The capacitor box must have some bleeder resistor in it that causes the 6 uA in one of the ranges. The current correspond exactly to 100 MOhm and is way to stable to be due to capacitor leakage. Also 100 MOhm would be an extremely bad insulation resistance for any type of polymer capacitor.
Shahriar, if you get a chance to interview some of the Keithley designers and ask them questions regarding basic design choices (within IP laws of course) that would be a very nice addition to your channel.
WoW 😲 it's absolutely fascinating what those SMU's can do, I was watching the video in absolute awe, that's how I last stared when I saw the capabilities of the new Huntron curve tracker. I just didn't understand the price, after watching your fantastic sample videos I'm not surprised at the price anymore, I just shed a tear again because the exchange rate of our currency makes SMU unattainable for me. 👍 Nice day everyone 🙂 Tom
Nice, more Keithley toys. Now TSP need 2651A or 2657A review ;) Sadly, the actual hardware or even just brief look of most interesting analog circuit board is not shown. I wish TSP would spend additional 5 minutes to do that, just like its done for all the countless wonderful RF equipment repair/teardowns. I believe just removing outer board for few nice shots wouldn't ruin the instrument. For example, those large ferrite coils with coax cable windings is isolated low-noise DC-DC to power analog domain. Keithley have patent published about that, similar like other LCD SMUs. Given the high cost of such instruments, it is nearly impossible to find any photos of inner boards of recent hardware such as these new SMUs, so TSP have excellent chance to help with this information void... Those read triaxial ports are not just usual triaxial, but SHV triax version with higher insulation ratings. It's easy to spot the high-voltage variant by protruding PTFE insulation rings.
I will probably do a more detailed teardown later on, I just didn't have the time during this review. There are many other tasks I still have to complete in the lab. :)
@@Thesignalpath Thanks. Such instruments deserve more than one video, I think I am not the only one with that idea. If you need some low current tests, I also have Keithley 6430 (the ultimate low current SMU) and just 30 minutes away from you :)
I think I recognize the auxiliary PSU as an XP-Power 100 W module. These are available in multiple output configurations but the formfactor is the same for all 100 W models as is the (apparent) PCB layout. Nice to see and a logical step as well. They are quite clean (EMI), and when used within spec last for years and years. Good to see them not reinventing the wheel. Great video!
Thanks for reminding me to update my DMM firmware! Sadly it didn't update into an SMU :( Also if anyone else here has the same issue as me with Keithley’s firmware downloads on their webpage asking you to update your profile details due to US legal requirements they’re referring to your details on the “Change Address” and “Change Translated Data” pages (whatever that means).
Hi Shahriar, I always enjoy your in depth review and testing, great video. I also have that SMU in my "Dream Lotto Lab" if I ever win the lotto!..lol Take care, Shahriar, and thank you for many years of education and insight on electronics.
I sold my house so that I could afford to buy a handful of triax connectors and cabling :) It really is a shame that Kickstart became pay as you go/ subscription and is a lot more expensive now. Just saying.
Kickstart is really nice, juts bought a perpetual suite license, but some drawbacks - I have a 2250 and two 2281 (battery tester and power supply), but cannot combine it into the IV Characterizer, Though the 2281 is a power supply its programmable, and could be used to form one source to step through the voltages (limit 0..20V ). On the other side the 2281 appears as power supply and battery tester of course, but the 2250 does not appear as power supply though it can be used as pure power supply ? -- so some improvements can be one in the Kickstart Software. -- The IV tracer license, is very high, higher than a perpetuall kickstart suite license which I dont understand as the IV Characterice (comes in teh suite cheaper), can do more. A pitty that APps cannot be programmed by themself only TP Scripts (ist aroun dthe same), buts this is a closed system- I requested for a APP developer SDK or so without getting response
Did you try high-capacitance output mode? The standard mode is designed to be used with no more than 20 nF. PS: I've fried a Keithley 2602 by loading it with electrolytic capacitors, causing oscillations in the SMU's control loop. Usually the measured current was way off during oscillations.
One thing that bothers me about the Keithley GUI is that anything from 100V up is shown in kV. I always have to stop and think when I see that and I find it quite distracting when there is a lot going on.
Would be nice for the Y axis to be logarithmic, and the display to be in volts rather than kV eg 900v instead of 0.9kV or 100MOhms instead of 0.1GOhms.
In the varistor measurement, when you are discussing the part of the curve where the current limit kicks in: Any thermal effect in the varistor shouldn't care about the polarity of the applied voltage. The curve should be folded back towards zero on both ends. However in your measurement the positive end curves forward, not back. The positive part is not a mirror image of the negative part. Hence I think what you saw was not a thermal effect in the varistor itself.
I think it was a thermal effect on the varistor here. As we had a cold start, the varistor starts at a low temperature and then increases its temperature to an intermediate level. At this temperature the clamped voltage has about the same magnitude whether negative or positive. At the end of the measurement the varistor already is at an intermediate temperature and gets stressed even further to its highest temperature at the end of the test. As a result, it also increases its clamped voltage and the curve goes to the right (or "forward", as you put it). So in both cases the result of the temperature increase was an increase of the magnitude of the clamped voltage.
Great instrument, but common, when are we going to stop doing stupid things such as licensing basic computer software. Those pieces of software are a great selling point for the device, but buying them is not good. They should be open sourced so people can modify them since they are usually buggy and lacking in features. Companies milk their customers way too hard, in my opinion they would be making more money if they didn't charge for such basic software.
@@jjoonathan7178 It used to be free with the units. The someone had a bright idea that makes the rest of us groan. I'd carry on using the original version which is Window7 only but I'm guessing that later firmware has closed that door. Sucks. $1K for a prepetual license that you know at some point they'll say isn't supported anymore so buy the new version again.
I wish the instead of their TSP language (TSL), used Python 3. The benefit of their TSL (Actually checking manualls TSP programming, a TSL is just Lua with some libraries to interact with the device and network), is that it actually runs on the device, and can be stored on the device, with much tighter timing and response to triggers and conditions, and can run offline. Of course you can use GPIB with any language you like with these (and other) devices from some computer or controller, it will be just be somehow slower, but more flexible. The benefit of Lua (TSL) over Python is that TSL is somehow faster and easier to isolate for faults than Python interpreter and all the libraries.
It's a shame Keithley didn't send you a spare set of PCBs to review... By the way isn't compressed air a bad way to blow out a flame due to it's flammability?
Most canned compressed air contains R-134a, which isn't very flammable under normal circumstances. Funny how the EPA fines people tens of thousands for venting their car refrigerant (R-134a) but doesn't mind you wasting hundreds of cans containing R-134a for trivial purposes.
@@charliemarie44 Personally I think there is so much open source software of excellent quality I prefer to support these projects by donation, for instance Kicad which is a PCB design CAD package that suits my needs perfectly. By contrast Altium has a broader range of features and is many thousands of pounds to buy. To summarize, not much incentive to buy any I would always look for an open source project that you can donate to and fits your needs.
I always wonder what the designers of such tools with their langauges think when they provide basically a new language instead of making it available as a library in python or so... people often have already some development environments and its easier to integrate, espeially if you are starting to write algorithms that you later then want to incorporate into a product. Also you can then access much more libraries for algorithms.
The capacitor box must have some bleeder resistor in it that causes the 6 uA in one of the ranges. The current correspond exactly to 100 MOhm and is way to stable to be due to capacitor leakage. Also 100 MOhm would be an extremely bad insulation resistance for any type of polymer capacitor.
Nice observation.
Shahriar, if you get a chance to interview some of the Keithley designers and ask them questions regarding basic design choices (within IP laws of course) that would be a very nice addition to your channel.
WoW 😲 it's absolutely fascinating what those SMU's can do, I was watching the video in absolute awe, that's how I last stared when I saw the capabilities of the new Huntron curve tracker. I just didn't understand the price, after watching your fantastic sample videos I'm not surprised at the price anymore, I just shed a tear again because the exchange rate of our currency makes SMU unattainable for me. 👍
Nice day everyone 🙂 Tom
Nice, more Keithley toys. Now TSP need 2651A or 2657A review ;) Sadly, the actual hardware or even just brief look of most interesting analog circuit board is not shown. I wish TSP would spend additional 5 minutes to do that, just like its done for all the countless wonderful RF equipment repair/teardowns. I believe just removing outer board for few nice shots wouldn't ruin the instrument. For example, those large ferrite coils with coax cable windings is isolated low-noise DC-DC to power analog domain. Keithley have patent published about that, similar like other LCD SMUs. Given the high cost of such instruments, it is nearly impossible to find any photos of inner boards of recent hardware such as these new SMUs, so TSP have excellent chance to help with this information void...
Those read triaxial ports are not just usual triaxial, but SHV triax version with higher insulation ratings. It's easy to spot the high-voltage variant by protruding PTFE insulation rings.
I will probably do a more detailed teardown later on, I just didn't have the time during this review. There are many other tasks I still have to complete in the lab. :)
@@Thesignalpath Thanks. Such instruments deserve more than one video, I think I am not the only one with that idea. If you need some low current tests, I also have Keithley 6430 (the ultimate low current SMU) and just 30 minutes away from you :)
I think I recognize the auxiliary PSU as an XP-Power 100 W module. These are available in multiple output configurations but the formfactor is the same for all 100 W models as is the (apparent) PCB layout. Nice to see and a logical step as well. They are quite clean (EMI), and when used within spec last for years and years. Good to see them not reinventing the wheel.
Great video!
Great review! Thanks for posting.
Thanks for reminding me to update my DMM firmware! Sadly it didn't update into an SMU :(
Also if anyone else here has the same issue as me with Keithley’s firmware downloads on their webpage asking you to update your profile details due to US legal requirements they’re referring to your details on the “Change Address” and “Change Translated Data” pages (whatever that means).
Hi Shahriar, I always enjoy your in depth review and testing, great video. I also have that SMU in my "Dream Lotto Lab" if I ever win the lotto!..lol Take care, Shahriar, and thank you for many years of education and insight on electronics.
Yeah SMUs are what I lust for the most
Thank you. :)
A look inside this device would be awesome!
I sold my house so that I could afford to buy a handful of triax connectors and cabling :) It really is a shame that Kickstart became pay as you go/ subscription and is a lot more expensive now. Just saying.
Last time I was this early, the review wasn't finished yet.
Beautiful SMU. If only I had a spare $10K burning a hole in my pocket.
... Just one??? I need TWO! (I would permanently connect them to my gas hot water cylinder as explained in the video)...
Kind regards - Jeff Bezos
Hello, how is assembled the interlock on the 2470 model?
Very informative. Thanks. Keithley is way out of my budget though. I like the experiment format.
@Kevin A. You get what you pay for. Quality has its price.
Kickstart is really nice, juts bought a perpetual suite license, but some drawbacks - I have a 2250 and two 2281 (battery tester and power supply), but cannot combine it into the IV Characterizer, Though the 2281 is a power supply its programmable, and could be used to form one source to step through the voltages (limit 0..20V ). On the other side the 2281 appears as power supply and battery tester of course, but the 2250 does not appear as power supply though it can be used as pure power supply ? -- so some improvements can be one in the Kickstart Software. -- The IV tracer license, is very high, higher than a perpetuall kickstart suite license which I dont understand as the IV Characterice (comes in teh suite cheaper), can do more. A pitty that APps cannot be programmed by themself only TP Scripts (ist aroun dthe same), buts this is a closed system- I requested for a APP developer SDK or so without getting response
Did you try high-capacitance output mode? The standard mode is designed to be used with no more than 20 nF.
PS: I've fried a Keithley 2602 by loading it with electrolytic capacitors, causing oscillations in the SMU's control loop. Usually the measured current was way off during oscillations.
Can you please show how to make a pulse voltage of particular pulse width using 2450?
One thing that bothers me about the Keithley GUI is that anything from 100V up is shown in kV. I always have to stop and think when I see that and I find it quite distracting when there is a lot going on.
I also prefer not showing things in kV.
So 14 February after dinner video 😎
Would be nice for the Y axis to be logarithmic, and the display to be in volts rather than kV eg 900v instead of 0.9kV or 100MOhms instead of 0.1GOhms.
I like it
In the varistor measurement, when you are discussing the part of the curve where the current limit kicks in: Any thermal effect in the varistor shouldn't care about the polarity of the applied voltage. The curve should be folded back towards zero on both ends. However in your measurement the positive end curves forward, not back. The positive part is not a mirror image of the negative part. Hence I think what you saw was not a thermal effect in the varistor itself.
I think it was a thermal effect on the varistor here. As we had a cold start, the varistor starts at a low temperature and then increases its temperature to an intermediate level. At this temperature the clamped voltage has about the same magnitude whether negative or positive.
At the end of the measurement the varistor already is at an intermediate temperature and gets stressed even further to its highest temperature at the end of the test. As a result, it also increases its clamped voltage and the curve goes to the right (or "forward", as you put it).
So in both cases the result of the temperature increase was an increase of the magnitude of the clamped voltage.
Great instrument, but common, when are we going to stop doing stupid things such as licensing basic computer software. Those pieces of software are a great selling point for the device, but buying them is not good. They should be open sourced so people can modify them since they are usually buggy and lacking in features. Companies milk their customers way too hard, in my opinion they would be making more money if they didn't charge for such basic software.
$10,000 for the SMU and it doesn't even include the software. Yikes.
@@jjoonathan7178 It used to be free with the units. The someone had a bright idea that makes the rest of us groan. I'd carry on using the original version which is Window7 only but I'm guessing that later firmware has closed that door. Sucks. $1K for a prepetual license that you know at some point they'll say isn't supported anymore so buy the new version again.
if i ate up a 2450 to ignite an arc lighter id loose my job, but i do wish i had the "im filming a video" excuse to not set up a shielded triax jig :P
Nice review! As usually. That software is a joke:D I'm using python for all of my instruments via GPIB.
You can still do that here as well.
I wish the instead of their TSP language (TSL), used Python 3. The benefit of their TSL (Actually checking manualls TSP programming, a TSL is just Lua with some libraries to interact with the device and network), is that it actually runs on the device, and can be stored on the device, with much tighter timing and response to triggers and conditions, and can run offline. Of course you can use GPIB with any language you like with these (and other) devices from some computer or controller, it will be just be somehow slower, but more flexible. The benefit of Lua (TSL) over Python is that TSL is somehow faster and easier to isolate for faults than Python interpreter and all the libraries.
that is a nice Curly Maple stylus
It's a shame Keithley didn't send you a spare set of PCBs to review...
By the way isn't compressed air a bad way to blow out a flame due to it's flammability?
This particular spray is non-flammable.
Most canned compressed air contains R-134a, which isn't very flammable under normal circumstances. Funny how the EPA fines people tens of thousands for venting their car refrigerant (R-134a) but doesn't mind you wasting hundreds of cans containing R-134a for trivial purposes.
I thought you had the Rohde & Schwarz SMU R&S®NGU
It will be reviewed as well.
NGU is more like glorified power supply / meter rather than true SMU. It's not in competition to 24**, leave alone 26** series of Keithley stuff.
Paying for software seems old fashioned now !...cheers.
I’m new to the industry and am looking into software options.. what would make purchasing software worth it in your opinion?
@@charliemarie44 Personally I think there is so much open source software of excellent quality I prefer to support these projects by donation, for instance Kicad which is a PCB design CAD package that suits my needs perfectly. By contrast Altium has a broader range of features and is many thousands of pounds to buy. To summarize, not much incentive to buy any I would always look for an open source project that you can donate to and fits your needs.
Wish list had a new entry
I always wonder what the designers of such tools with their langauges think when they provide basically a new language instead of making it available as a library in python or so... people often have already some development environments and its easier to integrate, espeially if you are starting to write algorithms that you later then want to incorporate into a product. Also you can then access much more libraries for algorithms.
$12,400 😭