Talk:Multivibrator

Poor wording

the sentences here are awful. can someone with a good knowledge of the English language pls have a look at it? --zeno

Completely rewritten 5/11/03 - i hope it meets with approval from former contributors GRAHAMUK 23:05, 4 Nov 2003 (UTC)

Astable

I've expanded this section considerably. Sorry to delete existing text, but I think it had some problems, particularly references to capacitors discharging when they were actually charging. TerraGreen 20:10, 10 January 2006 (UTC)

The changes are fine by me, though as the author of the original text you removed, I am pretty sure it was correct. Perhaps it wasn't clear, but it definitely wasn't wrong. Graham 22:10, 10 January 2006 (UTC)

In state 1, C1 is discharging, not charging as the text now says. The voltage is lower on the right side of C1, and the current is going to the left, reducing the voltage across the cap. And C2 is charging. Pfalstad 23:03, 10 January 2006 (UTC)

I don't think so - C1 is charging. One problem at present is that the circuit (which I drew, so my apologies for this) shows electrolytic capacitors, which might be misleading - a real circuit shouldn't use electrolytics here because they are charged in both directions. However the reverse-bias charge only ever reaches ~0.6V, so it will probably function... anyway, as far as I can see the description of the operation of the circuit is correct both now, and in my original version. Graham 01:35, 11 January 2006 (UTC)

Do you agree that when Q2 turns on, then C2's right side is ~0v and C2's left side is negative? And when Q2 turns off again, both sides of C2 are ~0v? Pfalstad 02:33, 11 January 2006 (UTC)

Yes, I agree with the first part. Before Q2 turns on, the left side of C2 must be at ~0.6V, and the right side at +V. When Q2 switches, the right side is brought to 0v. The left side has no choice but to follow, so at that instant its voltage will be 0.6 - V, which will be quite negative. This switches off Q1... C1 now goes to +V on the left side. The right side tries to follow but will be clipped at 0.6V by the BE junction of Q2 - previously C1 was charged to 0.6V the other way, now it will be charged to +V via R1, rather rapidly, since R1 >> R2. But when Q2 turns off again, C2 will be 0v on the right, and +0.6V on the left, which is what turns Q1 back on. It has charged to 0.6V via R3. This is the charging that the description is referring to - the important one, since it's what the time constant comes from. I think I can see what you're getting at actually - both capacitors are in fact 'charging', but at different rates and in different directions. That makes it hard to really talk about charging without specifying what direction we are talking about. The charging that sets the time constant is as the text describes it - the charging in the other direction isn't mentioned, which is sort of OK since it is really just getting rid of the previous charge and doesn't change the circuit behaviour. You know what would help? Some annotated waveforms and perhaps current arrows around the circuit. I think this seemingly simple circuit is actually a lot more subtle than it looks, and understanding it isn't always that easy. Graham 06:40, 11 January 2006 (UTC)

Well since C2's + side is on the right, it's got +V-.6 across it when Q2 swiches on and -.6V across it when Q2 switches off (going through 0V on the way), so it seems to be discharging when Q2 is on. Pfalstad 23:43, 11 January 2006 (UTC)

I'm having trouble understanding what you mean - please bear with me! When you say "when Q2 switches on" are you talking about it having just switched on, or just about to? This is where I think a steady-state explanation for the circuit falls down - it's a very dynamic circuit with everything changing all the time, so you have to be very precise about exactly which moment in time you mean. So, from what you say about the voltage on C2, I assume you mean the moment just before Q2 switches on. When Q2 switches on, the existing charge (+V - 0.6) on C2 is first dumped - so yes, it is discharged at that moment, quite rapidly. It then goes on to charge, more slowly, in the other direction via R3 to +0.6V, and it is this charging which matters. The fact that it is a charge opposite in sign to the earlier charge doesn't to my mind make it a 'discharge' - the capacitor is still storing charge, just of opposite sign to its previous state. As I said, showing electrolytics here doesn't help, since we are not used to the idea of charging these in reverse... I really must fix that I think, and maybe add some waveforms as well. Graham 00:16, 12 January 2006 (UTC)

