Siglent SPD3303X(-E) calibration FW1.01.01.02.07R2 - best practice

[BreakingOhmsLaw]

[Nov.11,2023 Updated telnet information]

Hello everybody,

I recently acquired a Siglent 3303X-E, hacked it to 1mV/mA resolution as per the known method.
The calibration on channel 1 was a bit off, so i tried the calibration in the service manual:
https://siglentna.com/wp-content/uploads/dlm_uploads/2019/08/SPD3303C-Service-Mannual.pdf

However, it does not work as stated in the manual, the procedures in the manual are flawed. I always ended up with either the displayed voltage or the actual voltage about 5% off.  After much trial and error, here are the correct procedures that ended up getting all values extremely accurate. I hope this saves some of you of going down that rabbit hole.

Some side notes:
- You cannot use EasyPower to send the SCPI commands. It seems to (intentionally?) botch the calibration commands. The best way is to use a telnet client (like https://putty.org/) to establish a telnet connection to port 5025. Or you can use NI VISA MAX instead. If you do, you must use "Write" in the NI VISAVisa Test Panel, not "Query".

 - If the unit beeps, it did not accept the calibration command. Check your input. (Use a period, not a comma in the voltage value.)
 - I got slightly more accurate results when doing the voltage calibration with a 100mA load rather than open load. It seems to me they are not tapping the sense lines close to the binding posts.
- Check the box at Configuration | I/O Settings | Enable termination character and select "Line Feed \n". Update: not necessary
 - you need to restart communication after the power cycles

Follow the procedures closely.

Display voltage calibration
First, clear the voltage calibration values for the channel you want to calibrate:
Channel 1:
*CALCLS 0
*CALCLS 1
Channel 2:
*CALCLS 4
*CALCLS 5
Store with
*CALST
...and power cycle the unit. Edit: Some users report this is not necessary.
Now, hook up the unit to a good multimeter and set the voltage to 1V output.
Both displayed and output voltage should be wrong. Measure the output value with your multimeter.
Set the calibration point 1:
CALibration:VOLTage ch1,1,<measured voltage>
or ch2 if you are calibrating CH2

Next, set the unit to 25V output and set calibration point 2:
CALibration:VOLTage ch1,2,<measured voltage>
or ch2 if you are calibrating CH2

That completes the calibration, but: the displayed value will still be wrong. Don't panic! You now need to change the output value to any other voltage, only then will it update the display using the new calibration value. This had me chasing rabbits for 2 hours.
Save with
*CALST
...and enjoy your calibrated PSU.



Current display:
Clear old calibration first:
*CALCLS 3 for CH1
*CALCLS 7 for CH2
*CALST
Power cycle the unit Edit: Some users report this is not necessary.
Set PSU channel to 10V, 3A.
Connect to an electronic load and set it to 100mA constant current. Check actual current with multimeter and enter into the command below:
CALibration:CURRent ch1,1,<measured current value> or ch2 for CH2
Change the electronic load to 2.5A constant current. Again, actual current with multimeter and enter into the command below:
CALibration:CURRent ch1,2,<measured current value> or ch2 for CH2
Store calibration with
*CALST

My unit had <0,5mA accuracy after this.

Current limit calibration:
The calibration procedure in the manual is flawed and will screw up the CC badly.
Do this instead:
*CALCLS 2 for CH1
*CALCLS 6 for CH2
*CALST
Power cycle the unit
Set the electronic load to 3A constant current.
Set the channel to 10V, 500mA  (NOT 100mA as written in the service manual!)
CALibration:CURRent ch1,3,<measured current value> or ch2 for CH2
Not set the crrent limit of the channel tp 2.5A
CALibration:CURRent ch1,4,<measured current value> or ch2 for CH2
Now set the current limit of the channel to 2.4A. This updates the CC and actual current should be very close to 2.4A.
Store calibration with
*CALST

The CC should now be within 0,5mA accuracy.

Please post results if you did this.

[jeraymond]

Hi, thanks for this info. I was following the guide and using 100mA for the current calibration and it was not working at all. Using 500mA as you suggested worked great!

Not set the crrent limit of the channel tp 2.5A
CALibration:CURRent ch1,4,<measured current value> or ch2 for CH2
Not set the current limit of the channel to 2.4A. This updates the CC and actual current should be very close to 2.4A.
Store calibration with
*CALST

Here I presume you mean 'Now set the current...'. I also presume you meant 2.5A throughout not both 2.5A and 2.4A.

