Forums » CircuitLab Support Forum

Hi,

should that work?

-christoph

@Midiman,

If you'd like specific help then you need to make public and then show or point us to your circuit.

:)

Search for "initial-conditions" to see some examples of how to kick circuits out of unstable equilibrium at the start of a simulation.

Hint: make the circuit slightly asymmetric. make the pullup or bias resistors slightly different values (even by only what you'd expect within the normal component tolerance).

You might look at this:

It's your basic astable I whipped up. It works!

Well, with a few rough edges, the simulation doesnt converge at a few spots, where you see the jiggles.

But the basic flip-flopping is there.

Note that if you change the collector resistors to be both 10K, the circuit ramps up evenly on both sides and doesn't oscillate! Even a 1% change in the resistors stops this madness. In real life the resistors are going to be plus or minus a few percent and the transistor leakages and betas much more variable-- many cheap transistors can have a beta range of 30 to 300, so in the real world these hardly ever latch up.

@arduinohacker,

I don't understand your reference to latch and the use of local negative feedback via the two collector base resistors. The classical two transistor astable multivibrator has no local negative feedback, only global positive feedback.

Your circuit does demonstrate what I was hinting at and the classical circuit below shows that less than 1% difference between the base resistors can set off oscillation quite happily.

:)

Sorry, I meant the reference in your circuit notes to "latchup" not "latch".

Well, I kinda lied. I was lazy, and for DC bias I used those two collector to base resistors. To do it right would require SIX resistors, two in the emitters and then two each as a voltage divider on each base. But I was lazy.

As a side benefit, the C-B resistors guarantee that the transistors are not going to ever saturate, at least at reasonable temperatures.

Remember LSTTL?

If you want to stop transistors saturating in a switching circuit then this is how it's done:

A link to the basic multivibrator:

http://en.wikipedia.org/wiki/Multivibrator

An article on edge speed improvement in the humble astable:

arxiv.org/pdf/1201.1819

And a way to ensure startup with slowly rising supplies:

http://www.davidbridgen.com/astable.htm

Good grief, they wrote that up and didn't know of or mention the earlier work done by HP? The HP frequency counters from 1955 to 1970 had fast flip flops with catching and steering diodes. These guys are 30 years too late.

Don't forget; not everyone can know everything.

Time was, an engineer built bridges, ships, railways, tunnels and now? We have Analogue Engineers, Digital Engineers, Signal Integrity Engineers ... the list goes on.

:)

I remember seeing things like this:

in Practical Wireless and Wireless World from the 1960s and 1970s.

I did some searching for links to those techniques but oddly, couldn't find any.

Tee hee!

Just posted that last and the number of posts to the forum reached 555.

Very apt I thought!

Guess its just me, getting old, but you gotta wonder, did those guys that "came up" with the speedup trick in the 1980's, did they just not know that frequency counters and flip-flops have been around for 50+ years, and they had ones that could count at 10MHz in 1951? A pair of hot 6AH6's, six diodes, and 5% resistors, per flip-flop? Old-hat enough to be in the textbooks by 1955.

I know people who did not know what a turntable was or how vinyl is played or how a magnetic pickup works even to not knowing about the little bit of stone being rattled around in the orthogonally modulated groove.

But they know stuff about OS's, webbyness, programming, deep submircon fabbing and stuff I struggle to comprehend let alone keep up with.

If they don't know about things from 50 years ago then a couple of comments:

i) Just how much has happened in those 50 years that you need to know about as compared to how much is handy to know?

ii) Who's fault is it that they don't know? Who should have taught them?

In the meanwhile, here's a bit of that history:

Runs dog slow so be patient and hope the world hasn't moved on too far to catch up with it ...

:)

Come to think of it, how much detail do any of us know about the technology (or anything else) that happened maybe 30 years before we were born?

Maybe I should post a Feature Request for a "Philosophical ramblings from old gits" section of the forum.

If I were teaching young engineers, I'd suggest:

(1) Don't assume you're the cleverest thing ever, just because you have access to SPICE and 3GHz 8-core CPUs. You might have heard that rather fancy things, like A-bombs and moon missions were done with slide rules and paper and a few punched cards running on 1MIPS computers.

(2) Look up your idea on Google Patents. You'll probably find someone's done it before, and better too.

(3) Read and understand every issue of HP Journal. A huge amount of really neat technology explained in there. Some stuff has yet to be bettered.

... and get off my lawn!

Thanks! I changed on base resistor, switched on the skip initial an it works!