AoE x-Chapters, High-Speed op-amps section, DRAFT

I've been tweaking my 500 MHz triggered Colpitts oscillator [1]. It's now common-collector, using either a bipolar NPN or an enhancement phemt. I'm seeing DC effects at startup.

If the device transfer curve is nonlinear upward, which they always are, an emitter follower can rectify even if there is no base/gate current. For a largish AC voltage, a mosfet source follower can rectify. The rectification will be a fraction, a percent maybe, of a bipolar.

[1] One of the several circuits that I have been futzing with for years. Turns out that the BFT25 likes to oscillate at un-observable frequencies in addition to the frequency it was told to oscillate at.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

There are some dual and quad 741 equivalents that are also crazy cheap, but aren't as bad as a 324.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

Yes. The 324 is more common thus cheaper as recycled parts. It doesn't get better or worse than a 324, depending on your pov.

NT

Am 01.05.19 um 22:08 schrieb John Larkin:

IIRC, there is a nice triggered oscillator in the interpolator of the HP5370 time interval counter. That's from the time when the circuits were in the manuals. Should be easy to find.

regards, Gerhard

I've studied that one. It's an ECL gate+delay line triggered oscillator. I think it's actually a ceramic hybrid.

It's not very temperature stable, so they keep it running all the time when it's not in use, and phase lock it to an OCXO. At trigger time, they quench it with a one-shot for 75 ns, then restart it. That adds jitter and delay. A bunch of our products are based on instant-start oscillators. The hard part is to keep the tempco and jitter down.

The 5370 is an astonishing instrument, especially considering the state of technology when it was designed. It uses a 6800 depletion-load 2-phase nmos CPU that takes, I recall, 2 us to execute a no-op instruction. The math is 8-bit adds and subs, no mul/div. I can't imagine how the code does what it does.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

Even that isn't so great. There is _nothing_ to recommend the '324, these days. Fifty years ago, sure. Today, nope. No way.

The price is the best it gets. The more there are in scrap, the less labour it takes to get them. And you get 4 for the same work as 2 with other opamps.

NT

Anyone using scrap deserves what they get. I won't even use a part that's fallen on the floor or has escaped my sight. It's just not worth the hassle to save a nickel (or less).

absolutely, they get to make a living & eat.

For you, no. Many aren't so lucky.

NT

Bullshit. They deserve scrap. Their customer is getting screwed.

Luck has nothing to do with it.

nope, that's what they get

since you don't know a thing about them

no, the customer gets a good deal.

Clue has nothing to do with it in your case.

NT

What do you do that's so price sensitive?

Do you have a lot of test failures form using reclaimed parts?

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

So you screw your customers. No surprise.

You *SAID* they were scrap, dumbass!

No, he gets screwed. You're a thief.

I've gleaned that he lives in India, where $15/week is a basic living. At that point it's hard to afford enough calories to survive, so anything that can be scavenged and turned into rupiah can change someone's life. krw probably thinks they're poor because they haven't shown enough initiative so they deserve it, but he's probably never visited. He probably can't spell "privilege" either.

Clifford Heath.

hat's the point. Price matters for a whole lot of people.

,

labour it takes to get them. And you get 4 for the same work as 2 with othe r opamps.

We educate. It's the assemblers who have to pay for parts, that's the prime cost sensitive area.

In electronics, basic consumer doodads, EE tools of the cheap kind, tools & equipment that are useful in the developing world. And much else outside e lectronics.

Most scrap electronic goods have a fairly high parts count, it only takes o ne failure for them to get chucked. Much of the time that failure isn't a p art anyway. So the failure rate from reuse is low but not zero. More failur es occur from assembly. If assembling stuff with 1000+ parts it would be a different story. 100% of units are tested. What symptom each part failure w ould produce is assessed beforehand so repairs are quick & easy.

NT

I see you're too clueless for this to be worth continuing.

NT

UK. Assembly is dotted about the developing world. At the extreme end some previously made as little as 8-10p a day, under 15 cents.

that's true at 10p a day. Also corn alone does not deliver the right nutrients.

Yes, but I do design these things to be reliable, safe & nontoxic, it's not just a free for all. I don't want to create more problems.

NT

Well, kudos for doing something meaningful to improve life in the 3rd world. You have my respect for it, even if some people would prefer to bomb them into oblivion or continue to cheat them into eternal servitude and subordination (as they believe god and Ayn Rand require them to).

I hadn't pegged you as another Slowman. Live and learn.