When I did my cal, I would toggle the channel output on/off after *CALCLS and to see the new calibration parameter in use. I didn't have to power cycle the device.

[BreakingOhmsLaw]

Hi, thanks for this info. I was following the guide and using 100mA for the current calibration and it was not working at all. Using 500mA as you suggested worked great!

Not set the crrent limit of the channel tp 2.5A
CALibration:CURRent ch1,4,<measured current value> or ch2 for CH2
Not set the current limit of the channel to 2.4A. This updates the CC and actual current should be very close to 2.4A.
Store calibration with
*CALST

Here I presume you mean 'Now set the current...'. I also presume you meant 2.5A throughout not both 2.5A and 2.4A.

When I did my cal, I would toggle the channel output on/off after *CALCLS and to see the new calibration parameter in use. I didn't have to power cycle the device.

Thanks, corrected "Not" to "Now".
Changing the current limit to 2.4A is only done to update the setting. You could also change it and go back to 2.5.

[HamDancer]

I did this with my SPD3303X-E upgraded to SPD3303X and it worked great! Thanks so much for putting this together. Despite multiple attempts I didn't quite get 0.5mV/mA accuracy across the entire range, but I definitely got 1mV/mA, and usually much better.

[eloso]

Hi,

I was more or less happy with my SPD3303X-E and didn't feel the need for additional resolution on voltage, but getting the additional resolution on current has finally proved a draw for me and I upgraded following the usual instructions (thanks everyone involved!).

Channel 2 was more or less spot-on but channel 1 was out by more than 10mV from the setpoint reading to the displayed reading of output which more or less agreed with the SDM3055 reading of the output. The current was bang-on on both channels.

I followed the calibration procedure just for voltage on Channel 1 and it worked fine the third time. The first twice I can't see that I did anything differently and the commands seemed to be accepted (no beeping) but whenever I went back to the 1V setting I was 200mV out. When it worked the third time I was bemused because I genuinely couln't see what I had done differently and in fact felt a bit foolish for doing something a third time when the first twice had not worked.

So it's a mystery. There will be an answer - there has to be a reason - but I think we will never know it.

Anyway, I have quit while I am ahead - the instrument seems spot on now although I can't comment on any longer term drift or cycling. Its certainly good enough for my purposes as a power supply and even if the absolute accuracy drifts after a while, the resolution is going to be very useful to have just to see short tem relative change in current in the region of mA.

Thanks for providing the extra info here.  I guess knowing that others had made this work is what made me perservere. That's why I am posting my own report - it doesn't add anything other than saying, yes, it can work so don't be dejected if it doesn't at first. Or at second....   

Regards


Eloso

[doppelgrau]

- You cannot use EasyPower to send the SCPI commands. It seems to (intentionally?) botch the calibration commands. Use NI VISA MAX instead.

Also telnet on port 5025 didn't work for me. Every *CALST resulted in freezing the power supply (became completely unresponsive).
But following it to the letter using my "windows backup device" helped.

Please post results if you did this.

Done, went really well.

[KG7AMV]

There was a guy on a facebook group or Instagram that was in process of writing a Labview (Community Edition) GUI for Calibration. I Cannot find him or what group he was in but I thought it was a good idea and so did a few others. Another guy made some interesting stuff also.

[Martin_B]

Thanks very much!
This works very well.

The mean thing is that the EasyPower.exe allows deleting the values, but not setting the new values.
I almost despaired until I found this thread.

The displayed and actual (V) values are now accurate to the mV on both channels.

I will take care of the calibration of the current values, but I need an electronic load...

[JOFlaherty]

Should the thread title reference SPD3303, rather than SDS3303?

[MathWizard]

Yeah it's SPD3303X for the 1mV/1mA version, and -E for the 10mV/10mA version. I hacked mine, and it's about 5 years old. And for very light loads, anyways, the voltage accuracy is doing great. I should do a few simple tests and write it down, see if it changes in a year. I don't use it all the time, so it should age slowing.

[XerXes777]

Thanks for this !!!
But i have a problem.
The Unit peeps after send a command like CALibration:VOLTage ch1,1,0.8591.

I have to say, that i delete all my calib data 
And i dont have a backup.

What could be the error?

EDIT: I use putty and choose telnet as connection typ and this doesnt work.
with NI MAX the commands are work.

could this be a problem with the syntax? when i use putty and i use this CALibration:VOLTage\sch1,1,0.8596 with an \s than the beep is gone but it doesnt work too.

[Bad_Driver]

Today I went trough the process with some hickups but sucessful at the end. As always with Siglent gear I used Putty as terminal.
I needed 3 or 4 attempts to accomplish a step. Reboot was not needed.
My SPD accepts for the reference measurements only 3 digit behind the period!

example: CAL:VOLT ch1,1,0.833

With 4 digits it beeps.

Thanks to BreakingOhmsLaw for his efforts!

[Tjuurko]

With firmware version 1.01.01.02.07R2 it beeps (any number of signs after 0.).
With firmware version 1.01.01.02.05, calibration commands are accepted without errors.

My calibration commands:
CAL:VOLT CH1,1,0.8195
CAL:VOLT CH1,2,25.5040
CAL:VOLT CH2,1,0.8060
CAL:VOLT CH2,2,25.5015

4 years ago they were:
CAL:VOLT CH1,1,0.82188
CAL:VOLT CH1,2,25.5025
CAL:VOLT CH2,1,0.80669
CAL:VOLT CH2,2,25.4942

[slugrustle]

Does anyone know if this procedure works for the SPD3303C?

I'm going to run performance verification shortly, and hopefully my unit doesn't need adjustment.  I'm just curious how to perform the adjustment if needed, and whether anyone has tried.  A little worried about erasing calibration info as the first step in case the subsequent steps aren't the same for the SPD3303C.

EDIT: Performance verification passed just fine.  Still would like to know how to run the adjustment.

[sy]

Connect to an electronic load and set it to 100mA constant current. Check actual current with multimeter and enter into the command below:
CALibration:CURRent ch1,1,<measured current value> or ch2 for CH2

Hello, I have just hacked my X-E and wish to calibrate my unit. I new to this, I was wondering why we need an electronic load and if we can just use the multimeter to measure current output? If not, what kind of electronic load did you use?

[Mortymore]

If the PSU is new, and you don't have a good multimeter and an eletronic Load, my suggestion is that you shouldn't be concern with the calibration for now, unless your PSU seems to be far off.
Let the PSU age a few months.

With only a multimeter (5.5 digits would be good) you can calibrate the voltage, and let alone the current calibration.

You will need an electronic load to draw the current to be set from the PSU, and then you can calibrate it.
I only calibrate my hacked PSU, 2 or 3 months after buying and using an SDL1020X-E load.

The calibration process is no easy task, and you have to be patient. Let all the devices on for 1h or so, before start the calibration process, take your time, and let each adjusted value set for a few seconds, at least.
 

[BreakingOhmsLaw]

Connect to an electronic load and set it to 100mA constant current. Check actual current with multimeter and enter into the command below:
CALibration:CURRent ch1,1,<measured current value> or ch2 for CH2

Hello, I have just hacked my X-E and wish to calibrate my unit. I new to this, I was wondering why we need an electronic load and if we can just use the multimeter to measure current output? If not, what kind of electronic load did you use?

You could just hook up beefy load resistors and measure the resulting current with a multimeter, that will work just fine. An electronic load just makes this much more easy. I use a Siglent SDL1020X-E electronic load and run the current though a calibrated R&S HMC8012 benchtop multimeter for added accuracy.

[martin1959]

Yesterday I had problems trying to calibrate the Current Display values of my SPD3303X-E (updated to SPD3303X, FW 1.01.01.03.10, HW V6.1). Fortunately I was able to resolve it, so I describe the problem and the solution here.

I first tried to enter the calibration commands using the "telnet" command line tool on my Linux computer. Until now never had a problem using this for SCPI commands or queries.

telnet spd3303x-e 5025
Trying 192.168.178.65... (your IP address may vary!!!)
Connected to spd3303x-e.
Escape character is '^]'.
SYST:ERR?
+0, No error
*CALCLS 3
*CALST
CAL:CURR CH1,1,0.100 (your value may vary!!!)
SYST:ERR?
-108,Parameter not allowed,
CAL:CURR CH1,2,2.500 (your value may vary!!!)
SYST:ERR?
-108,Parameter not allowed,
*CALST
^]
telnet> quit
Connection closed.

Now the Current Display was way off, e.g. showed ca. 111 mA with nothing connected at the output.  I tried again several times, but without luck.
It seems that the power supply accepts the commands to erase the calibration but does not accept the commands to enter the new calibration data - at least when using the "telnet" tool under Linux. So I was left with a power supply with huge error in output Current Display. 

After an hour of despair, I decided to repeat the same procedure, but now using the "Putty" software on MS Windows OS (configured for proper hostname or IP address and port 5025). Fortunately it worked fine this time.

I understand that "Putty" does not require the NI/VISA software. So in case "Putty" does not work it is most likely not related to NI/VISA.
But you do need the NI/VISA software for using Siglent's EasyPower software, e.g. for installing software updates.

I have no explanation why the power supply is so picky or about fine differences between "telnet" and "Putty". But I am glad that it finally worked, thanks to this forum.

I tried only to calibrate the Current Display. I found the existing Voltage and Current Limit calibration good enough.
Note that I allowed the equipment (Power Supply, DMM, Electronic Load) to warm up ca. 1 hour before performing the calibration. After calibration, the cold power supply initially shows 3 or 4 mA output current with nothing connected. This quickly reduces to 2 mA, and after 1 hour warm-up to 0 mA.

[sy]

I was searching for DC electronic loads and came across this device I found on aliexpress. I have had a look at the SDL1020X-E mentioned earlier, but I'm on a budget and don't think I'll ever be using this instrument again besides wanting to calibrate my power supply for current.

I am not sure if these things are suitable to calibrate the SPD3303x(-e) as some guy on the youtubes said it was moreso geared towards testing batteries? I was wondering if anyone else had experience using one of these and if they are reliable enough? I've had a look online but can't find much instruction on how to use it

[orzel]

I am not sure if these things are suitable to calibrate the SPD3303x(-e) as some guy on the youtubes said it was moreso geared towards testing batteries? I was wondering if anyone else had experience using one of these and if they are reliable enough? I've had a look online but can't find much instruction on how to use it

There's a thread about it. Basically, they ARE battery testers, can be used as electronic loads (I do and i'm happy, but my needs are very limited). The voltage sensor is surprisingly accurate for such a cheap gear. The only drawback is that the FET (the big chip dissipating heat under the heatsink) is often weak, either a bad component (fake) or badly driven or wrongly mounted or whatever. Lot of people burnt them. ymmv. I've used mine (rated for 180W) up to 150W for several minutes, and never had any problem. The fan started (which is rare in my tests) and temperature rose, but not so much.

[Mortymore]

...
I am not sure if these things are suitable to calibrate the SPD3303x(-e)
...

I have one Atorch DL24 and after connecting this Load to 3 linear PSUs in CC mode, I have to say that the load is not very stable. It was set to draw 500mA, but it was constantly adjusting the current, usually jumping between 497mA and 503mA.

For a budget solution I'd rather go for the method of the beefy load resistors suggested by BreakingOhmsLaw, on par with a good ampmeter and voltmeter. As long as you can measure accurately the load current, a resistor would be more stable.

[rolfdegen]

Hello friends of electronics and technology

I have had a Siglent SPD3303X-E for three years and with your help I was able to calibrate the power supply today. I used EasyPower for the calibration. It didn't work with PuTTY.
Thank you very much for your help here in the forum 

[rolfdegen]

Hallo Friends

I have an SPD3303X-E with firmware on SPD3303X. Today I did a firmware update to SPD3303X firmware - 1.01.01.02.07R2 (release date 02.05.21).
Worked well with no errors.I noticed that shortly after powering on and the screen logo appears, the relays click once. I didn't have that in the old firmware. Is that normal 

Video: https://youtu.be/zkDXdXI9CcU

[blurpy]

I have an SPD3303X-E with firmware on SPD3303X. Today I did a firmware update to SPD3303X firmware - 1.01.01.02.07R2 (release date 02.05.21).
Worked well with no errors.I noticed that shortly after powering on and the screen logo appears, the relays click once. I didn't have that in the old firmware. Is that normal 

Video: https://youtu.be/zkDXdXI9CcU

Mine does exactly the same. Guessing it's how it is supposed to work. I don't remember how it was with older firmware or if it changed.

[rolfdegen]

Thanks for your information