Siglent SDL1000X/SDL1000X-E Electronic Load

[Martin72]

I actually have everything together for the tests.
I also bought the small black power supply unit to have more than 3A available (10A).
Then an LED driver from meanwell with 1050mA constant current.
With the other power supplies I can test the maximum power, e.g. the 2x60V to 120V and 3 A are 360W.
Plus my Tektronix current clamp and a small siglent scope.
So it will start soon.
One thing I tried out earlier was the meanwell led driver.
You can actually see a significant ripple on the current.
More on this at the weekend.

[ddrl46]

Has anyone been successful in performing an adjustment on this unit? The procedure in the service manual is a bit confusing...

[SHF]

Hmm black power supply not nice   

[Martin72]

OK,
Then I already have a black one.

The procedure in the service manual is a bit confusing...

That's why Siglent withdrew it again, there were too many requests regarding the calibration.
However, it must also be said that calibration is only due when the load has run out of its specification.
This in turn can be tested and/or the "Verification Performance Test" from the service manual can be carried out.

[theblinkingman]

Can somebody please post the service manual with the calibration procedure again?

[theblinkingman]

I messed around a little bit with calibration and generally it seems to work like described.  The adjustment is a little bit coarse, so I didn't get it dialed all the way in. 

It's not mentioned in the manual, but you can save off the current calibration by running CAL:MEAS? and CAL:CTRL?. It returns something like 0.000668,1285,0.002859,1292,0.000084,884,0.000502,797 which are the "a" and "b" values you set with CAL:DATA <1 or 2>,"a","b".  1 is for the CTRL calibration and 2 is for the MEAS calibration.  The first two numbers are for the 36V CV range and the second two are for the 150V CV range.  I believe the last two a/b pairs are for CC mode for both ranges too.   

When setting the calibration, it applies to the voltage range that is currently active and currently active mode (CC or CV). 

Remember to CALCLS:VOLT before starting (and after you've saved your existing calibration).  CAL:ST stores and applies the calibration values.  The load doesn't like it when changing the CTRL values while it is enabled. 

Here's a quick script for calculating the a and b values for CV, but I found that I had to play around with the numbers afterwards to get a decent result. 

Code: [Select]

#!/usr/bin/env python3

# Voltage calibration in CV mode
# *_dmm is the voltage measurement done via external meter
# *_sdl is the voltage measurement displayed on the load
# y1 is CV at 1V and y2 is CV at 10V.  Set the power supply ~15V CC .1A
def calc(v_range, y1_dmm, y1_sdl, y2_dmm, y2_sdl):
    # start with voltage limit set to 1
    v_set = 1
    ctrl_const = 2000 if v_range == 36 else 1500
    # v_range = 36
    # v_range = 150

    ctrl_x1 = (65536 * v_set) / v_range
    meas_x1 = (65536 * y1_sdl) / v_range

    # next set voltage limit to 10
    v_set = 10
    ctrl_x2 = (65536 * v_set) / v_range
    meas_x2 = (65536 * y2_sdl) / v_range # Not used?

    ctrl_step = (y2_dmm - y1_dmm)/ (ctrl_x2 - ctrl_x1)
    ctrl_offset = ctrl_x1 - (y1_dmm / ctrl_step) + ctrl_const
    meas_step = (y2_dmm - y1_dmm) / (ctrl_x2 - ctrl_x1)
    meas_offset = meas_x1 - y1_dmm / meas_step

    print(f"CAL:DATA 1,{ctrl_step:.6f},{ctrl_offset:.0f}")
    print(f"CAL:DATA 2,{meas_step:.6f},{meas_offset:.0f}")


# First start with the 36V range
calc(36, 1.5111, 1.9507, 12.445, 10.934)
# Switch into 150V range and do it again (values should be different)
# calc(150, 1.5111, 1.9507, 12.445, 10.934)


[Mortymore]

My 6 month old Load is starting to look like this one.
Is this a known display problem?

After a month out for Siglent repair services, my SDL returned yesterday and under warranty the LCD and mainboard were replaced. New HW version 03-03-022
Welectron took care of all the shipping process through UPS. 

[tautech]

Request.

For those that are logging battery discharge please offer detailed instructions on setup and the discharge log capture.

Ideally to USB if supported and in csv format however via EasySDL is also acceptable.
A CSV for others to play with would be fantastic.
TIA

Didn't find such info after inspecting this complete thread......

[Mortymore]

The batteries I've been testing are those used in small UPS, Pb (AGM) 12v, 6Ah to 18Ah

As usual, the battery datasheet must be observed, but in general for 12V/7Ah, the settings are, I.discharge = 1.2A and V.stop=10.5v (1.75v/cell). Under this circumstances it is expected a 5h test duration and 6000mAh drawn from the battery, if in good condition.

Multiple test conditions could be adopted, but to my needs a 5h run to test a good battery with a discharge (6Ah) near the stated (7Ah) that would take 20h, is a good compromise.

Obs: The batteries are fully charged, prior to the discharge test be done

1) As an example, from the datasheet of the Ultracell UL 7-12 we can extract the test conditions to set (1.2A;10.5v), and what to expect (5h;6Ah)




Another test option could be for instance, the same V.stop=10.5V (1.75V/cell), but I.discharge=1.79A, and then should be expected for a good 12V/7Ah battery a runtime of 3hours, and 5.37Ah


2) As another example, according to the datasheet of the Ultracell UL18-12, with the discharged settings as I.discharge=3.06A and V.stop=10.5v (1.75V/cell), the test for a good battery should take 5h and the battery should provide about 15.3Ah (see attached datasheet Ultracell_UL18-12.pdf)


3) An example of a test to a Pb battery, 12v/7Ah in end of life



After 10 minutes discharge


After 40 minutes discharge


After 1 hour and 8 minutes discharge


EasySDL has a reported problem of the time axis, with the time stamp being half of the real time.


4) An example of a test to a Pb battery, 12v/7Ah as new (over 5h run and over 6Ah delivered)

[tautech]

EasySDL has a reported problem of the time axis, with the time stamp being half of the real time.

Thank you Mortymore 

Do you have EasySDL_V9.2 installed ?
https://int.siglent.com/upload_file/zip/software/DC%20Electronic%20Load/EasySDL_V9.2_EN.zip

I see you had Record engaged, did you set a path for it to be saved to disk, any disk ?
This is a feature I'd love to see fully explored so to be able to graph discharge curves from the recorded data.

Is it Winner-2_descarga_20230813.zip  ?

[Mortymore]

EasySDL has a reported problem of the time axis, with the time stamp being half of the real time.

Thank you Mortymore 

Do you have EasySDL_V9.2 installed ?
https://int.siglent.com/upload_file/zip/software/DC%20Electronic%20Load/EasySDL_V9.2_EN.zip

I see you had Record engaged, did you set a path for it to be saved to disk, any disk ?
This is a feature I'd love to see fully explored so to be able to graph discharge curves from the recorded data.

Is it Winner-2_descarga_20230813.zip  ?

Hi Rob

Yes I have the EasySDL v9.2 installed, and record csv files to the HardDrive.

The zip file previously attached contains a csv file "Winner-2_descarga_20230813.csv" that is recorded data from a battery test.

Now, attached is a zip containing a xlsx file, that has some text in portuguese but it's purpose is quite understandable, with some manipulated charts and graphs extracted from a couple of cvs files recorded from battery tests.

[tautech]

Excellent. 
I'll report the logging time bug right this instant.

Reported.....

[OP-SEC]

Hello All,

I've got a SDL1020X-E that I just purchased.  Today, I connected up a small 12V LiFEPO4 battery to do a battery test.  The battery is fully charged but, the SDL claims 0.002 - 0.003V is what it sees.  I have an inline load meter that is showing the battery at 13.3V.  If I enable CC at say 1A, the SDL puts a load of 1A on the battery and I can see that but, the voltage displayed by the SDL isn't right.  What am I doing wrong here?

[OP-SEC]

FWIW,  I spoke to Siglent Support and they indicated that I have a bad SDL.  New one should arrive on Wednesday.  I don't think they would have argued anyway but, when I told them that I was monitoring voltage at the input lugs of the SDL1020X-E with my SDM3065X and it agreed with my inline voltage/current monitor, it was pretty apparent that the SDL was the confused party.

[Johnny B Good]

 Yeah, a glance at the image told me all I needed to know to come to exactly that conclusion (even without your diagnostic testing results). 

[electr_peter]

On DC load screen there is a word "sense" in the top row. Can you check if sense terminals are enabled? If yes, then DC load would show 0V but still correctly sink current. Disable sense function if not needed.

[Johnny B Good]

In that case, it would be a fatal flaw foolish mistake (it's only fatal in the case of power supplies blessed with such remote sensing capability) if the design hadn't included internal bridging resistors (one to ten K each?) between the load and sense amp connections.

 I have an SDL1000X-E 200W (upgraded to 300W). I've never had cause to use the remote sensing feature so remain blissfully unaware of the possibility of such a design flaw and ICBA to test it out for myself. I'll leave that test for OP-SEC to try out and satisfy for himself whether that had been the problem in the first place.

[electr_peter]

In that case, it would be a fatal flaw foolish mistake (it's only fatal in the case of power supplies blessed with such remote sensing capability) if the design hadn't included internal bridging resistors (one to ten K each?) between the load and sense amp connections.

No, it is not a mistake. It functions exactly as described, many other DC loads do the same. There are other ways to integrate sense lines with main terminals, but there are implications to accuracy/complexity.
You may be confusing DC load sense with DC PSU sense, these are slightly different in function.

[Johnny B Good]

 Having considered your insight into this issue, I guess this choice of not providing internal resistor links to the sense amp, as is standard practice with PSUs, is simply to assure the user that the external sense connections are properly connected when this option has been selected.

 It hadn't occurred to me as to why such internal resistor connections would be omitted in the case of an electronic load. I still can't be bothered (at least not right now) to test this out for myself. Hopefully, OP-SEC will do this test with the replacement just to satisfy curiosity and post his result.

[martin1959]

Some weeks ago when I tried to update firmware from V1.1.1.23R1 to V1.1.1.23R2 the update failed with "Host update failed by udisk!" (using 1GB USB-Stick and filename SDL1000X_1.1.1.23R2.ADS).

Today I retried it with some different strategy and this time the firmware update WORKED:
- renamed the ADS file to 8.3 character MS-DOS compatible name "SDL1000X.ADS"
- ADS file saved to root directory of the USB-Stick
- using different USB-Stick, 8GB Sandisk USB-Stick, FAT32 format
- I found that the cluster size was 16384 bytes (instead of the 4096 bytes recommended by Siglent)

I think that renaming the ADS file to the short 8.3 filename and saving it to the root directory was the trick to make the update work.
I don't know whether the FAT32 cluster size in this case did make a difference or not. Normally I would go with the 4096 bytes recommended by Siglent, but I used that Sandisk 8GB USB-Stick out of the box and only afterwards checked the cluster size.

[tautech]

Some weeks ago when I tried to update firmware from V1.1.1.23R1 to V1.1.1.23R2 the update failed with "Host update failed by udisk!" (using 1GB USB-Stick and filename SDL1000X_1.1.1.23R2.ADS).

Today I retried it with some different strategy and this time the firmware update WORKED:
- renamed the ADS file to 8.3 character MS-DOS compatible name "SDL1000X.ADS"
- ADS file saved to root directory of the USB-Stick
- using different USB-Stick, 8GB Sandisk USB-Stick, FAT32 format
- I found that the cluster size was 16384 bytes (instead of the 4096 bytes recommended by Siglent)

I think that renaming the ADS file to the short 8.3 filename and saving it to the root directory was the trick to make the update work.
I don't know whether the FAT32 cluster size in this case did make a difference or not. Normally I would go with the 4096 bytes recommended by Siglent, but I used that Sandisk 8GB USB-Stick out of the box and only afterwards checked the cluster size.

Only this  ^^^ is required.

Long filename will work.
Installed from a folder works.
16k clusters also work.

Your old USB stick was the problem, for whatever reason, nothing else.

[Martin72]

I might want to install my load in my 19" rack - which Siglent rack kit would be suitable?

[tautech]

I might want to install my load in my 19" rack - which Siglent rack kit would be suitable?

SDG-RMK

Rack mount kit,  compatible with SDG800,  SDG1000X,  SDG2000X,  SDG5000, SDG6000X series generators and SDM digital multimeters, SDL1000X load, Height 3U