Author Topic: EEVblog #509 - Rigol 832 Lab Power Supply  (Read 123472 times)

0 Members and 1 Guest are viewing this topic.

Offline Electro Fan

  • Super Contributor
  • ***
  • Posts: 3160
Re: EEVblog #509 - Rigol 832 Lab Power Supply
« Reply #125 on: August 18, 2013, 08:09:14 pm »
Hi Guys,

First post here. Excuse my English, not my native language.

Hi Dave,

In the user manual for DP832 Rigol has a TIP:
 When powering on the Instrument after powering off it, make sure that the time interval between the two operations is greater than 5s.

Would it be possible for you to make another test and see if that spike is still there after 5s?

Thanks for the great reviews and info.

Hi g3org3,

Good find!  Hopefully Dave or someone here can try your suggestion and let us know if it has an impact on the spikes.

PS, your English is A-OK :)

Mike
 

Offline tinhead

  • Super Contributor
  • ***
  • Posts: 1918
  • Country: 00
    • If you like my hacks, send me a donation
Re: EEVblog #509 - Rigol 832 Lab Power Supply
« Reply #126 on: August 18, 2013, 08:31:46 pm »
at 38:48 there are 9sec with almost 5sec 10R load and Dave still got that spike, so no, this is not solution.
I don't want to be human! I want to see gamma rays, I want to hear X-rays, and I want to smell dark matter ...
I want to reach out with something other than these prehensile paws and feel the solar wind of a supernova flowing over me.
 

Offline MasterOfNone

  • Regular Contributor
  • *
  • Posts: 123
Re: EEVblog #509 - Rigol 832 Lab Power Supply
« Reply #127 on: August 18, 2013, 08:49:04 pm »
No matter how long you wait between power-ups on my PSU, the 2V pulse still occurs on chan1 and the 5V spike still occurs on chan2. I don’t see any issues with Chan3.
 

Offline leafi

  • Regular Contributor
  • *
  • Posts: 67
Re: EEVblog #509 - Rigol 832 Lab Power Supply
« Reply #128 on: August 18, 2013, 09:02:44 pm »
Dave thanks for the video this is making me looking at getting one.

The question I have for anyone is Dave did a start up transient with the output enable (not the power switch) and said it did not over shoot. If I'm not mistaken it was done at higher voltages.  Can anyone that has one of these confirm that they do not over shoot at lower voltages such as 1.2V, 1.8V, 3.3V 2.5V... I have a POS bk precision 1550 and it overshoots and their response is basically it is a crappy supply and to but a better one... WTF! Dont sell it if you all know it is a POS!
 

Offline MasterOfNone

  • Regular Contributor
  • *
  • Posts: 123
Re: EEVblog #509 - Rigol 832 Lab Power Supply
« Reply #129 on: August 18, 2013, 10:02:10 pm »
The question I have for anyone is Dave did a start up transient with the output enable (not the power switch) and said it did not over shoot. If I'm not mistaken it was done at higher voltages.  Can anyone that has one of these confirm that they do not over shoot at lower voltages such as 1.2V, 1.8V, 3.3V 2.5V... I have a POS bk precision 1550 and it overshoots and their response is basically it is a crappy supply and to but a better one... WTF! Dont sell it if you all know it is a POS!

You’ll have to wait for someone with equipment like an electronic load to get some good tests results for this PSU, but for 1.2, 1.8, 2.5 and 3.3 Volts, I didn’t see any overshoot on any of the channels (when each channels was enabled). The output was just the same type of smooth ramp you saw in the video. But that was with a 10 ohm load. With no load, sometimes it would ramp up the voltage even slower in a more linear fashion.
 

Offline Rexus

  • Newbie
  • Posts: 4
  • Country: gb
Re: EEVblog #509 - Rigol 832 Lab Power Supply
« Reply #130 on: August 18, 2013, 11:07:06 pm »
Having watched Dave's video I measured my tti ql355 power supply for switch on spikes, and was amazed to see 4 or 5 transients up to 20v, positive and negative, though only for about 100ns duration per transient.
 

Offline Electro Fan

  • Super Contributor
  • ***
  • Posts: 3160
Re: EEVblog #509 - Rigol 832 Lab Power Supply
« Reply #131 on: August 19, 2013, 12:08:25 am »
Having watched Dave's video I measured my tti ql355 power supply for switch on spikes, and was amazed to see 4 or 5 transients up to 20v, positive and negative, though only for about 100ns duration per transient.

Wow, TTi has been one of my holy grails / hopefuls on this issue, are you pretty sure?  (I'm sure you are sure, just checking :) )

Maybe we need a bit of written / clear-cut test protocol so that we are all looking at / testing the same things?

Once we get a simple test process then we could see if users of various PSs (Rigol, TTi, Agilent, and others) might be up for following the script and reporting the results.
 

Offline Electro Fan

  • Super Contributor
  • ***
  • Posts: 3160
Re: EEVblog #509 - Rigol 832 Lab Power Supply
« Reply #132 on: August 19, 2013, 12:54:21 am »
Just in case it helps anyone ID the problem/solution - and also test other 832's or other PS's, here is a quick recap of what (I think) were the tests and findings in the video starting at about 36:27:

1.   AC on, switch on Ch 1 no load – no noise
2.   AC on, switch on Ch 1 with 40 watt load – no noise
3.   AC on, switch on Ch 1 40 watt load - no overshoot 20ms/div
4.   AC on, switch on Ch 1 no load - good
5.   AC on, switch on Ch 3 5 volts no load - good
6.   AC off/on – 2V on output (first of these tests that seemed to get a different result than what was expected)
7.   AC off/on – 1.5V on output for 250ms
8.   AC off/on – with 10 ohm load ok, just ringing
9.   AC off/on – with no load some amount of transient output
10.   AC off/on – (2nd channel) negative 0.4 volts output
11.   AC off/on – (2nd channel) with 10 ohm load – some amount of transient output
12.   AC off/on – (2nd channel) with 10 ohm load – 4 volts transient for some amount of time

(Not all the test gear and displays were visible at all times.  This was the best I could do to transcribe the video based on what we could see and hear.  Feel free to add/correct any info.  Thx)

Just to "zoom out", it looks like the unit performs well when the AC power is already on and then an invdividual output is turned on; the issue seems to be when the the AC power is switched on - sometimes something leaks through the outputs?

Central to all of this is the question: Can a DC power supply output be expected to never leak DC when the AC power is turned on while the outputs are turned off?  Does anyone have such a power supply? 
« Last Edit: August 19, 2013, 01:15:02 am by Electro Fan »
 

Offline uwe

  • Newbie
  • Posts: 6
  • Country: de
Re: EEVblog #509 - Rigol 832 Lab Power Supply
« Reply #133 on: August 19, 2013, 07:16:24 am »
+1 for a teardown  :D i am thinking about ordering one, but if i get it before Dave has a teardown Video i have to disassemble mine...

:q! Uwe
 

Offline M0pmz

  • Newbie
  • Posts: 8
Re: EEVblog #509 - Rigol 832 Lab Power Supply
« Reply #134 on: August 19, 2013, 11:23:34 am »
Indeed. Does anyone have one? sounds like the right question:
An engineer I know actually services many m/f's Pro power supplies. Knows their various design intricacies - He has seen this before!

Word is Kikusui takes a lot of design care on kind of thing...  Quite upmarket gear!
Found one Kikusui DC supply in video here: 

...This m/f does a range of DC bench models as well apparently

Dave

...

Central to all of this is the question: Can a DC power supply output be expected to never leak DC when the AC power is turned on while the outputs are turned off?  Does anyone have such a power supply?
 

Online tom66

  • Super Contributor
  • ***
  • Posts: 6678
  • Country: gb
  • Electronics Hobbyist & FPGA/Embedded Systems EE
Re: EEVblog #509 - Rigol 832 Lab Power Supply
« Reply #135 on: August 19, 2013, 12:29:37 pm »
Having watched Dave's video I measured my tti ql355 power supply for switch on spikes, and was amazed to see 4 or 5 transients up to 20v, positive and negative, though only for about 100ns duration per transient.

