New entry-level R&S NGE100 Power Supply Series

[JanJansen]

Hi, i like to know : does this PSU still have problems today when you buy it ?
It seems nobody is buying it, i cant find many youtube movies.

[Rich@RohdeScopesUSA]

Hi,

I did a quick test with the scope: 20V output voltage, current limit set to 50mA. Putting a 330R resistor across the output and capturing the resulting voltage with the scope. That outcome can be seen in the image attached below. Pretty pathetic. The voltage stays at the 20V when the resistor is connected (around 60mA, remember 50mA limit is set) and then drops a little bit after ~150ms staying at one level for about 600ms before it drops to the 50mA it is set to. So from connecting the load to reaching the set current it "only" took about 1s. Pretty bad.

Now a few videos:

https://youtu.be/CVJEZAvv9l4
In this video I short the output with the amp range on the multimeter. Voltage set to 20V, current limit of 50mA. As soon as I connect the meter, the voltage drops to 0 volts and it takes about a second to rise up to the current limit of 50mA. You also see the current display wildly fluctuating until it sets to the 50mA, whereas the meter shows the 50mA as soon as the output rises.

https://youtu.be/VKatT8dCHn0
Here I have a current limit of 50mA with an output voltage of 12V. Again connecting the amp meter across the output. Same delay as before, but now it also supplies almost 80mA and is steady but slowly dropping. So again, total fail.

https://youtu.be/cyDhJy3BPSc
Now I'm using the 330R resistor again. Voltage set point is 12V, current limit is 30mA. When I connect the load 37mA are supplied without ever limiting to 30mA.

https://youtu.be/KFM-quOGaLY
Same settings as in video before. Shorting with amp meter. Now it supplies 60mA on a set point of 30mA.

Ok, so I registered for a youtube account only to be able to post those videos here, so please let me know if they don't work, or any other thing is wrong.
I think with this testing, the fate of the supply is sealed. It will go back and I will be looking for another brand.
About US$1000 is ridiculous for this supply, I only would pay the premium if I would have gotten R&S quality.

Anyways, hope this is informative.

Michael
 
Edit: Grammar...

Sorry for the late response to this - this issue has been resolved with FW 1.51.

-Rich

[Octane]

Hi Rich,

Thanks for pointing that out. Sadly too late for me now. But who knows, maybe I’ll buy it again some day.

But besides that, how can something like this be overlooked?
That’s the very most basic functionality of such a PSU.

Thanks,
Michael

[skench]

Hi Rich,

I have a NGE103 and the firmware upgrade page does not list FW 1.51 only 1.40.

I see the newer version the NGE103B has FW 1.51. Will the NGE103 be upgraded to the same level?

Thanks
Stephen

[Rich@RohdeScopesUSA]

Hi Rich,

I have a NGE103 and the firmware upgrade page does not list FW 1.51 only 1.40.

I see the newer version the NGE103B has FW 1.51. Will the NGE103 be upgraded to the same level?

Thanks
Stephen

Hi Stephen - yes, same FW between the two.  You can load 1.51 on your NGE100.

-Rich

[skench]

Thanks Rich,

Firmware updated to 1.51 and all appears to be working.

-Stephen

[Gearbuilder]

I have three units with 1.52 and they do have issues:

This is a great little supply but the digital IO option absolutely does not work! It has the following problems:

1) The IO lines latch up frequently and the digital IO needs to be manually disabled and then re-enabled.
2) The IO lines are very sensitive to RF and interference from nearby ESD events. Adding filters and shielding does not seem to help.
3) The IO lines draw current when they are disabled. They draw 44 ma when driven by 5 V.
4) The IO inputs cannot be enabled while they are being driven from another logic source. The input must be low when enabling an input. Otherwise the input latches up and draws current to a voltage part way between 0 and 5 V. This means that the drivers must be physically disconnected every time an input line is enabled.
5) The individual digital IO lines are always disabled when the unit is turned on. Enabling them is a long and involved (75-step) procedure. Issue 4 makes this even more problematic.
6) Recalling the digital IO settings turns the each of the digital IO lines off. Again, enabling them is a long and involved procedure. Issue 4 makes this even more problematic.
7) The digital IO fails even when no wires are attached. Connecting pin 1 (configured as an output) to pin 2 (configured as an input) will result in an interface that fails after a few minutes of operation.

The Ethernet, WiFi, and Digital IO options are now bundled together in a promotional package for about $995 through the end of 2019. Just beware of the fact that the Digital IO is completely non functional! It may still be a good deal, but I was expecting to be able to use the digital IO.

 

[thm_w]

2) The IO lines are very sensitive to RF and interference from nearby ESD events. Adding filters and shielding does not seem to help.
3) The IO lines draw current when they are disabled. They draw 44 ma when driven by 5 V.

So just one input is drawing 44mA, not three as you've shown tied together in the photo right?

Thats wild if it is drawing that much but somehow is still sensitive as an input.. Can't imagine what circuit would be there. I didn't see any specs in the manual other than "don't exceed 5.25V".


The Rigol is a bunch of resistors and transistors but I haven't actually tried using it yet: https://www.flickr.com/photos/eevblog/9607780666
A good input I would expect to have an opto-isolated I/O but that probably adds too much to the cost.

[Gearbuilder]

So just one input is drawing 44mA, not three as you've shown tied together in the photo right?

Thats wild if it is drawing that much but somehow is still sensitive as an input.. Can't imagine what circuit would be there. I didn't see any specs in the manual other than "don't exceed 5.25V".

The Rigol is a bunch of resistors and transistors but I haven't actually tried using it yet: https://www.flickr.com/photos/eevblog/9607780666
A good input I would expect to have an opto-isolated I/O but that probably adds too much to the cost.

Yes, each IO line draws 44 mA when disabled (if you attempt to drive it to 5 V).

I didn't have a photo of digital IO 1 driving digital IO 2. In the photo I was attempting to drive DIO2 and DIO3 from DIO1. The purpose was to trigger channels 2 and 3 when channel 1 turned on.

The biggest problem is that the inputs randomly start drawing current several minutes after being enabled. This can only be corrected by disabling the inputs, allowing them to drop to ground, and then re-enabling the inputs. If the inputs are actively driven high or are pulled up with a resistor, the inputs will not reset (they remain latched up, drawing current from an intermediate voltage level between 2 and 3 volts).

It is amazing that this interface doesn't work given the fact that you have to pay $150 to turn it on. You then have to enter a long license code that unlocks the feature. R&S spent more time figuring out how to license the feature than they ever spent on making sure that it works. Unbelievable! If I didn't have three units, I would think that I had a bad unit.

 I feel like I am the first person that ever tried to use this interface. 

[exe]

Yes, each IO line draws 44 mA when disabled (if you attempt to drive it to 5 V).

Sounds like they actively pull it down or something about ESD/protective diodes and internal rails. Did you try to drive from lower voltage? Like, from 3.3V?

I'd also suggest contact R&S.

[Rich@RohdeScopesUSA]

Hi Folks - just a quick update to this:

We're making a couple changes to the NGE's firmware based on Gearbuilder's feedback (should be released before September):

1.  The team had erred on the conservative side for the DIO pins and how they were enabled after recalling settings.  They were thinking of applications when a customer may not always check DIO or the channels' wiring when they play with the trigger settings or recalling totally different DIO settings from memory.  With that said, we've decided to implement a solution where the NGE will store the settings as they used to, but we're adding a button to activate all four Digital I/Os to make that simpler/faster.  I think Gearbuilder will have a beta soon for that.

2.  With respect to the current on the IO lines, there currently is a small amount of current flowing even if the port is not used. While this theoretically shouldn't be an issue, we understand the concern, so we're going to modify the firmware to eliminate this.  This will also be in the updated firmware currently expected before September.

-Rich

[thanasisk]

Hi,

Have the problems in the current limiting and overshooting and not working digital IO been resolved in the new NGE103B (presently bundled as NGE-COM3a)?

[Rich@RohdeScopesUSA]

Hi,

Have the problems in the current limiting and overshooting and not working digital IO been resolved in the new NGE103B (presently bundled as NGE-COM3a)?

Yes, those are fixed.  There is also a new firmware coming out soon (1.54) with additional updates/fixes.  As soon as I have those I'll update here as well.

-Rich

[goaty]

Just got the NGE103 running, nice compact unit, seems to work.

Besides the price, there are two design choices I really don´t understand:

- Why does the three channel model not also display the statistics like the two channel model ?
I do not need it, but seems to me like a really strange decision

- The tracking function is really a joke. I thought it links channels together 'permanently', but all it does
is let you modify the voltage and current momentarily. As you leave the setting and enter it again,
the tracking is not selected anymore.
I really wish R&S would improve that and let you link 2 or 3 channels permanently, and then whenever you change
voltage, both setpoints change. (Also in live mode, you need to remember you want channels to be tracked, if not, you just modify the first that comes up )

Also, the response of the unit to serial connected channels does not seem to work in all cases.
When limiting current to 100mA, and going into overcurrent condition by a certain load,
suddenly one channel displayed and drove 3A. [IMG_0713]
(Edit: This seems to happen only if one does not follow the recommendation in manual to set all currents to same value)
Seems there is at least one more bug. (Got 1.54 here)

And another bug when pressing Track as can be seen in image:
The knob modifies current, but the button "Voltage" is illuminated. (IMG_0711)

Please R&S: If you really must produce such low-quality software, at least give us the complete sourcecode so we can improve.

Edit: There´s more that drives you crazy: You cannot enable Fuse on an currently disabled channel if the output is already switched on with one other channel enabled.
First disable output, then select fuse on the new channel, then enable output. But I want the output to stay on. Damn.

Oh, and it doesn´t remember the last set voltages as well. It always resets to one of the memory settings. Who has thought that out ? Almost killed my circuit with that "feature"

Greetings from Munich/Germany
Thomas

[goaty]

Update: I swapped the NGE103 for a HMC8043 after not being too happy with the user interface of the NGE.
The HMC is more expensive, but the features and usability are much better. The options for interfaces and sensing all included, also same
features like fuse.
Only thing I wasn´t aware of is the reverse polarity protection, which is basically a diode across the terminals. Wasn´t aware of that. Be careful when connecting a battery for example, to connect it the right way around.

[eevbstedt]

There is also a new firmware coming out soon (1.54) with additional updates/fixes.  As soon as I have those I'll update here as well.

-Rich

Update?

[Alex P]

There is also a new firmware coming out soon (1.54) with additional updates/fixes.  As soon as I have those I'll update here as well.

-Rich

Update?

It was released March 16, 2020. See https://www.rohde-schwarz.com/nl/firmware/nge100b/

[alonsojar]

Hi, there is a new firmware 1.55 released since Friday. I've noticed this is deployed in the R&S Spanish download page https://www.rohde-schwarz.com/es/firmware/nge100b/ but not in the English one by now. I imagine it will be generally available soon, but if anyone wants to take a look.... . At the first sight looks very close to the previous one, except the fan behaviour that now has more hysteresis (less start-stop cycles).

[Electronics404]

When I set the current set point to 20mA it immediately went into CC mode without anything connected to the terminals.

Does anyone else have experience with this behavior?
When I got my NGE103 I noticed that behavior on CH1 and I got in contact with the R&S support.
I was told to send it back and just got a replacement unit delivered today. Unfortunately, it has the same issue but this time on CH3.
Video of the issue on my first unit: https://youtu.be/ltwoPL4OYWs

I also noticed that CH2 & CH3 enter CC-mode about 15mA before actually reaching the set current.
I use an active load to apply a constant resistive load to the supply's output.
As you can see in the video, I increase the load until about 85mA where it already shows cc-mode (set current is 100mA). But it looks like it's constantly trying to jump between CV and CC.
Only if I increase the load further, it will eventually limit to the set current and decrease the voltage without jumping around.
Video:

I don't know if this is a hardware issue or software related, I got the unit delivered with FW version 1.56 and tried updating to 1.57, but the same issue still persists.

Not sure what to do now, with this being the 2nd unit that has issues and reading previous post in this thread...
I mean currently I would trust my two Korad supplies over this any day, and they were like 1/10th the price 

EDIT:

I turned the supply on today and now I have problemes even with CH1, that I didn't have when I tested yesterday!

Video: https://youtu.be/CH3KNrPozws
Like I already mentioned above: CH3 stays at 0V, even with nothing connected to the output, until I increase the set current to above 15mA...
But now CH1 won't start limiting the current until around 40mA, with a set current of only 3mA??? It didn't do this yesterday!

Video: https://youtu.be/NoL4lNmjB5c
Another test with the set current at 100mA: It won't start limiting until about 135mA!

[exe]

> I use an active load to apply a constant resistive load to the supply's output.

Do you have an oscilloscope to check for oscillations? Which active load do you use?

What else you can do is to use resistor and gradually rise the voltage to see at which current the CC mode trips.

[Electronics404]

Do you have an oscilloscope to check for oscillations? Which active load do you use?

I use this active load: https://www.tekbox.com/product/tboh02-self-powered-active-load/
But I don't think that it has anything to do with the issue, since I did the same exact test with my Siglent SPD3303X supply, which behaved just like you'd expect.
I also tried using a decade resistance box instead of the active load, just to be sure and rule out the active load as a potential error source.

It switches to CC mode at around 85mA, which I also confirmed with an external multimeter. So what the NGE is showing, are the correct voltage and current values.
But like I described in my last post, and can be seen in the video, the output is not constant. It jumps between CV and CC mode while the load current is between ~85mA and 100mA.

I also looked at the output with an oscilloscope, looks clean.

I've also had issues come and go after rebooting the device:
- I've had CH2 working correctly, but instead CH1 wouldn't switch to cc mode until around 40mA above the set current.
- I've had the device boot up, but none of the buttons worked.

[exe]

Based on what you described, things don't look right to me. My internet is very bad, so I can't watch the videos you made. I hope R&S will  resolve the issue. Please post here if there are new findings.

One speculation: if the unit has multiple current ranges, it might happen that the upper range is not very precise when used at the bottom of the range. I wonder if the unit still misbehaves at higher currents and if the 15mA offset (when CC triggers at 85mA instead of 100mA) is consistent. I.e., if you set current to, say, 500mA, will it trigger CC at 485mA?

PS the load looks cute to me, and I think I can recognize the parts it's built from. I must resist the temptation to buy one...