That does not make sense to me. The two overlays show two different things (take a look at the plot titles). Or did you maybe not realise that there are two curve-sets in each figure because they overlap so much?
Those Ridens look good and are widely available.
(Edit: are there any plans to support the W-version, i.e. WiFi programming? I can imagine it will eliminate a potential ground loop. If not, does it hurt to get W version?)
I was looking at USB programmable PSUs on aliexpress this morning before I found your latest update. The Wanptek units are dirt cheap and look good. I wasn't able to find out though what you can program. One reviewer seems to have even gotten a unit where the USB port was not connected and didn't have the signals even on the PCB.
(Edit: are there any plans to support the W-version, i.e. WiFi programming? I can imagine it will eliminate a potential ground loop. If not, does it hurt to get W version?)
I was looking at USB programmable PSUs on aliexpress this morning before I found your latest update. The Wanptek units are dirt cheap and look good. I wasn't able to find out though what you can program. One reviewer seems to have even gotten a unit where the USB port was not connected and didn't have the signals even on the PCB.
Last edited:
A thing that springs to mind would be latency over WiFi. Unpredictable latency too depending on quite a few environmental factors.
Nice find!
PyPSUcurvetrace allows you to specify coefficients to calibrate the set and readback values of each PSU from external voltage and current calibrations. It should therefore be possible to get accurate setpoint and readback values with the Riden even at very low currents if you determine your own calibration data.
Uhm, I need to work on the documentation of this feature...
Software developers the world over feel your pain...Uhm, I need to work on the documentation of this feature...
Not yet. I am still thinking how and where to implement such parameter analysis.
However, you can do just about any analysis on the data files. Just load the ASCII data into your preferred number crunching software to calculate whatever parameter you like.
Looking at my own post #22 the Voltcraft PPS current resolution is maybe 20 mA or so, nothing to get excited about. I wouldn't recommend to get a new Voltcraft PPS just to use it for curve tracing. Other manufacturers like Riden, Korad/RND, B&K, etc. have much more suitable units. I only used the Voltcrafts because that's what was on my bench when I had the idea for PyPSUcurvetrace.
I got my hands on a Riden 6012P power supply. This thing goes up to 60 V and 12 A. The voltage resolution is 1 mV. The current resolution is 0.1 mA if using the "low-current" mode from 0...6 A, or 1 mA if using the "high-current" mode (12 A).
The Riden driver in PyPSUcurvetrace needed a few tweaks to work smoothly with the 6012P, but now it works nicely.
The Riden 6006P and 6012P allow high power and high resolution at the same time, and they don't cost an arm and a leg! Great stuff! I may end up selling some of my other (expensive!) power supplies...
The Riden driver in PyPSUcurvetrace needed a few tweaks to work smoothly with the 6012P, but now it works nicely.
The Riden 6006P and 6012P allow high power and high resolution at the same time, and they don't cost an arm and a leg! Great stuff! I may end up selling some of my other (expensive!) power supplies...
As I'm sat in the man cave drinking my coffee this morning this thread reminded me.
The Multicomp Pro power supply I have is a MP710081 and that has the command codes at the back of the operating manual: https://www.farnell.com/datasheets/2829738.pdf
I suspect these will be the same as the Voltcraft, and it looks that if it works for the Voltcraft then it will work for this Multicomp Pro power supply too (at least the commands are the same):
The 710081 also has a static 12Vdc and 3.3/5Vdc supply with a current meter (no current limit) for each on the display but I don't think that is exposed on the programming interface.
The issue is the resolution (probably similar to most of the mass produced models rebranded from the same basic design):
Voltage
So full range, 60V * 0.02 = ±120mV which is the ripple it specifies, then ±9mA not including the counts.. Not quite the Brymen 0.01mV 500,000 count!
I think you may be right with the cheaper power supplies, they may not give a particularly accurate supply or measurement - unless you run multiple runs (measuring the temps naturally) and then average.
The Multicomp Pro power supply I have is a MP710081 and that has the command codes at the back of the operating manual: https://www.farnell.com/datasheets/2829738.pdf
I suspect these will be the same as the Voltcraft, and it looks that if it works for the Voltcraft then it will work for this Multicomp Pro power supply too (at least the commands are the same):
GMAX - get max voltage and current supported
SOUT - enable/disable
VOLT - set voltage
CURR - set current
GETD - get voltage, current, CV mode and CC mode from the display
The 710081 also has a static 12Vdc and 3.3/5Vdc supply with a current meter (no current limit) for each on the display but I don't think that is exposed on the programming interface.