Wow, TTi has been one of my holy grails / hopefuls on this issue, are you pretty sure?  (I'm sure you are sure, just checking :) )

Maybe we need a bit of written / clear-cut test protocol so that we are all looking at / testing the same things?

Once we get a simple test process then we could see if users of various PSs (Rigol, TTi, Agilent, and others) might be up for following the script and reporting the results.

Most likely it is simply the spark produced by the switch creating a wideband RF noise source, your circuit will not see that, but a high impedance scope input will.
 

Online PA0PBZ

  • Super Contributor
  • ***
  • Posts: 5121
  • Country: nl
Re: EEVblog #509 - Rigol 832 Lab Power Supply
« Reply #136 on: August 19, 2013, 12:31:01 pm »
Central to all of this is the question: Can a DC power supply output be expected to never leak DC when the AC power is turned on while the outputs are turned off?  Does anyone have such a power supply?

The easy way would be to put a real relay in the output on/off switch, when you depend on solid state stuff there is always the uncertaincy when you power it up.
Keyboard error: Press F1 to continue.
 

Offline Rexus

  • Newbie
  • Posts: 4
  • Country: gb
Re: EEVblog #509 - Rigol 832 Lab Power Supply
« Reply #137 on: August 19, 2013, 02:03:15 pm »
Having watched Dave's video I measured my tti ql355 power supply for switch on spikes, and was amazed to see 4 or 5 transients up to 20v, positive and negative, though only for about 100ns duration per transient.

Wow, TTi has been one of my holy grails / hopefuls on this issue, are you pretty sure?  (I'm sure you are sure, just checking :) )

Maybe we need a bit of written / clear-cut test protocol so that we are all looking at / testing the same things?

Once we get a simple test process then we could see if users of various PSs (Rigol, TTi, Agilent, and others) might be up for following the script and reporting the results.

Most likely it is simply the spark produced by the switch creating a wideband RF noise source, your circuit will not see that, but a high impedance scope input will.

You may well be right. When I disconnect the scope probes from the PSU I still see the same transients when switching on, though they reduce in amplitude by about half, so RF from the switch looks the likely source, well spotted.
 

Offline xchip

  • Regular Contributor
  • *
  • Posts: 58
  • Country: es
Re: EEVblog #509 - Rigol 832 Lab Power Supply
« Reply #138 on: August 19, 2013, 10:39:06 pm »
A stupid question :)

What do you use such a power supply for? I am wondering what sort of projects are you working on.... Same for the your 20Mhz+ oscillos.

I seem to be happy with my USB scope, batteries and sometimes my PC power supply, I want to know what am I missing out for being such a cheap ass :P

Cheers!

Offline Corporate666

  • Supporter
  • ****
  • Posts: 2008
  • Country: us
  • Remember, you are unique, just like everybody else
Re: EEVblog #509 - Rigol 832 Lab Power Supply
« Reply #139 on: August 19, 2013, 10:45:36 pm »
A stupid question :)

What do you use such a power supply for? I am wondering what sort of projects are you working on.... Same for the your 20Mhz+ oscillos.

I seem to be happy with my USB scope, batteries and sometimes my PC power supply, I want to know what am I missing out for being such a cheap ass :P

Cheers!

First project with the DP832 was a voltage level switch...  3.3v supply to the MCU which was switching a higher voltage (between 5V and 15V) with the switching level adjustable via a 10-turn pot.  In that case, I needed at least 2 outputs and it was nice to be able to set the DP832 to, say 6.32 volts and see what the turn-on point was, then make adjustments in software to get everything just right.

It's not always the most popular person who gets the job done.
 

Offline dausmus

  • Newbie
  • Posts: 1
Re: EEVblog #509 - Rigol 832 Lab Power Supply
« Reply #140 on: August 20, 2013, 03:14:21 am »
Yep, I agree with PA0PBZ: consider a relay or FET that isolates the outputs just long enough for stable operation at 0.000 volts (perhaps use a low cost power monitoring chip for that purpose),  then turn on the outputs, thus blocking any output instabilities. Then, if this option was selected for power on, let the unit do it's normal clean ramp to the remembered output settings.

