Hi all, A (mostly clueless) physics type wants to do some RF stuff. (measuring nmr signals.. protons, ~5-25 MHz range) He's got something working, but construction is ugly. I was thinking of recommending an ARRL handbook, is any version better than others?
I learnt a lot from the 1970 ARRL handbook - can't speak on later issues.
Probably of more use for your colleague might be one of the books by Wes Hayward, like "Experimental Methods for RF Design" (quoting title from memory so could be slightly wrong).
Although not RF specific AoE is, of course, also obligatory reading :)
The _1986 Handbook_ is my oldest and the _2018 Handbook_ is my newest. The _2019 Handbook_'s apparently too wide for a single book, so ARRL made it a boxed set of six books. To answer George's question, newer's better. All four of my handbooks, from 1986 to 2018, contain a chapter on construction techniques. Among other things, the 2018 book replaces two pages of photosensitive copper-clad board paraphernalia content with a single paragraph that advocates on-line PCB board fabrication. In areas other than construction techniques, the _1986 Handbook_'s "Power Supplies" chapter only talks about transformer conversion with linear regulators. The _2018 Handbook_'s "Power Sources" chapter includes 1986 content and it adds an additional ten and a half pages about Switch Mode and high-voltage techniques.
Thank you, 73,
--
Don Kuenz KB7RPU
There was a young lady named Bright Whose speed was far faster than light;
I used to do a lot of NMR stuff, but mostly current controllers and pulsed gradient drivers and temperature controllers. Let me know if I can help there. I don't do much classic (sinewave, tuned circuit) RF.
My customers would use an ARB to push a reverse-FFT bandlimited current waveform into the RF coil, over some frequency range of interest, then listen and FFT.
--
John Larkin Highland Technology, Inc
picosecond timing precision measurement
You might search on Ebay and AliExpress for prototype pcbs. THere are some with multiple holes per pad that might be of some help. Try searching on Aliexpress for TMOEC pcb.
Although it's extremely tempting to buy the 2019 book, my resolution is to at least work my way through the 2018 book first. A new chapter on Computer-Aided Circuit Design appeared sometime between the 2006 book and the 2018 book. Some of the topics in the new chapter are identical to SED threads from the past few years:
Hobby versus Professional
Monte Carlo analysis
Gummel-Poon model
IMD Simulators
Harmonic Balance Simulators
Maxwell Without Tears
Surface Meshing-Method of Moments
OK. Although the new print chapter covers Maxwell the "With Tears" QEX reprint is actually in the downloadable eBook, which comes bundled with the book. Sometime between 2006 and 2018 ARRL stopped bundling physical media with the book.
Thank you, 73,
--
Don Kuenz KB7RPU
There was a young lady named Bright Whose speed was far faster than light;
The uglier the antenna, construction, or layout, the better it works.
I don't know. I have some old ARRL Handbook issues at home that I rarely read or use.
5-25 MHz is practically DC as far as breadboarding is concerned. As long as the design isn't extreme in some manner (very high gain, low noise figure, high power, very low power supply voltages, etc), then fairly sloppy construction methods will suffice. The problem is that nobody writes articles or books explaining how to build a sloppy or messy 3D prototype.
There are lots of breadboarding methods available. I rather like John Larkin's method of using PCB islands glued or soldered to a Cu clad PCB. The problem is that I've only built two such breadboards. The way I normally do it is with a single sided Cu clad 0.062 PCB. The copper is the common ground. The components that have one or more grounded leads are soldered directly to the ground. I use small value ceramic caps as standoffs, with the ungrounded lead pointing up and away from the ground plane. Active parts are connected between these proxy standoffs to form a 3D circuit. I like this method because the huge ground plane provides really good bypass capacitor grounding and device isolation, which keeps the various gain stages from becoming oscillatory. If I keep the leads short, there's a little ringing and little tendency for the leads to radiate. If you look carefully at the above photos, the breadboards are build in somewhat the same manner.
Normally, I would also post a photo, but I don't have any such prototypes in my palatial office. Maybe when (or if) I go home tonite[1].
[1] The water company called to inform me that I'll need to boil my water for the next 48 hrs because their redwood tank ran dry and sucked up the crud from the bottom of the tank.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
My own inclination would be to concentrate on building a circuit on (and as much as possible in) a multilayer board, with all the fast connections rea lised as controlled impedance microstrip or (better) buried striplines.
This is not traditional RF construction, and stuff with history (like the A RRL handbook) may not go in for it as the defaukt approach.
One of the responses does point out that the latest copy of the ARRL handbo ok talks in terms of getting printed circuit board made on-line, and KiCad lets you do the layout and create the Gerber files you'd need for that for free, not that the process isn't time consuming (but that's unavoidable).
the empty space has a SDcard holder hidden under it.
formatting link
double sided tape
peeseebees only make sense for mass production. Working this way is much faster, more reliable, easier to change and update, and cheaper. peeseebees is a religion, supported by peeseebee software sellers. All fanatics that want the end of this uniqueverse.
I can assure you that they exist, whether you believe or not. ;)
But much more fragile, especially those haywire contraptions of yours. Also perf board is much less reliable and much slower to build than dead bug. A bit of Cu-clad fit inside the lid of a die-cast Bud box and held down by BNC bulkhead connectors is my usual style--works great up to about a gigahertz.
I do a lot of dead bug stuff for proof-of-concept systems. Customers like it because it's fast and (comparatively) cheap, and I like it for the same reasons and also because I get to keep the money instead of giving it to CMs.
Yeah we're fairly well versed in nmr here. This colleague is trying to do CW nmr... (which is such a pain once you see how nice pulsed methods are.) The two CW techniques I know of are; a balanced bridge where sweeping through the resonance puts the bridge out of balance. And you monitor the reflected power. Or a marginal oscillator where the nmr coil is part of the oscillator circuit. (I only know the marginal oscillator 'in theory' and never built one.)
Right... but of concern is not just bread boarding, but in nmr you typically have a coil down the end of a probe/ transmission line and there is some thought needed to get good coupling into the coil.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.