Author Topic: New Rigol DC Load DL3000 Series  (Read 77661 times)

0 Members and 1 Guest are viewing this topic.

Offline jackenhack

  • Contributor
  • Posts: 47
  • Country: se
    • Jackenhack Blog
Re: New Rigol DC Load DL3000 Series
« Reply #125 on: September 08, 2017, 06:05:17 pm »
Yeah, it's kind of weird (at least for my feeble brain) that what is essentially a MOSFET drawing power should introduce noise?
 

Offline Smokey

  • Super Contributor
  • ***
  • Posts: 2512
  • Country: us
  • Not An Expert
Re: New Rigol DC Load DL3000 Series
« Reply #126 on: September 08, 2017, 08:42:30 pm »
Yeah, it's kind of weird (at least for my feeble brain) that what is essentially a MOSFET drawing power should introduce noise?

Loads are just control loops, and having an unknown dynamic source (which itself is a control loop in the case of a regulated power supply) can lead to tuning and stability issues.  Really hot loops inject noise sometimes.
 
The following users thanked this post: jackenhack

Offline schopi68

  • Regular Contributor
  • *
  • Posts: 123
  • Country: de
    • My stuff page
Re: New Rigol DC Load DL3000 Series
« Reply #127 on: September 09, 2017, 10:05:37 pm »
Okay, measuring the ripple/noise has been the decision-maker. Rigol has to go back.

I got an Maynuo M9812 for comparison. It has been sent with an factory calibration sheet with good calibration data, but this was useless: dated 2015-10-28. Now, two years later - this device now has completely different data and is out of spec: its current setting has an zero offset of 1,95 mA in the 3A range (Tolerance is +-1,5 mA).

But the ripple/noise test gave good results.

With the Maynuo i got a noise/ripple measurement of <0,17 mV(Ueff) at 10mA load and < 0,4mV(Ueff) at 5A load in the range of 20 Hz-4 MHz.

This was the point where Rigol had a total failure. Starting at 1A it showed oscillations with 3,5 MHz at 30 mV(Ueff). When switching the Power Source off and on again, the oscillation stopped and i could measure a ripple up to 4,3 mV(Ueff) at 5A. 10 times worse than the Maynuo.

Switching the Rigol Load input off and on restarted the oscillation. I tried this with three different Power supplies and different lengths of the connection cables, this did not change the oscillation behavior, so it is very likely that this is a problem of the Rigol.
« Last Edit: September 14, 2017, 08:51:36 am by schopi68 »
 
The following users thanked this post: thm_w, kostasb, nugglix

Offline schopi68

  • Regular Contributor
  • *
  • Posts: 123
  • Country: de
    • My stuff page
Re: New Rigol DC Load DL3000 Series
« Reply #128 on: September 21, 2017, 08:30:31 am »
Arghhh... the last days i did some further measurements.  Totally showstopper on the Manyuo.. |O

With the rigol the user can pre-select the slew-rate. But in the Manyuo spec-sheet you even rarely find any information about the slew-rate. As it was mentioned somewhere in the spec-sheet i was thinking of about 2,5A/?s. But going deeper in detail i could see that the maynuo do not have a constant slew-rate. Regardless what step-size is done, the step always takes around 12 ?s. So i.e for a 1A Step this would be 0,08A/?s.

A rule of thumb is that if you want to measure the Transient recovery time of an power supply you should inject a step with a rise-time of at least 1/5 of the expected recovery time. Even an old and rubbish HP E3615A Power supply has a recovery time of 50 ?s - and therefore cannot be tested correctly with this device.  :wtf:

Rigol seems to win this competition.  :-//
 

Offline 2N3055

  • Super Contributor
  • ***
  • Posts: 6407
  • Country: hr
Re: New Rigol DC Load DL3000 Series
« Reply #129 on: September 21, 2017, 09:16:50 am »
In dynamic mode you can set rise an fall time independently, from microsecond to few hours... ???
 

Offline schopi68

  • Regular Contributor
  • *
  • Posts: 123
  • Country: de
    • My stuff page
Re: New Rigol DC Load DL3000 Series
« Reply #130 on: September 21, 2017, 09:35:01 am »
In dynamic mode you can set rise an fall time independently, from microsecond to few hours... ???