Another poster posited (using other words) that perhaps this capability is there and just not working in firmware... could be an easy fix then.

Central to all of this is the question: Can a DC power supply output be expected to never leak DC when the AC power is turned on while the outputs are turned off?  Does anyone have such a power supply?

The easy way would be to put a real relay in the output on/off switch, when you depend on solid state stuff there is always the uncertaincy when you power it up.
 

Offline Crazy Ape

  • Regular Contributor
  • *
  • Posts: 181
Re: EEVblog #509 - Rigol 832 Lab Power Supply
« Reply #141 on: August 20, 2013, 03:32:53 am »
Central to all of this is the question: Can a DC power supply output be expected to never leak DC when the AC power is turned on while the outputs are turned off?  Does anyone have such a power supply?

The easy way would be to put a real relay in the output on/off switch, when you depend on solid state stuff there is always the uncertaincy when you power it up.
Yep, I agree with PA0PBZ: consider a relay or FET that isolates the outputs just long enough for stable operation at 0.000 volts (perhaps use a low cost power monitoring chip for that purpose),  then turn on the outputs, thus blocking any output instabilities. Then, if this option was selected for power on, let the unit do it's normal clean ramp to the remembered output settings.

Another poster posited (using other words) that perhaps this capability is there and just not working in firmware... could be an easy fix then.

Pretty sure this won't be fixed in firmware as the output glitch it likely caused by an unknown state before the onboard micro-controller has finished loading/booting/setup etc.
 

Offline Hypernova

  • Supporter
  • ****
  • Posts: 655
  • Country: tw
Re: EEVblog #509 - Rigol 832 Lab Power Supply
« Reply #142 on: August 20, 2013, 03:48:12 am »
Central to all of this is the question: Can a DC power supply output be expected to never leak DC when the AC power is turned on while the outputs are turned off?  Does anyone have such a power supply?

The easy way would be to put a real relay in the output on/off switch, when you depend on solid state stuff there is always the uncertaincy when you power it up.
Yep, I agree with PA0PBZ: consider a relay or FET that isolates the outputs just long enough for stable operation at 0.000 volts (perhaps use a low cost power monitoring chip for that purpose),  then turn on the outputs, thus blocking any output instabilities. Then, if this option was selected for power on, let the unit do it's normal clean ramp to the remembered output settings.

Another poster posited (using other words) that perhaps this capability is there and just not working in firmware... could be an easy fix then.

Pretty sure this won't be fixed in firmware as the output glitch it likely caused by an unknown state before the onboard micro-controller has finished loading/booting/setup etc.


Depending on how well they write the boot loader they can still minimize it. If they make sure that the very first instructions the MCU fetch on power on reset sets the GPIO pins to shut the FETs off they have a shot at cutting the pulse down to micro seconds.
 

Offline Dread

  • Regular Contributor
  • *
  • Posts: 122
  • Country: us
Re: EEVblog #509 - Rigol 832 Lab Power Supply
« Reply #143 on: August 20, 2013, 03:56:29 am »
I think some engineers and technicians get way too caught up in the test equipment on the bench and lose sight of what the hobby/profession is all about and the Rigol 832 is a perfect example of this.

I have two HP E3615A power supplies on my bench and I would not trade them for a 832 no matter what.  I made the mistake once of getting caught up in the Wiz bang Bling of a Chinese product when I bought my Owon 8202 (200 MHZ) scope to replace my older Tektronix Digital, this was long before it's short comings were known. To be fair it works OK but it's so dam time consuming to use and none of the features work perfectly or intuitively.

The 832 felt like a deja vu in that it does a lot of neat things but very few of them will I ever use in my daily work and most of them don't work perfectly or intuitively or give me a feeling of trust.   Since using the Owon and yes I know Dave said it's a POS but I also thought the scope he reviewed the following week was even worst but he was not as harsh with it as he was with the Owon but I guess that's pretty much how he felt at the time.

