Author Topic: EEVblog #512 - Rigol DP832 Bad Design Investigation  (Read 260356 times)

0 Members and 1 Guest are viewing this topic.

Offline ResR

  • Regular Contributor
  • *
  • Posts: 149
  • Country: ee
  • -Artificial_Intelligent-
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #125 on: August 30, 2013, 04:23:53 pm »
I suspect some grid noise causes resets, is there any mains noise filter (for ex. 15SRB1-Q) in that PSU? I also have seen too many times where the capacitors are too close to the heatsink, like just to make sure it brakes eventually within few years.
« Last Edit: August 30, 2013, 04:27:07 pm by ResR »
 

Offline IanJ

  • Supporter
  • ****
  • Posts: 1057
  • Country: scotland
  • Pro EE guy many years ago, now a hobby/home biz.
    • IanJohnston.com
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #126 on: August 30, 2013, 04:28:38 pm »
I still find it odd that the CPU was resetting even though we never actually see the 5V supply drop out - just a fairly minor increase in the (frankly awful) ripple voltage on the 5V supply.

Thats what I was thinking......can only think that there's a watchdog circuit in there that trips the Cpu at the smallest thing. Either that or it's a bad watchdog design.......didn't Rigol state they had fixed a number of things?.....so maybe more than the reg and possible transformer winding.

Ian.
Ian Johnston
www.ianjohnston.com
Manufacturer of the PDVS2 & PDVS2mini
 

Online free_electron

  • Super Contributor
  • ***
  • Posts: 7304
  • Country: us
    • SiliconValleyGarage
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #127 on: August 30, 2013, 04:44:16 pm »
its even more off since that motorola thing is not a 5 volts part .... it is designed to run on much lower voltages. so there must be an additional regulator on the front panel . probably a switcher ...

and there is ...

so that could be the trouble. the lm317 starts to buckle , the switcher is drawing pulsed currents , the input voltage droops enough so the switcher goes in protection , starving the cpu , the brownout detector trips and slams on the brakes-> reset.
Professional Electron Wrangler.
Any comments, or points of view expressed, are my own and not endorsed , induced or compensated by my employer(s).
 

alm

  • Guest
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #128 on: August 30, 2013, 05:00:12 pm »
Your theory sounds likely, but that still doesn't explain why the +5 V rail appeared to be perfectly stable.

It would be kind an odd choice: use a switcher to go from 5 V to 1.8-3.3 V at 100 mA or so, but use a linear regulator from 12 V to 5 V at 700 mA. Would they have been worried about interference on the analog power supply board?
 

Offline ElectroIrradiator

  • Frequent Contributor
  • **
  • Posts: 614
  • Country: dk
  • More analog than digital.
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #129 on: August 30, 2013, 05:12:05 pm »
Someone already mentioned and previously posted photographs of the transformer in their DP832 being a 500VA unit with a 9v winding while Dave's unit has a transformer marked 400VA with a 10v winding.

Can anybody measure their 500VA transformer and post the dimensions, please? Dave already provided the height and diameter for the 400VA unit. I'm curious to know if this is just a 'cosmetic' change of a sticker and one of the secondaries, or if the toroid core is actually somewhat larger in the 500VA units.
 

Offline LaurenceW

  • Frequent Contributor
  • **
  • Posts: 258
  • Country: gb
    • It's Time, Jim, but not as we know it
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #130 on: August 30, 2013, 05:25:56 pm »
A couple of points.

Dave, where are all the twelve year old trolls, who only a couple of videos back were claiming that you were somehow in the pay of the equipment suppliers, such as Rigol? Hellooooooooooo??? Nah, can't hear 'em... :-DD

Five watts is five watts, no matter how big the heatsink which dissipates it. This is poor design in the age of higher efficiency equipment. A very simple (LM2575?) switcher would have done the job much better. Less heat all round, and could have probably saved themselves the price of any heatsink, at all.

Dave, to get MAX heat dissipation in the main heatsinks, you'll have to measure the supply running at the LOWEST output voltage corresponding to the highest (Triac) transformer tap <you knew that!> - I think you said it was 22V. Then (~50v-22V) x 3A is 54W. And then all this from one output device??? Hmmm - my old school TTI Thurlby Thandor supply outputs up to 3A at 30V and spreads that load across no fewer than SIX 2N3055's - luverly :-+ I think you Oz folk have a saying including the words "Brick" and "Dunney" to describe this construction style...?

Last thing - check your blood pressure, mate. We worry about you.
« Last Edit: August 30, 2013, 05:30:38 pm by LaurenceW »
If you don't measure, you don't get.
 

Offline pickle9000

  • Super Contributor
  • ***
  • Posts: 2132
  • Country: ca
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #131 on: August 30, 2013, 05:49:15 pm »
So in this particular case assuming Rigol fixes the supply problem (and a few of the other software issues and poweron spike), how many people will buy the unit? Personally I prefer to know the faults my equipment has (within reason of course). What I don't like is to find an equipment fault after a major investment. 
 

Offline salviador

  • Regular Contributor
  • *
  • Posts: 61
  • Country: it
    • https://www.youtube.com/user/mancio92M
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #132 on: August 30, 2013, 05:50:38 pm »
So in this particular case assuming Rigol fixes the supply problem (and a few of the other software issues and poweron spike), how many people will buy the unit? Personally I prefer to know the faults my equipment has (within reason of course). What I don't like is to find an equipment fault after a major investment.

is what happened to me !!  :palm: |O |O
 

Offline c4757p

  • Super Contributor
  • ***
  • Posts: 7805
  • Country: us
  • adieu
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #133 on: August 30, 2013, 06:00:16 pm »
This is poor design in the age of higher efficiency equipment. A very simple (LM2575?) switcher would have done the job much better. Less heat all round, and could have probably saved themselves the price of any heatsink, at all.

Yes, this is bizarre to me as well. Why in god's name would you put 800mA or whatever it was through a linear regulator in 2013? It's obviously not in the output path, so who cares if it's "supposed to be a linear supply"? Dear lord, with a little care in design, a decent switcher isn't going to put any appreciable noise on the output...
No longer active here - try the IRC channel if you just can't be without me :)
 

Offline Phantomix

  • Contributor
  • Posts: 12
  • Country: de
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #134 on: August 30, 2013, 06:12:31 pm »
LOL

ROFL


without reading the whole thread, i can exactly say why they designed it like that

They're sitting in the US, where they only have half the input voltage! So, in the design process this was fully OK and wouldn't justify a switching regulator. Then, when it's gone to production, noone looked again at it, as it basically works. Maybe, some software developer had "random software resets" on his todo list before they finally found it... :D

 

Online Monkeh

  • Super Contributor
  • ***
  • Posts: 6090
  • Country: gb
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #135 on: August 30, 2013, 06:14:56 pm »
LOL

ROFL


without reading the whole thread, i can exactly say why they designed it like that

They're sitting in the US, where they only have half the input voltage! So, in the design process this was fully OK and wouldn't justify a switching regulator. Then, when it's gone to production, noone looked again at it, as it basically works. Maybe, some software developer had "random software resets" on his todo list before they finally found it... :D

No, that'll have nothing at all to do with it.
 

Online free_electron

  • Super Contributor
  • ***
  • Posts: 7304
  • Country: us
    • SiliconValleyGarage
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #136 on: August 30, 2013, 06:15:21 pm »
LOL

ROFL


without reading the whole thread, i can exactly say why they designed it like that

They're sitting in the US, where they only have half the input voltage! So, in the design process this was fully OK and wouldn't justify a switching regulator. Then, when it's gone to production, noone looked again at it, as it basically works. Maybe, some software developer had "random software resets" on his todo list before they finally found it... :D

what have you been smoking ?

- RIGOL is chinese
- China is 220 volts...
- the machine has a switch to select 110 /220 ...
Professional Electron Wrangler.
Any comments, or points of view expressed, are my own and not endorsed , induced or compensated by my employer(s).
 

Offline Phantomix

  • Contributor
  • Posts: 12
  • Country: de
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #137 on: August 30, 2013, 06:29:38 pm »
what have you been smoking ?

- RIGOL is chinese
- China is 220 volts...
- the machine has a switch to select 110 /220 ...
just wanted to be funny

You're right, they sit in china, but also in the US and Germany, according to their website.
Quote
RIGOL Headquartered in Beijing, China, with the branch offices in Cleveland, OH,and Munich, Germany,
Yes, maybe they have developed it in China, I would not bet on this.


About the 110/220 switch:
Oookay, maybe. With such a design flaw, I wouldn't be surprised to find out about a 2nd design flow and the problem still having to do with 110/220
 

Offline Paul Moir

  • Frequent Contributor
  • **
  • Posts: 875
  • Country: ca
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #138 on: August 30, 2013, 06:41:02 pm »
Maybe I'm being naive, but how do you blow 700ma on a power supply user interface?  I mean, I understand that it's a benchtop power supply and that low power consumption would be just about the last thing to strive for, but to me that's seriously getting up there.

 

Offline FrankBuss

  • Supporter
  • ****
  • Posts: 2311
  • Country: de
    • Frank Buss
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #139 on: August 30, 2013, 06:44:09 pm »
Maybe I'm being naive, but how do you blow 700ma on a power supply user interface?  I mean, I understand that it's a benchtop power supply and that low power consumption would be just about the last thing to strive for, but to me that's seriously getting up there.
The display and background light can use quite a lot of power, if powered from the same 5V.
So Long, and Thanks for All the Fish
Electronics, hiking, retro-computing, electronic music etc.: https://www.youtube.com/c/FrankBussProgrammer
 

Offline c4757p

  • Super Contributor
  • ***
  • Posts: 7805
  • Country: us
  • adieu
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #140 on: August 30, 2013, 06:48:54 pm »
Maybe I'm being naive, but how do you blow 700ma on a power supply user interface?

Well, that's another option. Base model works for five minutes. $100 more and it shuts off the 500mA dummy load...
No longer active here - try the IRC channel if you just can't be without me :)
 

Offline Zbig

  • Frequent Contributor
  • **
  • Posts: 861
  • Country: pl
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #141 on: August 30, 2013, 06:50:38 pm »
The display and background light can use quite a lot of power, if powered from the same 5V.

Is CCFL still that widely used for LCD backlight purposes? I'm surprised they haven't switched to white LEDs exclusively by now.
 

Offline SeanB

  • Super Contributor
  • ***
  • Posts: 15309
  • Country: za
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #142 on: August 30, 2013, 07:12:58 pm »
Most likely RGB LED's or white ones. Not going to save much power wise, you still need about 200mA at 5V to light them on a display of that size.
 

Offline Galaxyrise

  • Frequent Contributor
  • **
  • Posts: 526
  • Country: us
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #143 on: August 30, 2013, 07:37:56 pm »
Dave, to get MAX heat dissipation in the main heatsinks, you'll have to measure the supply running at the LOWEST output voltage corresponding to the highest (Triac) transformer tap <you knew that!> - I think you said it was 22V. Then (~50v-22V) x 3A is 54W.

I was wondering the same thing.  And why tap 50V when the max output is 30V? That seems like a needless voltage drop across the transistor.  It's almost as if the person specifying the transformer voltages was speaking peak-to-peak but they got implemented as RMS.
I am but an egg
 

Offline sync

  • Frequent Contributor
  • **
  • Posts: 799
  • Country: de
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #144 on: August 30, 2013, 07:48:18 pm »
I was wondering the same thing.  And why tap 50V when the max output is 30V? That seems like a needless voltage drop across the transistor.  It's almost as if the person specifying the transformer voltages was speaking peak-to-peak but they got implemented as RMS.

Lower mains voltage. It should still works at ~200V.
 

Offline IanJ

  • Supporter
  • ****
  • Posts: 1057
  • Country: scotland
  • Pro EE guy many years ago, now a hobby/home biz.
    • IanJohnston.com
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #145 on: August 30, 2013, 07:49:22 pm »
With these design issues I wouldn't be surprised if the 220/120vac sticker is reversed....... :scared:

.............ok, I'll get my coat.

Ian.
Ian Johnston
www.ianjohnston.com
Manufacturer of the PDVS2 & PDVS2mini
 

Offline nack

  • Regular Contributor
  • *
  • Posts: 75
  • Country: nl
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #146 on: August 30, 2013, 07:50:16 pm »
Maybe they use a 50v transformer tap for yet another model. I believe I've seen a DP811 or DP821 which has a 60V dc output.

It seem more and
Ore likely they use identical hardware for their entire model range.