The voltage across C2 is the same before and after Q2 switches on. When Q2 switches on, the absolute voltages on either side of C2 change rapidly, but the relative voltage across C2 does not change at all. At least that's how it works in the simulation I am looking at. C2 discharges rapidly? How rapidly.. What's the time constant, what does the current look like, etc.? Where is the current coming from to make this happen? Pfalstad 00:33, 12 January 2006 (UTC)

On further thought, I think I see why you want to call it a discharge... When Q2 switches on, the LEFT side of C2 goes negative (0.6 - V), but since the base of Q1 is reverse biased in this condition, it does look like a high impedance... so C2 ramps up from ~-V to 0v to +0.6v via R3... so yes, I would agree that you could call this a discharge for the most part, since it is a gradual reduction of negative charge... I just tended to look at this as a charge with opposite sign, but perhaps a discharge is more accurate. Urk, my head is spinning... Again, showing the waveforms will cut through all of this - perhaps a text explanation isn't such a great help. Does your simulation feature the ability to create a chart of the circuit waveforms? Graham 00:38, 12 January 2006 (UTC)

Well this could be prettier, but here is what the waveforms look like: [1] Green or white = voltage, yellow = current. Pfalstad 02:18, 12 January 2006 (UTC)

Actually, there is an error, though it's not the one you mention. It states "Q1 base at slightly below the supply voltage...". That can never be. Q1 or Q2 base can never exceed 0.6V, because the B-E junction of the transistor "looks like" a diode to ground. I think it will be best to simply remove those lines. Graham 01:39, 11 January 2006 (UTC)

OK, another little problem (I should have checked the change more carefully!). One of the keys to the astable operation is that when the circuit changes state (Q2 turns on, say), the sudden change at the collector from +Vs to ~0v is transmitted via C2 as a negative pulse to the base of Q1, thus rapidly switching it off. Otherwise there is no obvious reason why Q1 should switch off when Q2 switches on. This probably needs adding to the description of the operation of the circuit - I'll have a go. Graham 01:45, 11 January 2006 (UTC)

Thank you for the above explanations. My apologies to Graham if I actually introduced more problems when I thought I was "correcting" the text. I can now see where I was going wrong. I was thinking of the base of the transistor as being high impedance. Of course this is wrong, you are right the BE junction will look like a diode so will conduct if the base voltage exceeds 0.6V. That error was what led me to think the capacitor will be discharging, when it will in fact be reverse-charging as you say, and I think the other problems in what I wrote are all follow-on consequences of assuming zero base current. Sorry, it's too long since I've done any electronics, and maybe it was foolish of me to make such a major edit when I'm so rusty... Anyway, I wholeheartedly agree with you that the best thing would be some annotated plots of waveforms. If you are willing to do that, then that would be fantastic. TerraGreen 22:30, 11 January 2006 (UTC)

Another thing which will almost certainly still need changing is what I wrote about the output voltage during state 1, as this too was based on the assumption of zero Q1 base current. But I'll leave it to others to make the change, because I'm now getting muddled when trying to think about it. TerraGreen 23:01, 11 January 2006 (UTC)

Ah yes, if you assume the transistor has no base current then the working will seem rather different. In fact, I'm not 100% sure, but if that were the case I don't think it will oscillate - both transistors would just end up switched on. You can visualise this by putting enhancement-mode MOSFETs instead of bipolars for Q1 and Q2 - a stable state would be both Q1 and Q2 on, no oscillation. C1 and C2 would be both fully charged and there is no discharge path available to them, as it is the base current in the bipolar case that provides this. Graham 00:16, 12 January 2006 (UTC)

Please repair the Astable Multivibrator Circuit section. First paragraph speaks of "suggested values" when there are no "suggested values". Much further down the article are the "suggested values". 14:10, 19 February 2010 (UTC) —Preceding unsigned comment added by 68.11.173.188 (talk)

I've read this circuit, and I was a little confused as to why "C1" charges first. I eventually understood that the minute differences in resistors or capacitors are what causes one capacitor to charge first. If I'm wrong about this, please clarify - and would it be possible to include that in the article? Thanks. — Preceding unsigned comment added by 172.129.110.171 (talk) 01:54, 16 January 2013 (UTC)

Theory

Can someone take a look at this? I am willing to rewrite it but I don't know when I can get to it. I suggest just deleting the explanation of how an astable works if you can't nail it down.

