Author Topic: EEVblog #1146 - New Low Cost Rigol MSO5000 Oscilloscope  (Read 139882 times)

0 Members and 1 Guest are viewing this topic.

Offline amx

  • Contributor
  • Posts: 9
  • Country: ua
Re: EEVblog #1146 - New Low Cost Rigol MSO5000 Oscilloscope
« Reply #925 on: September 03, 2020, 04:52:33 am »
Hi CChin254 I do not understand why they use Spartan 6 for communication between Kintex and Zynq the both Kintex and Zynq have a Gigabit Transceivers the more efficiently would be use GTX for sending data between Zynq and Kintex...
Probably Spartan 6 placed for other purposes...
 
The following users thanked this post: CChin254

Offline danosimo

  • Contributor
  • Posts: 5
  • Country: es
Re: EEVblog #1146 - New Low Cost Rigol MSO5000 Oscilloscope
« Reply #926 on: September 23, 2020, 06:54:56 pm »
Hey Alexvg (and everyone),

Based on your nice work about lowering the temps on our MSO5k scopes I quickly put together a design for 3D printing that should resemble yours (I am not good with scissors, cut myself every time so this was easier). So this way we can all 3D print the structure to put the fan on. Attached is the .STL file to print. This one I have made for a 90mm fan, currently running on 5v (with a small regulator), though it is a 12v fan, but noise can be an issue depending on the RPM the fan you use. I have also put a 80mm (15mm thick) 5v fan for the inflow air, though after testing the most impact is clearly the reversed air on the main fan location with the structure forcing air current through the right places. I have confirmed that the temps can be lowered 10+C degrees, It was ridiculously hot on stock config (mine was showing 58c at the protocol analyzer connector in the front and some of the BNCs).

One thing that needs to be said is the placement of the scope on the bench, that is if you have obstacles behind it so that the hot air stays lingering or not. In my particular case the unit is against the wall (bad choice but no space) and so this may actually make a difference, leaving a bit of room behind will help. In any case I have confirmed with the attached design that the temps are really reasonable now (45c at the front connector). Clearly when they decided to rev up the main board one of the changes (if no the only) was simply powering the current fan setup from lower volts (is it 7v?) instead of 9v originally, and while it does make it quieter, it sure makes the unit suffer and the user very toasty!. Not good in summer time for sure. Between the 3D printer and the scope I think I am losing about 2 pounds a day sweeting like a pig..... Oh, and one more thing, I would not recommend to print this on PLA for obvious reasons, ABS seems to do well.

Anyway, so here is my two cents on this. Again, thank you for the help and bringing this to our attention and finding a very clever and effective solution. Kudos!
 

Offline danosimo

  • Contributor
  • Posts: 5
  • Country: es
Re: EEVblog #1146 - New Low Cost Rigol MSO5000 Oscilloscope
« Reply #927 on: September 23, 2020, 07:10:29 pm »
Just so people can see it before downloading and printing, here are a couple of snapshots of what it looks like. And please be kind, I know I am not the best designer by far...... I do what I can.

And can anyone help with being able to read the internal temps via SSH in the command line? I am sure we can get some more accurate temp inside this way. My Siglent bench unit can only handle 1 temp probe and I have also used an infrared for external temp reading, but it would be good to see the temps at the core of the CPU at least. But again, after 1 hour of continuous sampling (two channel @4GSa/s) it shows stable at 43c approx. in a 25c degree room, so it is a very effective solution. Huge improvement no doubt.

Anyone?
« Last Edit: September 23, 2020, 07:25:33 pm by danosimo »
 

Offline oliv3r

  • Regular Contributor
  • *
  • Posts: 121
  • Country: nl
Re: EEVblog #1146 - New Low Cost Rigol MSO5000 Oscilloscope
« Reply #928 on: October 24, 2020, 11:16:37 am »
I did a IC Identification for the back side of the Main PCB of this device:

Thanks CChin254, I guess I like your style a bit better then what I did here: https://gitlab.com/riglol/rigolee/-/wikis/MSO5000-teardown

Though as I want to keep it as one picture, I'll have to see how that can be done; but some new items here is good!

Btw, the reason my picture looks so 'weird' is that it's very high-res close ups and double sides overlayed. But for the chip identification, using a clear single sided picture probably works better :)

Also, I'm in agreement with
Hi CChin254 I do not understand why they use Spartan 6 for communication between Kintex and Zynq the both Kintex and Zynq have a Gigabit Transceivers the more efficiently would be use GTX for sending data between Zynq and Kintex...
Probably Spartan 6 placed for other purposes...
I don't think the Spartan sits there exactly like that, as we can very clearly see the high speed communication lines between the Zynq and the Kintex7 (especially on X-ray). So are you certain that the Spartan6 is not some sort of 'controller'. IIRC the spartan6 only has a (slow) SPI link ... It's been too long since looking at the architecture, and sure want to pick it up again :)
« Last Edit: October 24, 2020, 11:32:32 am by oliv3r »
 
The following users thanked this post: CChin254

Offline CChin254

  • Contributor
  • Posts: 38
  • Country: us
Re: EEVblog #1146 - New Low Cost Rigol MSO5000 Oscilloscope
« Reply #929 on: October 27, 2020, 04:22:10 am »
Yeah I need to figure this thing out, I'm just too busy now of days since school started (last year of High School). Hopefully I can figure this out and fix the block diagram soon enough. Thanks for letting me know anyways!
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf