New Siglent SPD4000X series power supply

[Furna]

Why not use single trigger on your scopes to test this?

Because I had no time this morning and I am just back from work

[sparky2024]

Scope trace of SPD4323X channel 2 with no load (hence voltage drop is prolonged).

Steps to reproduce:

1) Set channel 2 to a voltage (5v selected in this case), and switch on channel.  Switch off.
2) Set scope to Normal or Single trigger, timebase 1sec.
3) Switch off main SPD4323X power switch.

Note I have no load attached, but have seen LEDs flash on my projects.  The amplitude follows the setting.  Also this is only evident on ch 2 & 3.  Channels 1 & 4 behave as I would expect.  Serial or Parallel modes are disabled.

Measuring the level (zooming in a bit in this case), we see the level persists for 75ms, and starting from the rising edge is around 100ms total.  Sounds like software to me.

[Furna]

Scope trace of SPD4323X channel 2 with no load (hence voltage drop is prolonged).

Steps to reproduce:

1) Set channel 2 to a voltage (5v selected in this case), and switch on channel.  Switch off.
2) Set scope to Normal or Single trigger, timebase 1sec.
3) Switch off main SPD4323X power switch.

Note I have no load attached, but have seen LEDs flash on my projects.  The amplitude follows the setting.  Also this is only evident on ch 2 & 3.  Channels 1 & 4 behave as I would expect.  Serial or Parallel modes are disabled.

Measuring the level (zooming in a bit in this case), we see the level persists for 75ms, and starting from the rising edge is around 100ms total.  Sounds like software to me.

Just tried also with your settings and cannot reproduce; I have absolutinh nothinh even with trigger set to 30mV.
What FW do you run?

Indeed I have the following at master power on ... something I do not have at channel power on.
I usually disconnect cables to avoid any issue at every master power off/power on.

[electr_peter]

@sparky2024, what is the max voltage with mains switching off (if set not to 5V but to max V)? Can you also check maximum possible current with a load (resistor on output)?

[TheDefpom]

T hat looks like capacitor discharge to me, what about leaving the channels off, setting a high voltage, power of the main switch, and then power main switch on?

I suspect there would be no pulse in this instance.

I’m just waiting for my unit to arrive so I can’t test this myself yet.

[sparky2024]

Thank you for swift replies.

Peter, I see a peak of around 22v with a voltage setting of 32v.  This occurs identically with (img_3337) or without (img_3338) a high wattage 56 ohm resistor as load.  Ref. picture.

Defpom, I see the same response (as img_3338) without applying the “soft” on button i.e. power-up with prior settings, and just power down with the main switch.

There’s a definite capacitor discharge, but the relay in my opinion is charging that capacitor for 75-100ms.

[sparky2024]

Furna, my unit is running 4.1.2.8R1.  As per another earlier message, my unit arrived with 4.1.2.6.  I’ve been living in hope of a firmware fix for this.

Wondering if @tautech, or a Siglent rep is following this thread and may be able to comment on this?

I may be in touch with my distributor in the UK soon if all else fails.

Thanks to all.

[electr_peter]

Potential issue with output turning ON with high voltage and high current capability if mains switch is pressed should be looked into more closely (Is it a single unit/early FW/defective component or design specific?).
With complicated multi-channel PSUs outputing a "blip" with mains turn-off is not unheard off, but that's not critical when voltage/available current is limited to safe levels by design. If not limited, PSU can kill DUTs.

[Furna]

Potential issue with output turning ON with high voltage and high current capability if mains switch is pressed should be looked into more closely (Is it a single unit/early FW/defective component or design specific?).

I guess TheDefpom and tautech will help when they receive their units ... I understand sparky2024 might chose to return his unit.

With complicated multi-channel PSUs outputing a "blip" with mains turn-off is not unheard off, but that's not critical when voltage/available current is limited to safe levels by design. If not limited, PSU can kill DUTs.

I agree; also as I wrote a very simple even if annoying solution is to disconnect cables as an attitude at everty mains power in/off.

[mawyatt]

Potential issue with output turning ON with high voltage and high current capability if mains switch is pressed should be looked into more closely (Is it a single unit/early FW/defective component or design specific?).
With complicated multi-channel PSUs outputing a "blip" with mains turn-off is not unheard off, but that's not critical when voltage/available current is limited to safe levels by design. If not limited, PSU can kill DUTs.

Any overshoot at PS Turn ON/OFF with the Output Enable Buttons and also to the Main Power Switch ON/OFF is of concern to us. A few decades ago we lost a couple of very expensive custom chip sets to a rogue PS that had a glitch at Turn Off, ever sense we have been extremely sensitive to such

Best

[Furna]

Potential issue with output turning ON with high voltage and high current capability if mains switch is pressed should be looked into more closely (Is it a single unit/early FW/defective component or design specific?).
With complicated multi-channel PSUs outputing a "blip" with mains turn-off is not unheard off, but that's not critical when voltage/available current is limited to safe levels by design. If not limited, PSU can kill DUTs.

Any overshoot at PS Turn ON/OFF with the Output Enable Buttons and also to the Main Power Switch ON/OFF is of concern to us. A few decades ago we lost a couple of very expensive custom chip sets to a rogue PS that had a glitch at Turn Off, ever sense we have been extremely sensitive to such

Best

With SPD4000X series power supply you cannot Turn ON with the Output(s) Enabled; channels are Off even if they were On at previous PS Turn Off.
So you may now think the caputure I took, demonstrates even a worst situation than you tough.
Disconnect the cables and leave happy

[mawyatt]

With SPD4000X series power supply you cannot Turn ON with the Output(s) Enabled; channels are Off even if they were On at previous PS Turn Off.
So you may now think the caputure I took, demonstrates even a worst situation than you tough.
Disconnect the cables and leave happy

So what happens when you have the PS Main Power Switch ON and the Outputs OFF, sensitive circuits connected and a Main AC Power becomes disconnected/interruped 

You shouldn't have a PS that requires removal of the Banana Leads every time to "protect" your sensitive circuits, and then reinstall when you want to power up things again. For us this is totally unacceptable PS behavior and just a recipe for disaster at some point

Best

[Martin72]

You shouldn't have a PS that requires removal of the Banana Leads every time to "protect" your sensitive circuits

Sensitive circuits usually have overvoltage protection. If not, then you should take action yourself and implement it.

[blackdog]

My mantra is that a power supply should always protect its load.
And as Martin72 points out, that sensitive circuits should also already have protection themselves.

But the world is full of bean counters, so....
Then it still comes down to the users, this to know your equipment well and handle it with care.

If you have a Spectrum or an expensive VNA, when you make an initial measurement, there is always an attenuator on your input!
And if your noise floor is not that important, then leave that attenuator in place.

When developing electronics, I still regularly unplug banana sockets.
If I am testing something in the early stages of development, there is not always protection in place.

The world is not perfect, I certainly am not! 

With kind regards,

Bram

[mawyatt]

You shouldn't have a PS that requires removal of the Banana Leads every time to "protect" your sensitive circuits

Sensitive circuits usually have overvoltage protection. If not, then you should take action yourself and implement it.

Unless you can't as in a product development where you are relying on the chip ESD protection in early stages of said development (example using a Lab Supply to bypass the system or on PCB Power Supply and/or Regulators). A chip ESD will protect from the usual body discharge energy but maybe not be robust enough to handle a PS glitch which evidently some of the mentioned PSs can produce!

And what about the Mains Power interruption, how is one going to predict this in advance and remove the Banana Leads?? What if the PS is part of an automated test setup, here removing the leads seems impractical, altho one may have a UPS with this setup to protect against Mains Power Interruption, however this adds additional expense and complexity in the setup.

Coming from the analog side, and retiring from a career in electronics development, including IC Design, we can tell you that we would never ever allow any type Lab Supply that produces Significant Output Voltage Glitches upon Power Up or Down, whether by Mains Power Switch and/or Output Enable Button.

We can assure you that if this is the character of the new Siglent SPD4000X series, they will never end up on our lab bench as a Lab Supply until it's proven to be acceptable which the SPD3303X and GPP-4323 have met and performed well. Even our high current 30A/15A (read not for use with sensitive circuits) Korad KWH102 and 103 which are SMPS types don't misbehave in this regard!!

Anyway, as always YMMV.

Best

[mawyatt]

My mantra is that a power supply should always protect its load.
And as Martin72 points out, that sensitive circuits should also already have protection themselves.

Agree a Lab PS should not damage or destroy a load by misbehaving!!

Also agree, completed circuits/systems/products should have basic built-in protection. However this usually doesn't apply during early circuit/system/product developments, where one is chasing circuit/system glitches, oscillations, overheating, overcurrent, miswiring, and so on issues. Here, one needs to concentrate on the problems at hand and not Lab Power Supply induced issues

Best,

[mawyatt]

BTW another potential issue wrt the plugging in and out of the Banana leads involves the Lab Supply Output Capacitance. Many supplies have significant capacitance across the supply output terminals and if these are charged to the set point voltage as when the Output Enable is On, this can introduce a significant current "glitch" into the circuit, well beyond the Current Limit Set Point, when the Banana leads are engaged.

Of course one shouldn't plug the leads into a "Live Output", and must remember to make sure the Enable is OFF and no extraneous voltage is present before plugging in the leads, but just another potential source of user error where the result could be disastrous 

We make these comments based upon a 60+ year of electronics development and have many "teeth marks" to show, and have learned what works and what doesn't.

Learning by error is the most lasting impression, learning from others errors in lasting wisdom

Best,

[jjoonathan]

Yep, that's why IMO the first electronics lab class should always start by letting kids explode LEDs on lab power supplies.

  "But I set the current limit to 2mA!"

"Well, you see..." fun times all around.

[Furna]

With SPD4000X series power supply you cannot Turn ON with the Output(s) Enabled; channels are Off even if they were On at previous PS Turn Off.
So you may now think the caputure I took, demonstrates even a worst situation than you tough.
Disconnect the cables and leave happy

So what happens when you have the PS Main Power Switch ON and the Outputs OFF, sensitive circuits connected and a Main AC Power becomes disconnected/interruped 

You shouldn't have a PS that requires removal of the Banana Leads every time to "protect" your sensitive circuits, and then reinstall when you want to power up things again. For us this is totally unacceptable PS behavior and just a recipe for disaster at some point

Best

As you said an UPS could and should be used; also someone else wrote about protection circuits.
That said this is what I have now and I chose to live with it; if the issue will be addressed I will think about my action.

Best

[mawyatt]

Let's all hope Siglent addresses some of these early teething issues, especially the ones that can cause damage rather than just an inconvenience, then we'll consider evaluating one if the need arrises

It's interesting how our equipment needs seem to follow new introductions, like the superb SDH814X HD. We've explained this to our CEO/CFO/Boss as just a unique coincidence

Best

[KungFuJosh]

our CEO/CFO/Boss as just a unique coincidence

Does that mean "the wife?" 

[mawyatt]

 

[tautech]

Freshly arrived SPD4306X tests.
FW V4.1.2.8R1

Measurements taken with DP-35 differential probe @ 20:1 to eliminate any chance of Gnd bounce abnormalities.
Probe Check not run to remove any amplitude error.
Single shot captures.

Output set to OFF and for 2V 30mA on Ch1 (15V 1.5A max)
Unloaded output except for Diff probe
Mains power ON




Ch1 ON into 50 Ohm 3/4" banana load.

[mawyatt]

That looks good, so Siglent must have addressed these "Output Voltage Glitch" issues with the new firmware release

If so, this demonstrates Siglent is "listening" and responding quickly with important updates that address serious issues such as this, so kudos to Siglent

Hopefully others will follow up!!

Best

[sparky2024]

Hi @mawyatt, Unfortunately V4.1.2.8R1 did not solve the issue on my SPD4323X.  Does anyone else have the SPD4323X, and do they have the same issue with power glitch when mains power is turned off, at channels 2 & 3?

Thanks