The BJTs are current-operated devices. It is the displacement current through each cap and its connected base that switches Q1/Q2 on and off, not the base voltages going above and below Vt. The bases will nearly always be at the junction potential. You can build this circuit with FETs, but it requires some modification to make the transistors shut off. The way the description is currently worded suggests that the resistors must be tuned somehow to bias the bases around Vt. 69.207.129.45 (talk) 12:39, 1 October 2010 (UTC)

I agree with you that the explanation section is poor, misleading and somewhere even wrong. For example, the concept "R1 and R2 serve as parallel resistors feeding the base of Q2...and the resistance of the two parallel resistors R1+R2 is LESS than R2 alone" is misleading. The sentence "R1 charges FASTER than R3, causing Q2 to switch off before reverse charging can occur" and then "This leaves insufficient voltage to keep the base voltage of Q2 above the 0.6 V threshold and Q2 switches off" are wrong. R1 just restores C1 charge through the Q2 base-emitter junction but R2 is what keeps Q2 switched on. So, R1 is small enough to do this work (to provide a sufficient base current) and Q2 stays switched on until reverse charging occurs. Charging C1 (C2) is many times faster than discharging C2 (C1). Circuit dreamer (talk, contribs, email) 17:48, 1 October 2010 (UTC)

I would like to replace this "Basic Mode of Operation" section with a paraphrased version from the following source: http://www.electronics-tutorials.ws/waveforms/astable.html. The author of the content has emailed me his permission to do so, so long as an inline reference points back to his website as the source of the explanation. (Which is what Wikipedia prefers anyway.) What's helpful about his explanation is his description of the states of "plate A and B" for capacitor C1. In general his write up if far more explicable than the current, rather skeletal list of bullet statements. Take a look at his explanation and let me know. Dgiroux (talk) 18:23, 17 October 2010 (UTC)

I have read the explanation but I disagree with the first part, namely:

"Assume that transistor, TR1 has just switched "OFF" and its Collector voltage is rising towards Vcc, meanwhile transistor TR2 has just turned "ON". Plate "A" of the capacitor C1 is also rising towards the 6 volts supply rail of Vcc as it is connected to the Collector of TR1. The other side of capacitor, C1, plate "B", is connected to the Base terminal of transistor TR2 and is at 0.6v because transistor TR2 is conducting therefore, capacitor C1 has a potential difference of 5.4 volts across it, 6.0 - 0.6v, (its high value of charge). The instant that transistor, TR1 switches "ON" plate "A" of the capacitor immediately falls to 0.6 volts. This fall of voltage on plate "A" causes an equal and instantaneous fall in voltage on plate "B" therefore plate "B" of the capacitor C1 is pulled down to -5.4v (a reverse charge) and this negative voltage turns the transistor TR2 hard "OFF"." Circuit dreamer (talk, contribs, email) 20:02, 17 October 2010 (UTC)

Merge

Someone suggested merging the stubs into this article. This is a good idea. --Wtshymanski 02:20, 27 July 2006 (UTC)

Definition

A definition is mysteriously not there. A multivibrator is not defined as something thats used to implement a variety of simple two-state systems. This needs to be more explicitely defined in the intro. Fresheneesz 22:20, 25 October 2006 (UTC)

3-way

There's an interesting circuit here: [[2]] (about 2/3 down the page), which uses 3 transistors to have 3 states. Would be worth addin

Monostable

I added details to the monostable multivibrator and redefined it with One-Shot as well. I think a One-shot link should be created to redirect to this page if they are looking for the electronic definition of a one-shot. Dante 16:26, 22 June 2007 (UTC)

It's kind of a strange schema for the monostable, applying the trigger to the base of Q2 doesn't seem correct. Cmyalrm (talk) 18:29, 12 March 2010 (UTC)

Multivibrator equation

for the multivibrator equation f=1/(.693*R*C)

I'm wondering, is .693 more accurately represented by ln(2)? I'm no expert, just a hobbyist, but knowing that the capacitor equation is exponential, it seems to make sense.

?? Thanks in advance, and my apologies if I am mistaken. --Smiley325 20:23, 30 October 2007 (UTC)

Missing component values

Somewhere, somehow, apparently the component values for the circuit[s] were removed. It would be nice to give practical ranges for them.Nikevich (talk) 09:32, 24 November 2009 (UTC)

Derivation

Moved from User talk: Oli Filth:

I saw that you deleted a section calling it "...one particular example circuit". Isn't this a derivation of the time-constant? It would be nice to have the derivation in the article somewhere. Mikiemike (talk) 01:29, 28 December 2010 (UTC)

It's the derivation of the time-constant for one particular implementation of one particular type of multivibrator. IMHO, this derivation bloats the article with tangential information. Regards, Oli Filth^{(talk|contribs)} 09:42, 28 December 2010 (UTC)

I see you've restored the material, and you seem to have misquoted me in your edit summary! I don't claim that this is "an example", I claim it's the derivation for one particular example circuit. This derivation offers little insight into multivibrators or oscillators in general, so I don't see much purpose in it being here. That is why I removed it. Your new section headings don't help this situation! Oli Filth^{(talk|contribs)} 16:51, 28 December 2010 (UTC)

I have added a link to RC circuit (time domain). Maybe this is a more suitable place for this derivation? You have considered a classical problem - determining the charging time (0,7RC) of a capacitor to half the maximum voltage. Circuit dreamer (talk, contribs, email) 18:17, 28 December 2010 (UTC)

Dubious tags

The general flavor is WP:NOR.

This article makes several general statements about MVs based upon specific instances. The results are often false. Multivibrators must have two passive networks. Multivibrators must have a capacitor.

Ref 1, the etuttle blog, is not WP:RS.

Glrx (talk) 05:07, 24 January 2011 (UTC)

Yes, the article has been edited a lot lately; perhaps winding back a month or so would be a better starting point. Lecture notes don't make the clearest encyclopedia style. --Wtshymanski (talk) 05:35, 24 January 2011 (UTC)

Excellent work! It's a good idea to limit the scope of multivibrator only to the classical cross-coupled transistor pair. Circuit dreamer (talk, contribs, email) 14:18, 24 January 2011 (UTC)

Articles are always simpler when they stick to their subject. "Hysteretic" circuits, Christmas light flashers, and ballroom dancing are not multivibrators. If it's not a cross-coupled pair of amplifiers, it's not called "multivibrator" in the books. Wish I could find a clear description of relaxation oscillator but as usual that article has no references either. --Wtshymanski (talk) 14:54, 24 January 2011 (UTC)

It does now. --Chetvorno^TALK 22:04, 2 December 2016 (UTC)

Name

"The name "multivibrator" was initially applied to the free-running oscillator version of the circuit because its output waveform was rich in harmonics"

I always thought it was because earlier oscillators were electromechanical, i.e. they were literally vibrators. Gigs (talk) 17:53, 22 May 2012 (UTC)

Two gain devices

Is there a source that multivibrators must have two gain devices?

Why not one? Plenty of relaxation oscillators manage with that. Is there a sourceable statement that they can be excluded from this scope?

Why not five? Take an (odd) handful of inverters with a propagation delay, couple them into a ring and away they go. Again, is there a sourced statement that this cannot be counted as a multivibrator? Andy Dingley (talk) 15:14, 25 September 2012 (UTC)

Sadly, IEEE Std. 100 "Authoritative Dictionary of IEEE Standards Terms" has let me down because it says, in part

multivibrator A relaxation oscillator employing two electron tubes to obtain the in-phase feedback voltage by coupling the output of each to the input of the other through, typically, resistance-capacitance elements.

If they'd said two "gain devices" we'd be all set.

An (odd) number of inverters in a ring is a ring oscillator and is the Hello, World! of any new IC fabrication process. (IEEE also needs to update the definition of "ring oscillator" out of the vacuum tube age.) --Wtshymanski (talk) 20:54, 25 September 2012 (UTC)

All is lost. Horowitz and Hill think a 555 is a multivibrator, as do innumerable Tata-McGraw Hill authors. --Wtshymanski (talk) 21:20, 25 September 2012 (UTC)

So what are you going to do? Change the claim that the only multivibrator is the two device form, or cling defiantly to the IEEE and exclude the transistor ones too? Andy Dingley (talk) 21:25, 25 September 2012 (UTC)

Ohh, yes, that'll teach that Shockley-loving crowd. How about we leave it like it is and say two amplifiers cross-coupled with RC networks? Surely the important part of the IEEE standard definition is two RC coupled stages. With all the great folks on the Wikipedia working on it, I'm sure a suitable reference is available. Of course, common sense is not a Wikipedia policy. --Wtshymanski (talk) 02:33, 26 September 2012 (UTC)

Why should we entertain the idea that oscillators with <>2 gain devices are not multivibrators? A multivibrator (IMHO) has always been a functional description of what it did, not how it does it. There was no implication as to the circuit used, or number of devices. Admittedly the etymology used is obscure, because I recall pre-war valve circuits termed multivibrators, pre-dating the vibrating reed and similar power supply choppers. Just what does "multi-" mean here? References would be welcome here, but I see no good cause (and a single IEEE reference that also insists on valves doesn't cut it) to limit the definition to 2.

As a question of writing, I'd be happy to describe multivibrators as the two-device form, so long as we footnote the others and don't get all ontological about excluding them. Andy Dingley (talk) 10:04, 26 September 2012 (UTC)

What a great plan for simplifying the encyclopedia! Instead of all those confusing articles, let's just put everything that has four legs and fur under "Kitty". Or don't you think there's some educational and philosophical value in discriminating between a 555 and the classic two-device multivibrator? Let alone every other kind of relaxation oscillator (another term that I've yet to see properly discussed). I think the two-active-device cross-coupled topology needs its own article and all the other kinds of oscillators can be described elsewhere. It's not great nomenclature, but "multivibrator" is what people call it, not "that Eccles-Jordan-Abram-Bloch thing that makes square waves". --Wtshymanski (talk) 13:56, 26 September 2012 (UTC)

"I think the two-active-device cross-coupled topology needs its own article "

Fine, but maybe it can't sit quite so simply under the name multivibrator.

Maybe it can, but the article text shouldn't be quite so firm that circuits with other than two devices are no longer multivibrators. Andy Dingley (talk) 14:02, 26 September 2012 (UTC)

• 'Oppose two gain device / cross coupled / only capacitor definition. 555 supra. The one opamp / one capacitor / three resistor circuit is an astable multivibrator. I consider the blocking oscillator to be an instance of a multivibrator (as Harley is an instance of a linear oscillator). Also neon bulb relaxation oscillators, unijunction oscillators, and tunnel diode logic. I'd lump in a buzzer, too. Of course, somebody could define the term to be anything. Some dictionaries do the classic two device version, but others emphasize the harmonic content. To me, MVs involve sudden state changes. Glrx (talk) 17:58, 2 October 2012 (UTC)

== Comments on the article ==I am new to this so please excuse my lack of fines in suggesting a change to the section for the Astable frequency calculation.

In solving the log equation for t assumption is made that Vbe << Vcc. Since Vbe~0.6v & Vcc = 5V this is a too big assumption. However by including Vce_Q2(on) into the equation there is no need for the above approximation. Thus

Vcap(t) = Vbe_Q1 - Vce_Q2, 
Vcapinit = Vbe_Q1 - Vcc + Vce_Q2

Since Vbe ~ 0.6V & Vce_Q2 = 0.2V Vbe_Q1 - Vce_Q2 = 0.4V while Vbe_Q1 + Vce_Q2 = 0.8V which creates a ratio of 2 to Compliment Vcc/2Vcc Thus the equation naturally comes out at solution Ln (2)

1. "This circuit is also known as a one shot."

The sentence should end with a noun after the adjective and "one shot" should probably need a hyphen.

2. "Thus C1 restores its charge and prepares for the next State 1 when it will act again as a time-setting capacitor...and so on... (the next explanations are a mirror copy of the second part of Step 1)."

This is not a good sentence. The part from the first three periods should be fixed.

3. I disagree with the use of Vcc. It's inconsistent with Figure 1 which has V+. If Vcc is preferred, then the figures should be modified. I propose the second option.

4. "RB < β.RC"

Is that dot really meant to be a *? If it is there should be a a properly written equation for the expression.

5. I do not understand why the monostable multivibrator, aside from having a description, also has a link to a separate article explaining how the circuit works that is pretty much the clone for the same. I don't even think the article should exist since this article covers the monostable vibrator anyway.

6. It would definitely help to have actual values of R and C for simulation purposes to verify the functionality of the circuits.

ICE77 (talk) 02:13, 28 July 2015 (UTC)

re: 3) Vcc is the correct term, and consistent with the standard nomenclature Vce, Vbb, Vcer, Vces, Vceo etc, (for example V+ is obviously wrong with PNP transistors) , on the same topic it should mentioned that when Vcc exceeds about 7v, you will get zener action on Vbe, specifically when Veb exceeds V_BES this will make the MV run faster than theory, Note Veb measures the same node as Vbe except that Veb has the emitter positive and the base negative (i.e. so Veb < 7v or Vbe > -7v are equivalent, the order of the subscripts catches many new EE's out) Salbayeng (talk) 04:08, 14 February 2016 (UTC)

More comments

This article really needs a diagram showing the waveforms , it's much more obvious when you can see the base potential being negative for a long time, and ramping up to the trip point. like on this page schoolphysics.co.uk
"Multivibrator" definition is probably lost in history. You can make one with a pair of RC cross coupled relays, possibly someone did this back in the 1800's but failed to publish a web-page. While you can make a "vibrator" with just one relay Trembler_coil (Model T ford), two make it more accurate as a timer.
I would definitely define a multivibrator as requiring two cross coupled RC stages and two active inverting elements noting the classic multivibrator has two complementary outputs (this might not be a necessary requirement?) so for example you can alternately flash two light bulbs (as on a railway crossing)(a 555 can't do that). With two RC stages you can set each of the two delays independently of the other (a 555 can't do that). The 555 is definitely not a multivibrator, it belongs to the class of integrator + schmitt trigger oscillators, and has only one capacitor there are two basic topologies for the 555 , producing approximately sawtooth (classic relaxation oscillator) or triangle wave timing waveforms, the 555 explicitly sets the threshold voltages.
The other thing that isn't mentioned is that the output voltages on the collecters have a soft leading edge, and this may be undesirable in some applications (transistors get hot).
Historically, I have located Charles Adlers probably first patent on alternating flashing railway lights US patent 1622018 and this is not a multivibrator, just a mono-vibrator with a heavy weight. So no real evidence for my earlier suggestion of cross coupled relays being the precursor to multivibrators. Salbayeng (talk) 02:34, 14 February 2016 (UTC)

Consider updating with information from the Polish version of this article

Hi Has anyone considered adding some of the content that appears on the polish version of this article https://pl.wikipedia.org/wiki/Multiwibrator ?. It is about twice as long, and has more schematics including the triple-vibrator. Salbayeng (talk) 04:19, 14 February 2016 (UTC)