The mixed-signal idea is probably an accurate guess. There is certainly a open space on the front-panel where the probes could connect, with a straight-route to the interior of the device. I would agree on the device between the ADCs. It almost certainly does handle clocking the ADCs.
I'm pretty sure that most of the fans blow to the "bottom side" (side on which the middle part is mounted), so I think it is working other way around - fan is sucking air from the case and temperature sensor is close to the fan where air should be the hottest.
Yeah, those silicone covers are used to minimize handwork on power devices. They're worse than mica or thin silicone+thermal compound at conducting heat, but it only takes a second to pop on and assemble whereas applying silicone goop and little sheets of insulation is more time consuming and hence expensive.
I like your videos, when you speak, you are informative, and we can tell you have some knowledge, where as the "other guy" comes off as if he knows everything, and always saying "i dont know why they did that, that makes no sense, I wouldnt have done it that way" ok, now I am rambling... if I may say, thanks to your hour long review of the DS4014, I just bought one today.
@16:44 the masked IC is probably a clock driver for all of the RIGOL ADC chips (4 count compared to the just one on the 2000 series) you can see something similar on Dave's teardown @13:36. Another funny thing is that the 2000 series has a lot of probing points, the 4000 series seems to have almost none and a lot of unpopulated JTAG headers. I'm so tempted to try to figure out the Blackfin DSP, Actel & Spartan programming and configurations to alter my 2000 series (already unlocked of course)
I wonder if that connector is for a daughterboard for a planned mixed-signal version? There is also a cutout in the board where a connector could reach the front panel... The device between the ADCs is probably a high performance PLL clock generator chip.
@30:45 - If they could get the chip mfg's to send them unmarked die inside the chips, then I'm sure they could get them to mark only the pin 1 location or house-mark the chips for them as well... Rather than having to laser them off after the fact..
Great tear-down, I have the same unit and am just trying to find a tear-down on of a DS4054 and/or a fast rise time pulse gen to test the bandwidth one day I think it turned it off and on too fast or something. I think it lost an eeprom. Now the device thinks its a 4054 instead of the 4014.
Considering the PCB just specifies "DS4000", I'm not too sure what delineates between the various speed versions. They could have different front-end ASICs or stuff different parts for the different versions. Alternatively, it could just be a software flag for the front-end ASICs, which would open the window for hacks like the ones for the cheaper Rigol scopes. Anyways, I'll probably borrow a high-frequency signal generator from work, to do bandwidth testing at some point in the near future.
@14:53 by looking at Dave's teardown @21:33, the masked IC must be an Actel ProASIC3 A3P030 VQG100 or an upgraded one. Seems to be an FPGA to control and maybe to configure the spartan 6 FPGAs. Datasheet: www.actel.com/documents/PA3_DS.pdf. Interesting that the DS4000 series has double the Spartan 6s, but seems to have an extra one (total of 5) my take will be that the 5th one with the high speed connector is for an optional DSA input of shorts, but you will have to show what's on the other side of the board
Last comment for now, but it does seem this teardown indicates speeds way over 100MHz, unless the ASICs or whatever they are at @26:56 are designed to limit the bandwidth somehow, btw forgot to mention it earlier, great teardown, other than you forgot the other side of the board :)
At best, it would require a trip to the factory to install the connector that I think it would connect to. I don't see why the hardware couldn't support it, though.
Since you've already popped it open... You mentioned it's a standard 120mm fan? Ever considered dropping something like a noctua or coolermaster fan in it's place? Something that'll be real quiet? You can stuff the original in the box... you know, incase you ever need to send it back to Rigol for anything! :P
I know that Rigol is developing MSO and WAVEGEN modules for the UltraVision line right now, although I don't know if they'll backward compatible with the current models.
Probably a bit of both. I like your style though. The more i look at your unit and the review, to me it looks like they reworked it from a 2ch version to a 4ch version.
The mixed-signal idea is probably an accurate guess. There is certainly a open space on the front-panel where the probes could connect, with a straight-route to the interior of the device.
I would agree on the device between the ADCs. It almost certainly does handle clocking the ADCs.
I'm pretty sure that most of the fans blow to the "bottom side" (side on which the middle part is mounted), so I think it is working other way around - fan is sucking air from the case and temperature sensor is close to the fan where air should be the hottest.
Yeah, those silicone covers are used to minimize handwork on power devices. They're worse than mica or thin silicone+thermal compound at conducting heat, but it only takes a second to pop on and assemble whereas applying silicone goop and little sheets of insulation is more time consuming and hence expensive.
I like your videos, when you speak, you are informative, and we can tell you have some knowledge, where as the "other guy" comes off as if he knows everything, and always saying "i dont know why they did that, that makes no sense, I wouldnt have done it that way" ok, now I am rambling... if I may say, thanks to your hour long review of the DS4014, I just bought one today.
@16:44 the masked IC is probably a clock driver for all of the RIGOL ADC chips (4 count compared to the just one on the 2000 series) you can see something similar on Dave's teardown @13:36. Another funny thing is that the 2000 series has a lot of probing points, the 4000 series seems to have almost none and a lot of unpopulated JTAG headers. I'm so tempted to try to figure out the Blackfin DSP, Actel & Spartan programming and configurations to alter my 2000 series (already unlocked of course)
I wonder if that connector is for a daughterboard for a planned mixed-signal version? There is also a cutout in the board where a connector could reach the front panel...
The device between the ADCs is probably a high performance PLL clock generator chip.
15:30 That's rework is really ugly for a pro level scope!
@30:45 - If they could get the chip mfg's to send them unmarked die inside the chips, then I'm sure they could get them to mark only the pin 1 location or house-mark the chips for them as well... Rather than having to laser them off after the fact..
Great tear-down, I have the same unit and am just trying to find a tear-down on of a DS4054 and/or a fast rise time pulse gen to test the bandwidth
one day I think it turned it off and on too fast or something. I think it lost an eeprom. Now the device thinks its a 4054 instead of the 4014.
Considering the PCB just specifies "DS4000", I'm not too sure what delineates between the various speed versions. They could have different front-end ASICs or stuff different parts for the different versions. Alternatively, it could just be a software flag for the front-end ASICs, which would open the window for hacks like the ones for the cheaper Rigol scopes.
Anyways, I'll probably borrow a high-frequency signal generator from work, to do bandwidth testing at some point in the near future.
@14:53 by looking at Dave's teardown @21:33, the masked IC must be an Actel ProASIC3 A3P030 VQG100 or an upgraded one. Seems to be an FPGA to control and maybe to configure the spartan 6 FPGAs. Datasheet: www.actel.com/documents/PA3_DS.pdf. Interesting that the DS4000 series has double the Spartan 6s, but seems to have an extra one (total of 5) my take will be that the 5th one with the high speed connector is for an optional DSA input of shorts, but you will have to show what's on the other side of the board
Dude, I hope you're getting paid to do this, because I wouldn't risk making the warranty void lol.
Last comment for now, but it does seem this teardown indicates speeds way over 100MHz, unless the ASICs or whatever they are at @26:56 are designed to limit the bandwidth somehow, btw forgot to mention it earlier, great teardown, other than you forgot the other side of the board :)
At best, it would require a trip to the factory to install the connector that I think it would connect to.
I don't see why the hardware couldn't support it, though.
Is this channel a meeting place for the more interesting youtubers or something?
Since you've already popped it open... You mentioned it's a standard 120mm fan? Ever considered dropping something like a noctua or coolermaster fan in it's place? Something that'll be real quiet? You can stuff the original in the box... you know, incase you ever need to send it back to Rigol for anything! :P
It looks like they do it in software. posted a video of it on my channel
I wish I could get paid to do this.
Could it be I'm just that awesome?
Or more likely, there just aren't that many people doing in-depth teardowns on youtube.
I know that Rigol is developing MSO and WAVEGEN modules for the UltraVision line right now, although I don't know if they'll backward compatible with the current models.
Probably a bit of both. I like your style though.
The more i look at your unit and the review, to me it looks like they reworked it from a 2ch version to a 4ch version.
I do not know why you guys do not use a power screwdriver.
I see 2012 date codes.
That would be why it somewhat confused me.
no one from rigol is watching...
Well shit!
Does that mean I finally have a real youtube channel?
Maybe when EEVBlog subscribes too....
"Verticalize the tabs." Love your Engcheese.
18:50 That ic is 4051, 1 to 8 Analog Multiplexer/Demultiplexer bilateral switch
The Ic's you don't know are 4051 8-CHANNEL ANALOG MULTIPLEXER/DEMULTIPLEXER