LM555/LMC555 Timer Control Voltage Limits

[spec]

Hy all,

The other day I was doing a design which included an LM555 and LMC555 powered from a 12V supply.

I would have liked to have pulled the 'Control' pin down to 3.33V to alter the comparator references to 3.33V and 3.33V/2, but I could find nothing in the data sheet that defines the use of the Control pin or the voltage limits.

Does anyone have any information/experience in this area?

spec

 

[gophert]

Hy all,

The other day I was doing a design which included an LM555 and LMC555 powered from a 12V supply.

I would have liked to have pulled the 'Control' pin down to 3.33V to alter the comparator references to 3.33V and 3.33V/2, but I could find nothing in the data sheet that defines the use of the Control pin or the voltage limits.

Does anyone have any information/experience in this area?

spec

They are three 5k ohm in series. The risk would be in putting 8.66V through the top 5k resistor. The spec indicates that each can handle at least a 6V drop (inferred because 18v supply is allowed). I would also assume that the 5k resistor is not the weakest link and 8.66V is only a 50% above spec! I think it should work.

It will cost about $0.75 or less to find out.

 

[Les Jones]

Hi spec,
The schematic in the data sheet for the Nat semi LM555 shows the resistors in the divider chain as all being 5K ohm. I suspect this is just a nominal value. IF they are all 5K then I calculate that a 2.34 K resistor between the control pin and ground would give you the reference voltages that you want. (I've just realised that I did the calculation for 3.3 volts not 3.33 volts that you wanted.) I can't find the resistor values for the cmos version. The one schematic I found (TI) for the cmos version seems to use mosfets rather than resistors in the divider chain.

Les.

 

[MikeMl]

Take the 5K with a grain of salt. Since these are implemented in an IC process, they could be anywhere from a few thousand to tens of thousands of Ohms. They will, however, be well matched... I would take the 5K value as a minimum...

Setting the control voltage by shunting the internal resistors with a single external discrete resistor is likely to require several tries...
Not very temperature stable, either...

 

[spec]

Hy everybody,

Thx for your replies. I am aware that the two reference points are generated in different ways, depending on manufacturer and era of fabrication.

I should have been more specific: how low can you make the control voltage, by whatever means, and still have the chip function: basically the two comparators setting and clearing the bi stable flip flop.

Yes, the nominal value for the three equal resistors is 3K for the bipolar chip and 30K for the CMOS version. I would imagine that the CMOS version would be able to tolerate the lowest Control voltage.

You could test an individual chip and get an empirical answer, but the trouble is, the next chip you fit to the circuit may not be the same. Even the schematic you see in the TI data sheet, for example is notional, although it may be true for the TI (National Semi) version.

spec

 

[alec_t]

The datasheet specifies the maximum CONT voltage as Vcc but does not give a minimum value, implying that down to 0V is ok. Simulation, for what it's worth, shows that for Vcc=12V the 555 will oscillate with 2.1V<Vcont<=12V.

 

[gophert]

The datasheet specifies the maximum CONT voltage as Vcc but does not give a minimum value, implying that down to 0V is ok. Simulation, for what it's worth, shows that for Vcc=12V the 555 will oscillate with 2.1V<Vcont<=12V.

Thanks for the simulation result.

The cont voltage is connected to the base of a darlington used in a current mirror (w/10k emitter resistor) so I was going to say it will need something above the threshold voltage of the darlington plus a bit more to reach the threshold current of that darlington. I would have guessed something between 1.4 to 2.4V. Your sim says 2.1V. All is right in the world.

 

[schmitt trigger]

You could test an individual chip and get an empirical answer, but the trouble is, the next chip you fit to the circuit may not be the same. Even the schematic you see in the TI data sheet, for example is notional, although it may be true for the TI (National Semi) version.

spec

Exactly!

I once saw internal 555 schematics from four vendors. The differences were quite significant, and I would assume that when one takes into account the different fab processes, the differences would even be larger.
I feel that alec_t's simulation is as close as you'll get to a value. But your mileage may vary with different vendors.

 

[audioguru]

The schematic in the datasheet for the LMC555 shows 100k resistors in the divider. I betcha they match very well but have an awful tolerance (plus and minus 20%?).

 

[MrAl]

Hi,

A couple notes about using the control voltage pin.

First, once you use the control voltage the temperature characteristic of the circuit then depends entirely on the temperature dependence of the control voltage signal.

Second and for some applications a little more important, once you use the control voltage you loose the independence from power supply voltage variations. In normal operation the oscillation frequency is independent from the power supply voltage level, so in theory at 5v you'll get the same frequency output as at 10v. Once the control voltage pin is used to change the output frequency however, power supply variations begin to have an impact on the output frequency again due to the loss of the Vcc/3 discharge and 2*Vcc/3 charge symmetricalness of the charge and discharge threshold levels. As the power supply level changes so does the frequency, so some testing and/or simulation is a good idea. If your supply voltage does not vary that much it may not matter as much, but keep in mind that not all voltage regulator chips are perfect.

I did a plot on the control voltage vs output timing a long time ago and i think i posted to this site. Not sure where it is now but we can do that again if needed. One thing i remember is that it is not linear with control voltage input.

For a free running oscillator there are other ways to do it too which are not very hard to do, such as using an LM339 or similar and set up the three resistors yourself.

BTW, the CMOS version has the 100k resistors but i dont know if they are subject to change in any version, bipolar or CMOS.

LATER:
Here is a graph of the normalized output frequency vs normalized control voltage. Note that setting the control voltage to zero will mean there is no charge/discharge difference so it wont work well. This is for the 50 percent duty cycle version of the Astable oscillator connection (that's where the output drives one resistor to the capacitor).

 

[spec]

Hi,

A couple notes about using the control voltage pin.

First, once you use the control voltage the temperature characteristic of the circuit then depends entirely on the temperature dependence of the control voltage signal.

Second and for some applications a little more important, once you use the control voltage you loose the independence from power supply voltage variations. In normal operation the oscillation frequency is independent from the power supply voltage level, so in theory at 5v you'll get the same frequency output as at 10v. Once the control voltage pin is used to change the output frequency however, power supply variations begin to have an impact on the output frequency again due to the loss of the Vcc/3 discharge and 2*Vcc/3 charge symmetricalness of the charge and discharge threshold levels. As the power supply level changes so does the frequency, so some testing and/or simulation is a good idea. If your supply voltage does not vary that much it may not matter as much, but keep in mind that not all voltage regulator chips are perfect.

I did a plot on the control voltage vs output timing a long time ago and i think i posted to this site. Not sure where it is now but we can do that again if needed. One thing i remember is that it is not linear with control voltage input.

For a free running oscillator there are other ways to do it too which are not very hard to do, such as using an LM339 or similar and set up the three resistors yourself.

BTW, the CMOS version has the 100k resistors but i dont know if they are subject to change in any version, bipolar or CMOS.

LATER:
Here is a graph of the normalized output frequency vs normalized control voltage. Note that setting the control voltage to zero will mean there is no charge/discharge difference so it wont work well. This is for the 50 percent duty cycle version of the Astable oscillator connection (that's where the output drives one resistor to the capacitor).

Hy MrAl,

Thanks for data- yes I was aware of the implications of messing with the control voltage.

spec

 

[spec]

Hy,

Some interesting observations from you all.

Guess what, the answer to my question has been staring me in the face all along. It is on the datasheet- it is just that I have been reading it incorrectly.

The 'Control Voltage Level' shows the permitted control voltages, both for 5V and 15V supply lines. I had been reading this as the level that the control voltage would assume on this pin. It is not; it is the maximum range that the control voltage can be set too.

So, to answer the initial question, the lowest voltage that control can be made with a 15V supply is 9V and the lowest voltage that control can be made with a 5V supply is 2.6V. I bet, in practice, you could go lower than 9V with a 15V supply but that would be outside the specification stated low limit. Of course, the low voltage limit may be constrained by something as basic as the current flowing through the upper resistor in the divider chain.

Thanks once again for all your inputs.

spec

​

 

[MrAl]

Hy,

Some interesting observations from you all.

Guess what, the answer to my question has been staring me in the face all along. It is on the datasheet- it is just that I have been reading it incorrectly.

The 'Control Voltage Level' shows the permitted control voltages, both for 5V and 15V supply lines. I had been reading this as the level that the control voltage would assume on this pin. It is not; it is the maximum range that the control voltage can be set too.

So, to answer the initial question, the lowest voltage that control can be made with a 15V supply is 9V and the lowest voltage that control can be made with a 5V supply is 2.6V. I bet, in practice, you could go lower than 9V with a 15V supply but that would be outside the specification stated low limit. Of course, the low voltage limit may be constrained by something as basic as the current flowing through the upper resistor in the divider chain.

Thanks once again for all your inputs.

spec

​

Hello again,

Say whaaaaaa????

Where are you reading this spec from spec <chuckle> ?

The listing shows the possible OUTPUT voltage of the control pin doesnt it? It should be roughly Vcc*2/3 for any Vcc, so for Vcc=6v it should be 'around' 4v, but you can apply as INPUT any voltage from -0.3v to Vcc+0.3v without damage.

 

[spec]

Hello again,

Say whaaaaaa????

Where are you reading this spec from spec <chuckle> ?

The listing shows the possible OUTPUT voltage of the control pin doesnt it? It should be roughly Vcc*2/3 for any Vcc, so for Vcc=6v it should be 'around' 2v min and 4v max, but you can apply as INPUT any voltage from -0.3v to Vcc+0.3v without damage.

Hy MrAl,

When the Signetics NE555 timer first came out there was more documentation about using the control pin. Then National Semiconductors launched the LM555 clone and Texas Instruments subsequently acquired National. This is what is on the latest TI (National) data sheet (see link below):

(1) The Control pin is an input as stated in the first table in data sheet. I had forgotten this.

(2) If the Control pin is not used for frequency modulation, it should be decoupled to 0V by a 10nF capacitor, (no larger or the turn on current could damage the top resistor, no smaller because the decoupling would be insufficient). The statement in brackets has been removed from the latest data sheet, but a 10nF decoupling capacitor is still called up in the application information on the data sheet.

(3) When pulling the frequency by applying a voltage to the Control pin, the levels shown on the data sheet against 'Level' are guaranteed to work, but not outside those levels. This is what I think. A Phillips (Phillips bought Signetics) application report for the NE555 says that, 'a control voltage of between 90% and 45% of Vcc are safe, but wider voltage levels have been used successfully' (pretty definitive).

I'm not sure about putting -0.3V on the control pin, but that may be OK now if the delicate top resistor in the divider chain has been beefed up on modern 555s. It certainly was not the case . The limit on the value of the decoupling capacitor on the Control pin of 10nF tends to counter this though. Also, with a supply of 15V this would imply a current of 5mA through the top resistor (assuming a nominal 3K value) which in turn would mean a dissipation of 75mW which sounds a bit heavy for a substrate resistor. By the way, I see no information on the data sheet about allowing -0.3V or VCC +0.3V on the Control pin. I may have missed it though.

The point is, not what voltage can be safely placed on the Control pin without damage, but rather what voltage can be placed on the Control pin and have the 555 still function. In the application that lead to the original post, all that was required was for the 555 to operate on 12V but the control voltage be set to 0.666 *5V by a voltage source, to allow the 555 bistable flipflop to be toggled by a 5V square wave on the trigger and threshold pins.

There is certainly some confusion about the word 'Level' on the data sheet, which could mean input or output, but the data sheet quite clearly states that the Control pin is an input (first table). Also, the reason why I think it is an input is because if it were an output the wide range of voltages would imply an equally wide range of resistor values which would not make sense when the absolute frequency and mark to space ratio accuracy are considered. The other point though, is what possible use would the output voltage on the control pin be. On the NE555 data sheet the output voltage of the Control pin is give as around 0.666 * Vss +- 0.6V, but once again this error is not compatible with the other parameters listed on the NE555 data sheet.

The data sheet for the LM555 is particularly bad with only nominal values stated for critical parameters dictating absolute accuracy and deltas caused by changes of temperature and supply line voltage.

It would be interesting to hear any technical views/information about what I think above, because the information on the use of the Control pin is not that clear.

spec

spec's spec sheet reference :

https://www.ti.com/lit/ds/symlink/lm555.pdf

 

[MrAl]

Hello again,

That's very interesting, especially about the part of the limit on the size of the decoupling capacitor. What data sheet does that fact appear on?

The -0.3v to Vcc+0.3v comes from the data sheet, but i would not actually do that. In the 1980's i did apply 0v to +Vcc i am sure, but what i dont remember exactly is what the supply voltage was at the time. It could have been 5v to 15v, but probably more like 5v to 12v.

Also, where do you get the value of the upper resistor to be 3k instead of near 5k? Which data sheet is that on?
The CMOS versions have 100k BTW.

Obviously with a 15v supply and 5k upper the max power dissipation is about 45mw, but if that resistor can really be 3k id like to see that spelled out on a data sheet somewhere. I am thinking we could measure that too just to find out for at least one particular package.

The control voltage pin is both an input and an output because the pin is connected to the internal voltage divider. This means that by specifying the output voltage (nothing else connected) we know approximately what voltage we have to apply in order to NOT get any change in frequency. For example, if the supply is 6v and we intend to use the control pin that means the pin will be connected to some source of voltage, and if we dont want any change in frequency (from what the values give with NO control voltage applied) then we must supply 4v approximately to the pin. So to me the control level spec is what we might measure at the control pin after we turn on the power and probably with that 0.01uf cap connected.
Note that some of the spec's on that data sheet are 'existing' spec's, not specs that tell us what we can do or cant do. The power supply range is telling us what we can or can not do, but some of the other ones just state info about what the chip itself can do or does do. Like the accuracy on a crystal, we dont set that initial accuracy, the manufacturer just tells us that.

You will notice that the spec on the control voltage level only differs by about plus and minus 10 percent with a 15v supply voltage. It would be absurd to think that is the limit of the control voltage input. If that were true, then looking at the chart i provided earlier in this thread i see that a spec of plus and minus 10 percent would lead to a frequency change of about 1 or 2 percent, which would not do any application any good. Rhetorical: Who would want to use the control voltage to 'modulate' the output with such a tiny modulation index? I am not sure if that is what you were thinking, but just pointing this out anyway.

One other little point is that if the control voltage is close to zero, the oscillation will not be as stable because there is less differential voltage involved for the comparators. This means as the control voltage gets low the stability might be affected somewhat. I do think it will still work though, just not as stable.

I will look around some more also. One thing that always puzzled me about the 555 is that they never seem to show that frequency vs control voltage pin voltage graph like i posted earlier in this thread. It's like they arent sure themselves how it works

Are you perhaps interested in using a comparator chip to generate your frequency? This results in a circuit that we can specify more exactly and we end up with a very stable oscillator just like the 555.

LATER:
I accidentally posted the "control voltage vs oscillation period" graph instead of the "control voltage vs frequency" graph previously, here is the corrected graph (assuming you dont mind going from 0v to +Vcc with the control voltage). Since there are at least two different ways of doing the astable, i posted the circuit with it which also shows the internals of the 555 for convenience.

 

[MrAl]

Hello again,

I found this on one of the data sheets, although it uses a 5v power supply.

 

[spec]

Hy,

I too have been looking around for data on the 555 families: there is a lot of it. I had forgotten about the round tin can version that we used for the first few years.

As a result of the information in this ETO thread, and on the net, my personal conservative rules for using the control pin will now tend to be:
(1) Nominal value of resistors in chain: 5K Ohms
(2) Maximum input voltage (operate): Vcc-1V
(3) Minimum input voltage (operate): 1.7V
(4) Linear voltage to frequency range Vcntrl: To be defined, but probably a delta of +-1.5V

But, I needn't have bothered about the Control input levels. While searching the net, I came across this circuit which does the job of turning a 555 into a buffer with a low level drive, without any critical aspects, and for any supply line voltage: http://electronicdesign.com/power/lm555-makes-inexpensive-power-driver

Here are some links that may be of interest:
(1) http://en.wikipedia.org/wiki/555_timer_IC
(2) http://en.wikipedia.org/wiki/Hans_R._Camenzind
(3) http://www.semiconductormuseum.com/Transistors/LectureHall/Camenzind/Camenzind_Page3.htm
(4) http://www.taydaelectronics.com/datasheets/A-249.pdf (more definitive about parameters, especially control pin)
(5) **broken link removed**

Additional
(1) /1979_Signetics_Analog_Applications.pdf (an old friend and a treasured document from way back)
(2) http://archive.org/details/Signetics555556Timers (cute little booklet)

This data flowed functional block diagram, modified from the net, sums up the 555 nicely:

 

[MrAl]

Hy,

I too have been looking around for data on the 555 families: there is a lot of it. I had forgotten about the round tin can version that we used for the first few years.

As a result of the information in this ETO thread, and on the net, my personal conservative rules for using the control pin will now tend to be:
(1) Nominal value of resistor chain: 5K Ohms
(2) Maximum input voltage (operate): Vcc-1V
(3) Minimum input voltage (operate): 1.7V
(4) Linear voltage to frequency range Vcntrl: To be defined, but probably a delta of +-1.5V

But, I needn't have bothered about the Control input levels. While searching the net, I came across this circuit which does the job of turning a 555 into a buffer with a low level drive, without any critical aspects, and for any supply line voltage: http://electronicdesign.com/power/lm555-makes-inexpensive-power-driver

Here are some links that may be of interest:
(1) http://en.wikipedia.org/wiki/555_timer_IC
(2) http://en.wikipedia.org/wiki/Hans_R._Camenzind
(3) http://www.semiconductormuseum.com/Transistors/LectureHall/Camenzind/Camenzind_Page3.htm
(4) http://www.taydaelectronics.com/datasheets/A-249.pdf (more definitive about parameters, especially control pin)
(5) **broken link removed**

Additional
(1) /1979_Signetics_Analog_Applications.pdf (an old friend and a treasured document from way back)
(2) http://archive.org/details/Signetics555556Timers (cute little booklet)

This data flowed functional block diagram, modified from the net, sums up the 555 nicely:

View attachment 99656​

Hi again,

That's a pretty nice set of links for the 555 and other stuff too. Interesting reads from the past, and also the 'buffer' app was interesting too. I might have to look at that more closely for an I2C logic level translator which has to be bidirectional.

The linearity of the frequency vs control voltage can be seen by looking at the graph i posted a couple posts back. It's nearly linear from about 0v to about 85 percent of Vcc, but it may be too unstable at low voltages like below 1v so yeah some limit on that could be imposed just to make sure the frequency stays stable. If we get too low a lot of things can affect the output frequency.