Lab Power Supply Turn ON and OFF Characteristics

[tautech]

With the new Siglent SPD4000 series we find it interesting that the results from CH1 and CH2 are different wrt to the conditions discussed here as shown by Furna. We don't see this with the older SPD3303X as both channels behave the same as expected!! Also, haven't seen any post that indicate other supplies have different behavior with 2 main channels with same specs as one would "expect" the detailed design and implementation of the channels would be identical. 

Would be interesting to find out why, as this is quite puzzling?

Best

For whom is not following the Siglent SPD4000 series thread, my measurements of Siglent SPD4121X are available at
https://www.eevblog.com/forum/testgear/new-siglent-spd4000x-series-power-supply/msg5667709/#msg5667709
https://www.eevblog.com/forum/testgear/new-siglent-spd4000x-series-power-supply/msg5667823/#msg5667823

Siglent SPD4000 series has 4 channels that are inerhently different ...
SPD4121X in particular has
CH1-CH4 15V 1.5A are independent
CH2-CH3 12V 10A can be setup in series/parallel and both also have 4wire sense

That's unusual that CH1 and CH4 are the same and CH2 and CH3 are the same but different from CH1 and CH4, this explains the difference since CH1 (15V 1.5A) is not the same specification as CH2 (12V 10A) and thus one might expect a different response from such!!

Best,

All three SDP4000X models have quite different ratings.
Ch 1&4 are equal and likewise Ch 2&3 but different for each model.
Check out P4
https://int.siglent.com/u_file/download/24_08_14/SPD4000X_DataSheet_EN01B.pdf

[mawyatt]

With the new Siglent SPD4000 series we find it interesting that the results from CH1 and CH2 are different wrt to the conditions discussed here as shown by Furna. We don't see this with the older SPD3303X as both channels behave the same as expected!! Also, haven't seen any post that indicate other supplies have different behavior with 2 main channels with same specs as one would "expect" the detailed design and implementation of the channels would be identical. 

Would be interesting to find out why, as this is quite puzzling?

Best

For whom is not following the Siglent SPD4000 series thread, my measurements of Siglent SPD4121X are available at
https://www.eevblog.com/forum/testgear/new-siglent-spd4000x-series-power-supply/msg5667709/#msg5667709
https://www.eevblog.com/forum/testgear/new-siglent-spd4000x-series-power-supply/msg5667823/#msg5667823

Siglent SPD4000 series has 4 channels that are inerhently different ...
SPD4121X in particular has
CH1-CH4 15V 1.5A are independent
CH2-CH3 12V 10A can be setup in series/parallel and both also have 4wire sense

That's unusual that CH1 and CH4 are the same and CH2 and CH3 are the same but different from CH1 and CH4, this explains the difference since CH1 (15V 1.5A) is not the same specification as CH2 (12V 10A) and thus one might expect a different response from such!!

Best,

All three SDP4000X models have quite different ratings.
Ch 1&4 are equal and likewise Ch 2&3 but different for each model.
Check out P4
https://int.siglent.com/u_file/download/24_08_14/SPD4000X_DataSheet_EN01B.pdf

See edited notes after your post.

If a supply has equal dual pairs of channels they should always be arranged in a 1,2 and 3,4 fashion, not in 1,4 and 2,3 arrangement IMO. These should be displayed in a 1,2 on top and 3,4 on bottom so one can easily compare like channels.
 
Best,

[thm_w]

Rigol DP832 was shown above, as I mentioned in the other thread, it does have some small overshoot to 15-20mA so its relevant if you are current limited sub 10mA, which is very low.

Rigol DP932 Channel 1 - no overshoot:





R&S NGMO high speed power supply - Overshoots to 8V:

[tautech]

With the new Siglent SPD4000 series we find it interesting that the results from CH1 and CH2 are different wrt to the conditions discussed here as shown by Furna. We don't see this with the older SPD3303X as both channels behave the same as expected!! Also, haven't seen any post that indicate other supplies have different behavior with 2 main channels with same specs as one would "expect" the detailed design and implementation of the channels would be identical. 

Would be interesting to find out why, as this is quite puzzling?