Anyway when I am working I want a power supply that I can 100% trust and is easy to use, these half baked Chinese products always try to Mimic the advanced features of their much more expensive rivals and they always end up doing a piss poor job at it and on top of that end up with core features that are unreliable eg the overshoot.  Would I trust a 832 to Bias a $300 RF FET, No Way!   Do I want to be constantly going into Menus and trying to remember how each piece of counter intuitive equipment on my bench works, No Way! All of this stuff is a distraction from the thoughts I am having as I trouble shoot a design or try to figure out why something is not working and that's the problem with a lot of this equipment, people spend more time playing with it than actually making or fixing things with it.

 BTW another example of good equipment made right is the Fluke 87V. I love my Fluke 87V, I can trust it and I must admit I find it amusing when Dave puts the 87V to the back and puts his Agilent front and Center whenever a critical test comes up and then we get to see the least two significant digits on the Agilent moving around like a slot machine.  Anyway I love Dave's videos I just wish that when he did equipment reviews they were done with a little less ratings emphasis on the rarely used features and concentrated more on the accuracy and usability of the basic features that most of us use 99% of the time.

Just my 2 cents.
« Last Edit: August 20, 2013, 03:59:46 am 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!
 

Offline ElectroIrradiator

  • Frequent Contributor
  • **
  • Posts: 614
  • Country: dk
  • More analog than digital.
Re: EEVblog #509 - Rigol 832 Lab Power Supply
« Reply #144 on: August 20, 2013, 08:35:48 am »
I think some engineers and technicians get way too caught up in the test equipment on the bench and lose sight of what the hobby/profession is all about and the Rigol 832 is a perfect example of this.

My thoughts exactly, but put into words better than I could have done.

The problem for Rigol and many similar low end manufacturers is, that a huge fraction of their potential customers are not very skilled when it comes to analog techniques and technologies. In particular many beginners tend to focus exclusively on the pure digital domain, as they can rather easily and quickly get something worthwhile accomplished. Who need this old school analog stuff anyway? We'll just throw a good A/D at it, and do everything in the programming, right? ;)

So at first blush it makes sense for Rigol to focus on the cheap frills, which can easily and inexpensively be added to the digital functionality. Add some LEDs, an LCD, some buttons and rotary encoders, and you have a huge opportunity for 'value added extras'. Better still, from the point of view of Rigol marketing, the beginner has a lot to talk about when 'evaluating' a product with many digitally implemented bells and whistles, and feel they can contribute meaningfully to the public debate.

Take my question earlier in the thread about whether the +5V rail ought to have had a +6V limit. Apparently people don't actually do any *testing* or *troubleshooting*, as opposed to just potentially wanting to use the PSU to power stuff. In my little world it is a 'done' thing to raise and lower rail voltages slightly away from nominal, as a natural part of testing and troubleshooting. I do that even for my hobby stuff, as it is easy to do. So to me having a low voltage rail with a limit of 'only' +5V feels a bit ... odd? :-//

Similarly, thinking of your $300 RF FET here, no-one but you has yet questioned how well this PSU holds up to a bit of RF riding shotgun on the power rails. As I suspect you will be aware, then some PSUs can go completely nuts, if they see just a hair of RF on their outputs. On the other hand it is a near certainty that Agilent's HP's Greybeards will have tested their PSUs to absolute destruction during development of their E36xx series and all the rest. They will know that their potential customers will use the PSUs to power RF equipment with the covers off etc.
 

Online PA0PBZ

  • Super Contributor
  • ***
  • Posts: 5121
  • Country: nl
Re: EEVblog #509 - Rigol 832 Lab Power Supply
« Reply #145 on: August 20, 2013, 08:42:34 am »
Depending on how well they write the boot loader they can still minimize it. If they make sure that the very first instructions the MCU fetch on power on reset sets the GPIO pins to shut the FETs off they have a shot at cutting the pulse down to micro seconds.

You can't do that in a few uSecs: the thing is powered from a 15-20 mS pulsed voltage which has to charge the caps, stabilize and activate the reset circuit of the MCU. Only after that you can switch off the output.
Keyboard error: Press F1 to continue.
 

Offline xchip

  • Regular Contributor
  • *
  • Posts: 58
  • Country: es
Re: EEVblog #509 - Rigol 832 Lab Power Supply
« Reply #146 on: August 20, 2013, 10:03:03 am »
Would I trust a 832 to Bias a $300 RF FET, No Way!