The issue is the resolution (probably similar to most of the mass produced models rebranded from the same basic design):
Voltage
5 counts for range V<5V
±0.2% +5 counts for range V≥5V
Current15 counts for range I≤1A
±0.5% +6 counts for range I>1A
So full range, 60V * 0.02 = ±120mV which is the ripple it specifies, then ±9mA not including the counts.. Not quite the Brymen 0.01mV 500,000 count!
I think you may be right with the cheaper power supplies, they may not give a particularly accurate supply or measurement - unless you run multiple runs (measuring the temps naturally) and then average.
Last edited:
There's other programmable Korad KEL-102 and 103 units that are quite affordable (i.e. 20% more than an assembled 6006P) and seem to have similar specs. What I like about them is that while they only have 150 resp. 300 W, they offer up to 120 V and up to 30 A (!).
https://www.aliexpress.com/item/33011465500.html?spm=a2g0o.best.moretolove.22.5d522c25rnxS5O&gps-id=pcBestMore2Love&scm=1007.40000.317776.0&scm_id=1007.40000.317776.0&scm-url=1007.40000.317776.0&pvid=8dad56e9-4c75-4c93-ae3c-7d441de669e4&_t=gps-idcBestMore2Love,scm-url:1007.40000.317776.0,pvid:8dad56e9-4c75-4c93-ae3c-7d441de669e4,tpp_buckets:668#2846#8109#205&pdp_ext_f={"sku_id":"12000024681976122","sceneId":"30000"}&pdp_npi=2@dis!USD!249.89!209.91!!!!!@211b88ee16749938888047813eca48!12000024681976122!rec
https://www.aliexpress.com/item/33011465500.html?spm=a2g0o.best.moretolove.22.5d522c25rnxS5O&gps-id=pcBestMore2Love&scm=1007.40000.317776.0&scm_id=1007.40000.317776.0&scm-url=1007.40000.317776.0&pvid=8dad56e9-4c75-4c93-ae3c-7d441de669e4&_t=gps-id
cBestMore2Love,scm-url:1007.40000.317776.0,pvid:8dad56e9-4c75-4c93-ae3c-7d441de669e4,tpp_buckets:668#2846#8109#205&pdp_ext_f={"sku_id":"12000024681976122","sceneId":"30000"}&pdp_npi=2@dis!USD!249.89!209.91!!!!!@211b88ee16749938888047813eca48!12000024681976122!rec
Hmm, these units are described as "electronic loads", not as power supplies. Will they work as programmable voltage or current sources?
That said, the Korad / RND power supplies are good choices for use with PyPSUcurvetrace. My Korad/RND KA3005P and KWR103 units work great for curve tracing!
That said, the Korad / RND power supplies are good choices for use with PyPSUcurvetrace. My Korad/RND KA3005P and KWR103 units work great for curve tracing!
The last few days were quite busy with work on pypsucurvetrace!
(A) I expaneded the code to do more stuff:
Some of the examples have a slight Linux smell (such as the serial port names in the PSU configs). This is simply because I have a Linux machine, and also because I have absolutely no clue about Windows. It would be nice if someone could provide some Windows or macOS equivalents.
(C) I made it easy to install pypsucurvetrace as a Python package. Just enter one single command, et voila!
I decided to keep pypsucurvetrace at the command line, because it's extremely useful to use the commands for scripting. The programs are pretty simple, so I don't see how a GUI would provide an advantage.
(A) I expaneded the code to do more stuff:
- Improved plotting of the curve data
- Added a tool for extracting parameters (gain, output conductance, etc.)
- Added a tool for "curve matching" of parts by calculating the "difference between two curve sets"
Some of the examples have a slight Linux smell (such as the serial port names in the PSU configs). This is simply because I have a Linux machine, and also because I have absolutely no clue about Windows. It would be nice if someone could provide some Windows or macOS equivalents.
(C) I made it easy to install pypsucurvetrace as a Python package. Just enter one single command, et voila!
I decided to keep pypsucurvetrace at the command line, because it's extremely useful to use the commands for scripting. The programs are pretty simple, so I don't see how a GUI would provide an advantage.
Hi Guys. Have been following this thread for a little while and look very exciting. I have the locky curve tracer but did not find it useful for my main application. Like many audio repair engineers, I want a convenient way to check if power BJTs and FETs are fake, but without resorting to buying a second hand 1970s tectronix 576 or similair, as most are forced to do. Do you think the The Riden 6012P & PyPSUcurvetrace could be used for this pupose?