The smalles step that is executed in dynamic mode is 40 microseconds (even with an rise- or falltime-setting of 0). This seems to be the internal minimum setting time of the device.

A look at the spec-sheet gives the information that T1&T2 is in the range of 60 mikroseconds to 999s.

The 12 mikroseconds i've measured are fro 10% to 90%. Measuring from the visible start of the change to it's end on the trace (where it's flattened) are very close to 20 mikroseconds.

Maximum frequency of transition list is 25 kHz - this would be exactly 20 microseconds for rise or fall within a full sine wave.
« Last Edit: September 21, 2017, 09:41:45 am by schopi68 »
 

Offline 2N3055

  • Super Contributor
  • ***
  • Posts: 6407
  • Country: hr
Re: New Rigol DC Load DL3000 Series
« Reply #131 on: September 21, 2017, 11:25:53 am »
Sorry  if it sounds obvious, but how are you measuring this? Using current output on the back of unit?
 

Offline schopi68

  • Regular Contributor
  • *
  • Posts: 123
  • Country: de
    • My stuff page
Re: New Rigol DC Load DL3000 Series
« Reply #132 on: September 21, 2017, 01:07:59 pm »
I measured it on the back of the unit and as described in the BK precision application note https://bkpmedia.s3.amazonaws.com/downloads/guides/en-us/dc_electronic_load_application_note.pdf

(which is similar to this video)


Is there a know bandwidth-limitation on the back of the unit?

Assuming that the BK Precision devices are built electronically similar or equal to the Maynuo the same parameters will apply. BK says itself that the slewrate the devices are rated at are measured for a change from zero to full load. If there is a smaller change the slew-rates are "significant lower".

The BK8500 seems to be the equivalent to the Maynuo M9812 - according to the BK documentation it is rated at 0,5 A/microsecond - even worse than what i measured with the Maynuo.
( Page 20 of https://bkpmedia.s3.amazonaws.com/downloads/manuals/en-us/85xx_manual.pdf )

The BK8600 range has user selectable slew-rates, so it is not comparable to the Maynuo.
« Last Edit: September 21, 2017, 01:19:31 pm by schopi68 »
 

Offline 2N3055

  • Super Contributor
  • ***
  • Posts: 6407
  • Country: hr
Re: New Rigol DC Load DL3000 Series
« Reply #133 on: September 21, 2017, 02:30:16 pm »
I just checked with mine M9812..

Fastest rise time I achieved was 5 µsecs risetime. Funny thing is that you specify rise time, regardless how much current the jump is...

So I did 5 usec with 15A (3 A/usec), 10A (2 A/usec) and also  15 usec with 0,1A (0,006 A/usec).

So specification is true.. Analog part is obviously capable of sinking large currents fast..
But it cannot keep same A/sec speed at high and low currents.. On low currents it keeps same rise time.

 

Offline MotoDan

  • Contributor
  • Posts: 38
Re: New Rigol DC Load DL3000 Series
« Reply #134 on: September 21, 2017, 11:23:08 pm »
Just received my 3021A and am already disappointed after only 30 minutes of use. The first thing I did was load up the latest firmware (v00.01.01.00.09   2017-08-014).

Like Dave already pointed out, not allowing the CR and CP modes for the built-in battery test function is nuts! Have the Rigol engineers never seen a typical battery capacity test curve? Many capacity tests are done using constant resistance - not just constant current.

While doing some basic testing on a 12V/20A supply, I discovered that there appears to be an 8.0V upper voltage protection when using the CR mode in the 15 ohm range - regardless of what resistance is selected. The only way out of that loop is to remove or reduce the load to less than 8V.

I will continue testing the load and let you know what else I find.

Fortunately, the seller offers a 30-day money back guarantee. I realize that this is a relatively new piece of test gear and I really was looking forward to being happy with it, but if these bugs and blatant omissions are not fixed within my 30 days, then I'm afraid this puppy is going back and I'll just go buy one of the proven, less frilly $300 loads.





« Last Edit: September 26, 2017, 10:21:16 pm by MotoDan »
 

Offline schopi68

  • Regular Contributor
  • *
  • Posts: 123
  • Country: de
    • My stuff page
Re: New Rigol DC Load DL3000 Series
« Reply #135 on: September 22, 2017, 12:17:45 am »
I just checked with mine M9812..

Fastest rise time I achieved was 5 µsecs risetime. Funny thing is that you specify rise time, regardless how much current the jump is...

So I did 5 usec with 15A (3 A/usec), 10A (2 A/usec) and also  15 usec with 0,1A (0,006 A/usec).

So specification is true.. Analog part is obviously capable of sinking large currents fast..
But it cannot keep same A/sec speed at high and low currents.. On low currents it keeps same rise time.

In the meanwhile i made a somewhat better test setup for my M9812. Direct wiring, no cable longer than 15 cm.
I used a measurement shunt with 1 mOhm and one with 100 mOhm. The result was amplified by an HP 400F AC Voltmeter to scale in the range of 0,5 Volt/div to get a low noise result.

What i observed:
there seems to be an capacitor in the path. With the 0,1 Ohm resistor the rise times has been:
~80 microseconds at 0,01A
15 microseconds at 0,1A
10 microseconds at 1A
17 microseconds at 10A
21 microseconds at 15A

And now the values with the 1 mOhm resistor:
~10 microseconds at 0,1A
~2 microseconds at 1A
~2 microseconds at 10A
~3 microseconds at 15A

But: with the 1 mOhm resistor i had a huge overshot to almost twice the current value that was selected. After 10 to 20 microseconds the correct current was reached. Even with a current-selection of 20A it overshots up to 40A.

Another observation was, that the pulse-width is not constant and/or wrong.
My setup has been:
LevelA: 0
WidthA: 100ms
Rising: 0
LevelB: 1A
WidthB: 0.1ms <or 5ms>

with a width setting of 5ms i get a pulse-width of 6.1 ms in triggered pulse mode. In continuos mode i get 4.98ms.
with a width setting of 0.1ms i get a triggered pulse width of 1,2ms. in continuos mode it is at 0,96ms.

when i cange LevelB to 0.1A i can observe that the pulse length in continuos mode is fluctuating. With a widthB setting of 0.1ms the pulse length fluctuates between 0.02ms and 0.92ms. :(
A slight fluctuation is also visible at 1A - between 0.87 an 0.97 ms.

MotoDan: the battery-test in the Mayuno does also just CC. :( (or did i oversee some setting?)
 

Offline MotoDan

  • Contributor
  • Posts: 38
Re: New Rigol DC Load DL3000 Series
« Reply #136 on: September 22, 2017, 03:42:37 pm »
schopi68, re the absence of CR mode in the new Rigol loads, why not add a very useful and common battery test? These are full featured (and price) units and offer may options over the baseline loads. I think it was just an oversight on Rigol's part. Hopefully it will be added soon.
 

Offline SparkyM8

  • Newbie
  • Posts: 7
  • Country: gb
Re: New Rigol DC Load DL3000 Series
« Reply #137 on: September 22, 2017, 10:45:28 pm »
Anyone else notice the metering shunt value is only R0005 or 0.5mohm.
So 1mA resolution corresponds to 500nV. That doesn't sound too sensible to me.
By comparison, the 30A Maynuo uses a 16mohm shunt from 2 lengths of 2.5mm diam. Manganin wire positioned where the fan will keep it cool at higher currents.
Wouldn't be at all surprised if the DL3000 suffers from noise and accuracy issues at low currents.
« Last Edit: September 22, 2017, 11:17:22 pm by SparkyM8 »
 

Offline k4rlhp

  • Supporter
  • ****
  • Posts: 45
Re: New Rigol DC Load DL3000 Series
« Reply #138 on: September 25, 2017, 01:04:05 pm »
then I'm afraid this puppy is going back and I'll just go buy one of the proven, less frilly $300 loads.

Hey there!