Well but this is an extreme example and not really representative of what people around here work on.. Also raising the voltage to 6V as ElectroIrradiator is not a good case since there are many simple ways to do it (add one battery?)

I'm looking for examples that would justify buying these scopes and power supplies, and you guys are perfect for that since you may be fully utilizing such equipment.. Could you guys give any examples?

I'd like to list 10 reasons/projects where I'd need such lab equipment (rather than a PC power supply unit)

Cheers




Offline Electro Fan

  • Super Contributor
  • ***
  • Posts: 3160
Re: EEVblog #509 - Rigol 832 Lab Power Supply
« Reply #147 on: August 20, 2013, 11:08:12 am »
Would I trust a 832 to Bias a $300 RF FET, No Way!

Well but this is an extreme example and not really representative of what people around here work on.. Also raising the voltage to 6V as ElectroIrradiator is not a good case since there are many simple ways to do it (add one battery?)

I'm looking for examples that would justify buying these scopes and power supplies, and you guys are perfect for that since you may be fully utilizing such equipment.. Could you guys give any examples?

I'd like to list 10 reasons/projects where I'd need such lab equipment (rather than a PC power supply unit)

Cheers

With your PC power supply how do you set the voltage for 3.3 volts and 0.25 amps?  And then what do you have to do to set it to 5.0 volts and 1.0 amp?  Or 20.5 volts and 4.5 amps?  How long does it take to make these settings and what do you have to do to display the volts and amps as you make these changes (connect 2 DMMs?  How accurate are those DMMs and how much did they cost?)  And if you want to toggle the PS back and forth between 1.0v and 5.0v for 80ms on each setting for 12 intervals how do you do that with your PC power supply?  With the PC power supply how quickly and confidently can you set volts to 1/100 of a volt or amps to 1/100th of an amp?  Ever strive to dial-in a millivolt or a milliamp?  How does your PC power supply deal with switching between constant voltage and constant current?  What indication does it give you that it's moving between CV and CC?   And when you first turn on the PC power supply how quickly does it hit the target voltage and with how much if any overshoot?  And how do you control the output from the PC power supply (do you have to turn off the AC input each time you want to shut off the output)?  If your PC power supply handles all this stuff pretty accurately and easily - and more importantly if it handles whatever you are doing with it - I'd say stick with it.

PS, how many batteries do you use in a year doing tests?  If the 1.5 volt batteries cost 40 cents and 9 volt batteries cost $2.00 and you go through 4 of each per month in about a year you could have a $100 power supply and in 3 years you could have a $300 power supply, and each year thereafter you'd be $100 ahead.   But if you aren't consuming batteries and the PC power supply has you covered then you are all set.
« Last Edit: August 20, 2013, 11:45:00 am by Electro Fan »
 

Offline madires

  • Super Contributor
  • ***
  • Posts: 7693
  • Country: de
  • A qualified hobbyist ;)
Re: EEVblog #509 - Rigol 832 Lab Power Supply
« Reply #148 on: August 20, 2013, 11:55:34 am »
Depending on how well they write the boot loader they can still minimize it. If they make sure that the very first instructions the MCU fetch on power on reset sets the GPIO pins to shut the FETs off they have a shot at cutting the pulse down to micro seconds.

A pulldown resistor for the MOSFET's gate (n-channel) is a very simple, reliable and cheap solution. There are things which can be done in hardware much better than in software  >:D
 

Offline xchip

  • Regular Contributor
  • *
  • Posts: 58
  • Country: es
Re: EEVblog #509 - Rigol 832 Lab Power Supply
« Reply #149 on: August 20, 2013, 12:03:45 pm »
With your PC power supply how do you set the voltage for 3.3 volts and 0.25 amps?  And then what do you have to do to set it to 5.0 volts and 1.0 amp?  Or 20.5 volts and 4.5 amps? 

Agreed but again give me real life examples where you needed that.. Or use cases that hobbiist and are likely to find in their projects.

Examples examples examples :)


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf