[QUOTE=Chryseus;44265807]I dunno about Eagle but you usually need to forward annotate the schematic changes to the PCB in order for it to properly deal with new parts or changes to the netlist.
Your layout is also too cramped, there is a lot of wasted space at the bottom and top areas, there is no need to line your components up perfectly.[/QUOTE]
Eagle does this automatically if you have both files opened. It will show a big message at the top of the window if it's not annotating.
[QUOTE=Chryseus;44265807]I dunno about Eagle but you usually need to forward annotate the schematic changes to the PCB in order for it to properly deal with new parts or changes to the netlist.
Your layout is also too cramped, there is a lot of wasted space at the bottom and top areas, there is no need to line your components up perfectly.[/QUOTE]
Think so? I was squishing everything together so I could fit it on a 5x5cm board. Right now it's like 47x40mm widthxheight so I only have a couple mm to spare.
Also I realized I was putting the inductor in the wrong place. So stupid...I blame staying up all night working on it.
[editline]17th March 2014[/editline]
Okay that's done other than maybe some silkscreen stuff...now please everyone tell me how much it sucks, it's my first one. Final dimensions are 50x40mm WxH
Top
[img]http://i.imgur.com/dcNm42y.png[/img]
Bottom
[img]http://i.imgur.com/XC7teIe.png[/img]
Did you actually measure the caps and stuff ?
Experience has taught me to never trust any footprint other than one I've made myself.
[QUOTE=Chryseus;44266576]Did you actually measure the caps and stuff ?
Experience has taught me to never trust any footprint other than one I've made myself.[/QUOTE]
Yeah I checked everything. They're tantalum caps that look like match heads, not rectangular things like my cap symbols may imply.
[QUOTE=Chryseus;44265251][img]http://u.cubeupload.com/Chryseus/MOqJJw.jpg[/img]
It's time gentlemen, to venture in to the weird and wonderful world of switching power supplies.
I can smell the magic smoke already.[/QUOTE]
I may just join you in that!
[URL=http://i.imgur.com/oUVDtFE.jpg][IMG]http://i.imgur.com/oUVDtFEl.jpg[/IMG][/URL]
Ah man, I too would like to venture to the world of smps, however I feel like that I would die quick.
[QUOTE=alexaz;44267939]Ah man, I too would like to venture to the world of smps, however I feel like that I would die quick.[/QUOTE]
It's not that complicated, it just involves a lot of digging for information, experimentation and careful simulation.
Single transistor flyback converters are pretty much the easiest isolated topology to start with or a non-isolated buck / boost converter, you will of course need a scope at the minimum.
[QUOTE=Chryseus;44268073]It's not that complicated, it just involves a lot of digging for information, experimentation and careful simulation.
Single transistor flyback converters are pretty much the easiest isolated topology to start with or a non-isolated buck / boost converter, you will of course need a scope at the minimum.[/QUOTE]
The biggest thing is equations really...
I have a big pile of app nots that are really useful, anyone wany a zip file of them?
[editline]17th March 2014[/editline]
Tough shit, your getting it anyway.
[url]https://mega.co.nz/#!VAACSTyJ!gCza6bG5skoBFC2_BiwFFU1vn0GxNYFCHvNhsVAOC0A[/url]
Woohooo! Managed to get the DCU (digital control unit) of this HP 8660 synthesizer to work!
[media]http://www.youtube.com/watch?v=-h_D6uDbOJM[/media]
(Yes that hum is the transformer in the power supply of this thing, its noisy).
Now onto the fun part, the RF plugin...
[QUOTE=ddrl46;44269372]Woohooo! Managed to get the DCU (digital control unit) of this HP 8660 synthesizer to work!
[media]http://www.youtube.com/watch?v=-h_D6uDbOJM[/media]
(Yes that hum is the transformer in the power supply of this thing, its noisy).
Now onto the fun part, the RF plugin...[/QUOTE]
I think thats upside down... best turn it the right way up or all the electrons will fall out!
I have no idea what I am doing with making an antenna for this am transmitter. At one point it was transmitting ~50m with a random wire attached to a decoupling capacitor but now that it is soldered it only goes about a meter.
Heres the final stage of the radio. I don't have a tank resonator but that shouldn't cause it's range to be that small should it?
[IMG]http://i.imgur.com/lzd2BKE.png[/IMG]
Actually what you have there is more or less a wide band RF power amp, which is rather bad ju-ju as you can amplify stray signals and really clutter up the AM band (Not to mention harmonics).
A tank resonator is not explicitly necessary, but a pi filter with an inductor load on the collector will help control what exactly gets amplified as well as antenna matching. In this case, you need a few more caps to filter as you're in the AM band (which would make your coils rather large).
[URL="http://www.sentex.ca/~mec1995/circ/broadcast.html"]This[/URL] may be of use, some of the other guys here are more seasoned with RF than I am, so I wish you luck. (I recommend using 2N2222A/2N2907A instead of 3904/3906s, respectively)
Also it's important to keep in mind the current gain of a transistor drops with increasing frequency, as a result the input impedance also drops which results in attenuation of the input signal, it's also a good idea to feed the 12V supply through a radio frequency choke (large inductor) to prevent the signal leaking in to the 12V supply, the common emitter configuration also suffers from the early effect and miller capacitance which can degrade performance, this can be reduced by using a common base or cascode configuration.
I spend like 1.5 hours playing with a shift register and an arduino, and only now I notice there's a function to write values in binary, exactly as I needed :v:
[QUOTE=Chryseus;44277451]Also it's important to keep in mind the current gain of a transistor drops with increasing frequency, as a result the input impedance also drops which results in attenuation of the input signal, it's also a good idea to feed the 12V supply through a radio frequency choke (large inductor) to prevent the signal leaking in to the 12V supply, the common emitter configuration also suffers from the early effect and miller capacitance which can degrade performance, this can be reduced by using a common base or cascode configuration.[/QUOTE]
I'll try adding a choke - looking on an oscilloscope I'm using in my teacher's shop there is ringing in my VCC line. I had thought that was just from having a poorly regulated power supply. I've looked at the transistor's datasheet and at 800kHz the gain has barely dropped off; I'll keep the other things in midn though, thank you.
[QUOTE=LoneWolf_Recon;44272216]Actually what you have there is more or less a wide band RF power amp, which is rather bad ju-ju as you can amplify stray signals and really clutter up the AM band (Not to mention harmonics).
A tank resonator is not explicitly necessary, but a pi filter with an inductor load on the collector will help control what exactly gets amplified as well as antenna matching. In this case, you need a few more caps to filter as you're in the AM band (which would make your coils rather large).
[URL="http://www.sentex.ca/~mec1995/circ/broadcast.html"]This[/URL] may be of use, some of the other guys here are more seasoned with RF than I am, so I wish you luck. (I recommend using 2N2222A/2N2907A instead of 3904/3906s, respectively)[/QUOTE]
That was helpful, thank you. And you're right about the harmonics, I low pass the signal earlier int he circuit but the harmonics are definitely there. I don't have a spectrum analyzer but I can hear the signal up to the 3rd harmonic on my handheld radio so I've been trying not to keep it on very long. And I'll try ordering some 2N2222A's.
I have a question about capacitors. Often times I see a very small value capacitor, like say .01uF in parallel with a large filter cap. In theory, shouldn't the large filter cap be adequate for eliminating all stray AC signals?
[QUOTE=No_Excuses;44284474]I have a question about capacitors. Often times I see a very small value capacitor, like say .01uF in parallel with a large filter cap. In theory, shouldn't the large filter cap be adequate for eliminating all stray AC signals?[/QUOTE]
As it happens. no, this is because all capacitors have reactance. each AC signal will be affected differently by the capacitors.
Xc = 1/(2*pi*F*C)
Xc is similer to resistance, but for AC and it only means anything to the Frequency it reacts to.
It's largely due to impedance as nuttyboffin said.
Aluminium and tantalum caps have a much higher ESR (equivalent series resistance) compared with other low value types such as ceramic, film, etc.
Since impedance is Z = sqrt( X^2 + R^2 ) you generally get better high frequency rejection with a low ESR cap.
[QUOTE=Chryseus;44287644]It's largely due to impedance as nuttyboffin said.
Aluminium and tantalum caps have a much higher ESR (equivalent series resistance) compared with other low value types such as ceramic, film, etc.
Since impedance is Z = sqrt( X^2 + R^2 ) you generally get better high frequency rejection with a low ESR cap.[/QUOTE]
Uhm, Z = R + j*X
[QUOTE=DrDevil;44288441]Uhm, Z = R + j*X[/QUOTE]
My math is not so good but from what I understand based on the Pythagorean theorem is that sqrt(R^2 + X^2) gives the absolute magnitude including phase where as R+jX does not factor in the phase angle between R and X.
[img]http://hydrogen.physik.uni-wuppertal.de/hyperphysics/hyperphysics/hbase/electric/imgele/cimp10.gif[/img]
[b]Edit[/b]
[t]http://u.cubeupload.com/Chryseus/P6raYA.jpg[/t]
[t]http://u.cubeupload.com/Chryseus/BInBWx.png[/t]
Forward converter is alive!
Oh my goodness golly gee, I found a 7 seg display in an old computer case for the "TURBO HURRR" button and the back of it looks like this
[t]https://dl.dropboxusercontent.com/u/23344247/20140226_205336.jpg[/t]
I've never seen this before! Kinda neat!
[QUOTE=Chryseus;44288884]My math is not so good but from what I understand based on the Pythagorean theorem is that sqrt(R^2 + X^2) gives the absolute magnitude including phase where as R+jX does not factor in the phase angle between R and X.
[img]http://hydrogen.physik.uni-wuppertal.de/hyperphysics/hyperphysics/hbase/electric/imgele/cimp10.gif[/img]
[/QUOTE]
What you calculated is the absolute of Z, also known as |Z|
[QUOTE=DrDevil;44292943]What you calculated is the absolute of Z, also known as |Z|[/QUOTE]
Isn't that what we want in most cases?
[QUOTE=Chryseus;44293031]Isn't that what we want in most cases?[/QUOTE]
No. You usually want Z = R + j*X, as that contains all the phase stuff.
[QUOTE=nuttyboffin;44285616]As it happens. no, this is because all capacitors have reactance. each AC signal will be affected differently by the capacitors.
Xc = 1/(2*pi*F*C)
Xc is similer to resistance, but for AC and it only means anything to the Frequency it reacts to.[/QUOTE]
Yeah but a large C causes a low Xc. Using a small cap would just cause a large Xc and therefore not be as effective.
[QUOTE=Chryseus;44287644]It's largely due to impedance as nuttyboffin said.
Aluminium and tantalum caps have a much higher ESR (equivalent series resistance) compared with other low value types such as ceramic, film, etc.
Since impedance is Z = sqrt( X^2 + R^2 ) you generally get better high frequency rejection with a low ESR cap.[/QUOTE]
So it's just the resistive component as I suspected. Thanks.
[QUOTE=DrDevil;44293067]No. You usually want Z = R + j*X, as that contains all the phase stuff.[/QUOTE]
Just include the phase angle and it's fine. It can be represented as polar or rectangular coordinates.
[QUOTE=No_Excuses;44294380]Yeah but a large C causes a low Xc. Using a small cap would just cause a large Xc and therefore not be as effective.
So it's just the resistive component as I suspected. Thanks.
Just include the phase angle and it's fine. It can be represented as polar or rectangular coordinates.[/QUOTE]
Of course, you can also write it as |Z| * e^(j*phi), but what's the point? You're needlessly complicating things.
Z = R + j*X is the way to go, as you can easily represent inductivity and capacity with it in the form of j*omega*L and 1/j*omega*C.
(I wish facepunch had LaTeX support)
[QUOTE=DrDevil;44295898]Of course, you can also write it as |Z| * e^(j*phi), but what's the point? You're needlessly complicating things.
Z = R + j*X is the way to go, as you can easily represent inductivity and capacity with it in the form of j*omega*L and 1/j*omega*C.
(I wish facepunch had LaTeX support)[/QUOTE]
You can just write it as Z = x, phi = y degrees.
It's not complicating things either. If you have to do hand calculations, it's easier to do with polar coordinates when it comes to division and multiplication. Rectangular are easier for addition and subtraction.
[QUOTE=No_Excuses;44296457]You can just write it as Z = x, phi = y degrees.
It's not complicating things either. If you have to do hand calculations, it's easier to do with polar coordinates when it comes to division and multiplication. Rectangular are easier for addition and subtraction.[/QUOTE]
You can also write it as Z = 3 potatoes, pie = pie degrees.
The point is that with Z = R + j*X you can easily do all the maths in one step in your calculator instead of seperating it all out into adding/multiplying the absolute and then adding/multiplying the angles. Also, with this notation you can easily enter your elements into a Smith Chart, whereas with discrete absolutes and angles you have to convert it to a complex number anyway.
I guess you've never done calculations with polar coordinates then? You enter the magnitude and angle at the same time in the calculator. It too, is done in one step.
I just think either way has pros and cons.
[img]http://i.imgur.com/yS2uSZl.png[/img]
Well, shit... Didn't think 20QFN is that small.
Sorry, you need to Log In to post a reply to this thread.
