Keithley 2280s power supply

[Circuitous]

I saw a few mentions of the new Keithly 2280S power supply in the forum, but nothing substantial. 
http://www.testequity.com/products/5282/

Has anyone tried out one of these or have any feedback about them?

[Circuitous]

Any feedback on this PSU?   
Is there much fan noise?

[HKJ]

I have been looking at it, but I need raw socket or telnet connection and it does not support it according to Keithley.

[biot]

I have been looking at it, but I need raw socket or telnet connection and it does not support it according to Keithley.

It supports standard SCPI commands over several transports, including TCP (LXI). What's wrong with that?

[HKJ]

I have been looking at it, but I need raw socket or telnet connection and it does not support it according to Keithley.

It supports standard SCPI commands over several transports, including TCP (LXI). What's wrong with that?

Only via LXI interface, you cannot send it directly to a socket. In the manual to the 2450 (SMU) it says that Keithley always support raw socket and telnet, but that chapter is missing from the 2280 manual and a question to Keithley got the answer that it is not supported in the 2280

[Stella]

I saw a few mentions of the new Keithly 2280S power supply in the forum, but nothing substantial. 
http://www.testequity.com/products/5282/

Has anyone tried out one of these or have any feedback about them?

Unlike traditional power supply, 2280S can monitor very small current changes when source the DUT, because it embedded a 6 ½- digit DMM. And 2280S has a big color screen and GUI interface. So it is easy to change settings and watch the voltage and current waveform from the display.
Besides, 2280S supports LXI web interface. It is very easy to control the instrument remotely.

[Circuitous]

I want to thank you all for the feedback.  I did order one, and got a decent price from testequity.com 

Here's a quick and dirty first look at the unit for those who haven't seen one.  It really has a beautiful display.  This review doesn't do justice to the many features of this thing; but I plan to follow-up once I have used it a bit more.  I am really interested in the triggering features and the high resolution readings.

The video has a few glitches, but by the time I noticed, I couldn't re-shoot the video due to other things going on.  The next one will be smoother.

[dr.diesel]

I specifically asked the OS question in regards to the K2450, WinCE/Embedded or Linux etc.  They responded it runs neither, wouldn't further specify.  I assume the 2280 being a somewhat close cousin is likely the same.

[rx8pilot]

I want that PSU pretty bad now. Going to check if they have a slightly higher current version. 10A or so....

[HighVoltage]

Nice review, thanks for that.
Added to the wishlist.

[brycheiniog]

Only via LXI interface, you cannot send it directly to a socket. In the manual to the 2450 (SMU) it says that Keithley always support raw socket and telnet, but that chapter is missing from the 2280 manual and a question to Keithley got the answer that it is not supported in the 2280

I managed to get a raw socket connection going to my 2280. For some reason Keithley have blocked SCPI connection attempts which don't conform to a particular handshake after the socket has been opened. If you use Wireshark to capture the traffic between a PC and the 2280 when you open the SCPI console in the web browser you can clearly see the handshake after the socket has been opened. I replicated this extra step in my PC application and it worked fine after that.

Sorry I cannot post any code as it was developed for work..

[kwass]

I want to thank you all for the feedback.  I did order one, and got a decent price from testequity.com 

Thank you for your very nice video review!  Can you comment on the fan noise (if any) from the 2280S?

[brycheiniog]

My 2280s is not what you would call quiet. I will stick a sound level meter on it on Monday.

[HKJ]

I managed to get a raw socket connection going to my 2280. For some reason Keithley have blocked SCPI connection attempts which don't conform to a particular handshake after the socket has been opened. If you use Wireshark to capture the traffic between a PC and the 2280 when you open the SCPI console in the web browser you can clearly see the handshake after the socket has been opened. I replicated this extra step in my PC application and it worked fine after that.

I expected something like that, the problem is it requires time to decode and write the code and I am always low on that. I also expect Rigol is exactly the same.

[brycheiniog]

I managed to get a raw socket connection going to my 2280. For some reason Keithley have blocked SCPI connection attempts which don't conform to a particular handshake after the socket has been opened. If you use Wireshark to capture the traffic between a PC and the 2280 when you open the SCPI console in the web browser you can clearly see the handshake after the socket has been opened. I replicated this extra step in my PC application and it worked fine after that.

Ok I dug the packet capture out and it looks like when the TCP socket is opened (on port 5051) the 2280 sends an message with 'ACCEPT' in the payload to which the response is: {0xf0,0xf0,0xf0,0xf0,0x04,0x01,0x06,0x00}. After this the normal SCPI commands seem to work fine.

Jonathan

[Circuitous]

Can you comment on the fan noise (if any) from the 2280S?

Not bad, particularly at low load.  At 1 amp load, it's quieter than my Agilent 34410A DMM.  It gets a bit noisy as I get into the 3-6A load range.  But, even that is probably quieter than the Rigol DS4024 scope. 

Overall, the noise doesn't stand out when other equipment is running, until you push it into the 100+ watt range.

[kwass]

Circuitous and brycheiniog,  Thank you for your comments on fan noise.  I was hopping it would be nearly silent at low loads, but that doens't seem to be the case.  Oh well, I'll keep looking.

[Macbeth]

£1600 for a single channel PSU. 

Yeah, I am so happy that I got a brand new Rigol DP832 with 3 channels for a fraction of that. Much nicer interface and quicker boot time too.

For a single channel supply I am more than happy with my second hand HP 6632B that was only £100 notes. I'm amazed at how often I use it. I love it's 2 quadrant feature and voltmeter mode when in "Dis" - I wish my Rigol would do that rather than just grey out the display and read 0.000. Why not show what is happening with those sense wires? (maybe there is a way to do this I am missing?)

ETA: To be fair, if you wanted a "new" Keysight 6632B from Farnell, it will set you back over £2150 with the VAT.   

[es]

Just found a teardown video of it:

[Circuitous]

Thanks! 

That transformer is huge, no wonder the thing weighs a ton.

[majbthrd]

Very interesting, but it seems like there is room for improvement in the user interface.

It baffles me why the x-axis on the graph is labeled in (Sample) "Number", but yet the "Sheet" screen shows time (e.g. it is tracking the time of each sample).  Displaying time on the x-axis of the graph would be a lot more helpful to an engineer, I think, than having to remember what the sampling interval is currently set to and convert sample # to time in one's head.

The other annoyance I saw in the teardown video was that when the measurement accuracy was set to 4 digits for higher sampling rate, the interval of time shown on the Graph was woefully small relative to the activity of the device being monitored.

So, the first thing an engineer faced with this would want to do would be to change the time base to zoom out.

The reference manual says: "Only when buffer is in never mode, you can configure position and zoom to specify the readings that are displayed on the screen. Otherwise, if buffer is in next or always mode, the customized position and zoom settings cannot be applied to graph."

So, you have to stop acquisition to change the zoom!?    It seems like the firmware should be capable of allowing the user to change this dynamically.

[Circuitous]

I must have missed a notification on this thread...

Yeah, I am so happy that I got a brand new Rigol DP832 with 3 channels for a fraction of that. Much nicer interface and quicker boot time too.

I have a 3 channel Rigol PSU, and a similar 1 channel (DP1308A & DP1116A).  They work fine, but I wasn't looking for a regular bench supply.  I wanted to characterize the power consumption of low power devices, with their sleep and wake cycles.
I found the Rigol really couldn't do that on its own. 

Some of the controls/menus are a bit awkward, but perhaps they'll improve, or I will just get used to it.  You'll see me fumble around a couple of times while setting up the external trigger.  And my camera wasn't giving sharp focus on some shots... and zoomed too closely a few times and cut out part of the display...and more errors I'm sure!

[Macbeth]

I have a 3 channel Rigol PSU, and a similar 1 channel (DP1308A & DP1116A).  They work fine, but I wasn't looking for a regular bench supply.  I wanted to characterize the power consumption of low power devices, with their sleep and wake cycles.
I found the Rigol really couldn't do that on its own. 

Ahh, I see. Great vid. That is a PSU on steroids, I can see how it's useful for such low power devices.

Also, those RF modules are amazingly cheap. I've got a few HC05/HC06 Bluetooth UART transceivers coming from China, just a few dollars each.

[majbthrd]

It appears that Keithley recently released v1.06 firmware for the 2280S.

One of the features added is raw socket access on port 5025.  I just tried the upgrade and can confirm raw socket access.

[tuju]

Just found a teardown video of it:

He actually said in 10:30 ”....our operating system is Linux ...”. Thanks, I was wondering does it have showstopper Windows but now I can consider buying it.