Author Topic: EEVblog #512 - Rigol DP832 Bad Design Investigation  (Read 322042 times)

0 Members and 1 Guest are viewing this topic.

alm

  • Guest
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #475 on: November 09, 2013, 01:35:12 am »
Running one power supply in CV and the other in CC is a perfectly standard way of operating parallel power supplies, so I don't see the problem.
 

Offline AndersAnd

  • Frequent Contributor
  • **
  • Posts: 572
  • Country: dk
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #476 on: November 12, 2013, 01:13:08 pm »
Just watched the video and short of replacing the LM317 with a bigger regulator and/or bigger heatsink the better solution would be to replace it with a switched supply (those 2$ ones on ebay should work well after a thorough check) as a preregulator to 6.5-7V and the heatsinked LM317 after it. It's a 15 mins and 3$ job compared to sending the unit over or causing it to fail and having to wait for another.
At least that's what I would do.
That's definitely not going to work.
LM317 has a 3 V drop-out voltage, which means you need > 8 V in (worst case scenario) to get a proper regulated 5 V out.
If you have anything < 8 V going in to a 5 V linear regulator you need to replace the LM317 with a pin-to-pin compatible adjustable LDO (Low Drop-Out) regulator like LD1086, LM1086-ADJ or LT1086 with a dropout voltage of max. 1.5 V.

Datasheets:
ST LM317 http://www.st.com/web/catalog/sense_power/FM142/CL1015/SC315/PF63704
TI LM317 http://www.ti.com/product/LM317
ST LD1086 http://www.st.com/web/catalog/sense_power/FM142/CL1015/SC312/PF259679
TI LM1086 http://www.ti.com/product/LM1086
Linear LT1086 http://www.linear.com/product/LT1086
 

Offline Monkeh

  • Super Contributor
  • ***
  • Posts: 7990
  • Country: gb
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #477 on: November 13, 2013, 05:22:35 pm »
LM317 has a 3 V drop-out voltage

No, no it doesn't.
 

Offline Rufus

  • Super Contributor
  • ***
  • Posts: 2095
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #478 on: November 13, 2013, 07:24:01 pm »
LM317 has a 3 V drop-out voltage

No, no it doesn't.

There are no specifications for operation with less than 3v  input to output. A datasheet might give you a plot of typical drop-out voltages against temperature and current which doesn't tell you how bad performance gets as you approach drop out or what maximums are. I wouldn't design a 317 into anything expecting less than 3v drop out.
 

Offline Monkeh

  • Super Contributor
  • ***
  • Posts: 7990
  • Country: gb
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #479 on: November 13, 2013, 07:43:13 pm »
LM317 has a 3 V drop-out voltage

No, no it doesn't.

There are no specifications for operation with less than 3v  input to output. A datasheet might give you a plot of typical drop-out voltages against temperature and current which doesn't tell you how bad performance gets as you approach drop out or what maximums are. I wouldn't design a 317 into anything expecting less than 3v drop out.

And the 3V figure isn't listed anywhere (the specifications are actually at Vin - Vout = 5V), nor is any fixed dropout voltage. The dropout voltage varies and how close you can get depends on how you're using the regulator.
 

Offline AndersAnd

  • Frequent Contributor
  • **
  • Posts: 572
  • Country: dk
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #480 on: November 13, 2013, 07:50:33 pm »
LM317 has a 3 V drop-out voltage
No, no it doesn't.

2.5 - 2.7 V @ full load (1.5 A) and max. junction temperature. Pretty close to 3 V.

But other specifications in the datasheets are often only specified at a range between 3.0 V <= VI-VO <= 40 V. Sometimes 2.5 V <= VI-VO <= 40 V.
Going under this regulation might get a lot worse than what's specified. Otherwise they would probably specify it down to a lower dropout voltage.
 
So you can tell what the specifications will be like below 3 V.

The dropout specs varies a bit depending on which manufactures datasheet you look in.

