jackinnj said:
Comparing pairs (and maybe multiples) of devices is currently the highest-priority add-on for my curve tracer. I was just starting to work on that when I interrupted it to first basically re-design the whole unit, which was initially supposed to be just for "manufacturability". For that, I was initially only planning to move the main board to be parallel to (and just behind) the front panel, so all of the controls could be PCB-mounted instead of hand wired. I also eliminated (almost) all other soldered-wire connections. But, during that process, I decided to also get every subsystem of the whole unit into LT-Spice and then optimize all of it. (I got "a little carried away" with that, but learned a tremendous amount, and am finally satisfied with the performance of everything in the current set of features.)
Back then, I was planning to simply use switching to alternate between two or more sets of DUT connections, so the displayed curves could easily be compared, by eye. I thought I'd probably include both manual and automatic switching, possibly with a variable or selectable automatic switching rate. The switching method itself was my main area of investigation, at the time I stopped working on it to do the re-design.
Now, considering the future PC connectivity, I guess I'll at least need to also be able to provide information to the PC about which set of DUT connections are currently active (and maybe eventually also provide a means for the PC to select the DUT).
jackinnj said:
The 22 kHz repetition rate is the fastest my curve tracer currently provides. So I was using that speed just to try to get an estimate of how fast the sampling might have to be. But the curve tracer has six selectable sweep rates, all the way down to 60 Hz. And the actual sweep waveform shape is selectable as sawtooth, triangle, or sine. The sweep's p-p amplitude also has six selectable settings, from 1Vpp to 30Vpp. But I don't currently have a way to insert a delay between sweep cycles.
Perhaps, in the more-digital future, I should consider also having a "staircased" type of waveform available for the sweep (i.e. the DUT + to - voltage), so the DUT could settle at each level and then be sampled. At that point, I'd probably want to just go with all-digital, for the sweep generator at least. And I guess it wouldn't even really have to "sweep", if it was a "PC-required" type of curve tracer, since software could contruct the curves no matter how fast or in what order the data was gathered. I initially thought that it might be valuable to be able to see the device characteristics with different sweep frequencies and slew rates, as my curve tracer can, now. But I don't actually have any data about how useful that is.
- Tom Gootee
http://www.fullnet.com/~tomg/index.html
-
If you could develop a slow sawtooth, say 10Hz, and catch your zero V, this could establish a nice starting point. Then read the current with the ADC at precise time intervals, say 100 readings in the 100ms time frame from 0 to Peak. A PIC's timing is very accurate and 1ms per reading is a breeze. Likely you can get the job done with a 100-400Hz sawtooth.
Ideally, a single ramp is all you need, either step controlled as you were thinking before, or timing controlled as above.
Ideally, a single ramp is all you need, either step controlled as you were thinking before, or timing controlled as above.
So guys are the usb curve tracers acurrate?
Im planning to purchase an curve tracer but dont kow what to purchase.
I noticed that the madell ones are just 600US and are not too heavy, i have an problem if purchase old Tek because are too heavy and im from Argentina and is not possible to ship that kind of products via USPS.
Thanks in advance, for your help, i need to purchase an curve tracer asap but want to be acurrate.
Im planning to purchase an curve tracer but dont kow what to purchase.
I noticed that the madell ones are just 600US and are not too heavy, i have an problem if purchase old Tek because are too heavy and im from Argentina and is not possible to ship that kind of products via USPS.
Thanks in advance, for your help, i need to purchase an curve tracer asap but want to be acurrate.
Hi Chris,
Busy is good and we are now back on track with the CT100. We have tested a new prototype to over 100ma with USB power only. While we work out new software changes and inclusion of a new CPU with higher resolution ADC's, we are going to test the limits of the existing board components to see if we can also add an optional wall-wart power supply. We project that a few 100mA's more are possible with an external power supply. Stay tuned.
Busy is good and we are now back on track with the CT100. We have tested a new prototype to over 100ma with USB power only. While we work out new software changes and inclusion of a new CPU with higher resolution ADC's, we are going to test the limits of the existing board components to see if we can also add an optional wall-wart power supply. We project that a few 100mA's more are possible with an external power supply. Stay tuned.
While the CT100 USB is accurate, it's purpose is not to create datasheets. Rather, it is best utilized for comparison of dynamic charateristics of semiconductors. The advantage of a USB PC-based system is that data can be represented on the screen at various resolutions and also be analyzed relatively easily, depending on software.
Hi wds,
Right on! I am looking forward to the next version.
Now, a request for a feature. Will it be possible to include two sockets so a prospective pair of <whatever, BJT / FET> may be tested and the curves overlaid? Failing this, can an external test fixture be made to accomplish the same thing? I'm also thinking a DIY'er may want to include a jig with a heat sink for larger devices.
On another train of thought. How easy would it be to interface your product to an external, active jig that may supply higher currents and voltages? There is no end to where you can take this device as the heart of a testing system. Heck, my Heathkit IT-18 tests smt's these days. I'm sure it's designers didn't even foresee the package!
I guess that one would only really need an interface port and the specifications to come up with some nifty devices. Even down to a tube tester. Why not?
-Chris
Right on! I am looking forward to the next version.
Now, a request for a feature. Will it be possible to include two sockets so a prospective pair of <whatever, BJT / FET> may be tested and the curves overlaid? Failing this, can an external test fixture be made to accomplish the same thing? I'm also thinking a DIY'er may want to include a jig with a heat sink for larger devices.
On another train of thought. How easy would it be to interface your product to an external, active jig that may supply higher currents and voltages? There is no end to where you can take this device as the heart of a testing system. Heck, my Heathkit IT-18 tests smt's these days. I'm sure it's designers didn't even foresee the package!
I guess that one would only really need an interface port and the specifications to come up with some nifty devices. Even down to a tube tester. Why not?
Oh, I am!
-Chris
I have designed and complete a "PC intelligent semiconductor characteristic curve Tracers"
the link is below
http://www.diyaudio.com/forums/showthread.php?s=&threadid=85076
measure the 2SC3519 curve Vce->Ic (Ib from 1mA to 9mA)
The whole circuit in post #6
the link is below
http://www.diyaudio.com/forums/showthread.php?s=&threadid=85076
measure the 2SC3519 curve Vce->Ic (Ib from 1mA to 9mA)
An externally hosted image should be here but it was not working when we last tested it.
The whole circuit in post #6
Are we into computer data acquisition my friends ?
Wishing that it would be simpler to get data in/out of the PC ?
Well, same here. I was fed up with not finding anything suitable on the Net for my needs so I am designing this :
http://home.peufeu.com/nik/fpga/board_v02/Layout1.png
Basically you have a 100 Mbps ehernet port, 16 MB of fast 32-bit SDRAM (should get about 150-200 MB/s throughput), a Spartan-3E-500 FPGA which will contain a full CPU, peripherals, etc, clock, flash for storing the firmware and FPGA bitstream, and a Hirose connector with about 30 I/0 pins and 30 input-only pins. It is designed to sustain the full Ethernet bandwidth full-duplex without choking. I estimate PCB + parts cost at about $130 but obviously it has a few drawbacks : it doesn't exist yet, and I have to solder 3 big fat super fine pitch SMDs (FPGA has 208 pins) plus about 60 passives in 0805 package.
UDP over Ethernet don't need drivers.
Basically it's a Suzaku SZ-130 module DONE RIGHT.
Applications are anything where a PC wants to control some remote peripheral, like : streaming audio to DAC, digital scope, data acquisition, curve trace, etc. Having a FPGA is cool since you can use HDL to code prettu much any data interface.
For applications requiring much lower bandwidth, obviously a cheap PIC or Atmel with lwip or USB is more suitable.
Check my site http://audio.peufeu.com : I'll create a new section for this when I build it.
Wishing that it would be simpler to get data in/out of the PC ?
Well, same here. I was fed up with not finding anything suitable on the Net for my needs so I am designing this :
http://home.peufeu.com/nik/fpga/board_v02/Layout1.png
Basically you have a 100 Mbps ehernet port, 16 MB of fast 32-bit SDRAM (should get about 150-200 MB/s throughput), a Spartan-3E-500 FPGA which will contain a full CPU, peripherals, etc, clock, flash for storing the firmware and FPGA bitstream, and a Hirose connector with about 30 I/0 pins and 30 input-only pins. It is designed to sustain the full Ethernet bandwidth full-duplex without choking. I estimate PCB + parts cost at about $130 but obviously it has a few drawbacks : it doesn't exist yet, and I have to solder 3 big fat super fine pitch SMDs (FPGA has 208 pins) plus about 60 passives in 0805 package.
UDP over Ethernet don't need drivers.
Basically it's a Suzaku SZ-130 module DONE RIGHT.
Applications are anything where a PC wants to control some remote peripheral, like : streaming audio to DAC, digital scope, data acquisition, curve trace, etc. Having a FPGA is cool since you can use HDL to code prettu much any data interface.
For applications requiring much lower bandwidth, obviously a cheap PIC or Atmel with lwip or USB is more suitable.
Check my site http://audio.peufeu.com : I'll create a new section for this when I build it.
On all of those points I apologize. Max I is 22ma for Sweep voltages of +/-15V. We have a small hack with 2 resistors that increases the CT100 current range to 44ma.
We are a small company with great ideas, good engineering, big plans, and it's slow progress for us right now.
The CT200 is still in development. We are applying some innovative techniques to increase voltage and current capabilities, and test flexibility, using USB power only. Also, the new board will feature the ability to use an external power supply to increase the range of V and I tests.
In the meantime, the CT100 is still available for sale. For $119.00 it's a great deal to test and match smaller components.
We are a small company with great ideas, good engineering, big plans, and it's slow progress for us right now.
The CT200 is still in development. We are applying some innovative techniques to increase voltage and current capabilities, and test flexibility, using USB power only. Also, the new board will feature the ability to use an external power supply to increase the range of V and I tests.
In the meantime, the CT100 is still available for sale. For $119.00 it's a great deal to test and match smaller components.
Hi wds,
If you want a beta tester, I'd be happy to help out. I am completely serious. Same for your USB scope thingy.
I'd like to see you guys do better. You have mail from me also.
Members, have a look at wds' web site. They have lots of PIC programming stuff and LCD displays at good prices on top of what has been talked about here.
-Chris
If you want a beta tester, I'd be happy to help out. I am completely serious. Same for your USB scope thingy.
I'd like to see you guys do better. You have mail from me also.
Members, have a look at wds' web site. They have lots of PIC programming stuff and LCD displays at good prices on top of what has been talked about here.
-Chris
Thanks Chris, I'll keep it in mind.
The USB Scope is made in Europe and we only distribute it. It's very slow as scopes go and not good for watching signals at several hundred KHz or more. The reason we have it is it has some great serial protocol analysis features that are useful to microcontroller developers.
The USB Scope is made in Europe and we only distribute it. It's very slow as scopes go and not good for watching signals at several hundred KHz or more. The reason we have it is it has some great serial protocol analysis features that are useful to microcontroller developers.
Hi wds,
That's what I was thinking when I saw them. I have just bought a pile of PIC chips and a programmer. Along with some books, I'm going to try and learn this stuff. It's a long way from the FORTRAN I learned in university.
Your 'scope might be excellent for audio work. Is it a 12 bit? The software looks much better than what I have, FFT especially.
-Chris
That's what I was thinking when I saw them. I have just bought a pile of PIC chips and a programmer. Along with some books, I'm going to try and learn this stuff. It's a long way from the FORTRAN I learned in university.
Your 'scope might be excellent for audio work. Is it a 12 bit? The software looks much better than what I have, FFT especially.
-Chris
By the way... you mentioned your interest in PIC microcontrollers. I have posted some interesting videos on YouTube that are demos with LCD's and LED Matrices connected directly to a PIC. Some of it is very unique:
http://www.youtube.com/profile_videos?user=circuited
http://www.youtube.com/profile_videos?user=circuited
Hi,
That would be great.
I know of a product / company that needs serious help with their 'scope software. Far east so I guess that is a lost cause.
-Chris
Edit: Hey, thank you. I'll check the videos out.
Edit (again!!): Some videos didn't load, some of the "Ultimate 64's". The touch screen stuff is really cool, and the odometer is really well done. The digital dash just got more analog. I like it.
That would be great.
I know of a product / company that needs serious help with their 'scope software. Far east so I guess that is a lost cause.
-Chris
Edit: Hey, thank you. I'll check the videos out.
Edit (again!!): Some videos didn't load, some of the "Ultimate 64's". The touch screen stuff is really cool, and the odometer is really well done. The digital dash just got more analog. I like it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.