Which one specifically did you have in mind? For around 300...400
 

Offline MotoDan

  • Contributor
  • Posts: 38
Re: New Rigol DC Load DL3000 Series
« Reply #139 on: September 26, 2017, 10:20:52 pm »
I'm trying to get my DL 3021A back to a firmware version earlier than the current 00.01.01.00.09 which I believe might be 00.01.01.00.03. Does anyone happen to have the GEL file for this version? I'm hoping it might not have the 8V OVP limit bug that shows up when selecting the 15 ohm range while in CR mode.
« Last Edit: September 26, 2017, 10:28:52 pm by MotoDan »
 

Offline Svuppe

  • Regular Contributor
  • *
  • Posts: 92
  • Country: dk
Re: New Rigol DC Load DL3000 Series
« Reply #140 on: September 27, 2017, 06:56:40 am »
I'm trying to get my DL 3021A back to a firmware version earlier than the current 00.01.01.00.09 which I believe might be 00.01.01.00.03. Does anyone happen to have the GEL file for this version?
I think I have one with .04 somewhere in the number. I will have a look for it later today.
 

Offline Svuppe

  • Regular Contributor
  • *
  • Posts: 92
  • Country: dk
Re: New Rigol DC Load DL3000 Series
« Reply #141 on: September 27, 2017, 02:48:27 pm »
Found it. It is version 00.01.00.04.10, but the filename is missing a .00. It is still available at Rigol, even though the link to it may be gone. Grab it here:
http://int.rigol.com/File/ModelSoftWare/20170604/DL3000%20Firmware_00.01.04.10.rar
 

Offline MotoDan

  • Contributor
  • Posts: 38
Re: New Rigol DC Load DL3000 Series
« Reply #142 on: September 27, 2017, 04:58:49 pm »
Thanks for the link Svuppe. Unfortunately that firmware also has the same 8V OVP issue. Wonder if anyone else has experienced this problem?
 

Offline TurboTom

  • Super Contributor
  • ***
  • Posts: 1380
  • Country: de
Re: New Rigol DC Load DL3000 Series
« Reply #143 on: October 07, 2017, 01:51:19 pm »
Got my hands on a DL3021 yesterday and was able to compare it to my Maynuo 9712. Initially I started experimenting just with a multimeter in diode test mode at its terminals. Immediately I found some severe shortcomings that shouldn´t happen to an experienced test equipment manufacturer, especially since already two firmware updates had been released:

In constant current mode, the output wouldn't switch off if you select it to via the on/off button, as long as the input current is approx. less than 5mA. A further investigations shows that this relates to zero current being displayed on the DL3021´s panel to be flowing in the external circuit. In this case, the instrument wouldn't switch off the passbank transistors. Since the instrument´s current measurement deadband is considerable at 3...5mA, you may end up connecting the device in "off load" mode to a circuit while the passbank is at a dead short. So far, i didn´t check how much current will flow upon connecting a powerful source and how long it will take the instrument to open the passbank and if it makes any difference, regarding the round about 4µF of capacitance across the DL3021´s input terminals that need to be charged. But anyway, I think this bug is an absolute no-go, off means off, regardless of the amount of external current flowing. I guess this can probably be addressed by a firmware update. Accuracy-wise in the low current ranges, the Maynuo wins against the Rigol hands-down. I wonder who came up with that brilliant idea to use a 500µOhms shunt resistor on that instrument...

Yet, as long as "macroscopic" currents are flowing, the load appears to work as it should (as far as I can tell).

Edit: What I really like about the load is the user interface (excluded the catastrophic front panel design). Configuring dynamic patterns is a no-brainer due to the elaborate graphic templates. Also, the current/voltage graph is a nice gimmick, yet I would have liked both current and voltage graph displayed contemporary in different colours on the same screen. Actually, that might be added in future via firmware update. Also, the choice to trigger the graph display properly would be a big plus (either by voltage or current)

What I don´t understand at all is that Rigol doesn´t appear to offer the user to chose which parameter (s)he wants to vary in [con]tinuous mode: As far as I cunderstood (I hope I´m wrong...), only current can be varied, neither voltage, resistance or power? Bummer! Rigol, I'm expecting a substantial firmware update. Along with the add-ons in battery test mode requested by contributors before.

In its current condition, I wouldn´t recommend to purchase the DL30XX. It´s got potential, no doubt, but this requires still a lot of initiative by Rigol, preferably in close contact with first-hour users of the instrument. If Rigol gets all these issues sorted, the DL30XX can turn into the "killer-load" and dominate the main stream market in this power class. But it´s still a considerable way to go I would say.


