SCR ring counter - sanity check please?

Go To Last Post
17 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I've done some modifications from the circuit I posted the other day in any other business... basically, turned the whole thing upside down so it can have an active high count which I need later. I've never used an SCR so any advice gratefully received; the device in question has a current limit of 200mA; trigger current of 200uA, trigger voltage of 0.8v.

Here's the theory:

At power on, none of the scrs are turned on because the gate drain resistors R1, 4, 7 etc keep the gates off. A high pulse via R32/C21/R11 triggers SCR1 on. At that point, LED 1 is lit and the voltage at the cathode of SCR1 is close to the positive rail. Along comes a positive going trigger pulse. At every SCR which is currently not conducting, it finds that it has a clamp to ground via the LED and limiting resistor (hmm, might need to move the LED above the SCR?) but at SCR1 it (a) takes the cathode above the anode (which is somewhat below the rail due to R31) thus turning it off, and also driving through the coupling cap C3 and turning on SCR2.

The more I think about it the more I think the LED has to be above the SCR...

And as for values - I haven't a clue. It needs to clock at 50H, so C1R1 has to have a time constant somewhat less, and I think R3 should be lowish compared to R2 - R2 is easy, of course, it's just the usual LED current limit, in series with R31. R1 can be quite high, I think. And I'm not sure quite what C2 is doing, save that there was one in the design I modified :)

Assuming a 15v rail (because I have a 15v power supply!) and 5mA for the LEDs gives a total load resistance of ~12/5ma ~= 2.5k - so let's make a guess at 2k load resistor R2 and 470 for the standoff resistor R31. Say 470 again for R3 and - finger in the air - 10k for R1? Sound sensible?

It's been too long since I did analogue...

Neil

Attachment(s): 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
It's been too long since I did analogue...
It's been too long since I saw a ring counter circuit!!! :shock:

From the 1969 GE transistor manual 7th edition. The zip file has a larger file for better clarity.

Attachment(s): 

John Samperi

Ampertronics Pty. Ltd.

www.ampertronics.com.au

* Electronic Design * Custom Products * Contract Assembly

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Me, I think I would take a Johnson counter (don't remember the 4xxx number) and drive the SCRs without bothering with turning the SCRs into a ring, themselves. Or, if it is not a 10-ring, I would make one out of JK flipflops and go from there.

Lazy? Maybe! Stuck in a rut? Maybe? Don't like SCRs? For sure!

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Ah Jim, but the idea is to do the job with SCRs - if I wanted to be efficient I'd just use an AVR to generate all the pulses. :mrgreen:

John, thanks, I'd seen those designs already; they're some of the few which can be found on the net.

Having slept on it, I'm more convinced it won't work. The design I modified it from is below (itself modified from a NASA bidirectional counter patent with no component values.)

Here, it's more obvious how the shift occurs; the sink resistor on the common cathodes holds the common rail probably a volt or two above ground. On the rising edge of the clock only the diode to the right of the scr which is turned on can conduct as all the others are reverse biased; that turns on the next SCR. At the same time, I think, the positive pulse turns off the first SCR... I'm not entirely sure. The NASA patent is here: http://ntrs.nasa.gov/search.jsp?...

Neil

Attachment(s): 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi Neil,
I smell a SCR clock coming.....
I think your counter could work, but when I made
experiments with my "transistor-trigger" I always had
to tweak the circuits a little or had forgotten some
mechanism. So you should probably test them.

I think certain tests could also be done using the
thyristor-tetrode connection of 2 bipolar transistors
making one SCR.

You may also have a look a patent DE1788067A, it
shows an "anode coupled" type of counter.

Or look at US3432684A which cites three other.
Which (free) patent service do you use ?
I use German "depatisnet".

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
they're some of the few which can be found on the net.
But I have the REAL book, $2.00!

Can I suggest that you just show the first 2 SCR and the last one with dotted lines in between?
It gets very hard to try and see what's happening with such a large image.

John Samperi

Ampertronics Pty. Ltd.

www.ampertronics.com.au

* Electronic Design * Custom Products * Contract Assembly

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Yeah, meant to do that, John, but the real world intruded :)

Can't find either of those patents, Ossi - obviously I'm doing something silly. Got a direct link? And yes, you do sniff correctly...

I'll get some TO92 SCRs in my next Farnells order; they're only eightpence each.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

try this one

http://depatisnet.dpma.de/Depati...

Can you order some more, just in case I want some ?

Would self-built devices from bipolar npn/pnp transistors also be allowed in your design ?

For dividing 50Hz to 1 Hz also unijunction
transistors could probably be used.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Got it now, thanks. Sadly, it's the same circuit; the active circuit goes low on trigger, the rest remain high. The transformer in the middle is interesting but doesn't really help me - I think I need a small transistor to get an inverted output. Heigh-ho, it's only a couple more resistors, a transistor, and another scr to be able to drive a nixie directly.

I suppose you'll want circuit diagrams if I make it work? :P

How many do you want? I'll probably get 25 - they're only 5.5 pence each then... problem is I need another eighteen and a half quids' worth of bits to get the free postage. Plus VAT, of course...

Self-built npnpnp junctions equally allowed - this is an open source design and anyone can play. But I seek elegance, and have to impose limits on myself; final plan is for everything possible to be surface mount, except the nixies, and maybe inductors (smps) or transformer.

Ossi, are you anywhere near Berlin on Sunday 21st or Friday 26, Sat 27?

Neil

Edit: Hmm, a thought: I wonder if I can put the nixie in directly as the anode load? Might need a stuffing big trigger pulse, mind...

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Could you use NE2 lamps in a 7segment layout for a modern/retro old is new design? I've never seen that, and if it is in fact an original idea, I hereby donate it to the public domain. (Aint I a nice guy?)

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Ain't you just!

I was thinking about using neons to test the drive, and there's no reason it shouldn't work... but you'd need to find a way of driving neons in series (lots of volts, probably) and of course to decode the decimal output to seven segment. Many diodes, I think - though it looks as though Ossi solved that one in his transistor counter.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi Neil,
on 28.th of September it's Berlin Marathon, so at that
weekend you probably can meet many many people in
Berlin, but not me.

I think for some experiments 10 thyristors for me
would do it.

If you need nixies, you may contact daqq, he sent me
some and probably has still some left.

A design using a sixty stage ring for seconds and
minutes and a 12 stage ring for hours. Each stage has an LED or neon. Then put them in big round circles and
you have the digital clock with analog-style display.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If you're going to do that, I think you should have a 60-lamp ring for hours as well; increment it every twelve minutes. :mrgreen:

Ask Jan can also provide nixies; ZM1210 are my favourite.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I got Z570M from daqq at a very good price.
I have also some ZM1100 and some smaller ones
NEC D8007 and JRC B-5853ST. In the
far futere perhaps I want to built a "pocket-watch"
using them.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

http://www.glowbug.nl/neon/94642...

http://www.glowbug.nl/neon/HowTo...

A similar BRY39 circuit appeared in ELEKTOR 1970.
I think BRY39 is hard to get, but with MPSA42 or 92
you could make high-voltage tetrodes.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Got a 'no trespassing' on the gif... I've seen the glowbug scr drive but that's not how I'm doing it; instead, feed the anodes with rectified 220v and drive the gate all the time I want the digit lit; it turns of fifty times a second and back on again next cycle if the scr is on.

I was on NeoNixie for a long time, before I was on here, though I haven't been back for a while:
http://www.nailed-barnacle.co.uk...
http://www.nailed-barnacle.co.uk...

Neil

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Just wondering:
Most of the nixie clocks I've seen have been running off of the power line frequency. The other end of the spectrum is running the nixie clock off of a caesium standard. That leaves quite a gap.

Not wanting to use modern quartz crystals (high frequency, annoying dividers, too ordinary), how about basing an oscillator on PAL piezo acoustic delay lines? For those who don't remember, they were (pre rohs) made using lead sulfide glass, achieving low tempco, and delay the signal by 64us.
No dinky little crystals that can't be exposed to air but real big, chunky slices of glass where you can trim the frequency by sanding the edges. I envision a monostable triggering itself through the delay line, with a NTC somewhere in the supply to compensate for remaining tempco.

It could even be phase locked to an external PPS signal using about 4 transistors..

/Kasper