Listen to Latest SolderSmoke Podcast

Tuesday, April 20, 2010

Astable Multi-NON-Vibrator (Help!)

Having made good progress on my Ugly, Discrete Component, DC recever for 30 meters, this week I turned my attention to the little QRSS transmitter that I mentioned last week. The idea is to run this thing out at the country house. The inspiration here comes from Hans Summer's ultra-simple QRSS transmitter:

I ran into trouble this morning. The oscillator is running fine. But the little two transistor astable multivibator that is there to put some FSK on the signal is not vibrating. What am I doing wrong here? Instead of the BC-107s or 108s used by Hans, I'm using 2N2222s. Hans used 22uF caps. I didn't have any, so I just put two 10uF electrolytic caps in parallel. But nothing happens. It just sits there. Any ideas? Hans was running his rig off bleach batteries and was struggling to keep current low -- I don't have this constraint, so maybe by using lower value resistors to the collectors and bases?

Help me! Help me!

(I apologize in advance if this posting results in ads for other kinds of vibrators!)

8 comments:

  1. looking at the high values of the collector and base resistors.it is important to have low capacitor DC leakage and hi beta gain on the astable oscillator transistors.I would check the beta of the transistors first.(at least 150 or more)
    73 wayne VA7AT

    ReplyDelete
  2. Bill I agree with Wayne the gain of the 2N2222 transistors is probably too low. The 680K has to be able to drive the transistor into saturation for the oscillator to work. Without the 22uF coupling capacitors you should be able to measure if this is happening. If you have more than a couple of hundred mV on the collectors you are in trouble.
    You could reduce the 680K and 220K resistors but then you will have to increase the 22uF capacitors to maintain the same oscillation frequency.
    BC108/BC548 transistors are good ones for the junk box and lots of projects as they have reasonable gain and are inexpensive. I could send you a couple of BC548's but I doubt they would get to you in under a week.
    73's Mark VK6WV.

    ReplyDelete
  3. Bill me again. I have done a few calculation and it appears that even with the worst case gain of a 2N2222 that you should be able to achieve saturation and hence oscillation. I now suspect that Waynes second choise of leakage through the capacitors may be the problem. How about for a test replacing them with 0.1UF non electrolytics. The oscillation frequency will be way off however if the circuit does oscillate then you will at least know that the electrolytics are the problem.
    73's again Mark VK6WV

    ReplyDelete
  4. Hi Bill,

    This circuit changes output state about once every ten seconds. [T=ln(2)*680k*22uF]. I'm in agreement with other posters - with such a long period you should probably use low-leakage capacitors. Polystyrene and polypropylene (film or metal-film) caps have low-leakage, but at 22uF they're hard to find and expensive.

    In the two transistor circuit reducing the capacitance and increasing the resistance is pushing things. You might want to try using a smaller value metal film cap with a larger resistor in a 74HCXX logic-based CMOS astable multivibrator. There are lots of examples of these on the Web. You can use NANDs, NORs, Schmidt triggers, etc.

    For fun, go here are look for the file: astable1.asc

    http://tech.groups.yahoo.com/group/LTspice/files/%20Examples/Apps/

    This is a rigorous LTSpice simulation of a two BJT astable multivibrator. Add capacitor parasitics to the model (right-click on the capacitors) to see the effects with a long time constant.

    Other alternatives: Run the astable at a higher frequency and divide it down with a counter. Use a micro-controller, a little ATtiny 8-Pin PDIP part would be ideal for this, plus you can program it with a unique output sequence. Look at 555 timers and 7400 logic multivibrators. Get one of those dunking bird toys and make it toggle a flip flop; don't forget to keep the glass topped-up with water.

    Best 73's, David WB4ONA

    Hey my CAPTCHA word is "stable", too bad it isn't "astable"!

    ReplyDelete
  5. Bill -

    Here is a link to info on computing the free-running frequency for an astable multivibrator.

    http://www.qrp.pops.net/multivibrator.asp

    I am not comfortable with the paralleled capacitors and in troubleshooting, would substitute a single capacitor of 10uf or slightly less for each pair. They do not have to be matched in value, as it will affect only the on-off timing ratio of each transistor.

    A second comment - ususally the ratio of base resistor to collector resistor will be higher - on the order of 10:1 as mentioned at the web site.

    As long as the dc-beta of your transistors is greater than 100 (say about 150, or so), the selection of transistors should not be much of a problem - especially since you are not running either to a significant load.

    Best wishes -
    Bruce - KK0S

    ReplyDelete
  6. This may seem too obvious, but be careful with capacitor polarity. The "+" side should be connected to the collectors. If you reverse the capacitors, the leakage will be too high.

    73, Pablo EA4FUK

    ReplyDelete
  7. For some reason, I don't know why - I sometimes have trouble with astables when the cap/resistor combo is exactly the same value. In these cases touching the circuit, or grounding one sides collector momentarily may kickstart it.

    73 de KB8UFP Robert

    ReplyDelete
  8. Bill,
    You could try a simple 555 timer ic
    oscillator.

    ReplyDelete

Designer: Douglas Bowman | Dimodifikasi oleh Abdul Munir Original Posting Rounders 3 Column