Cheers,
Tom
« Last Edit: October 07, 2017, 09:22:22 pm by TurboTom »
 
The following users thanked this post: thm_w

Offline schopi68

  • Regular Contributor
  • *
  • Posts: 123
  • Country: de
    • My stuff page
Re: New Rigol DC Load DL3000 Series
« Reply #144 on: October 09, 2017, 10:40:02 am »
Hi Tom,

could you do some dynamic measurements to compare it with the Maynuo (as in the pictures shown in the post below)?
https://www.eevblog.com/forum/blog/eevblog-1023-rigol-dl3021-electronic-load-teardown/msg1316074/#msg1316074
 

Offline TurboTom

  • Super Contributor
  • ***
  • Posts: 1380
  • Country: de
Re: New Rigol DC Load DL3000 Series
« Reply #145 on: October 09, 2017, 07:36:34 pm »
I´ll do my best but unfortunately that won´t happen before next weekend when I´ll again have access to the load.

Cheers,
Tom
 

Offline MotoDan

  • Contributor
  • Posts: 38
Re: New Rigol DC Load DL3000 Series
« Reply #146 on: October 10, 2017, 07:05:23 pm »
TurboTom,

I also have a new DL3021A load and am very disappointed with it. This is the higher priced unit which has the high resolution option as well as a color display. As you stated, if you are using it in a simple constant current mode at relatively high currents it seems to work OK. It has an issue with the constant resistance mode where it will throw up an over voltage protection fault when trying to select the 15 ohm range with a DC load over 8.0 VDC. The resolution is terrible (erratic) for the lower level loads. The unit will also not apply a load of less than about 4ma. I'm still waiting for a firmware fix for the OVP problem, but am probably going to have to return the unit as it is just not able to meet my initial (and basic) needs in its current state. I agree that the DL3000 series loads have the potential to be a really great piece of gear, however my guess is that it will take more than some code changes to get there. Might want to wait for the "B" models!
« Last Edit: October 11, 2017, 04:15:44 pm by MotoDan »
 

Offline Rbastler

  • Frequent Contributor
  • **
  • Posts: 286
  • Country: it
  • Wörk Wörk
    • Rbastlers Blog
Re: New Rigol DC Load DL3000 Series
« Reply #147 on: October 10, 2017, 07:06:59 pm »
Now I'm really glad I didnt buy it.

I will probably go for the Maynuo M9710.
http://rbastlerblog.jimdo.com/
Gamma spectrometer works. Now some yellow crystals need regenerating and testing.
 

Offline tzanetos

  • Newbie
  • Posts: 2
  • Country: us
Re: New Rigol DC Load DL3000 Series
« Reply #148 on: October 12, 2017, 04:10:14 am »
Rigol has released a new firmware update.

https://www.rigolna.com/products/dc-power-loads/dl3000/
[Updated Contents]
v00.01.01.00.09   2017-08-014
     - Optimize the waveform adjustment mode for Roll and Fast modes
     - Optimize Roll and Fast mode screenshots to support saved CSV format files
     - Roll mode to supports the waveform playback function
     - Roll mode to supports the waveform record function
     - Fast mode supports the selection of voltage and current curves
     - Support CC+CV function
     - Battery mode support W.H units
     - List mode show the results of the last loop after the run completes 
     - List mode support the recording and playback function ( Thumb drive), save the measurement results of each cycle, save the file format to CSV
     - The SCPI command increases the waveform mode adjustment and the readback command of the waveform data

It seems the two big problems are still:
1) The hot regulator/heatsink
2) The noisy loading which wasn't addressed by Dave, but shown by one of the forum members.

Are there any comparable programmable loads on the market in the same price range?
 

Offline schopi68

  • Regular Contributor
  • *
  • Posts: 123
  • Country: de
    • My stuff page
Re: New Rigol DC Load DL3000 Series
« Reply #149 on: October 12, 2017, 08:23:40 am »
Oh... thats interesting to see. On their website Rigol claims "Calibration Certificate included". The load i got one month ago had been shipped without a calibration certificate.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf