Actually to thaumaturge, a question on the RF oscillation of a fixed-length end-connected, insulated loop.
While I understand the concept of
F = c/kλ … where
c = 299,792,458 m/s
k = 1.03 or so for insulated small-gauge Cu wire
λ = 2 D … two times scale-length (D diameter) of loop
F = 299,792,458 / (1.03 × 0.3048 ft/m • (2 × 18" (D) ÷ 12 in/ft ))
F = 318,308,000 Hz
Pretty close to the 327 MHz, right? (I actually measured the speed of light in microphone cable in college, oh… back when Dinosaurs roamed the streets of UCBerkeley). Got it.
But when “the loop” is all twisted co-linearly, does it still resonate? At least to any significant degree that isn't self-cancelling 'cuz of the twist-over-length and because the “loop” is topologically degenerate (being a co-linear pair of wires)?
That is my long, long standing question, to which I defer to those Thaumaturge, DF96, others readily and heartily. I just don't remember anymore. 50+ years since all those electrical engineering classes and HAM RF work.
Thanks in advance,
GoatGuy
While I understand the concept of
F = c/kλ … where
c = 299,792,458 m/s
k = 1.03 or so for insulated small-gauge Cu wire
λ = 2 D … two times scale-length (D diameter) of loop
F = 299,792,458 / (1.03 × 0.3048 ft/m • (2 × 18" (D) ÷ 12 in/ft ))
F = 318,308,000 Hz
But when “the loop” is all twisted co-linearly, does it still resonate? At least to any significant degree that isn't self-cancelling 'cuz of the twist-over-length and because the “loop” is topologically degenerate (being a co-linear pair of wires)?
That is my long, long standing question, to which I defer to those Thaumaturge, DF96, others readily and heartily. I just don't remember anymore. 50+ years since all those electrical engineering classes and HAM RF work.
Thanks in advance,
GoatGuy
Thank you. Perhaps at times I'm over reactive about things. Comes from being overly weary over arguments I wanted nothing to do with in the first place.
In this case the learned expertise applied directly to myself, back when I built my first good amp. I was just assistant broadcast engineer to my first real mentor. He drew me up a design for a great amp. I gathered the parts and built it. I was pretty green and made the exact same error as this thread covers in my construction. I had lots of a smaller guage wire than was called for, so I thought I would just double the leads up. I mean I doubled pretty much everything but the input signal leads. Power supply lines, output leads, dven transformer wiring. Mentor pointed out that I had loops everywhere.
Well, the amp sounded phenomenal, but best noise floor it reached was -85 db S/N. It only acheived that because I happened to use fully sheilded can transformers. A few years later I thought I'd see what actual effect all those loops were having, so I rewired everything using just single conductor wiring. Same parts. Same layout. S/N dropped to -90db.
Hard to argue with that.
Doc
¶1 — Yep, I know the feeling.
¶2 — Very interesting result.
Worth remembering, to be sure.
PS: thaumaturge = "worker of miracles and extraordinary magical tricks"... fits.
Thx.
GoatGuy
Grew up in Alameda.
Took my bike, 6 days a week to Oakland Airport (summers, high school)
Worked at General Aviation Supply Company "GenAvCo". Emmons… hard nut.
Next door to Mike Quinn's Electronics' quonset hut. Just like GenAvCo's.
All nature of obscure, arcane, mostly-quixotic WW2 and more current surplus.
And chips. And discrete transistors. And LEDs and other fun stuff.
But most importantly, Bear. The gargantuan elf behind the counter.
Totally honest, without knowing the providence of his Art. Bear. Bear
Haltek was a bit of a longer trip, not really possible by bike.
But that didn't mean I couldn't tag along with “friends”.
Great place. Few things like it.
The bike ride was literally "against the wind, both ways".
In the AM, blowing against me, Central-valley clearance of bay.
In the PM, blowing against me, Central valley sucking it back in.
Whew… by my Sr. year, I was clearly THE strongest runner.
No one — even the guy that went onto the Olympics trials, could keep up.
But I was into tinkering. Didn't try out for the track or other teams.
No time. Too many transistors to discover. Most barely marked.
No problem… just "invented" my own transistor tester.
Had limiters and all sorts of cool things not to kill the poor critters.
Anyway, good times.
Take care, fellow "old timer".
GoatGuy
Not meaning to hijack thread, but started as cal tech in cal lab at Applied Technology (Later div Litton). Worked up to metrology engineer when Litton bought them from Itek. Litton sucked to work for. Left and became a bench tech at Haltek alkng with another sdnior tech from cal lab.
Up to then Haltek used equipment side was rather iffy. They'd buy lots of equipment on pallets and just sell it as was. Return rate was heavy.
Suddenly between Mike and I the used equipment side got hot. We fully checked out and actually fixed things before shipping them. Returns dropped to nill and orders came flooding in. Soon equipment sales began exceeding retail store sales by magnitudes. Haltek spun of "Test Lab Company". We shifted from two bench backroom over to upstairs full sized 8 bench lab in adjacent warehouse which emptied out and became big test equipment showroom feeding sales all over US.
Was heaven. We mostly got to pick and choose our own equipment. Built a primary standardd station that included a Cesium beam atomic clock and a Rubidium working standard. A Lear Solatron system DMM and cream if the crop LN standard resistors with oil bath.
On my own bench I had a Tek internal built 2465B with every option. A Systron Donner branded Daytron 1082 DMM and a Fluke 6061A 1GHz synth just to hit high spots. I left after 4 years as Sr. Metrologjst and started my own consulting business.(which eventually tanked sending me back into broadcasting.)
Mike bestowed me with the "Doc" nickname when we were back at first cal lab. Thaumaturge I came up with myself. Thaumaturge is what companies expect of me... Doc is what they get.
I fully grok the uphill both ways voncept. I ride an average 100 miles a week down there near sea level. A 3 MPH breeze felt like a 5 degree grade.
Take care buddy.
Doc
We now return this to our regularly scheduled thread.
Up to then Haltek used equipment side was rather iffy. They'd buy lots of equipment on pallets and just sell it as was. Return rate was heavy.
Suddenly between Mike and I the used equipment side got hot. We fully checked out and actually fixed things before shipping them. Returns dropped to nill and orders came flooding in. Soon equipment sales began exceeding retail store sales by magnitudes. Haltek spun of "Test Lab Company". We shifted from two bench backroom over to upstairs full sized 8 bench lab in adjacent warehouse which emptied out and became big test equipment showroom feeding sales all over US.
Was heaven. We mostly got to pick and choose our own equipment. Built a primary standardd station that included a Cesium beam atomic clock and a Rubidium working standard. A Lear Solatron system DMM and cream if the crop LN standard resistors with oil bath.
On my own bench I had a Tek internal built 2465B with every option. A Systron Donner branded Daytron 1082 DMM and a Fluke 6061A 1GHz synth just to hit high spots. I left after 4 years as Sr. Metrologjst and started my own consulting business.(which eventually tanked sending me back into broadcasting.)
Mike bestowed me with the "Doc" nickname when we were back at first cal lab. Thaumaturge I came up with myself. Thaumaturge is what companies expect of me... Doc is what they get.
I fully grok the uphill both ways voncept. I ride an average 100 miles a week down there near sea level. A 3 MPH breeze felt like a 5 degree grade.
Take care buddy.
Doc
We now return this to our regularly scheduled thread.
Last edited:
I find it hard to believe that an entire industry misuses such a clear term as 'ground loop' so that it include non-ground loops. I find it easier to believe that some individuals within an industry somehow picked up the wrong meaning early in their careers and either nobody corrected them or they didn't hear the corrections. Anyway, there are others on here from that industry so maybe they can enlighten us.thaumaturge said:
It will still resonate, but more as a dipole than a loop. The loop area is small and has numerous phase swaps so will not couple well to external fields - which is why we do twisting.GoatGuy said:
Difficult to remove wires and leave everything else exactly as it was. Some change in S/N not surprising, but not necessarily due to removal of loops.thaumaturge said:
I've long felt (in concurrence to your riff) the same way. But this whole enterprise (your pick: valve steampunk electronics, discrete not-so-steampunk, digital gee-whiz stuff) is FULL of people who are incredibly enthusiastic and veritable terms-of-Art magnets. And parrots. And flapjacks. So, yah. I agree.
Yes, “more as a dipole”, this I get. That's what the external braid is there for too. To shield “the dipole component” from ambient RF field induction 'cept for the ends. Which, if short enough (and ironically, if not an odd sub-multiple of the total dipole length!) markedly reduce induced resonance.
Yep, I apparently and normally privately, channel your way of thinking. My first thought was “well, the layout changed upon replacing 100% of the twisted multi-wire with single conductor stuff”. Even fairly modest changes in wiring topology can have measurable and often acoustic effects on an otherwise identical layout.
I, as a for-instance, learned the “dâhmned hard way” about that in one of the University of California's graduate EE classes. A lovely class. Huge bed-of-sockets breadboards for every student team. Yep, even in the '70s, we were paired up more-or-less-at-random for our projects. I suppose it kept us from individually ransacking the EE supply guys guarded cage.
Anyway, having a really-really-really crazy Father who insisted on eat-off-the-floor-and-label-one's-sock-drawer cleanliness, learned to value precision in layout, clean lines, 90° corners, that kind of thing. Had a project to lay out a full 12 bit CPU … kind of a bear … with 7400, 74S00, 74LS00 and some 74F00 5 volt logic. 250 chips or so. For a class that lasted 10 weeks.
Since my lab partner was a total theory-geek and didn't know how to strip wires, I ended up taking over building the thing. Used roll after roll of 26 gauge telephone company wire, everywhere. All wires measured, all co-linear, all with 90° bends. Absolutely textbook beautiful. When static testing demonstrated it all was really working as planned (i.e. 0 Hz clock, momentary contact switch as clock), I dedicated a Saturday-before-grading-it to actually wire-tie all the wires. Neat and clean became a showpiece.
Still passed the DC test.
However, it became the Pig of the Lab, MHz wise. Ramping up the clock, we simply couldn't get the darn thing to get past 140 kHz. Even then, it wasn't really stable. 100 kHz was totally stable. Meanwhile, everyone else's (few of which actually worked, mind you) were clocking well above 2000 kHz. Why?
Got an 'A' for the project in spite of it. Prof Cory Hunter asked that I bring the project to his lab and “volunteer this Summer” to work the lab, unpaid, and to get the thing working at 5 MHz with his tutelage. Thus… the hardest dâhmned 8 weeks-of-Summer in my time at University. Discovered all about resonant lines-as-dipoles, about shielding, about distributed capacitance, inductance; about 90° corners-as-chokes, about the hard need for there to be ground-planes between HF lines. And, by September 1, the very same board was clocking at 4.8 MHz, the fastest in the lab.
Those kind of lessons, unpaid, 12 hour days, 6 days a week, all Summer … are the ones you remember forever. Every last hard-won finding. Needless to say, it also was almost the last wire-tied project I did, too. LOL
Again, thanks for the memories.
Back to our usual programming.
Just saying,
GoatGuy ✓
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.