Transformer articles posted

[dissonantstring]

CJ , please remeber that many members are not california dudes; sometimes it's very difficult to grok your lingo ???

hi abbey,
i always have trouble understanding CJ's posts and i live in bay area of california too!  ;D 
i've never considered CJ nor myself the typical "california dude", but who knows what people perceive via the forum.  i've met CJ a few times and have no problems understanding him in person.  CJ even helped me build and test a couple of 2503-type output xfrmrs at his shop a few years back (at a time when they were difficult to get) and they reside in a hard working dual 312 pre til this day.

CJ's posts are more likely a creative outlet for him as well as a source of useful info for the rest of us.  "your entertainment value from his posts may vary."
you just have to work to decipher it most times. ;D  makes this place a little more interesting.  let's all just live and let live as long as we're civil, right?
or is that civil rights?
sorry for the rant.
kind regards,
grant

 

[lassoharp]

Sorry for backtracking but could someone please tell me how to access the article links at the top of the thread.

When I click, it takes me to the main DIY gallery page.  Don't really know how to properly search it from there.

Thanks

 

[CJ]

what do yo want to build?

we could take an application and design a transformer around it, you just pick the project,

that might be the best way to learn transformers, try to design a few,

i tried talkin normal, but it was like reading a textbook from McGraw Hill, youknowwhati'msayin? :

see spot. see  spot run.  E = IR. 

if people want that they will go to the liebrary. 

 

[lassoharp]

Jeeez no wonder - just realized the original post was 3 years old!

What I'm tryin to research is exactly how Q of a coil affects an LC resonance circuit in a passive EQ (MEQs and EQP-1As specifically).  In terms of boosting amount, distortion, and curve broadness/narrowness

Also the theory behind why the common reactance is usually around 2K in resonant circuits.

 

[CJ]

  Q

1) an OEM equipment supplier for 007 movies


2)

so that is a good, simple start, we can expand on that for our purposes>

wait, hold on, too much giberish already, lets translate this engineer's coffee rush into this>

(flip the equation over and put Q on the left so we can actually use it, i love lazy authors  )

Q = pultec band frequency / frequency  endpoints  (where you only have half the signal level)

so lets say we want a filter at 1 k. so,

Q = 1,000 / bandwidth  , so if we want a 200 hz  notch, thats 100 hz off each side of the 1 k,

Q = 1,000 / 200 = 50.  thats a pretty high Q, higher than my IQ, 

so we see that the pultec  might be spread out to maybe a 400 hz notch?

Q v=  1,000 / 400 = 25, that is more reasonable.

are you with me so far, let everybody say yeah!

"yeah"

let everybody say yay,

"yay"

ok, this time, i want all the girls to go...

don't you love it when concert acts pull that aged crap?  ;D

 

[lassoharp]

Perfect!!  Mucho thanks CJ!

That explains why my curves with some junk-box coils were spreading out wider than a cheerleaders split - kinda cool though.

 

[CJ]

you can simulate families of curves with variable R to really see it.

ok, so on the 2 K thing, i think you want to stay away from too high an impedance with these inductors,

if they pick up noise and hum, it does not short end to end, but has to go through your coil to null itself,

this induces some nice stuff if you are into noise and canibal corpse or coffin, or oxbow , or i'm on crack, 

is nick cage still alive?    edward s robinson?

 

[sodderboy]

Slenderchap reminded me that the original links were no longer valid.  I corrected it in the original post.  There are a total of 7 groups of articles.
Enjoy!
Mike

 

[lassoharp]

you can simulate families of curves with variable R to really see it.

ok, so on the 2 K thing, i think you want to stay away from too high an impedance with these inductors,

if they pick up noise and hum, it does not short end to end, but has to go through your coil to null itself,

this induces some nice stuff if you are into noise and canibal corpse or coffin, or oxbow , or i'm on crack, 

is nick cage still alive?    edward s robinson?

Yeah, it looks like for the lower freq boosting the common reactance is about 720ohms.

The coils I've been tinkering with have Qs around 3 to 5 mostly, some 10s.  Using some old 70.7V speaker

to lines - the ones they used for intercom feeds.  Handy for having a bunch of multiple taps.  So using the

same test jig, some ind/cap combos for a specific freq give audibly higher boost amounts than others.  

One other thing I'm not sure about is the 'slope' of the curve for inductance vs freq of a coil.  In other

words, if a coil measures 1.2H @ 1K and 638mH @ 120Hz, then what's it doing at 60Hz?  If another coil

measures roughly the same at 1K, what influences what it will be at 120Hz and 60Hz?  So far I've been

using a ratio - 1mH:1.5Hz in this case for a crude estimate, but I've no idea what function dictates

that particular curve.  

 

[phishman13]

thanks for updating the link to those articles!  It looks like there is some info that is EXACTLY what i'm looking for.  The harmonic distortion in iron core transformers! 

Awesome!!!  ;D

 

[CJ]

lasso, 2 pi f L is a linear equation, so all you do is draw a line at  45 degree angle and put down some units.

for a cap, you get a 1/x curve, which is an ellipse/hyperbola type curve, very bent compared to the inductor.

so there are probably some funny waveforms from the non linear vs linear fight going on over the ac signal.

 

[PRR]

> if a coil measures 1.2H @ 1K and 638mH @ 120Hz, then what's it doing

No way to know.

A perfect inductor has the same inductance from DC to infinity. (And as CJ notes, impedance rises from zero to infinity.)

A perfect inductor has the same inductance at any signal level.

There are no perfect inductors.

Iron-core coils are exceptionally lousy inductors. Their "inductance" is not constant with frequency or level.

> 1.2H @ 1K and 638mH @ 120Hz

That seems wrong to me. Iron-core audio coil inductance tends to drop above 400Hz (incresing eddy current forces flux out of the core).

But fooey on theory. You have a part, you have a practical application, just do it. Get a signal generator, an R, your L, and a C, wired in your proposed topology, driven at your proposed levels (from teeny to +8dBu or whutever). Hear and see what comes out. Fiddle C for desired freq, R for bandwidth.

Don't be shocked if you can't make it work, or not without going to absurd low impedance or wimpy Q. High-Z high-Q filters require coils designed for the purpose, or exceptionally lucky "finds", not cheapo hotel speaker-iron.

 

[lassoharp]

Correct - had that stated backwards.  Should read 1.2H @ 120HZ - 638mH @ 1K

 

[CJ]

what are you using to measure inductance?

watch out for dr j.

































































operator.  ;D

 

[lassoharp]

LCR meter.  Reads at 2 different freqs. and gives a Q.

Then  I sweep through a sig gen into a scope to verify/compare what the calculations give.  Then listen

to program material.  Looks for the most part that I've got some coils good enough for making a couple of

passive 'color boxes' for assorted boosting/cutting - I can see them being useful during mixdown.

 

[dougsta]

The links do not seem to work. Any chance to see the articles?

et voila

 

[CJ]

here is a cool article>

https://www.sowter.co.uk/pdf/GAVS.pdf