Help me understand LC Lowpass filter theory bls

>>1187966
1. Those kinds of filters are very rarely calculated the way you seem to assume. The normal way is to either use a filter design program or scale pre-calculated prototype filters suitably. Pre-calculation is practical, because those filters are chosen from a (small) bunch of well-known topologies, like Butterworth or Chebyshev.
https://en.wikipedia.org/wiki/Chebyshev_filter
2. It is there to match the fet's output impedance to the filter (and load).

The center-tap is a way to flip the polarity of that stage's output.

File: hitler just reich.jpg (26KB, 500x514px) Image search: [Google]

26KB, 500x514px

>>1187986
>1. Those kinds of filters are very rarely calculated the way you seem to assume. The normal way is to either use a filter design program or scale pre-calculated prototype filters suitably. Pre-calculation is practical, because those filters are chosen from a (small) bunch of well-known topologies, like Butterworth or Chebyshev.

FOUT which comes out between L4 and L5 is fed into a peak detector which is buffered by an op amp and fed to the adjustable buck converter, the output of the filter is connected to the Metcal soldering iron, during heat-up it absorbs all of the RF power, but once it reaches the curie temperature set by the tip it begins reflecting the RF back, which, according to the document, leads to an increase in voltage between L4 and L5, when this happens the buck converter decreases the voltage to keep the voltage between L4 and L5 relatively constant (when energy is reflected, voltage increases, buck converter decreases input voltage, voltage decreases, and thus is kept constant)

Observing the component values I can see that the capacitors on each side of L5 are equal and that L5 is 400n instead of 1u like the rest of the inductors, and since the tap comes out from it, this leads me to believe that it's special, but I don't understand the basis or theory behind why it is this way

No filter design program gave me 400nH for the last indcutor.

>It is there to match the fet's output impedance to the filter (and load).

How is it calculated?

>>1188004

for what purpose

Sorry for not including them from the start, but this is the full schematic: https://my.mixtape.moe/gzogbm.pdf

and this is its description

https://my.mixtape.moe/pecbeq.pdf

>for what purpose

Who knows. I suspect much of the Metcal design is unnecessary and amounts to obfuscation.

>>1188239
>obfuscation

I don't think so, the design was not intended to be public.

Here's the original Mx-500 (from the 90s I think)

http://scopetechniques.com/Metcal/MX-500P-11.pdf

>>1188261
The problem with electronics is you just open them up and see what's inside.

>>1188264
Everything I've seen does indeed suggest that the publicly available schematic is unnecessarily complicated, but the one I linked in >>1188231 was made (simplified pretty much) by a hobbyist, so nothing there is obfuscation, but he did indeed copy pretty much the entire power output section from the original

>>1187966
This looks like it was derp engineered by tacking shit together on some copper clad until it worked, which honestly is not uncommon in RF stuff, especially something like a RF iron that's not extremely precise like a radio set. The 47nF caps are pass caps, which remove any DC component from the signal. The 22pF and 56pF are a high-ish frequency path to ground to get rid of spikes and harmonics. The inductors remove lower frequency noise. Add it all together and you get something that probably works good enough for the application.

My advice is to get LTSpice and simulate this filter network out with various inputs. This definitely strikes me as fairly imprecise, back of the napkin engineering, however. I've been in the electronics field for 10 years and it's amazing how you can fabri-cobble shit together with minimal math and calculation that just fucking works.

>>1188279
>fabri-cobble shit together
Is that the EE equivalent for "spaghetti code" we have in programming?

>>1188299
Kind of. People tend to think that electronics design is a super precise job with tons of math and crap that goes into every decision. That's true in certain instances, but in a lot of cases with knowhow and experience, you can throw components at a problem and generally get a decent result. 13 MHz is quite slow and easy to deal with, in my book things really only get challenging in the microwave spectrum.

Remember that this is a soldering iron meant for production/rework on a shop floor, not some ultra precise radio amplifier. This to me looks like a combination of EE knowhow and bench top prototyping/testing, not crazy math.

>>1188231
The FOUT feedback loop is quite clever. I always wondered how these things managed to regulate temperature.

File: graph.png (38KB, 1927x838px) Image search: [Google]

38KB, 1927x838px

>>1188313
>I always wondered how these things managed to regulate temperature

They regulate temperature via induction heating. There's a coating of a material with a curie temperature exactly the temperature the soldering tip is rated for, during heat up it absorbs pretty much all of the RF energy, but once it reaches its curie point is practically disappears magnetically and current is instead induced into the copper core (which acts as a shorted turn) which radiates it back into the coil and back into the supply, which makes the voltage rise. At least that's how I understand it.

Reflected power is apparently caused by a load mismatch, which I'm not sure I understand completely, so I'm mostly sticking to the explanation above. (I'm not really a radio guy, feel free to educate me)

The feedback loop is only there to stop the voltage at the MOSFET drain from getting too high from reflected current, on the EEVBlog thread people had 100+ peaks at 30 volts input, the mosfet is only rated for 100 volts at its drain. (At least that's how I understand it)

>>1188279
>pic related
first time I use this software, went with pretty basic settings. the output is from the point where L3 and C3 meet.

If the graph is correct and I didn't fuck shit up, it looks like a fairly crude filter (it doesn't have that smooth drop I've seen low pass filters have) but what I'm interested in is why the point between L2 and L3 (in the screenie) was used to regulate the input power, I'm pretty sure the last filter stage being symmetric has something to do with it.

File: graph 2.png (25KB, 1896x450px) Image search: [Google]

25KB, 1896x450px

>>1188334
Here's a graph showing both the point between L2 and L3 (green) and the output (blue)

Looks like it just increases the amplitude at the frequency range we're interested in and decreases it near the end.

AC amplitude at the voltage source is 5V btw.

What does the metcal tip look like electrically?

>>1188510
From the scope techniques pdf linked above...

>The MX-500P Power Unit provides RF energy at 13.560MHz to the Soldering Tip Cartridge, which contains an induction
heater consisting of an 18 turn AWG33 wire coil wound around a 0.11" diameter by 0.5" long slug. The slug is composed of
a copper core, clad in a thin magnetic alloy having a curie point equal to the desired soldering tip temperature. The magnetic
alloy absorbs RF energy from the coil, causing the slug to heat up until the curie temperature is reached. At this point
absorption stops and heating ceases, because the RF energy is now reflected back to the power unit by the copper core.

Its impedance is around 50 ohms at 13.56 MHz.

From http://www.randomfunprojects.co.uk/metcal.html

>There's a gradual sweep from cold to warm that passes through a near perfect 50 Ohms, then there's a quick transition at the Curie temperature to a low impedance with a lot of inductance. Notice that even at this point return loss is still better than -3dB - more than half the power is absorbed, this is the reason for the current source drive - constant voltage drive from 50 Ohms would only give a 50% reduction in power when hot, not enough.

>>1188510
check patents 4,745,264 and 4,877,944

>>1188510
Also, the tip holder (handle) has a capacitor connected in series with the handle, I assume to provide matching in an LC tank circuit

File: 1495122079276.png (42KB, 500x322px) Image search: [Google]

42KB, 500x322px

>>1188334
>20 point AC analysis
how slow is your computer?

>>1188684
Intel 4790. The tutorial video I saw didn't tell me to change that.

>>1188334
up your number of steps significantly, you'll get a better plot. Also I think there's a way to plot in dB which would be my preference because it gives you a quick way to ID the power at each frequency, although voltage is also fine for understanding the filter.