Siglent SPD3303X-E Hi-Resolution Output Unlock?

[mawyatt]

My experience doing this kind of cal stuff with gear is that sometimes they don’t actually use a “0” cal point, but something other than 0, it sounds like you should be loading the unit up with 100mA and then doing that part of the cal so it sets its 100mA point when it is actually supplying 100mA, and then moving on from there.

I haven’t read the cal procedure for these units, but from what you have written I am 99% sure that is what is wrong.

The cal procedure should fit a simple straight line fit, or Y=MX+B. Per manual the two fits are at 100ma and 2.5A. The CC 100ma offset is not exactly 100ma, more like 85ma on ch1 and 90ma on ch2, and 115ma on ch1 and 122ma on ch2 in the other unit. On the 3303X that is good (I'm not about to try and calibrate it!!) this is a really good fit on both voltage and current, within +-1ma or +-1mv over the entire range as measured with our KS34465A!! The CC setting are also spot on over the entire range, as you would expect with almost 0 offset. Think Siglent is using 24 bit delta-sigma ADC and DAC (kind used for electronic weight scales), so some real precision here!! We've used this 3303X as a variable reference on many occasions, it's accurate, stable and low noise, sure it's not a precision calibration reference but when you need say 1.250V, 2.500V, or 5.000V this puppy is spot on! So hat's off to Siglent for using this kind of ADC and DAC!!

However, the other two 3303X have not been able to get the CC offset back to around 0 where it should be like the one mentioned above.

Many may not care about the current limit CC setting and just set it above some arbitrary limit, as long as the voltage & current readings are accurate. However in our use, especially when dealing with expensive semiconductors in a development system, having a precise CC or current limit can help protect things from overcurrent. I've used this method for over 40 years when we had very accurate current limit lab supplies (HP, Lambda and others), and this has saved the day often!!

Have another SPD3303X-E arriving 7/17, and debating whether to try and "enhance" it to a 3303X, if we do don't think I'll try a cal procedure just yet. If I find some time today will try fiddling with this some more on the unit that's hosed up.

I'm wondering if somehow "enhancing" a 3303X-E to a 3303X has something to do with this, as I'm sure the factory or others would have run across the CC offset problem, or maybe it's bug in the latest firmware release related to the "enhancement". Was able to convert back to a 3303X-E from an enhanced 3303X on one unit, but the offset remained and still could not be removed, so getting back to the initial state of the 3303X-E has failed.

Best,

[mawyatt]

Ok, I am reading the service manual right now…

Shown for Ch1, change channel to Ch.2 once ch1 done.

This sets the 100mA cal point: CALibration:CURRent ch1,3,NR3 (with NR3 being the actual current present with the load set to 100mA)

This sets the 2.5A cal point: CALibration:CURRent ch1,4,NR3 (with NR3 being the actual current present with the load set to 2.5A)

It is a bit confusing in the manual where it also has a DISPLAYING calibration section too, which is basically identical apart from the ch1,3 being ch1,1 and ch1,4, being ch1,2 instead.

Agree, this is a bit confusing and takes some time to figure out what this cal procedure is trying to do. You have a "Display" value for voltage & current and a "Setting" value for both. This is how I think it's working, the Display value is what's being measured by the internal ADC, for Setting value it's what's being produced by the internal DAC. When you do a calibration for voltage the output Setting is what you set, 1 or 25 volts per manual, then you enter the DMM actual reading in the CALibration:VOLTage command. With two Settings of 1volt and 25 volts, the Y=MX+B equation is fit which matches the Setting and Display values to the entered DMM readings.

When you do a calibration for current the output Setting is what you set on the EXTERNAL LOAD, 100ma or 2.5 amps per manual, then you enter the DMM actual current reading in the CALibration:CURRent command. With two LOAD Settings of 100ma and 2.5A, the Y=MX+B equation is fit which matches the Display & Setting values to the entered DMM readings.

A little algebra shows Y1=M*X1+B and Y2=M*X2+B, so M=(Y2-Y1)/X2-X1) and should yield ~1. B= [Y1+Y2 -M(X1+X2)]/2 which should be ~0 since X1+X2 ~ Y1+Y2 and M~1.

When you do a voltage cal, the Current must be set to higher than the load current which is just the DMM, anything above ~10ma should work. When you do a current cal the Voltage must be set to something higher than the DMM & Load voltage requirement, usually greater than a few volts is sufficient.

