Twin T circuit wanted

I loved your bell circuit John! I didn't quite get how the inductor was working. But still I got the idea. Thanks.

George H.

This is kind of a cute circuit. I first designed it when I needed a very frequency and amplitude-stable sine wave to drive a Talyvel LVDT-like inclinometer, part of the Boresight Alignment Kit for the C5A. We had to measure level to arc-seconds of accuracy.

It's a transformer with a resonant tank in the collector and a positive feedback drive winding into the emitter. The emitter feedback is just a couple of tenths of a volt p-p.

The cool thing is that the collector swing is almost exactly 2xVcc peak-to-peak. As the amplitude builds up, at the negative swing peak the emitter goes a little bit negative, to get out of the way, and the collector swings to just about ground. That forward-biases the c-b junction and discharges the base cap, reducing transistor base current hence gain. So it has a built-in peak detecting AGC amplitude levelling loop with close to zero TC. All from 5 parts. Or sometimes six.

I tested the system on a 55 gallon drum full of sand, with a huge steel plate on top. We built a platform that pivoted on ball bearings and we moved a long lever arm with a micrometer, to tweak the sensor angle. I had to tell people not to walk nearby, because their weight would flex the concrete slab of the building.

I learned a bunch on this project: HeNe laser power supplies, synchronous detectors, optics. I've always sort of liked this oscillator.

John

ftp://jjlarkin.lmi.net/Ships_Bell.JPG

With your explanation, the oscillator is much more attractive. Some folks can't stomach magnetics, perhaps that's part of the criticism. Anyway, ignore JF and JT when they get bitchy, it must be something they ate, or is age catching up with 'em?

Our old friend, Tony Williams, R.I.P., would have been pleased.

. John Larkin's LC oscillator, supply-V sets amplitude . _________________ . | | | . | _|_C E || L . Rb --- E || . | | E || . | | * | || . | +-----' || . | | || . | C gnd || . +- B | || . _|_ E E || . --- | * | || . | +-----' || . gnd | . '--- out

Tell us more about the inductor / transformer, turns ratio, etc. Did I get the winding polarities right? I suppose you have to keep the emitter's reverse-voltage under say 5V breakdown, so 10Vpp max output? What are the prospects for making this into a high-power oscillator? Drive a speaker directly?

--
 Thanks,
    - Win

ng

t e

it,

r

g.

Thanks for the explanation. Why didn't you take the output from the collector if you've got so much voltage swing there? Would that have killed the oscillations?

Fun, Reminds me of our interferometer, if you move a 1/4-20 screw on the 1/2" aluminum bread bread board you can see the fringe pattern shift. (See it electroncially I mean.)

George H.

I'd be curious to see values. It isn't "necessary" but I would like to know what core, type, size and material you used. Did you tweak the tuning? Running with banks of them to simulate a real bell, did you calculate and tune for specific harmonics?

My basic interest in the twinT was simplicity, and easy to adjust frequency and decay time, but your design is simpler.

Like you said, if you were to really try for an authentic bell sound, you'd probably find some weird frequencies and their amplitudes changing as it rang. A million year ago I played drums in concert marching and dance bands. With large cymbals and gongs you can see the resonant modes fading in and out and chasing around the circumference - a blur, but visible. From the sound of large bells, I'd guess they also behave that way. (then there's generally a bell tower with its own resonant columns adding to bring out and attenuate harmonics)

Age is a terrible thing. You have to fight it, or at least fool it, for as long as you can.

Polarity is right.

The voltage in the emitter winding is small, just a couple tenths of a volt p-p at full output, so there's no stress on the e-b junction. The collector swings to 2*Vcc, so the transistor Vce has to be rated for that. If you want a big output, take it off the collector or a third winding.

I vaguely recall doing this with permalloy powder cores and with slug-tuned pot cores. I used to use a lot of magnetics for stuff like this, before active filters and DSP got to be easier.

There is a tiny flat-top on the negative collector swing, when the "peak detector" works. The more base bias current, the bigger it is. If it matters (it didn't for the inclinometer) add a small resistor in series with the collector.

John

OK not my circuit, but try whatever they build audio transformers out of. The LC on the collector sets the freqeuncy so that pretty much determines your L. (Guess I figure a C of 0.01uF to maybe 1uF.??) Then John said that the emmiter sees only a few tenths of a volt where the collector is going +/- 12, so I'd start with maybe 50 or 100 to 1 turns ratio.

Can spice do oscillators? (I remeber trying once and having a hard time getting it started.)

George H.

Here's a quick implementation:

Vcc =3D +5v --+-------+------+-- | | | | | |_ || | | _)|| .-. --- L1a _)|| Rb | | C1 --- 1mH _)||

100k | | 1uF | _)|| '-' | _)|| | | *| || | .-+------' || | |/ || +---| Q1 || | |>. 2n3904 || | | * || C2 --- +---------. || 1uF --- | L1b _)|| | | 25uH _)|| | | | =3D=3D=3D | =3D=3D=3D | '-------------> output

(5KHz values shown)

collector swing ~=3D 2*(Vcc+0.6v) emitter swing =3D Vc * sqrt(L1b/L1a)

1 f(out) =3D ------------------- 2*pi*sqrt(L1a * C1)

That's a 1rst cut--I may have left L1b a little hot...

Simulates really nicely--sine waves.

-- Cheers, James Arthur

:

Cool thanks James, Say can you take it up to 10's of MHz?

George H.

I was thinking about it from a more practical perspective - like do I really want to build 12 of these? Frequency of oscillation would be where the inductive and capacitance reactance are equal - but from a theoretical perspective that could be with any inductance like 6" of wire and large C to work in the audio range . . . in theory at least.

If I have to come up with 12 laminated cores that's harder than say 12 ferrite toriod cores (which I may have on hand). But I guess all I really need is a handful of iron wires and a coil around them . . . that I could do with what is in my junk collection.

[snip] [snip]

Win, Did I say _anything_ about Larkin's LC oscillator... other than he's a continual pompous ass?

"Pimp" Larkin seems to keep lumping me with JF anytime JF points out the fallacies of Larkin's (mostly crap) circuits.

As for age, I'm still designing chips, at around four-per-year... though 2010 looks like it'll be a "sixer" :-)

I think you and JL are the ones showing "age"... and it ain't like wine... more like vinegar ;-) ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
      The only thing bipartisan in this country is hypocrisy

After 35 years, I can remember schematics but I can't remember parts values. I do recall that a few guys stood around and critiqued the bell sound, and I changed capacitors to tweak the frequencies. I'm tin-eared, totally unmusical, so I get advice on stuff like this.

I recall doing some of these kinds of oscillators using slug-tuned pot cores, to trim frequency, but the ship's bell thing was pretty crude, and once we got some cap values we just built them, no tuning.

It would be interesting to Spice this oscillator. Complex things happen as the collector voltage dips down near zero, and I wonder how accurately Spice would handle that. Probably pretty well.

I think a bell has several resonant modes close in frequency, and they tend to beat. And it has not-quite harmonics of those modes, which fade out faster than the fundamentals.

I know a guy, a PhD ME, who owns a small machine shop as a side business. He makes handbell sets. It's hard to buy the right "bell brass" which he says is the only stuff that works right. He casts the basic forms and finishes them on an n/c lathe. He has a hammer that whacks the bells, a microphone, a lot of FFT type processing in a PC, and he closes the loop on machining. He claims his bells are as good as hand-tuned-filed ones.

John

On Mon, 07 Jun 2010 20:33:23 -0700, John Larkin wrote: [snip]

[snip]

"designed"??

Quite _exactly_ like you'd find in common AM radios as the LO, once transistors came upon the scene....

First one I hit upon surfing. I'll look for _my_ first radio that used CK760's :-) ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
      The only thing bipartisan in this country is hypocrisy

te:

k

de

gs

r t

Sure. You need to change transistors...

Vcc =3D +5v --+-------+--------+-- | | | | | |_ || | | _)|| .-. --- L1a _)|| Rb | | C1 --- 2.2uH _)|| 100k | | 22pF | _)|| '-' | _)|| | | *| || | .-+-------' || | |/ || +---| Q1 || | |>. MPSH10 || | | * || C2 --- +-----------. || 10nF --- | L1b _)|| | | 100nH _)|| | | | =3D=3D=3D | =3D=3D=3D | '-------------> 22MHz output

The oscillation frequency starts being greatly affected by the transistor parameters though. Much higher than this, and other circuits are used.

This variation's good to UHF (100Mhz shown):

Vcc =3D +5v --+--------+--------+-- | | |_ | | _)|| .-. --- L1 _)|| Rb | | C1 --- 100nH _)|| 100k | | 22pF | _)|| '-' | _)|| | | *| | .--+--------+ | |/ | +---| Q1 --- C3 | |>. MPSH10 --- 2.2pF | | | C2 --- +-----------' 1nF --- | | .-. | | | Re =3D=3D=3D | | 1k '-' | =3D=3D=3D

Here, C3 replaces L1b in providing the positive feedback. It's made as small as possible, to avoid loading the collector tank. At really high frequencies, Q1's collector-emitter capacitance often suffices.

C1 is often replaced with two caps in series forming a divider, with low-z output taken from the divider tap. Or, you can tap the inductor.

-- Cheers, James Arthur

Cute, and familiar, yes, but that one doesn't have John's c-b conduction agc mechanism. That's a pretty neat feature--have you seen it elsewhere?

-- Cheers, James Arthur

Right, it's an L, not a transformer. Inductors can be wound from the same ferrite cores as transformers, etc., but with one *big* difference: They need an intentional air gap, to set a predictable A_L value, and insure a stable inductance.

--
 Thanks,
    - Win

Hey Win,

Have you looked at blocking oscillators? I've gotten pretty good at keeping them oscillating, which seems to be the first trouble with them. Meanwhile, delivering substantial power output with high regulation from just a couple active components. :)

Tim

--
Deep Friar: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms

He has a pretty inexact definition of "_exactly_"

It seems to mean "not very similar at all."

John

What a pile of nonsense. I'll leave it to the newbies to do some reading on their own to see how that is done _all_the_time_... otherwise (and unfortunately) they won't know shit about how anything works.

But, hey!!! It's good for my business if newbies can't make anything _really_ work :-) ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
      The only thing bipartisan in this country is hypocrisy