Also note Fairchild specs are only for a junction temperature between 0 and 125 deg C, while ST Micro is between -50 and 150 deg C.

« Last Edit: November 13, 2013, 08:02:51 pm by AndersAnd »
 

Offline AndersAnd

  • Frequent Contributor
  • **
  • Posts: 572
  • Country: dk
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #481 on: November 13, 2013, 08:37:07 pm »
And the 3V figure isn't listed anywhere...
It is at the top of page 3 in TI's LM317 datasheet: http://www.ti.com/lit/ds/symlink/lm317.pdf
Quote
Recommended Operating Conditions
VI – VO Input-to-output differential voltage: min. 3 V | max. 40 V
 

Offline Monkeh

  • Super Contributor
  • ***
  • Posts: 7990
  • Country: gb
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #482 on: November 13, 2013, 08:56:17 pm »
And the 3V figure isn't listed anywhere...
It is at the top of page 3 in TI's LM317 datasheet: http://www.ti.com/lit/ds/symlink/lm317.pdf
Quote
Recommended Operating Conditions
VI – VO Input-to-output differential voltage: min. 3 V | max. 40 V

And yet, not the datasheet for the LM317T, which is the TI part I typically use in TO-220.

Can we just agree that rather than blindly following figures which vary wildly between manufacturers, across the same manufacturer, and with the phase of the moon, we should actually design and test things?
 

Offline AndersAnd

  • Frequent Contributor
  • **
  • Posts: 572
  • Country: dk
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #483 on: November 13, 2013, 08:57:35 pm »
LM317's (ST brand) I have bought recently also have the new thin-style tap design. Copper is expensive these days I presume...
Some LM317 TO-220 single gauge vs. dual gauge part numbers:

Fairchild www.fairchildsemi.com/ds/LM/LM317.pdf
LM317T = Single Gauge
KA317TU = Dual Gauge

ST Micro www.st.com/st-web-ui/static/active/en/resource/technical/document/datasheet/CD00000455.pdf
LM317T = Single Gauge
LM317T-DG = Dual Gauge

TI  www.ti.com/lit/ds/symlink/lm317.pdf
LM317KCT = Single Gauge
LM317KCS = Dual Gauge
 

Offline AndersAnd

  • Frequent Contributor
  • **
  • Posts: 572
  • Country: dk
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #484 on: November 13, 2013, 09:01:57 pm »
And yet, not the datasheet for the LM317T, which is the TI part I typically use in TO-220.
Do you have a link for the TI LM317T datasheet? I can't find it at TI's website.
 

Offline Monkeh

  • Super Contributor
  • ***
  • Posts: 7990
  • Country: gb
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #485 on: November 13, 2013, 09:02:59 pm »
And yet, not the datasheet for the LM317T, which is the TI part I typically use in TO-220.
Do you have a link for the TI LM317T datasheet? I can't find it at TI's website.

It's under 117. http://www.ti.com/lit/ds/symlink/lm117.pdf

It does reference 3V for some of the specs, having looked closer, but not in relation to dropout.

Test it and see is the best solution. Or move to a more modern regulator.
 

Offline Rufus

  • Super Contributor
  • ***
  • Posts: 2095
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #486 on: November 13, 2013, 09:06:32 pm »
Also note Fairchild specs are only for a junction temperature between 0 and 125 deg C, while ST Micro is between -50 and 150 deg C.

They are not specifications they are typical characteristics giving guidance on what may happen if you operate outside the specifications.
 

Offline AndersAnd

  • Frequent Contributor
  • **
  • Posts: 572
  • Country: dk
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #487 on: November 13, 2013, 09:11:23 pm »
Also note Fairchild specs are only for a junction temperature between 0 and 125 deg C, while ST Micro is between -50 and 150 deg C.

They are not specifications they are typical characteristics giving guidance on what may happen if you operate outside the specifications.
Correct. Junction temperature isn't specified to more than 125 deg C even though the curve goes all the way up to 150 deg. C.
 