Best,

[2N3055]

Ok, I am reading the service manual right now…

Shown for Ch1, change channel to Ch.2 once ch1 done.

This sets the 100mA cal point: CALibration:CURRent ch1,3,NR3 (with NR3 being the actual current present with the load set to 100mA)

This sets the 2.5A cal point: CALibration:CURRent ch1,4,NR3 (with NR3 being the actual current present with the load set to 2.5A)

It is a bit confusing in the manual where it also has a DISPLAYING calibration section too, which is basically identical apart from the ch1,3 being ch1,1 and ch1,4, being ch1,2 instead.

Agree, this is a bit confusing and takes some time to figure out what this cal procedure is trying to do. You have a "Display" value for voltage & current and a "Setting" value for both. This is how I think it's working, the Display value is what's being measured by the internal ADC, for Setting value it's what's being produced by the internal DAC. When you do a calibration for voltage the output Setting is what you set, 1 or 25 volts per manual, then you enter the DMM actual reading in the CALibration:VOLTage command. With two Settings of 1volt and 25 volts, the Y=MX+B equation is fit which matches the Setting and Display values to the entered DMM readings.

When you do a calibration for current the output Setting is what you set on the EXTERNAL LOAD, 100ma or 2.5 amps per manual, then you enter the DMM actual current reading in the CALibration:CURRent command. With two LOAD Settings of 100ma and 2.5A, the Y=MX+B equation is fit which matches the Display & Setting values to the entered DMM readings.

A little algebra shows Y1=M*X1+B and Y2=M*X2+B, so M=(Y2-Y1)/X2-X1) and should yield ~1. B= [Y1+Y2 -M(X1+X2)]/2 which should be ~0 since X1+X2 ~ Y1+Y2 and M~1.

When you do a voltage cal, the Current must be set to higher than the load current which is just the DMM, anything above ~10ma should work. When you do a current cal the Voltage must be set to something higher than the DMM & Load voltage requirement, usually greater than a few volts is sufficient.

Best,

Mike,
 I believe that is the error.   

Procedure is to calibrate 4 points:  100 mA and 2.5A for current set and readout, as you said.
But external load is set for 3A constant all the time. You set the PSU for 100 mA and 2.5A points and measure with external meter.

Load set at 3A could probably be replaced with dead short, but this way you have safe, current limited short that you know won't overload PSU even if calibration is completely off..

I don't have the PSU, but could you try this way?


Connect output to load. Set load to CC 3A (this is constant trough all 4 measurements)

(Calibration of readout of 100mA point)
Set PSU voltage to 10v, set current to 100mA
Measure current with external meter.
CALibration:CURRent ch1,1,"measured current"

(Calibration of readout of 2.5 A point)
Set PSU for 2.5A
Measure current with external meter.
CALibration:CURRent ch1,2,"measured current"

(Calibration of current limit of 100mA point)
Set PSU voltage to 10v, set current to 100mA
Measure current with external meter.
CALibration:CURRent ch1,3,"measured current"
(Calibration of current limit of 2.5 A point)
Set PSU for 2.5A
Measure current with external meter.
CALibration:CURRent ch1,4,"measured current"


And that for both channels.

If I'm right, i might write down the procedure in clearer language and submit it to Siglent..

Regards,

Siniša

[mawyatt]

Ok, I am reading the service manual right now…

Shown for Ch1, change channel to Ch.2 once ch1 done.

This sets the 100mA cal point: CALibration:CURRent ch1,3,NR3 (with NR3 being the actual current present with the load set to 100mA)

This sets the 2.5A cal point: CALibration:CURRent ch1,4,NR3 (with NR3 being the actual current present with the load set to 2.5A)

It is a bit confusing in the manual where it also has a DISPLAYING calibration section too, which is basically identical apart from the ch1,3 being ch1,1 and ch1,4, being ch1,2 instead.

Agree, this is a bit confusing and takes some time to figure out what this cal procedure is trying to do. You have a "Display" value for voltage & current and a "Setting" value for both. This is how I think it's working, the Display value is what's being measured by the internal ADC, for Setting value it's what's being produced by the internal DAC. When you do a calibration for voltage the output Setting is what you set, 1 or 25 volts per manual, then you enter the DMM actual reading in the CALibration:VOLTage command. With two Settings of 1volt and 25 volts, the Y=MX+B equation is fit which matches the Setting and Display values to the entered DMM readings.

When you do a calibration for current the output Setting is what you set on the EXTERNAL LOAD, 100ma or 2.5 amps per manual, then you enter the DMM actual current reading in the CALibration:CURRent command. With two LOAD Settings of 100ma and 2.5A, the Y=MX+B equation is fit which matches the Display & Setting values to the entered DMM readings.

A little algebra shows Y1=M*X1+B and Y2=M*X2+B, so M=(Y2-Y1)/X2-X1) and should yield ~1. B= [Y1+Y2 -M(X1+X2)]/2 which should be ~0 since X1+X2 ~ Y1+Y2 and M~1.

When you do a voltage cal, the Current must be set to higher than the load current which is just the DMM, anything above ~10ma should work. When you do a current cal the Voltage must be set to something higher than the DMM & Load voltage requirement, usually greater than a few volts is sufficient.

Best,

Mike,
 I believe that is the error.   

Procedure is to calibrate 4 points:  100 mA and 2.5A for current set and readout, as you said.
But external load is set for 3A constant all the time. You set the PSU for 100 mA and 2.5A points and measure with external meter.

Load set at 3A could probably be replaced with dead short, but this way you have safe, current limited short that you know won't overload PSU even if calibration is completely off..

I don't have the PSU, but could you try this way?


Connect output to load. Set load to CC 3A (this is constant trough all 4 measurements)

(Calibration of readout of 100mA point)
Set PSU voltage to 10v, set current to 100mA
Measure current with external meter.
CALibration:CURRent ch1,1,"measured current"

(Calibration of readout of 2.5 A point)
Set PSU for 2.5A
Measure current with external meter.
CALibration:CURRent ch1,2,"measured current"

(Calibration of current limit of 100mA point)
Set PSU voltage to 10v, set current to 100mA
Measure current with external meter.
CALibration:CURRent ch1,3,"measured current"
(Calibration of current limit of 2.5 A point)
Set PSU for 2.5A
Measure current with external meter.
CALibration:CURRent ch1,4,"measured current"


And that for both channels.

If I'm right, i might write down the procedure in clearer language and submit it to Siglent..

Regards,

Siniša

Reading the procedure again it does read as you indicate, thanks for pointing that out. However following this path of setting the load to 3A and using the 3303X Current setting for 0.1 and 2.5 amps for the calibration points still does not remove the current offset. Just tried this 4 times, and reset the current reading and settings a couple times, then the calibration procedure keeping the load at 3A, then issued a complete coefficient reset with the *CALCLS 8 command a couple times and went thru the same current calibration procedure a couple times. Results are the current offset remains intact, nothing seems to affect it!!

BTW either using the load as the current control, or the 3303X current setting, results in the same excellent current reading throughout the range from a few ma all the way to 3.2 amps, accurate with +-1ma!! Just wish I could figure out how to get rid of this !@#$ current offset, this has become a real pain!!

Thanks for spotting this,

Best,

[blurpy]

Did you try this procedure on the unit you converted back to X-E as well?

[mawyatt]

No that unit was returned. A new unit should arrive today, so we'll have 3. One the gold standard is spot on in voltage, current and CC setting without any offset, the other is what we've been trying to fix the offset. The conversion file for going back to X-E from X is for the older firmware, not sure new firmware version (07R2) is available, if I can find it may give this a try.

The original usage plan was to have a couple 3303X to create various voltages for a ongoing project, which involved creating HV very accurate (13~14 bit) precision bipolar waveforms, the required dual HV sources are 120V Bench supplies which work OK, but are switchers and produce noise which will impede the measurements. We need to measure arbitrary waveform +-peaks within 13~14 precision from +-1V to +-70V, and have offsets of a few millivolts over the entire range, and this is for 64 independent channels minimum!!

Was going to get a couple clean linear supplies to replace the 120V switchers, but after looking around and some reviews wasn't impressed with what was available at reasonable cost and good build quality. Decided to use the 3303X in series mode to yield +-65 volts which is good enough to make the measurements and "infer" the +-70 waveform performance. This is a very complex system which has me completely stressed out without with these PS issues!! Add insult to injury my right eye was just diagnosed with non-repairable retina damage caused by reduced blood flow likely due to a stroke which explains my depth and peripheral vision problems On top of this is some significant on-going family problems, so my stress level is maxed out

Anyway, I decided to try and use a custom +-HV switchmode PS designed for a dual HV boost amplifier that works with standard AWG, this was developed to allow precise cause and effect measurements on the Phase Array system this HV 64 channel controller is to support, by allowing high voltage bipolar arbitrary waveforms to be created from a standard AWG like the SDG2042X. It worked very well, with little added noise, significantly lower than the 120V Bench supplies and will become the foundation for the custom +-HV supply for the 64 channel controller. This bipolar HV supply uses a technique we hadn't tried before, where the - voltage is created by transferring charge current at the same time the single inductor current is ramping up, so the FET switch current is the sum of the two currents and controlled. Works well and the +-HV are very close in magnitude and controlled. Here's what the HV amp looks like, the fan is optional. We have another design which is higher current/power based around the PA443 , and another with much higher currents but lower voltages based upon the LM3886. Still need to create the 3D printed end plates for the case.


Best,

[tv84]

Is it possible that since you calibrated with the wrong current value, the device now isn't able to rollback from that calibration? Like, for example, you exceeded a calib range (offset)... from which the software doesn't have a value to recover.

(For example, the device is able to recover from up to 100 mA offset, but not the 100 mA value itself.)

And now, probably only Siglent or a full flash would be able to reset the device.

[blurpy]

I am not aware of any previous example of a rollback for the newest firmware. But possibly the attached file would work.
Note this is completely untested! I wouldn't try it without tv84 verifying it first.

Log from creating it:

$ python3 ../convert.py SPD3303X-E_V1.01.01.02.07R2.ADS
SPD3303X-E_V1.01.01.02.07R2.ADS  /  CRC32: 260B1F7F
File Header Size: 00000070
File Checksum: FD844256 [00000004-00094C43] (with only the File Header decrypted)  CKSM True
File Size: 00094BD4 (without 0x70 bytes of the File Header)
Product_ID: 65 / SPD3303X-E
Vendor/Brand: SIGLENT
USB Host Controller: ISP1763
****************************************************
New productId: 64 / SPD3303X
New checksum: FD844257
Writing file: SPD3303X-E_V1.01.01.02.07R2/SPD3303X-E_V1.01.01.02.07R2_ConvertFromSPD3303X.ads

Decrypt it:

$ python3 decrypt.py SPD3303X-E_V1.01.01.02.07R2_ConvertFromSPD3303X.ads
SPD3303X-E_V1.01.01.02.07R2_ConvertFromSPD3303X.ads  /  CRC32: 73030026
File Header Size: 00000070
File Checksum: FD844257 [00000004-00094C43] (with only the File Header decrypted)  CKSM OK
File Size: 00094BD4 (without 0x70 bytes of the File Header)
Product_ID: 64
Vendor/Brand: SIGLENT
USB Host Controller: ISP1763
****************************************************
Decrypting the 0x2800 and 0x1400 blocks...
Reversing file...
XORing with 0xFF (incrementing pattern)...
XORing with 0xFF from 0x0004A5EA until 0x00094BD3
****************************************************
Section Checksum: FA2ABBDC
Section Size: 00094BA0 [00000034-00094BD3]  CKSM OK
Section # 00000007
***** BIN file *****
Writing file: SPD3303X-E_V1.01.01.02.07R2_ConvertFromSPD3303X/decrypted_section7.bin
****************************************************
  File Processed OK

Decrypt the original:

$ python3 decrypt.py SPD3303X-E_V1.01.01.02.07R2.ADS
SPD3303X-E_V1.01.01.02.07R2.ADS  /  CRC32: 260B1F7F
File Header Size: 00000070
File Checksum: FD844256 [00000004-00094C43] (with only the File Header decrypted)  CKSM OK
File Size: 00094BD4 (without 0x70 bytes of the File Header)
Product_ID: 65
Vendor/Brand: SIGLENT
USB Host Controller: ISP1763
****************************************************
Decrypting the 0x2800 and 0x1400 blocks...
Reversing file...
XORing with 0xFF (incrementing pattern)...
XORing with 0xFF from 0x0004A5EA until 0x00094BD3
****************************************************
Section Checksum: FA2ABBDC
Section Size: 00094BA0 [00000034-00094BD3]  CKSM OK
Section # 00000007
***** BIN file *****
Writing file: SPD3303X-E_V1.01.01.02.07R2/decrypted_section7.bin
****************************************************
  File Processed OK

Compare the decrypted files:

$ diff SPD3303X-E_V1.01.01.02.07R2_ConvertFromSPD3303X/decrypted_section7.bin SPD3303X-E_V1.01.01.02.07R2/decrypted_section7.bin
$

The only difference between the original and the modified firmware should be the product id in the header.

[mawyatt]

Is it possible that since you calibrated with the wrong current value, the device now isn't able to rollback from that calibration? Like, for example, you exceeded a calib range (offset)... from which the software doesn't have a value to recover.

(For example, the device is able to recover from up to 100 mA offset, but not the 100 mA value itself.)

And now, probably only Siglent or a full flash would be able to reset the device.

This is likely, in my rush to get these supplies converted and calibrated, I calibrated the way I thought it should work after glancing over the manual rather than read the manual over and over to be absolutely sure I calibrated they way Siglent specified. Thanks to 2N3055 for pointing this out!!

What is interesting is the current reading is spot on from a couple ma all the way to 3A, but the darn CC offset remains!! I'm debating whether to give the one scheduled to arrive today a try, if I do, will certainly follow the cal manual procedure correctly before I attempt an enhancement, and if it behaves the same with a CC offset, then this points to a serious bug if it's not easily recoverable!!

A nice thing about the 3303X or X-E is not only the use of 24 bit delta sigma ADC and DAC (believe this is correct), but the internal reference seems to be very stable, and the 3rd output is completely isolated, not common to ch1 or ch2. If one were designing for a low cost, then the 3rd channel would be common to either ch1 or 2, which saves a few components (rectifier bridge, filter caps) and an extra transformer winding. So even tho it's not completely settable, being a floating output is nice!!

I'm burnt to crisp now stresswise, so may wait on the attempt when the unit arrivers later today!!

The worldwide shortage of semiconductors is wreaking havoc on the design process, I've had to redesign things over and over as components become available then disappear, so now if we just think a component might be useful we buy a bunch and just store them. Did so with some expensive DACs and HV amps, over $3K in just two part types!! But they are here in my hands 

Hand soldering the DAC 100 pin 0.5mm pitch package was difficult, tedious, but worked. Got one DAC up and have 4 channels of the 64 running, beautiful, clean and precise HV waveforms, now I need to get the other 60 channels of this stuff done before my eye goes south.

Hey, and thanks so much for all the help you, blurpy, 2N3055 and tautech have provided. You guys are great and really appreciated

Best,

[2N3055]

 
Any time!

[boggis the cat]

The worldwide shortage of semiconductors is wreaking havoc on the design process, I've had to redesign things over and over as components become available then disappear, so now if we just think a component might be useful we buy a bunch and just store them. Did so with some expensive DACs and HV amps, over $3K in just two part types!! But they are here in my hands 

'Just in time' has always been horse-crap.  The manufacturer where I had my first full-time job decided to save money by moving to producing exact component counts and JIT supply chains.  No order could ever be completed on time, and some were cancelled due to failure to meet contract.  So they ended up warehousing expensive finished product until they could find a buyer.

You always have a reject / failure rate, and letting a sales dweeb order your manufacturing supply instead of a production engineer is never going to work smoothly.  Particularly when they find a cheaper supplier, and order a load of counterfeit components.

[mawyatt]

Agree, always thought JIT was a recipe for disaster, so many opportunities for a kink in a single component delivery to stop the entire product production. Unless you build up an inventory of all components to allow for a issue with JIT you are at risk, then why bother with JIT if you maintain a backup inventory!!!

This is a developmental control system to support a new technology for Phased Array Systems, hopefully leading to a full custom chip design. So not exactly a case for JIT, but the shortage is very real and causing lots of issues. This has caused the design approach to shift from using the best design & components for the purpose, to designing for a generic solution which just mets the needs and uses generic components, or worse having to alter the requirements to met the available components.....not exactly the best design approach IMO

Best,

[mawyatt]

I am not aware of any previous example of a rollback for the newest firmware. But possibly the attached file would work.
Note this is completely untested! I wouldn't try it without tv84 verifying it first.

Thanks, I'll wait for verification since I don't understand how this firmware works. I've already messed things up enough

Best,

[tv84]

I am not aware of any previous example of a rollback for the newest firmware. But possibly the attached file would work.
Note this is completely untested! I wouldn't try it without tv84 verifying it first.

The file is correct. 

For future reference, both files are attached. As blurpy said, untested.

[mawyatt]

Just did the downgrade to the X-E from the X version and the CC offset remains. Tried a couple cal routines all unsuccessful in removing the offset. So will just have to live with this unit having a CC offset of 100ma for now I guess.

Thanks for the help.

Edit: Forgot to mention that both files work and allow transitioning from X-E to X and from X to X-E versions for the 07R2 firmware.

Best,

[blurpy]

The file is correct. 

For future reference, both files are attached. As blurpy said, untested.

Nice, thanks! Good to know I did the script correctly.

Just did the downgrade to the X-E from the X version and the CC offset remains. Tried a couple cal routines all unsuccessful in removing the offset. So will just have to live with this unit having a CC offset of 100ma for now I guess.

Hmm, just throwing out some thoughts here. I notice the SCPI command "CLREEPROM", mentioned in the first post here:
https://www.eevblog.com/forum/testgear/how-to-calibrate-siglent-spd1000x-spd3303x-spd3303x-e-series-power-supplies/

We can guess from the name that it clears the eeprom. So what's the consequence? What's stored in the eeprom? Settings probably, calibration probably, serial number possibly. Anything else?

Worst case it removes something you wont be able to get back outside the factory, and bricks the unit. Best case, it resets your settings, calibration and the weird offset.

Siglent knows if it's safe to run and whether it will fix anything, but will they tell us? Or maybe someone understands the code and can say what it does.

[tv84]

Edit: Forgot to mention that both files work and allow transitioning from X-E to X and from X to X-E versions for the 07R2 firmware.

Thanks for testing.

Nice, thanks! Good to know I did the script correctly.

I'll be providing an award!  You're the first guy that recreated part of my parser using many of my parsing messages, in Python!! (Have I told you that I hate Python? ? ) I even had some doubts if I was looking at my parsing or another guy's parsing. 

I notice the SCPI command "CLREEPROM", mentioned in the first post here:
https://www.eevblog.com/forum/testgear/how-to-calibrate-siglent-spd1000x-spd3303x-spd3303x-e-series-power-supplies/

We can guess from the name that it clears the eeprom. So what's the consequence? What's stored in the eeprom? Settings probably, calibration probably, serial number possibly. Anything else?

I agree, this is probably the command that is needed and the feature that I talked about before.  Since the type of device is determined by the software, it shouldn't create a problem there.

It should delete all settings (LAN, S/N, etc.). It would be great if it deleted the calib offset. The S/N can be re-inserted via the SRLN command.

Maybe tautech could ask Siglent if this could be the way to go.

[tautech]

Maybe tautech could ask Siglent if this could be the way to go.

Mike wrote up the CC offset problem for me in detail last week which I forwarded to Siglent....still waiting for a reply. 

[blurpy]

I'll be providing an award!  You're the first guy that recreated part of my parser using many of my parsing messages, in Python!! (Have I told you that I hate Python? ? ) I even had some doubts if I was looking at my parsing or another guy's parsing. 

Thank you
The many examples of output from your script that you have posted was very helpful in figuring out how to parse the firmware. I used your formatting for easy comparison!

I don't think I have heard of your hate for Python before

[mawyatt]

No luck using the CLREEPROM command, CC offset still remains!!

Best,

[tv84]

No luck using the CLREEPROM command, CC offset still remains!!

Did you notice that anything else got reset?

[mawyatt]

Used the CLREEPROM command followed by the CALCLS 8 command hoping to clear the CC offset. The model and serial number were not reset, so not sure what was reset with the CLREEPROM command.

Best,

[tv84]

Mike, last call:

Send the command

FACTORY ON

before issuing the  CLREEPROM.

Report if it changed anything.

[mawyatt]

Tried the FACTORY ON, then the CLREEPROM, no change. Then tried FACTORY (mentioned in thread blurpy pointed to) and then CLREEPROM, no change.

Best,

[blurpy]

Running out of ideas here. Curious to hear what Siglent has to say. It's probably fixable in firmware. Seems likely as mentioned earlier that it's an offset that's out of expected range, and the code doesn't handle it.

It could be that CLREEPROM either does nothing at all, or tries to call the code behind CALCLS which doesn't work.