Best

For whom is not following the Siglent SPD4000 series thread, my measurements of Siglent SPD4121X are available at
https://www.eevblog.com/forum/testgear/new-siglent-spd4000x-series-power-supply/msg5667709/#msg5667709
https://www.eevblog.com/forum/testgear/new-siglent-spd4000x-series-power-supply/msg5667823/#msg5667823

Siglent SPD4000 series has 4 channels that are inerhently different ...
SPD4121X in particular has
CH1-CH4 15V 1.5A are independent
CH2-CH3 12V 10A can be setup in series/parallel and both also have 4wire sense

That's unusual that CH1 and CH4 are the same and CH2 and CH3 are the same but different from CH1 and CH4, this explains the difference since CH1 (15V 1.5A) is not the same specification as CH2 (12V 10A) and thus one might expect a different response from such!!

Now looking at the Siglent page for the SPD4121X shown, this arrangement looks OK with terminals for CH1 logically on left then the channels proceed to right as 2 thru 4 (not like GPP 4323 where Instek placed CH4 on left, then 2 thru 4 towards right 

The display shows 1 to 2 top and 3 to 4 below left to right which makes sense, however because of the arrangement the similar channels are diagonal not left to right as one would expect!!

Imagine using a typical analog supply arrangement of +-15V (or +-12V), then the important display parameters (V and I) are diagonal not left to right for the active channels. So glancing at the active channels in this common use case requires determining the parameters on diagonals rather than side by side!!

This makes no sense IMO, would like to hear the logic behind this arrangement from Siglent designers!!!

https://siglentna.com/power-supplies/spd4000x-series-programmable-linear-dc-power-supply/

Best,

All three SDP4000X models have quite different ratings.
Ch 1&4 are equal and likewise Ch 2&3 but different for each model.
Check out P4
https://int.siglent.com/u_file/download/24_08_14/SPD4000X_DataSheet_EN01B.pdf

See edited notes after your post.

If a supply has equal dual pairs of channels they should always be arranged in a 1,2 and 3,4 fashion, not in 1,4 and 2,3 arrangement IMO. These should be displayed in a 1,2 on top and 3,4 on bottom so one can easily compare like channels.
 
Best,

Added.

While I don't speak for them, what they have done does make some sense to me.

You've used a split rail + argument to support your POV yet in Ch 2-3 Series mode one can pull 0V from the center connection for dual tracking supply needs without any need for external cabling and only 2&3 support Series and Parallel modes and Sense.

As Ch 1&4 don't provide Series or Parallel modes or Sense it's not logical they need be grouped together even for split rail use as they can't be Series'd for dual tracking supply use.

This does however bring additional thoughts forward, should Ch 1&4 have instead been paired and given Series and Parallel capability some major changes to the design and UI would be needed, otherwise how would one know which channels were being Paralleled or Series'd ?

@Furna
What is the display indication when Ch 2&3 are used in Series and Parallel modes ?

[mawyatt]

While I don't speak for them, what they have done does make some sense to me.

You've used a split rail + argument to support your POV yet in Ch 2-3 Series mode one can pull 0V from the center connection for dual tracking supply needs without any need for external cabling and only 2&3 support Series and Parallel modes and Sense.

As Ch 1&4 don't provide Series or Parallel modes or Sense it's not logical they need be grouped together even for split rail use as they can't be Series'd for dual tracking supply use.

This does however bring additional thoughts forward, should Ch 1&4 have instead been paired and given Series and Parallel capability some major changes to the design and UI would be needed, otherwise how would one know which channels were being Paralleled or Series'd ?

@Furna
What is the display indication when Ch 2&3 are used in Series and Parallel modes ?

The SPD3303X and GPP4323 both allow creating a Dual Voltage Tracking Supply utilizing the Series Mode on CH1 and CH2 without any "extra" cabling also, altho the SPD3303X doesn't read the current thru/from CH2 properly (serious flaw IMO). Example drawing a different current from CH2 while in the Series Mode doesn't show in the CH2 current display. Apparently this is just displayed as a replica of the CH1 current 

The GPP4323 reads this condition properly and displays the correct current drawn from both CH1 and CH2 while in Series Mode, whether thru or from both channels or some current "tapped" off from CH1 or CH2.

This dialog has allowed us to discover a potential issue with the SPD3303X that we didn't know, as we don't recall ever seeing this discussed before. Maybe others can do this test in addition to the ON/OFF characteristics?

If you or Furna have the time please check the SPD4000X in Series Mode and see if CH2 and CH3 display the correct current drawn respectively. Just place a resistor across either CH2 or CH3 and see if it registers properly on the respective channel display, then attach across the full series span of CH2 and CH3, this should show twice the current and displayed in both channels as flowing thru both.

Anyway, good discussions

Best

[tautech]

Sorry Mike, no SPD4kX here yet. 
A few weeks....

[mawyatt]

Maybe Furna can help

BTW do you recall this issue with the SPD3303X mentioned?

Best

[tautech]

BTW do you recall this issue with the SPD3303X mentioned?

Split rail current measurement, no.....maybe it's the advancing years. 

[BennoG]

Ok I tested all my power supply's to see if this is a common problem.
Agilent 6645A      Ok
EA PS 2084-10B  Ok
Keithley 2231A-30-3 CH1  (overshoot)
Keithley 2231A-30-3 CH2  (overshoot)
Keithley 2231A-30-3 CH3  Ok
Keithley 2400                  Ok
Agilent N6700B CH1-CH4  Ok
Riden DPS5005  Ok
Riden RD6012    Ok
Riden RK6006    Ok
Spellman SL2000 (overshoot)  not fair because this is a 1kV 2A 2kW psu
TDK Genh60-12.5  Ok

Benno

[mhsprang]

I found an article about overshoots in power supplies, dependant on the priorities of the voltage control loop and current control loop:

https://www.electronicdesign.com/technologies/power/article/21801598/getting-your-priorities-straight-managing-overshoots

While I have never written a control loop for a power supply, as an embedded programmer I keep wondering why you can't combine those loops or at least observe crossing the V and I limits at the same time and act accordingly?

[Geofrey]

Hi, I attached the results for my BCS6401. CH1 is on a 4wire configuration, CH2 is not. It did not cross test to see what factor is causing the overshoot as I would need some time to change the back wiring.
Overall I would say fast response, but with some overshoot, unless I do not set the "fast speed" setting but "normal speed" instead. Interestingly, a set rise time comparable to the "normal speed" do result in overshoot.

Also attached the results of my beloved  Multicomp MP710087 (Owon P4305 rebadge) that never fails to impress me. The settling isn't at 40mA but at 101mA as I cannot calibrate the current offset to less than roughly 80mA. So I calibrated it to 100mA offset just to get somewhat legible reading on its screen.

[_Wim_]

While I have never written a control loop for a power supply, as an embedded programmer I keep wondering why you can't combine those loops or at least observe crossing the V and I limits at the same time and act accordingly?

You can. In industry we call this an "overide" control (also called sometimes select control). This is used when you have multiple PVs (process values) and only a single CV (control value). In our case we have measured voltage AND measured current as "PVs", and typically output voltage as "CV" (but this could be output current also as on the Yokogawa GS200 above). This article briefly describes override control (https://www.pid-tuner.com/wp-content/uploads/2019/07/Override_for_all.pdf), but once you know how it is called, there is tons of information available on the net.

[AVGresponding]

HP 66311B on and off:





Siglent 3303X (hacked from RS Pro 3303X-E) on and off:





Will try HP 6633B and TTi EX354 later

[Furna]

@Furna
What is the display indication when Ch 2&3 are used in Series and Parallel modes ?

See images

[tautech]

@Furna
What is the display indication when Ch 2&3 are used in Series and Parallel modes ?

See images

Thanks.

Nice, I didn't notice the bicolor LED lit Series/Parallel button in the short play I had with one at HQ. 

[Furna]

If you or Furna have the time please check the SPD4000X in Series Mode and see if CH2 and CH3 display the correct current drawn respectively. Just place a resistor across either CH2 or CH3 and see if it registers properly on the respective channel display, then attach across the full series span of CH2 and CH3, this should show twice the current and displayed in both channels as flowing thru both.

I do not have much time now; but
https://www.eevblog.com/forum/testgear/new-siglent-spd4000x-series-power-supply/msg5630671/#msg5630671

Maybe tomorrow I can go through this better.

[Furna]

CH2 in series with CH3
CH2 is "master"; you cannot setup CH3
CH2 set to 5V and 5Ω

01_IMG_20241008_230601.png
10Ω across CH2 (red and black cables)
CH2 shows total volatge 5V (CH2+CH3) 0.247A demonstrates CH2 outputs 2.5V
CH3 outputs 2.5V but note green cable disconnected => 0A

02_IMG_20241008_225937.png
10Ω across CH2 (red and black cables)
50Ω (2x100Ω in parallel) (blue and green cables)
CH2 shows total voltage 5V (CH2+CH3) Amper is again the sum (CH2+CH3) 0.346A
CH3 shows 2.5V and 0.1A of CH3

03_IMG_20241008_235241.png
20Ω across CH2 (red&black)
50Ω across CH3 (yello&green)
10Ω across the series (red&green)

[mawyatt]

@ Furna

Thanks for the post, could you repost the above over here:

https://www.eevblog.com/forum/testgear/power-supply-series-mode-for-split-dual-tracking-use/

We have response prepared but realized it's better in the above thread for folks to follow, and didn't want to copy your post without permission, and didn't want to clutter this thread up since it's dealing with ON/OFF characteristics.

Best

[KungFuJosh]

could you repost the above over here

You can quote him from this thread in the other thread with a little copy-paste magic.

[colorburst]

• Harrison 865C
• Mastech HY3005D-3
• Electro Industries DIGI-35A

[AVGresponding]

HP 66311B on and off:





Siglent 3303X (hacked from RS Pro 3303X-E) on and off:

Revisited this, as I realised I'd used a 4R7 resistor instead of a 47R...   

3303X:




66311B:




Apologies for any confusion my blind-assed dumbshittery may have caused   

[Furna]

@ Furna
Thanks for the post, could you repost the above over here:
https://www.eevblog.com/forum/testgear/power-supply-series-mode-for-split-dual-tracking-use/
We have response prepared but realized it's better in the above thread for folks to follow, and didn't want to copy your post without permission, and didn't want to clutter this thread up since it's dealing with ON/OFF characteristics.

Answered there

[mawyatt]

HP 66311B on and off:





Siglent 3303X (hacked from RS Pro 3303X-E) on and off:

Revisited this, as I realised I'd used a 4R7 resistor instead of a 47R...   

3303X:




66311B:




Apologies for any confusion my blind-assed dumbshittery may have caused   

No worries, we've all been there done that especially with color coded components, heck we've even done this with numeric coded components

Your 3303X looks like ours wrt the ON waveform. Seems the short duration "glitch" at initial ON condition is independent of the PS settings, either Current limit or Voltage setting.

Anyway, thanks for update

Best,

[KungFuJosh]

Your 3303X looks like ours wrt the ON waveform. Seems the short duration "glitch" at initial ON condition is independent of the PS settings, either Current limit or Voltage setting.

I had a thought about this that's slightly weird I guess. But the current setting is a setting, not really a limit. So I tried adjusting OCP (hold down the right arrow under the knob on the 3303X), and that had absolutely no effect on the ~29mA turn on spike when using 10mA as the setting & OCP limit. I haven't tried higher levels yet.

[AVGresponding]

No worries, we've all been there done that especially with color coded components, heck we've even done this with numeric coded components

Your 3303X looks like ours wrt the ON waveform. Seems the short duration "glitch" at initial ON condition is independent of the PS settings, either Current limit or Voltage setting.

Anyway, thanks for update

Best,

Oh, it's worse than that; these were 10W epoxy coated wirewound, marked 4R7... 

bd139 suggested it might be the inductance (not knowing of my mistake), so I paralleled six 270Ω carbon composition resistors together, and got the results you see in my updated post. It was some minutes of head scratching until I realised what I'd done.



EDIT: Some recent experiments with some NOS 1N827's has shown me that using these psu's as a precision current source just doesn't work; the output is too unstable. The only way to get a stable output current is in CV mode by using a limiting resistor and getting the current you want by means of setting the voltage appropriately and avoiding going anywhere near CC mode.