Offline AndersAnd

  • Frequent Contributor
  • **
  • Posts: 572
  • Country: dk
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #488 on: November 13, 2013, 09:14:14 pm »
And yet, not the datasheet for the LM317T, which is the TI part I typically use in TO-220.
Do you have a link for the TI LM317T datasheet? I can't find it at TI's website.

It's under 117. http://www.ti.com/lit/ds/symlink/lm117.pdf

It does reference 3V for some of the specs, having looked closer, but not in relation to dropout.

Test it and see is the best solution. Or move to a more modern regulator.
I haven't really looked into it, but what's the difference between LM317 and LM317T?

Is it a case of one being a design from the old National Semiconductor and the other being a design from the old TI?
« Last Edit: November 13, 2013, 09:24:15 pm by AndersAnd »
 

Offline Monkeh

  • Super Contributor
  • ***
  • Posts: 7990
  • Country: gb
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #489 on: November 13, 2013, 09:22:54 pm »
And yet, not the datasheet for the LM317T, which is the TI part I typically use in TO-220.
Do you have a link for the TI LM317T datasheet? I can't find it at TI's website.

It's under 117. http://www.ti.com/lit/ds/symlink/lm117.pdf

It does reference 3V for some of the specs, having looked closer, but not in relation to dropout.

Test it and see is the best solution. Or move to a more modern regulator.
I haven't really looked into it, but what's the difference between LM317 and LM317T?

Is it a case of one being a design from the old National Semiconductor and the other being a design from the old TI?

Your guess is as good as mine. There are a lot of parts like that from TI, few have any difference in specification at all.

I've bought the T/AT because it often seems more readily available than the KCS, so I get a proper tab. I gave up on the lottery of ST parts from most distributers, and my experience with Fairchild has been popped parts and wild variation in specification compared to other manufacturers.
 

Offline AndersAnd

  • Frequent Contributor
  • **
  • Posts: 572
  • Country: dk
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #490 on: November 13, 2013, 09:26:46 pm »
In the LM117 datasheet there's one table for LM117 and another table for LM317A + LM317-N.

I guess LM317-N just a placeholder for all the LM317 versions in the datasheet other than LM317A? E.g. LM317T, LM 317H, LM317K LM317M etc.
 

Offline Monkeh

  • Super Contributor
  • ***
  • Posts: 7990
  • Country: gb
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #491 on: November 13, 2013, 09:30:15 pm »
In the LM117 datasheet there's one table for LM117 and another table for LM317A + LM317-N.

I guess LM317-N just a placeholder for all the LM317 versions in the datasheet other than LM317A? E.g. LM317T, LM 317H, LM317K LM317M etc.

I believe so. The A is specified down to -40C, the other versions are not.
 

Offline AndersAnd

  • Frequent Contributor
  • **
  • Posts: 572
  • Country: dk
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #492 on: November 13, 2013, 10:23:18 pm »
Can we just agree that rather than blindly following figures which vary wildly between manufacturers, across the same manufacturer, and with the phase of the moon, we should actually design and test things?
Rigol strongly disagrees  >:D :-DD  :-BROKE

But I wouldn't run a part outside it's recommended operating range in a commercial product even if it tests fine there. You never know if it will be the same for the next batch. And you wouldn't be able make any kind of warranty claim against the manufacturer if you run them outside specs.

I know Rigol overclocked 40 MHz ADCs to 100 MHz with great success in the DS1xxxE scopes. They did so because there's 10 ADC in each device and lot of money to be saved for the 40 MHz vs. 100 MHz versions.
But they are from the same die as the 100 MHz parts. Not sure if Rigol tested and sorted each of them before use or if almost all or all of them worked overclocked and they just took the risk.
 

Offline AndersAnd

  • Frequent Contributor
  • **
  • Posts: 572
  • Country: dk
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #493 on: November 21, 2013, 02:55:04 pm »
Hi - I have some pictures of the new boards.

For Australia and New Zealand customers we are expecting the new boards to be leaving China around October 17th. I'm not sure on the rest of the world.






Looks like Rigol now use the same hug TO-220 heatsink as they use in the DS1000E series power supply (except for the missing black anodizing):

https://www.eevblog.com/forum/projects/rigol-1052e-fan-replacement/msg104898/#msg104898
« Last Edit: November 21, 2013, 02:58:59 pm by AndersAnd »
 

Offline AndersAnd

  • Frequent Contributor
  • **
  • Posts: 572
  • Country: dk
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #494 on: November 21, 2013, 03:01:15 pm »
 

Offline Skimask

  • Super Contributor
  • ***
  • Posts: 1433
  • Country: us
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #495 on: November 26, 2013, 01:14:01 am »
Bought my DP832 direct from Rigol on Nov 7th.  Used it a dozen times or so.  Powered it up 2 days ago, Channel 1 failed.  Wouldn't source any current.  Channel 2 & 3 work just fine.  Just the way things go.  Either they fail right away or they last for decades, right?

Contacted Rigol this morning (~12 hours ago) via email, explained the issue, etc.  Within a few hours, got an RMA and shipping label.  Dropped it off for shipping a couple hours ago.  I've already got a tracking number for the new unit that's shipping back to me.

 :-+You can't beat that kind of service...Can you? :-+
Ok, sure, maybe it could NOT have failed within the first couple of weeks.  But other than a handful of minutes packing up the old unit, email'ing, dropping it off, etc., how much am I out?
I didn't take it apart.
I turned it on.

The only stupid question is, well, most of them...

Save a fuse...Blow an electrician.
 

Offline BravoV

  • Super Contributor
  • ***
  • Posts: 7547
  • Country: 00
  • +++ ATH1
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #496 on: November 26, 2013, 01:44:09 am »
:-+You can't beat that kind of service...Can you? :-+
Ok, sure, maybe it could NOT have failed within the first couple of weeks.  But other than a handful of minutes packing up the old unit, email'ing, dropping it off, etc., how much am I out?

Not if you purchased dozens or hundredths of these that you're going to use in an industrial/commercial usage or worst to be used in your production benches testing lines, imagine how many productivity hours lost because of that.

I think all these DP832 fuss summarized it self pretty well, its suitable only for personal or enthusiast/hobbyist grade usage only, and definitely should be avoided in pro/commercial environment that need a product that just work reliably out of the box when newly bought, not too demanding I guess.  ::)

Offline AndersAnd

  • Frequent Contributor
  • **
  • Posts: 572
  • Country: dk
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #497 on: November 26, 2013, 09:10:41 am »
Still nothing for EU owners. I emailed them again, no response.
Who did you e-mail, http://eu.rigol.com or a European distibutor?
 

Offline mcinque

  • Supporter
  • ****
  • Posts: 1129
  • Country: it
  • I know that I know nothing
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #498 on: November 26, 2013, 12:25:30 pm »
Not if you purchased dozens or hundredths of these that you're going to use in an industrial/commercial usage or worst to be used in your production benches testing lines, imagine how many productivity hours lost because of that.

The same example is valid also for other manufacturers.
 

Offline Skimask

  • Super Contributor
  • ***
  • Posts: 1433
  • Country: us
Re: EEVblog #512 - Rigol DP832 Bad Design Investigation
« Reply #499 on: November 26, 2013, 01:16:09 pm »
Not if you purchased dozens or hundredths of these that you're going to use in an industrial/commercial usage or worst to be used in your production benches testing lines, imagine how many productivity hours lost because of that.

I think all these DP832 fuss summarized it self pretty well, its suitable only for personal or enthusiast/hobbyist grade usage only, and definitely should be avoided in pro/commercial environment that need a product that just work reliably out of the box when newly bought, not too demanding I guess.  ::)
Well sure...can't argue that point.
But it wasn't and I'm happy...so...
I didn't take it apart.
I turned it on.

The only stupid question is, well, most of them...

Save a fuse...Blow an electrician.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf