Re: Astable Multivibrator help!



Tam wrote:

"Fred Bloggs" <nospam@xxxxxxxxxx> wrote in message news:4799E924.3070409@xxxxxxxxxxxxx



This is supposed to be simple, and I've built these in the past (10-12
years ago in school), but I'm having a heck of a time getting it
working now.

I'm following this schematic: http://www.qrp.pops.net/images/FLASHER.GIF

The components are exactly the same, and I've quadruple checked the
orientation of everything, swapped components with new ones, and torn
down rebuilt it countless times, and all I can get are two lit LED's
at the same time, both with between 1.6 and 1.7 volts across the
LED's.


The circuit may be locked up with both transistors stuck in saturation.


Can anyone give me any ideas on what my problem might be? They should
be blinking at a rate of about 1 cycle per second. I'm really at a
loss, and it could be because I've spent too much time working on it,
but each time I sit down to work on it, I end up going in circles.


In order to guarantee oscillation, you need to add a diode logic OR function which provides base drive only if one OR the other of the transistors is OFF, like shown below, a lock-up cannot occur because there will be insufficient base drive. You get better switching by placing the LEDs in the emitters, and you get around the breakdown problem by reducing the voltage through a common resistor in series with the battery lead, in this case it is a R1xIled=150x0.015=2.25V drop. The LED currents are the same, about 15mA, because the total resistance in series with Vbatt is the same. The frequency is about 2Hz:
View in a fixed-width font such as Courier.

.
. Vbatt
. |
. [150]
. R1|
. .----------+----------.
. | |
. [330] [330]
. R2| D1 D2 R3|
. +---|>|-+-----+-|<|---+
. | 1N914| |1N914 |
. .---+ | | +---.
. | | [39k] [39k] | |
. +| \| R4| |R5 |/ |+
. === Q1 |-----+ +-----|Q2 ===
. 10u|C1 <| | | |> C2|10u
. | | 2N3904| | 2N3904| |
. '---|-------|-----' | |
. | '-------------|---'
. | |
. |LD1 LD2|
. --- ---
. \ / ~~ \ / ~~
. --- ---
. | |
. '----------+----------'
. |
. com


I got bored and ran an SWCad simulation on this. Oddly, with the two diodes, it takes the circuit about 20 seconds to start up. The 150 Ohms has no effect on this. If you just connect the 39Ks to the supply, startup is about 200 ms.

You can't go by that, the circuit will fire the LEDs up immediately and start the alternation in one or two seconds. Try imbalancing the circuit a bit by changing component values within say 1% on either side and then see what happens.


With no diodes, and no 150 Ohms, the 3905 VBE reverse is 5.9V. About 4V with the 150R

Even at 9V, which I assume means an alkaline battery that will spend most of its time in the 7.2-8.4V range, you have the caps charging to 9-2.25( the 150 ohm v-drop)-2.2 ( the Vbe + LED drop of the ON transistor) for 9-4.45=4.65V. Then the OFF base sees this minus the Vbe+Vled (2.2V) of the ON transistor for 2.2-4.65=-2.45V, and this is what my SPICE sim shows the reverse peak to be.
Another feature, which is more of an accident than anything else, is the oscillation period is dead on within a few milliseconds for battery variation in the 9->6V range. The LED current goes from 15mA to 8mA though, so it does dim slightly.
The diode ORing is nothing new, people have been using it for over half century. And placing the LEDs in the emitter allows you to increase the ac-loop switching gain due to their relatively large capacitance rather than decreasing it with the collector placement, the drawback is that 39K is just about the maximum base resistance you can use because the base junction requires about 1.8V to fire things off.

.

Loading