it, uh, the back will pop off. Then we'll have uh, big shielding at the uh on the top. Half the shielded power supply as well.

that'll all pop open. We should be able to access the main board, but could be, uh, fairly involved in terms of uh, getting the RF stuff apart and things like that. But um, I Don't expect this to be a particularly, uh, you know, like a high-end Electronics as far as the RF goes. no fancy black magic stuff in terms of, you know, uh, PCB uh, etched PCB filters and traps and things like that because this only goes up to 1.5 gig.

it's not as far as um RF um Spectrum analyzers are concerned. this thing is, you know, is pretty much a baby it's basically working at DC as far as the RF Engineers are concerned and we have a sticker on top demo equipment. hey, we can fix that with a de slice of the knife. There we go, no problems and of course we have the obligatory warranty void sticker.

and if other bits of Ryo gear or anything to go by especially their new stuff, this should be pretty well designed and built. So cuz they're really, uh well. they've always been quite reasonable in their construction Ryo but as of late, they, uh, the new series of gear seems to be designed and engineered very, very well indeed. So I don't expect to see any shtin inside this thing I expect uh, really decent build quality and uh, nice clean PCB probably washed nicely.

all that sort of jazz. All right, this puppy should just lift straight off. Maybe if I've actually got all the damn screws out, maybe not. There we go.

Tada and yeah, the obligatory metal shielded back on this thing power supply under here and that'll also be shielded from the rest of the circuitry. and we've got a couple of uh headers down here. They're uh, polarized with the missing pin there. so there some sort of uh programming or Andor test interface something like that got the nice shielding Down here, in the uh ethernet connector and the USB.

You can see the shielded lugs there. really nice. um Emi shielding on that and uh, of course we've also got a little Mount and shield for the Main's connector there. Neat.
and it looks like this whole thing is going to lift off rather than just, uh, this up here. So get our screwdriver under there. We can lever that that whole thing off and that power supply is going to come off in one section and the obligatory big power cable going over. Oh, looks nice in there.

looks nice. you'll see it in a second. Hold on to your hat and here we have it, folks who look at this beautiful look at the huge diecast uh box we've got here for all the RF circuitry and all of this juicy stuff we've got under here. As you'd expect, we got a Spartan 6 Fpga and we'll go into details.

We've got an Actel Pro A63 by the looks of it, some flash memory and a black fin DSP processor down there. Some power supplies up here and uh, interestingly some. a small little angled heat sink there with to 220 package with packages with some uh seal pads under there for uh, isolation and this Tada Check it out out. they're a little bit proud of it.

Design in China not designed in China So yeah, right. good one. Chinglish And we're actually looking at a Uh version 1.05 uh, looks like the 30th of the 12th 2011 and here we've got a fairly pedestrian I guess you could call it Spartan 6 Fpga It's the XC 6s Lx25 and it's only like a 24, uh ,000 gate device. so it's not particularly high density at all.

But it does have built-in um, almost 1 Meg of ram. So they're using this for the display processor clearly. Um, there. Here's the external memory as well and you can tell it's essentially the display processor because, well, it's hooked into.

you guessed it, the display connector down there and you can see all the uh uh termination resistors on the bus there to drive the long cable off to the LCD. And once again, the bastards. look, they've put this uh Magic tape or laser edch or whatever it is on this chip here like they've done in other ones. and uh, this is, uh, it looks like the Um ADC because if you have a look here, there's this controlled impedance Trace here running up through that via there which probably comes out of the RF section here.

runs over to here, down through a few Choice inductors there and through. Looks like there's a little I don't know, is that some sort of little common mode Transformer or something? I'm not sure, but that's probably um, the ADC I would be assuming with maybe some uh, local voltage regulation here perhaps. And of course, that flows parallel output bus um flowing into the Fpga there, straight into the display Fpga processor I Mean that Fpga is probably doing other stuff as well as the display processor and they've gone to town on the silk screen here, they've identified um, all of these various um differential pair lines in here I mean we've got uh DN clock we've got done. We've got TX in uh, Txp sync, negative sync, positive Rxp init program various lines there Um, once again, directly coupled into that main Spartan 6 Fpga So that's obviously the Uh capture and processing as well as the display engine.
And of course that's got its own JTAG connector down in there, which they, uh, use to program this thing and or debug or do um in system, um, test or something like that. And there's that chip up close. and yep, it does appear to be laser etched off. You can almost see the lines going across the thing and it is definitely like stepped down into it.

so they've done something to really etch out the top of that chip. Bastards. Pretty darn effective though. And down closer towards the main Uh processor which is a black fin.

I'll show you in a minute. We got another Fpga it's an acil proasic3 and it's rather interesting. Why? Why they use uh, you know, multiple vendor Fpgas in this thing. They obviously had a specific requirement based on cost or whatever, but they love to chop and change for various things here.

Anyway, we got a 64 megabit uh spansion flash memory. We've got a fast SRAM here and up here. we've got a Um Cypress USB uh microcontroller and no surprises Of course for finding a black fin DSP processor in here same as what they use in the Riyo series scopes. Of course you know they've already got the Uh development system.

Lot of code. reuse it with the Blackf processor and uh, next to it. Here we have a Um Micron um sorry, a uh Mcel uh Ethernet controller. It's a Ksz 8051.

so this thing actually has Um Ethernet built in. but they've decided to use an external Ethernet controller here and this section here just above the processor. They've got a Linear Technology LTC 7328 LX 2phase Synchronous regulator and you can see above it just there. You can see the Um two external mosfets as well, the two external N Channel mosfets used for the synchronous rectification.

and uh, it's probably um, wouldn't be a stretch to Guess that they've uh, got the Um switching frequency. This thing locked to the Uh sampling frequency so that doesn't interfere with the acquisition part of the RF front end. No surprises for finding a whole bunch of uh, low Dropout linear Regulators Here you know, triple 17s. Um, you know, sort of like the industry standard, we've got 5volt rails.

They've got all sorts of little test points here. 15 volts you go right over to here. we're talking 32 volts up here and they got various other Uh rails and they're all hooked into nice little test points all down here. Very well labeled board, very well laid out.

Now there's one thing I'm not seen around this part of the circuitry here. I mean here's the 10 MHz reference out external 10 MHz reference in and um, you know, apart from a 32 khz uh watch crystal up there for the time date I don't um see any obvious Crystal oscillator module on this thing and I I Expect it to be a reasonable quality one, but uh, I assume it must be on the bottom side of the board cuz unless I'm blind I can't see it on the top here. And if you're wondering what the To220 devices are, well, sorry. nothing exciting.
just a bunch of LM 317s and 337s. So that's a very neatly presented and neatly laid out and well engineered. uh main PCB there I mean it's going to extend uh under here. it's quite possibly the full width of this whole unit right under the diecast.

Shield here where the RS front end is? Um, but yeah, you know it's super clean. There's no flux residue on here at all. Well laid out, well silk screen. It's got all the JTAG interfaces and various things if you want to hack and play around with it.

We got our Uh LCD connected down here. This will be our keyboard up here I'm assuming and uh, well, looks like we're going to have to take this thing apart. look at all the screws on there. so I'm not sure how this thing lifts out.

I mean there's a couple of screws on here for the main board but I think if I just take those out, it's not just going to lift out and but then I'm not sure if I take out these ones first. Maybe there's a couple of them that hold it in down lower. I Don't know, but you can bet your bottom dollar there's some porn under there. and here we go.

That was an awful lot of screws and there did seem to be two lengths here, which was, um, rather annoying. So I only got to record a couple of them. But anyway, here we go. let's lift the skirt.

I Expect to see some RF gasket in, perhaps on the bottom of this. oh, one big board. Oh no, there's no RF gask it in. look.

it's just, uh, directly. like, uh, you know, directly. Um, just the uh, machined aluminium backing on that. Go figure.

So yeah. anyway, this is a bottom of the Ranger Spectrum analyzer. so I guess you don't expect to get all that goodness. but oh, look at that once again.

As with every RF Spectrum analyzer, you'll get all there couple of filters and stuff on the board. I was wrong I was wrong. We'll take a look at these I Didn't expect to see um, well, much if anything in the way of PCB filters there because this thing's only a 1.5 gig scope. But there you go.

It's like it's very modularized. Um I Don't see any labels on the individual modules at this stage, but uh ah, you know if you know your Um RS Spectrum analyzer Uh, you know the workings of it. the topology, the top level block diagrams. you could figure out what each section does and I just popped out the main board and that was really easy.

I Mean it's there's nothing in, there's nothing holding in there at all. once you do off all those top screws and there's the matching shield for the bottom set of that. And of course the uh end connectors just pull straight out of there and the board down here is completely integrated. It's really very nice and look what we have here.
We have another zyink Spartan down there, but we'll take a look at each Uh section in a little bit more detail. I mean geez, we can spend hours analyzing this thing I Mean, really and is very typical of most products with these, uh, large hand soldered uh BNC connectors. You see some flux residue around there on the board. Nothing major though, and let's see if if we can follow this thing.

Shall we? now? I'm not going to go into a full system breakdown of a spectrum analyzer and absolutely nail e mo each module on the head. I'll leave that to the viewers I'll be posting highres photos on the site by the way. So if you want to, uh, check out the board in some more detail, there should be some um highres photos there for you to play with. Now here's our input.

of course here's our uh end connector down here. straight in. Now when you Pro um probing around these things, just be care very careful about ESD because there is some sensitive stuff on here which you really don't want to uh, kill at all. So anyway, we've got our input there directly AC Coupled straight in.

hence the minimum uh, you know bandwidth of this thing is 9 khz because well, there a series cap on the input you can't measure down to DC on these things, these are not Dynamic signal analyzers. they are RF Spectrum analyzers straight in. We've obviously got some sort of uh buffer or something like that. not sure what the other two chips are might have a look in more detail and and of course all of this is controlled impedance stuff as you know and as I've talked about before and also check out all the heavily stitched Vas all the way around each individual module is each Um circuit.

Each system module is um heavily via stitched basically to lower the ground inductance to ensure that the ground in inductance in all places around the module is. well. it's basically um DC at RF Inductance is absolutely everything. So what have we got? We've got our input here.

It doesn't look like there's any. There might be some attenuation happening in there I'm not sure. but anyway, we're jumping through the wall here because there's a diecast um Shield here. So this one, some of them go under.

like like this one here is jumping under. You can see the controlled impedance tracing there, go through AA then actually jumps under through a middle layer and back up through to the top side there. This one doesn't bother doing that, it goes straight across. I Guess they wanted the minimum amount of loss there cuz when you include a V like that, veers have inductance so you know if you want the utmost in.

Signal Integrity You try and keep everything on the top layer here. now. um, if there's any components on the bottom side of the board of any note. Um, with each module, I'll tell you now.

here's a chip. Once again, bastards have laser, um, etched the number off there. Now that possibly looks like some sort of uh, input, uh, filter. Maybe some attenuation, but those resistor values are all the same there by the looks of it.
Then we've got some sort of amp and then we I'm not sure what's going on there. we're going into another that's possibly no, that might be part that might be the mixer perhaps. Let me, uh, check the bottom side and let's back up a little. Back to the input here.

Check it out. Tl072 Hardly. Um. RF amps or anything like that.

So they're just doing some, uh, very basic stuff with those T 72s. Nothing RF there at all. And then we go over into this network here. and yeah, they are all the same value resistors.

Yep, they are. And then we're jumping in. There's a low voltage. um, local.

There's a low Dropout linear regulator on the other side of that. Um, just some local voltage regulation. Then we've got looks like a multi-stage LC filter there and going into another device. Not sure what that one's doing there.

and once again, they've lasered the number off that one. Bastards. So we're going into some sort of small network over to here. But over at this point is coming in from this filter here.

so let's check that out. Now this funny looking pattern etched onto the board here. this is just a St a multi um element stub filter. Each one of those like that big pad on the end of that is actually acting as a capacitor and then the trace going to it is actually acting as an inductor depending on what size and shape you make them based on the controlled impedance uh value and the you know the distance to the ground plan underneath and all that regular um, uh, highspeed signal Integrity stuff So you know it might look like magic, but it's just um, a basic um LC stub filter pretty much and that's coming from this circuitry over here I'm not sure what that's doing at all, but if you flip the board over, that's circuitry which is under there actually which is under round about.

there comes from pretty much the um Fpga on the bottom. there so looks like we've got our Fpga feeding this around here which then jumps over to there as we've looked at before, goes through this um stub filter in here, through this controll impedance trace and then it drops down to the bottom and it actually I believe it actually connects over to cuz I can't see anywhere else it would connect you so it's jumping over there into there which then mixes so that might be the mixer perhaps and this might be something to do with the local oscillator although it doesn't seem to be enough parts around there to do that. Um, but anyway, it's going through another Network up here, through another LC filter up here and into another magic looking device over here. but this is nothing, uh, fancy at all.

It's fairly, uh, common. it's just a band pass filter and once again, depending on the size and shape of the various traces on here, based on the thickness and type of material of the dialectric, etc etc. this is just a multi-element band pass filter. but as I said at the start, why they've done that on only a 1.5 gig um.
bandwidth Uh Spectrum Analyzer I Don't know cuz you can see once again, there's a just a regular component multi-stage LC filter there going into another band pass filter here. which why they didn't do that with uh, regular Parts at this sort of Um bandwidth I Don't know? Um, you'd have to talk to the system designer about that one I'm afraid. So this multi-element resonant stub Fielder here is really. you know it looks kind of weird, but there you have a capacitor going to ground.

essentially. then you have another inductor here in series. Then you have another capacitor going to ground. Then you have another inductor here and you can look at there's an inductor there, there's a capacitor to ground in, tiny little amount of inductance, capacitor to ground, and basically just a multi-element LC filter.

Now I think I Might actually have a bit of a bet that this section down here is the local oscillator next to the input there and if I flip the board over, there's the base of it. Once again got another laser chip there. The bastards. Um so I think that's possibly the local oscillator there once again going up into this, which then possibly goes and mixes up near the top somewhere because if you know your spectrum analyzer topology, that's basically what you've got.

You've got an input and you'll have some sort of input attenuator. Then you'll have some uh filtering, and then basically it goes into a mixer somewhere. so a mixer has to be up there somewhere. Uh, and then there's a local oscillator driving the mixer.

So possibly local oscillator here up there driving the mixer, which is somewhere up there? I Don't know. I Could you can spend hours analyzing each one of these blocks and drawing the complete system diagram? You'll also notice some routed slots in there like that and uh, they've really went well. We really don't want any coupling between those sections, so let's route out some slots. Good way to do it.

And once we come out of that B pass filter, there, we go into another couple of blocks. There's some more filtering there. I'm assuming we've got ourselves an RF relay there by the looks of it and going over to some more filtering. I Mean, there's bloody filtering everywhere on this thing.

Everywhere you look, there's a whole bunch of LC fil hey hello. Check out that uh top uh inductor there. Oopsie, It's on an angle, so you know, like there look. There's another section of LC filtering All right.

I've got a basic block diagram of how a typical Spectrum analyzer works here. Now it's very simplistic. If you want to go into more details, there's plenty of info out there just Google it. You'll find how Spectrum analyzers work in great detail and this is going to be quite similar.
so let's check it out. We've got our input here goes into input attenuator, so input on the BNC um the end connector down here uh, input attenuator is going to be around here somewhere. There's going to be some uh, you know, amplification and stuff like buffering and stuff like that. our Um input uh filter is going to be you know, around here somewhere.

Then we've got a local oscillator which I thought was down around here somewhere and then it goes into a mixer and that mixes the two signals together and then that goes into a Um the resolution bandwidth filter and I You know that that is actually adjustable. so um I'm not sure um at you know at what part bandwidth filter up there has anything to do with it I don't know it's in there somewhere and it basically then pops out and goes into your a traditional analog uh Spectrum Analyzer is going to have these additional stages as well. It's going to have an if Uh filter adjustable if filter, then it's going to have a log rhythmic amplifier. It's going to have envelope detector and it's going to have a Um low pass video filter and then goes into the ADC.

But the DSA 815 has basically doing all that in digital so it's basically got a hide. So the ADC is connected directly to there like that and it samples at high frequency and it's able to do all this traditional stuff in as you know, um, comput stuff either inside the um uh Spartan Fpga or inside the DSP processor itself. But all that is done digitally that you used to do analog in a spectrum analyzer. but you can see I Mean you know, we've only got um, five basic Uh blocks here for this for all this RS Spectrum part and you can see all you know.

There's like you know, more than a dozen different blocks on there. So think like this mixer. Um, you can get multiple mixers, uh, and things like that so you know there might be more than one mixer in there, and uh, all sorts of additional filtering and control and all sorts of stuff. But anyway, that's just a very quick breakdown.

I Mean, if somebody wanted to go through, you could map out all the individual Uh modules and how they fit in with your traditional um RF Spectrum analyzer front end like that. And by the way, this um ADC here doesn't have to be a particularly high-speed one. it's going to be uh High Precision one. But you know, the Um output of your resolution, bandwidth, uh, filter, and your mixer here is going to be you know, in the order of tens of megahertz.

So it's you Know, it's not like the ADC is. um, put bandwidth of several gigahertz or something like that. It doesn't because the local oscillator mixes that down to a lower frequency and that's why they call this part of it. uh, down conversion because it down converts it in frequency so that um, you can actually use a lower bandwidth ADC to do all of this um, extra stuff in digital.
So, as we saw before, this is almost certainly our ADC down here coupled in to our main Spartan 6 Fpga because here's the output of it here. as we saw before, it's actually jumping to an inner layer there as I mentioned before, so it, uh Ducks under so to speak, the um, uh ground shield around there with the die car shield and everything else jumps to a lower layer there. So this is the final output of our final filter here just before the ADC and there's another T72 down there. They just love using those on here cuz you know they're only working at uh, you know, sub one.

mahz, they're not um, working particularly fast at all. Maybe they're um, you know, buffering some sort of, uh, low frequency signal or something like that. But clearly we have a couple of uh RF Um amps in there. Those are a typical Um RF amplifier package I Have no idea of the part number.

um, but you know they. they're basically um, you know, Supply ground input output, but that's a very typical Package for you know, a uh, low noise amplifier, uh buffer or something like that. We got another package there which doesn't seem to be lasered off like the other ones. It's sort of.

It's almost like it's got some epoxy or something on top of that that. I'm not sure what's going on there. and there's our tracking generator output there. There's the big end connector.

We've got a multi-stage LC filter there. As you can see, there's no Uh PCB magic filtering happening there. and we've got yet another lasered off chip there. They really don't want you to know what ones they're using in there.

They're almost certainly uh, commercial chips. It's not like these are custom Asic or something, but they just stopping people reverse engineering this and it's probably not hard if you spent, you know, a few hours at it to, uh, find out what the device is based on the Uh block diagram of a what it should be and be. you know the uh package type and the number of pin you can. um, you know, narrow it down in not much time at all.

So I'm not sure what that one's doing there. Some obviously looks like some sort of selectable filter and on the back side of the board, yet another laser marked chip. And there's our Um Spartan Fpga So I'm not entirely sure what that's doing in the Um RF Uh section. aha I Found our local oscillator There it is.

It's an Analog Devices ADF 4106 and uh, that's a 6 gig uh, bandwidth? um uh. frequency synthesizer which they're using it as a local oscillator here. So let's have a look here to see if we can follow it. Here's our local Oscillator that Analog Devices part.

Let's flip the board around and it's on the back of here this section here. so it looks like that's doing some sort of uh output, uh, buffering and there's some probably some filtering in there and it's output in through the there's two. There's like a transmission line going out there, so let's take a look at where that pops through on the other side. Uh, no, it's obviously going to an inner layer, so it looks like uh, where are we? It looks like that pops up there like that.
It's going into here. Yeah. so we have local oscillator on the back here going up here. probably a mixer after, uh, some of the input, uh, filtering up here and then our resolution bandwidth filter is the rest of that going out to our ADC over here.

Aha I Finally found the main 10 MHz oscillator. There It is there and it's in the RF section here. So it's um, it's generating both the um uh system clock by the looks of it for all the digital section as well as of course all stuff for the RF section. So then it just pops out here and we've got some additional circuitry over here and it eventually comes on over to clock all of the digital stuff as well because there is no local oscillator here no pun intended for the Uh DSP processor and the rest of the digital stuff.

And if you have a look at the backside of the local oscillator circuitry here, you can see some more Uh PCB components going on. We've got diodes here and look, it goes through this long controlled impedance. Trace Over to the V stitching over here. why do they do that? Because that is a controlled impedance component on either side of that and in this block here just above the Uh main oscillator down here I Found an Analog Devices A549 dual out dual current output uh Dack with some Uh companion Teal 07 uh 2 devices there to uh convert the current output into voltage.

So that's been a relatively quick overview of the DSA 815 main board and it's um, relatively impressive. I Really like it there. It's well engineered, um, and you know it's only a 1.5 gig Spectrum analyzer. but they, you know, they really seem to have, uh, gone to town here and it's not just, you know, some gimmick.

That's why the performance of this thing isn't bad and uh, well, is you know is pretty spectacular for the price actually. and they've done a lot of work integrating this thing into a single board. I Really like it cuz there's nothing else in this uh product, just this main board and the power supply and as the connectors bolted directly on there. Brilliant! I like it.

Thumbs up and there's a quick uh peek inside the power supply. um, it's You know they're doing all the right things. They got input protection, they got input filtering they're using um EP cost uh brand capacitors which aren't too bad at all. They've sasti a lot of it down.

they got output filtering. yeah, it's all happened. They got high voltage isolation slots there. Ah, we won't go into it.

um in too much detail because uh, really? um, it's not that interesting and we've done that in previous videos like the oscilloscope. uh, tear down. but yes, it is a good quality. Looks like a quite a reliable or should be quite a reliable power supply.
well engineered. Um, it is Ryol branded there. but um, yeah, they Pro they may have, um, subbed it out to somebody else. but uh yeah, its model number is LPS Smps 80 version 1.1 and of course the rest of the guts under there.

not particularly interested in that, uh at all. It's just a display and a keyboard will have some uh, local circuitry on there for the keyboard. Ah, boring is the proverbial bat poo. So there you go, it's back together again and uh yeah, that was really interesting.

I mean got our ADC here and pretty much I think that Spartan uh 6 Fpga there is probably doing like the Um or your traditional uh sweeping functions and whacking it all uh, straight into memory, doing some processing and stuff. perhaps not sure how much of the heavy lifting the Um black fin DSP processor is doing down there. Probably not much like in the uh um, other Ryos. it's probably just doing some, you know, just your you know, regular gooey interface type stuff and things like that.

So maybe all the uh Grump work is done inside that Spartan 6. But as I said, that's not a particularly large unit. So, but it's at least handling the ADC input uh there. So it is doing that.

and quite possibly it looks like it's driving out here as well. like it's Genera like there's all these uh, clocks and um and control signals and stuff. So it's probably going out here and controlling the uh, you know, the local oscillator and doing all sorts of uh Jazz like that, so who knows. So there you go.

the warranty has been voided Gonsky and let's power this thing up and uh, see if she still works and it looks like we have a winner folks. Check it out there we go. Little spikes from my Uh RF remote here Beauty and uh yeah, that was an interesting look inside the Rogo DSA 815 very well engineered bit of Kit especially for the price I mean this is practically, uh, the lowest price spectr metalizer like this on the market by far. Um, yeah, it's a nicely designed engineered unit, but that's uh, what we've come to expect from Ryal these days and they certainly didn't disappoint.

I think it's a winner, so if you want to discuss it I've almost certainly um, skipped over stuff and missed stuff and maybe got the odd thing uh, back to front, but the best place to discuss it is over at the Eev blog. Forum Catch you next time.

Avatar photo

By YTB

25 thoughts on “Eevblog #391 – rigol dsa815 spectrum analyser teardown”
  1. Avataaar/Circle Created with python_avatars Paul Green says:

    Old post and so hope someone is looking… I might have damaged the Tracking Gen by soldering a coax to a Return Loss Bridge unit without disconnecting from the powered up DSA815. I used an old Weller PS20 Station. Late Night and tired!! The RLB doesn't seem to be doing what it should anymore. Could the voltage from the solder iron kill the scope??

  2. Avataaar/Circle Created with python_avatars jingji lau says:

    hahaha chinglish2333

  3. Avataaar/Circle Created with python_avatars Flordeliza Honrejas says:

    Thank you very much for your videos! Some years ago i used a HO 8753D VNA. I was able to measure SWR directly with it. Is it possible to do samething
    with this unit?

  4. Avataaar/Circle Created with python_avatars Gerald Chouinard says:

    I remember this video post on the review of the Rigol DSA 815 TG Spectrum Analyzer. I bought one a couple years ago and I have been quite satisfied with it except that lately, the clock has started to run intermittently at 1/8 the normal speed. It does not seem to have upset the normal operation of the analyzer but it makes a mess of the date and time of the stored files that I transfer on the USB memory device, besides the fact that I cannot rely on the clock on the device. Recently, I found a Rigol Web page where one could ask for an updated version of the firmware. Thinking that this could solve the problem, I sent my request and never got any answer; no very good in terms of customer support … Do you ave any suggestion as to a way to cure this problem or a Web site where I could get some support and answer from Rigol? Thank you.

  5. Avataaar/Circle Created with python_avatars Dapuppyz says:

    very impressive. i hope to be as cool as you one day

  6. Avataaar/Circle Created with python_avatars HeavyMetalGaming says:

    I love how you call the laser etched chips "bastards" haha. Also, I was thinking they did the bandwidth filter that way(unnecessarily in this particular device) so they can keep the general PCB design the same for future additions and improvements in the spectrum analyser. It may not be necessary in this edition of it, but will be when it is upgraded to a newer edition. That way they don't have to re-manufacture the PCB thus making it more cost-effective.

  7. Avataaar/Circle Created with python_avatars Luiz Carlos says:

    Power Saver (saving electricity) Show what's in it! thank you.

  8. Avataaar/Circle Created with python_avatars Peter Kinnear says:

    I think the Blackfin has the MAC on die, but not the phy, so that's what the other little chip is. Great Job on the review 🙂

  9. Avataaar/Circle Created with python_avatars OK2BCK says:

    WANTZ

  10. Avataaar/Circle Created with python_avatars Joop van der Velden says:

    The lasered chip is probably a fast A/D converter like a LTC2216

  11. Avataaar/Circle Created with python_avatars G zxiao says:

    I have a question  i can't find the reference clock of the LO, we can find the vco and pll,but where is the reference clock,how did it produce a sweep source

  12. Avataaar/Circle Created with python_avatars Linas K says:

    TL072 is for diode bias switching by the looks of it, it is always near sot-23 diode (in some cases)

  13. Avataaar/Circle Created with python_avatars TableAvocado says:

    Do you put them back together?

  14. Avataaar/Circle Created with python_avatars 1975RStefan says:

    Is the pcb soldered pb free? It would be a point of intresst, because pb free soldered pcb's are not really the best choice.
    I analyzed a defective pcb of an electronic ballast for HID light sources, this was soldered with Sn99Ag1 solder. It was intressting to see, that a lot soldering points lost its mechanical stability.
    Around the solder points I saw a lot of tin whisker needles on the tin surface.

  15. Avataaar/Circle Created with python_avatars ableite says:

    15:18 are these tracks made of gold or copper?

  16. Avataaar/Circle Created with python_avatars Izdat Afamas says:

    I would like to see inside a network analyzer

  17. Avataaar/Circle Created with python_avatars pigpen says:

    Its to keep the traces equal length on a parallel bus

  18. Avataaar/Circle Created with python_avatars Draco Xul says:

    I want to know about the wiggly traces next to the spartan as well. Any ideas?

  19. Avataaar/Circle Created with python_avatars Randy Lott says:

    That doesn't seem all that bad for a spectrum analyzer.

  20. Avataaar/Circle Created with python_avatars Billy says:

    at 5:23 you can see the traces coming down to the spartan are wiggly, why is this?

  21. Avataaar/Circle Created with python_avatars MFeinstein says:

    What about a deep teardown 😉
    h t t p s : / / w w w . youtube .com/watch?v=mT1FStxAVz4

  22. Avataaar/Circle Created with python_avatars Graafvaag says:

    From the JTAG header, one could auto-analyse the JTAG chain to see which ID codes come back. Some of the lasered IC's might be in the test chain right?

  23. Avataaar/Circle Created with python_avatars EEVblog says:

    $1500

  24. Avataaar/Circle Created with python_avatars AntiProtonBoy says:

    What's the wallet damage for something like this?

  25. Avataaar/Circle Created with python_avatars EEVblog says:

    I accept zero responsibility!

Leave a Reply

Your email address will not be published. Required fields are marked *