Edit: dp821a
http://www.rigol.com/prodserv/DP821A/
« Last Edit: August 30, 2013, 07:55:08 pm by nack »
 

Offline ant452

  • Contributor
  • Posts: 8
  • Country: au
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #147 on: August 30, 2013, 08:21:52 pm »
A lab power supply in which its power supply has a design issue....  Not good.
 

Offline Dread

  • Regular Contributor
  • *
  • Posts: 113
  • Country: jm
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #148 on: August 30, 2013, 08:48:32 pm »
A couple of points.
Dave, where are all the twelve year old trolls, who only a couple of videos back were claiming that you were somehow in the pay of the equipment suppliers, such as Rigol? Hellooooooooooo??? Nah, can't hear 'em... :-DD

While I was not one of those "twelve year old trolls" and I have made No such accusations about Dave, I would point out that this Video certainly had a different feel about it than Dave’s normal video's and I am pretty sure it was partly meant to address those people directly! You can read into that what you want, in any case I liked the teardown a lot.

Since this whole DP832 series started I have been saying that I would rather have my 2x (HP E3615A) power supplies any day of the week over a DP832 and that was met with a few negative replies.  The fact is that Dave and many of you guys talk a lot about the building things to a price point, using a $1 part versus $2.50 part but then the whole notion of that seems to go out the window when your buying equipment.  You need to use that same logic when buying test gear and reflect that same logic onto the company your thinking of buying from.

If your Rigol and your making a power supply, the first thing you do is work out a basic list of features and then you set a price point per unit so the engineers know what kind of budget they are designing around.   If you put that into a nice Pie chart and start to slice it up, in the case of the DP832 you need to take out a nice little slice for the microcontroller related section and include the costs involved for all the software development and the extra shielding etc.  For a power supply that sells for around $420 it probably costs about $220 to manufacture and at this low a price point once you take out the expenses involved in all the Microcontroller stuff and that nice Toroid your not left with a whole lot of money to make a proper linear power supply with high margins and a solid interior and exterior construction.

When you compare that to say an older HP/Agilent supply they used most of that pie chart on the Voltage/Current regulation design and put in a crap load of money testing those designs under all sorts of conditions.  The rest of the money was used on making a solid exterior.  On top of that your talking about a $800 unit in early 1990's money.

Every time I look at a piece of gear from China I look at how many bells and whistles they have added and I try to subtract that from what I figure the cost price is and then calculate out how much money they must have spent on the core engine of the device.

To me the Core is what really matters, what’s the point of having a Precision Lab Power supply with 100 bells and whistles if it’s output is not really precise or reliable?

When I saw the DP832 I knew the core engine had to be made on a very low budget, the rotary encoder knob was the first indication that money was very tight and it only got worst as I saw more.  I think if Dave was to look around further I think he would find that regulator is probably just one of many shortcuts they took to save money.  I would bet good money that almost every part in that unit from the regulators to the Caps are all picked from the lowest bidder and very little margin is built into the overal design.

BTW it was nice to see the Fluke 87V were it belonged, front and center.

Great Video Dave.
« Last Edit: August 30, 2013, 08:52:18 pm by Dread »
The Optimist says the glass is half full, the Pessimist says its half empty, an engineer only see's a glass that’s twice as big as it needs to be!
 

Online AndyC_772

  • Super Contributor
  • ***
  • Posts: 3523
  • Country: gb
  • Professional design engineer
    • Cawte Engineering | Reliable Electronics
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #149 on: August 30, 2013, 09:45:24 pm »
Your theory sounds likely, but that still doesn't explain why the +5 V rail appeared to be perfectly stable.

It would be kind an odd choice: use a switcher to go from 5 V to 1.8-3.3 V at 100 mA or so, but use a linear regulator from 12 V to 5 V at 700 mA. Would they have been worried about interference on the analog power supply board?
There are plenty of very small and efficient dc-dc converters which are designed for low voltage use and which can only accept inputs up to 5V or thereabouts. I use them all the time as my 'standard' way to generate all the low voltage rails needed for CPUs, FPGAs and the like, and other than that the QFN packages are a swine to solder and inspect properly, they do a superb job.

Check out www.enpirion.com for some good examples. The larger parts that can run off 12V are new, and over sized for many applications. The smaller parts, which have been around for longer and which are ideal for powering logic, can only tolerate up to 5.5V or so - hence the need for a linear pre-toaster.


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf