Is S.E.D actually sci.electronics.DUMMIES ??

In my haste I left out a few points...

(1) The namby-pamby's who never post a technical question/answer are always after me to offer a technical post. Here 'tis! Where's your response?

(2) Part of my "Looks like crap to me ;-)" comment was directed at the amateurs who always have to go back to voltage-between-stages mode. They can't cope with current as a signal.

(3) I'm off to get my (eye) lens replacement. Probably won't be back for at least 4 hours. My bet is there will still be no analysis.

Wimpy wooses ;-)

...Jim Thompson

"Jim Thompson" wrote in message news: snipped-for-privacy@4ax.com...

I'd not noticed that post till now when I had to go search for it, I dont look here for a few days sometimes, I could end up spending most of the day reading and answering posts here, wich I feel like ive ended up doing on a few occasions, It amases me how the likes of Win and several others make so many posts wich go into such detail, maybe they read/think/type a lot faster than me ?

but this looks like a piece of pie to me, the input stage is just a differential pair made with Sziklai darlington the rest is just taking the ratio of the right resistors .... there easy as cake.

Colin =^.^=

So write an equation. Then the defect/poor-"design" will be obvious.

...Jim Thompson

"Jim Thompson" wrote in message news: snipped-for-privacy@4ax.com...

Wel ok if you insist, im kinda busy with my lightspeed converter so il just give it a quick going over, in the first stage ignoring emitter resitance as its a darlington, current is due to the input voltage apearing accros the resistance between emitters, wich is (r2+r6) in parallell with (R9+vr1) this current flow through the collector load consisting of (r4+r8) in parallel with (R10,R11) and produces a voltage - good old ohms law again, this differential voltage is amplified by the op amp stage in the ratio of its feedback resistors, ie r13/r10. to produce a single ended op voltage.

see - its just a question of resistance ratios. now when do I get my pie or cake ?

Colin =^.^=

So write it as a single equation. You're waffling ;-)

...Jim Thompson

Sigh!

...Jim Thompson

Yep it is.... sci.electronics.DUMMIES

Where are all the gurus when you want a real engineering answer? Sucking on their thumbs and bowing to Gore ?:-)

...Jim Thompson

"Jim Thompson" wrote in message news: snipped-for-privacy@4ax.com...

You seem to be deluded. This isn't sci.electronics.design.answers.are.compulsory you know.

LIkely whatever post you are talking about was lost in the clutter of off-topic junk that people like you think fits the newsgroup because they see it as a hangout.

There is so much junk in here, it is increasingly not worth paying much attention to. I know a number of times lately I've just skipped over all the messages without paying any attention, because it's not worth the effort to go through them and see if there is anything of value.

Michael

Titled: "fully differential opamp implementation" ???

It didn't get lost. There's nothing but hot air blow-hards here when it comes to technology.

...Jim Thompson

"Jim Thompson" wrote in message news: snipped-for-privacy@4ax.com...

In your case you are correct!

Some of us have had real work to do these last few days, I for example spent all day Saturday, in an exhausting stint solving a vexing sp/dif interface problem, and all day Sunday trying to rebuild my wife's computer, now 80% done. Today I helped a fellow who wants 1kV at 38MHz, and struggled with bringing up more of my batch of eight handbuilt 400MHz DDS boards.

Also, frankly, not all of us enjoy talking to you, when you regularly insult us with your threatening daydream posts, re what should happen to the majority who can't stand GW Bush, or his cronies, or their policies.

I imagine you find the preamp, although functional, a poor example of streamlined design. Yes, sure, it falls down on many counts when evaluated as to what it might have been. (Working from memory now because my time is short: I'm supposed to be home.) I previously mentioned the appeal of current sources and optionally a mirror. Futher along the path, if the currents from the beefed-up input-pair had been sent directly to the output amplifier, four parts, including two ugly electrolytics, could have been eliminated. In addition the gain equation would look cleaner. The electrolytic in the input emitter pair is especially ugly. One reason is its time constant is hugely affected by the choice of gain, pushing up toward the minute region at low gains, IIRC. That's awkward at power-up, and makes second-stage ac coupling mandatory. Better to remove all three and implement the high-pass differently. I'd consider a different variable-gain scheme. Proper feedback to the input-pair emitters would be appealing. There's more, but that's a start.

Yes, certainly the circuit could be improved, but that doesn't make it crap.

Get an rf signal generator, 100 MHz maybe, near an oscilloscope. Run a long coax to a laser (pointer type is fine) from the generator. Run another long coax from a pin photodetector (with maybe a NON AGC amplifier) back to the scope. Start with the laser close to the pin and measure phase shift. Move them apart, ditto.

This would work with a pulse generator, measuring arrival time, too.

John

but the signal traveling down the coax is governed by the speed of light, by the time the received signal is brought next to the transmited signal it has undergone a roundtrip, moving the optical devices apart just alters the total trip, not just the trip in one direction only.

Colin =^.^=

Yep, I had an interesting weekend, too. Toting tables, chairs, and cases of cookies in my pickup for the Girl Scouts... I'm the troop "cookie mom" ;-)

Writing a couple of quotations and one proposal.

And replacing a PC power supply (grinding fan).

All the while working with just one functioning eye... fixed at noon today with zap-zap-zap from a YAG laser ;-)

I guess that makes us even... I loath leftist weenie policies.

Yep. But more than anything else I was trying to stir up design thought. But I failed. Just one attempt at analyzing the circuit... and that was feeble.

Nobody seems to want to understand how circuits actually work... or care :-(

I was hoping that enough analysis would occur that the limitations would be obvious... in other words a teaching exercise.

I suspect that most lurkers here _are_not_ practicing circuit designers... just cut-and-pasters :-(

...Jim Thompson

that isnt 1KV into 50ohms is it?

I lost momentum in fixing my PC with the flash bios chip upgrade failure, a new chip got lost in the post twice ! so im stil using an old one, buts its not that bad for basic stuff. PIC compiler stil runs almost fast enough. if I need to run a simulation i might find the enthusiasm to fix it.

I dont like electrolytic decoupling caps, theyr not the most reliable components IME, they dont seem anymore reliable with no dc bias either.

you could probably short the 1000uf but trim the offset instead, id probably be tempted to do away with the other caps and put a feedback path to act as a DC trim, but I expect the audiphiles wouldnt like this due to the extra noise it might introduce.

I dont feel comfortable with no DC blocking anywhere in an audio system though.

yeah, the gain pot is kinda screwy as far as gain relationship, but its got some imunity to common mode noise pickup and gives a good tradeoff between max signal headroom and noise performance at extremnes of gain.

as suggested several times now, the resistors in series with the op amp nodes could be shorted. as for the current sources/mirror im not sure any improvment would be notciable, although it would make a most effective point to introduce dc nulling feedback.

Colin =^.^=

The coaxes add a constant delay, which you just subtract out.

C = deltaDistance/deltaTime

Past that, you're getting philosophical.

John

View in a fixed-width font such as Courier.

. . . .-----. . | | . --- | . R2 . | ve -ve . +------+----------+---Rg---- . | | | . | | gm1vbe1 | . hie1 /|\\ | . vd | \\v/ | . -- >----' | | . 2 | | . +-----. | . | | | . R3 hie2 /|\\ gm2vbe2 . | | \\v/ . | | | . +-----+----+----> -vo/2 . | . | . R4 . | . | . --- . . . . R4 . ---------- . vo R2||(Rg/2) . Av= -- = ------------------------------- . vd 1 1 . 1+ --------- x ------------------- . R2||(Rg/2) gm1(1+gm2(R3||hie2)) . . . . .

Well, it depends on what you are trying to do. Are you looking for currents in the once presumed ether?

You could take two highly stable synchronized time standards, A and B, and move B to the destination end of the experiment, and turn on the laser when A reaches a predefined second, and record when the light pulse comes to B after that same predefined second.

If you are worried about relativistic effects on the time standards, you can bring A and B back together to recheck their synchronism.

-Chuck