You're a wise man. I've gone down those roads myself in other software environments where I create software tools for myself, people find out about them, want copies for themselves, and you can tell them until you're blue in the mouth that you're not going to support the software, but the expectation is always there with one or more people, and then the people will get angry, demanding.. Some will insist that you release the source code as open source, but no one has the right to demand that of anyone else.
Interesting history. I remember our department being billed for time usage on the 11-780. What a revolution we've seen. Working through labs in "Electronic Devices and Circuit Theory" which incorporates spice activities. Installed free Pspice from TI. So many tools are free or very affordable for home use these days.
I think we had one 286 6MHz IBM AT when we bought PSPICE. It was built as a packed executable but you had to rebuild it to add parts to the library using MS Fortran. I rebuilt it unpacked, then used Debug to examine how their protection worked (required the floppy to be installed).
These are really good videos of the how a VNA works and what to do with it. I think writing the software clarified and made it easier to explain even more? Also, this video was suggested under "consumer electronics" to me, but have subscribed a long time ago.
Great idea to import a real component to simulation model. Also shiftting the 100 steps thru the span is clever. Instead of supply source code, could it be supplied a flowchart of the program on paper or starUML ?
Remember the original Touchstone offered by EESoF? Ya ... the one that ran under DOS and used a parallel port dongle ... those dongles could be imitated by some discrete TTL logic and the proper response 'codes' burned into a 2716 EPROM. I also used EESoF's ESYN to create some magnificent filters, like the 160 Meter Broadcast-Band Brickwall Filter ...
I had to look it up. I've ran into my fair share of those hardware keys. I have some software that I still use from time to time that supports one but my PC doesn't have a port. Took some time to get the company to supply me with a non-dongle version of their software.
As the statute of limitations has probably run out, I can say that a couple of us reverse-engineered the data the dongles exchanged with the PC ... don't know what they used for hardware inside those dongles, as we simply mimicked what the responses were that we saw with an (now old) HP 1600 logic analyzer. We were cracker-jack digital designers back then too!
@@uploadJ I have resorted to reverse engineering a few of these as well. There is a video made for a CPU that I designed for an FPGA and I talk about the one hardware key I placed inside the FPGA as part of the design. Basically, using the manufactures tools to design the hardware key for their tools... I owned the tools but am not a fan of those printer port keys. The earliest ones I had looked at used a counter with some basic logic was all.
5 ปีที่แล้ว
I haven’t used spice since uni, but I’m planning on doing some small project soonish. Are there some issues with Ngspice for using microcap instead of it?
Nice video, great job on the youtube. However i'm confused, is crystal really a very nonlinear device? Feeding single tone in to non linear device, would create harmonics on the output? VNA wouldn't work correctly for such component? VNA does extract amplitude and phase only from present frequency through the sweep at the given moment. VNA can't look at other parts of the spectrum at the same time? For example diode is very non linear device. Every component, which produce harmonics and does mixing is not linear. Component which does single tone on the output, from single tone on the input, have to be linear? bdw, i'm not an expert in electronics or RF.. i think through analogy from control systems..
Crystal is linear, but, has unusual properties, like, really high Q factor ... Crystal filters are used in IF strips, and as "roofing" (the first filter) in HF comm rigs/radios.
Hello Bartosz, I hope this helps (and is correct), in most electronic conversations non linear means with respect to signal amplitude. So for a very limited case a 1K ohm 1/4 watt resistor with one Volt across it has 1 mA of current. Increase the voltage to 2 volts and the current will double. The circuit is linear. Put a 1uF capacitor in series and use a 1 Volt AC source at 1Hz and measure the voltage across the capacitor and you will measure very very close to 1 volt AC. Double the voltage source and you will measure double. Increase the frequency to 1MHz and with a 1 Volt source you will now measure noticeably less than 1 Volt AC across the capacitor. Double the voltage and you will measure noticeably less than 2 Volts AC by the same ratio. The circuit responds linearly with respect to the amplitude. We are not talking about response as a function of frequency. In summary the adjective linear referees to the response to different size signals and also has built into it the assumption that you do not for instance feed an amplifier so much signal that it would have to put put more voltage than it's power supply voltage allows. The electrical model for a crystal is mostly a capacitor in parallel with an inductor. There is some very very small resistance which is why a crystal is high Q (High quality factor meaning that the amount of power lost per cycle of oscillation is much smaller than the peak power stored in the reactive elements. The R the L and the C are all linear. Hope this helps.
An additional note about "Time invariant". There is often an unspoken assumption in conversations about circuits that the component values do not change with time. For all these assumptions there are exceptions and so it is best to remember conversations are only approximations. Let's make up a case. The 1/4 watt 1 ohm resistor I postulated above. At one volt it dissipates V^2/R = 1*1/1000 or 1 mWatt. Well below it's rated power. It will warm up only a little and has only a small temperature coefficient to the resistance. So you can pretty much ignore the time variance. But put 32 volts on it and you have 32*32/1000 = 1 watt. The part is now going to get very hot and may change resistance some what. It may even change resistance permanently when allowed to cool. Put 100 volts on it and you have 10000/1000 =10 watt and the part will rapidly heat to glow red and then probably fail open. It is certainly not time invariant now. In this overload application the resistor became non linear and time variant. Most (All?) parts will become nonlinear and time variant if overloaded but that is not built into the assumption during conversations about normal circuit operation. Again I hope this helps.
One of the better choices I made was learning it. I've posted about it in the past: www.eevblog.com/forum/chat/community-version-of-labview-coming/ www.eevblog.com/forum/testgear/national-instruments-labview-home-bundle/
joe smith thats B’s. After viewing the first 2 vids you tell the software wont be available. That makes no sense to me at all. First making 2 vids where everybody thinks hey great software , then tell everyone no i keep it for myself.
Mr. Smith, I admire your exceptional knowledge.
Merry Christmas!
You're a wise man. I've gone down those roads myself in other software environments where I create software tools for myself, people find out about them, want copies for themselves, and you can tell them until you're blue in the mouth that you're not going to support the software, but the expectation is always there with one or more people, and then the people will get angry, demanding.. Some will insist that you release the source code as open source, but no one has the right to demand that of anyone else.
github.com/joeqsmith/NanoVNA_V2Plus
Interesting history. I remember our department being billed for time usage on the 11-780. What a revolution we've seen.
Working through labs in "Electronic Devices and Circuit Theory" which incorporates spice activities. Installed free Pspice from TI. So many tools are free or very affordable for home use these days.
Jeff, does the TI simulator support Touchstone files? If not that Micro-Cap simulator is now available for free.
I've used PSpice for DOS too in Pisa faculty of electronic engineer in the '90s on my 80486DX.
I think we had one 286 6MHz IBM AT when we bought PSPICE. It was built as a packed executable but you had to rebuild it to add parts to the library using MS Fortran. I rebuilt it unpacked, then used Debug to examine how their protection worked (required the floppy to be installed).
Fantastic video Mr. Smith! Thank you and a very happy New Year to you and your loved ones.
Thanks for these videos on the NanoVNA. Very useful.
These are really good videos of the how a VNA works and what to do with it. I think writing the software clarified and made it easier to explain even more? Also, this video was suggested under "consumer electronics" to me, but have subscribed a long time ago.
Yes, working the math has been a good refresher for me.
Great idea to import a real component to simulation model. Also shiftting the 100 steps thru the span is clever. Instead of supply source code, could it be supplied a flowchart of the program on paper or starUML ?
Remember the original Touchstone offered by EESoF?
Ya ... the one that ran under DOS and used a parallel port dongle ... those dongles could be imitated by some discrete TTL logic and the proper response 'codes' burned into a 2716 EPROM. I also used EESoF's ESYN to create some magnificent filters, like the 160 Meter Broadcast-Band Brickwall Filter ...
I had to look it up. I've ran into my fair share of those hardware keys. I have some software that I still use from time to time that supports one but my PC doesn't have a port. Took some time to get the company to supply me with a non-dongle version of their software.
As the statute of limitations has probably run out, I can say that a couple of us reverse-engineered the data the dongles exchanged with the PC ... don't know what they used for hardware inside those dongles, as we simply mimicked what the responses were that we saw with an (now old) HP 1600 logic analyzer. We were cracker-jack digital designers back then too!
@@uploadJ I have resorted to reverse engineering a few of these as well. There is a video made for a CPU that I designed for an FPGA and I talk about the one hardware key I placed inside the FPGA as part of the design. Basically, using the manufactures tools to design the hardware key for their tools... I owned the tools but am not a fan of those printer port keys. The earliest ones I had looked at used a counter with some basic logic was all.
I haven’t used spice since uni, but I’m planning on doing some small project soonish. Are there some issues with Ngspice for using microcap instead of it?
Thanks Joe for information. very good
Nice video, great job on the youtube.
However i'm confused, is crystal really a very nonlinear device?
Feeding single tone in to non linear device, would create harmonics on the output?
VNA wouldn't work correctly for such component?
VNA does extract amplitude and phase only from present frequency through the sweep at the given moment. VNA can't look at other parts of the spectrum at the same time?
For example diode is very non linear device.
Every component, which produce harmonics and does mixing is not linear.
Component which does single tone on the output, from single tone on the input, have to be linear?
bdw, i'm not an expert in electronics or RF.. i think through analogy from control systems..
Crystal is linear, but, has unusual properties, like, really high Q factor ...
Crystal filters are used in IF strips, and as "roofing" (the first filter) in HF comm rigs/radios.
th-cam.com/video/CK0zl34nZJc/w-d-xo.html
Hello Bartosz,
I hope this helps (and is correct), in most electronic conversations non linear means with respect to signal amplitude. So for a very limited case a 1K ohm 1/4 watt resistor with one Volt across it has 1 mA of current. Increase the voltage to 2 volts and the current will double.
The circuit is linear. Put a 1uF capacitor in series and use a 1 Volt AC source at 1Hz and measure the voltage across the capacitor and you will measure very very close to 1 volt AC. Double the voltage source and you will measure double.
Increase the frequency to 1MHz and with a 1 Volt source you will now measure noticeably less than 1 Volt AC across the capacitor. Double the voltage and you will measure noticeably less than 2 Volts AC by the same ratio. The circuit responds linearly with respect to the amplitude. We are not talking about response as a function of frequency.
In summary the adjective linear referees to the response to different size signals and also has built into it the assumption that you do not for instance feed an amplifier so much signal that it would have to put put more voltage than it's power supply voltage allows.
The electrical model for a crystal is mostly a capacitor in parallel with an inductor. There is some very very small resistance which is why a crystal is high Q (High quality factor meaning that the amount of power lost per cycle of oscillation is much smaller than the peak power stored in the reactive elements. The R the L and the C are all linear.
Hope this helps.
An additional note about "Time invariant". There is often an unspoken assumption in conversations about circuits that the component values do not change with time. For all these assumptions there are exceptions and so it is best to remember conversations are only approximations.
Let's make up a case.
The 1/4 watt 1 ohm resistor I postulated above. At one volt it dissipates V^2/R = 1*1/1000 or 1 mWatt. Well below it's rated power. It will warm up only a little and has only a small temperature coefficient to the resistance. So you can pretty much ignore the time variance.
But put 32 volts on it and you have 32*32/1000 = 1 watt. The part is now going to get very hot and may change resistance some what. It may even change resistance permanently when allowed to cool.
Put 100 volts on it and you have 10000/1000 =10 watt and the part will rapidly heat to glow red and then probably fail open. It is certainly not time invariant now. In this overload application the resistor became non linear and time variant. Most (All?) parts will become nonlinear and time variant if overloaded but that is not built into the assumption during conversations about normal circuit operation.
Again I hope this helps.
Joe was wrong. Indeed a crystal is linear up to the point you exceed the power rating.
Great stuff! Thank you!
I think QUCS supported touchstone, so that could always available to hobbyists. Unfortunately I never did get it working though.
I looked at it at the time I was looking at Micro-Cap. I didn't spend enough time with it to try and model anything with Touchstone.
Wife works quite well with touchstone files. I've modeled quite a few projects with it so far and am a happy user
This is good stuff. Thank you!
LabView oh no.
One of the better choices I made was learning it. I've posted about it in the past: www.eevblog.com/forum/chat/community-version-of-labview-coming/
www.eevblog.com/forum/testgear/national-instruments-labview-home-bundle/
Why show off and make videos demontstrating software that is not available ? That makes no sense at all
Is your thumb raw from sucking on it? Don't like it, don't watch.
joe smith thats B’s. After viewing the first 2 vids you tell the software wont be available. That makes no sense to me at all. First making 2 vids where everybody thinks hey great software , then tell everyone no i keep it for myself.
@@LanWolf Guessing it's bleeding by now.
@@LanWolf This is a way to show that Smith is the dominant male.
thank you for those videos!
Awesome info .