555 Timer. Astable, 50% duty cycle, 100 kHz.

Thread Starter

dllu

Joined Aug 3, 2017
10
I need help wiring my 555 timer in an astable configuration with a 50% duty cycle at 100 kHz. Here is a link to the timer I am using:

https://www.digikey.com/product-detail/en/fairchild-on-semiconductor/LM555CN/LM555CNFS-ND/458696

I am applying a constant DC 6-volt input to the timer. I originally wired the timer as shown in this link:

http://www.learnabout-electronics.org/Oscillators/osc44.php

However, while I am getting an AC square wave from the timer, the duty cycle is much closer to ~66%. My resistor and capacitor values are:
R1 = 72 ohms
C1 = 10 nF
C2 = 100 nF
based on Figure 4.4.6 in the link above. Can someone tell me what's happening? Thanks.
 

dl324

Joined Mar 30, 2015
10,912
Welcome to AAC!

Make it easier for us to help you.

Post a schematic of how you have the timer wired, including what the output is driving.
 

Thread Starter

dllu

Joined Aug 3, 2017
10
I have attached a picture of my circuit schematic. The ultimate goal of the full circuit is to convert a 7.6 volt DC signal into a 2000 volt AC signal at 100 kHz. I actually have multiple issues with circuit, which I believe stem from the timer wiring, but may also stem from the MOSFET wiring. When I probe the output (pin 3) of the timer, right before the gate of the MOSFET, I obtain an AC square wave with an amplitude of approximately 6 volts, but the duty cycle is much too long. Any ideas on what the problem might be?
 

Attachments

crutschow

Joined Mar 14, 2008
25,102
The output cannot drive 72 ohms.
You need to look at the 555 data sheet.
Here are the output current limits:
upload_2017-8-3_16-16-12.png
So you need to increase R1 to at least 1.2kΩ.

Are you using that transformer as a flyback device, since that's how it's connected?
 

Thread Starter

dllu

Joined Aug 3, 2017
10
The output cannot drive 72 ohms.
You need to look at the 555 data sheet.
Here are the output current limits:
View attachment 132209
So you need to increase R1 to at least 1.2kΩ.

Are you using that transformer as a flyback device, since that's how it's connected?
Oh I see! I will change the value of that resistor. Does this mean I should also change the value of the capacitor to maintain a 100 kHz frequency?

Also, yes, I purposely wired the transformer in that fashion, although I have not gotten to testing that part of the circuit since my MOSFET is not currently working. I will try and fix the timer issue and get back to you on the transformer. Thank you!
 

MrChips

Joined Oct 2, 2009
21,281
For 50% duty cycle, set the 555-timer to generate a signal at twice the frequency and follow that with a toggle flip-flop.
 

Thread Starter

dllu

Joined Aug 3, 2017
10
For 50% duty cycle, set the 555-timer to generate a signal at twice the frequency and follow that with a toggle flip-flop.
Sorry, I'm a little confused here. How would I wire the 555 in that case? I would have to up the frequency to 200 kHz but what about the duty cycle, since that is also a parameter in consideration? Thanks for your reply.
 

MrChips

Joined Oct 2, 2009
21,281
Sorry, I'm a little confused here. How would I wire the 555 in that case? I would have to up the frequency to 200 kHz but what about the duty cycle, since that is also a parameter in consideration? Thanks for your reply.
You can set the 555-timer output to any duty-cycle. The flip-flop doesn't care. The output of the divide-by-two flip-flop will be exactly 50% duty-cycle.
 

dl324

Joined Mar 30, 2015
10,912
I will change the value of that resistor. Does this mean I should also change the value of the capacitor to maintain a 100 kHz frequency?
There was no need for the R and C to be so small. You mentioned 100kHz, but the values you chose would have given you almost 2MHz; which a 555 timer can't do.

Choose some sane values, keeping in mind that the tolerance of capacitors will limit your frequency accuracy more than the resistor.

Also keep in mind that the configuration you're using isn't guaranteed to give you a 50% duty cycle. It depends on the output voltage swing and worst case for HIGH output is VCC-2.25V. For a 50% duty cycle, it's better to use @MrChips suggestion of doubling the frequency and using a flip flop.
 

AnalogKid

Joined Aug 1, 2013
8,486
Two things make your schematic not 50% duty cycle. One, as above, the 555's effective output impedance, which is in series with the timing resistor, is not symmetrical. That is, it has two different values, one when pulling up and a different one when pulling down. Second, the 555 output voltage relative to the two power rails is different for the two rails. With a relatively low operating voltage, this is probably the larger contributor to your error. The output stage can swing pretty close to GND, but only to within 1 V (or more) of Vcc. So even though the internal resistor string sets the trip points to almost exactly 0.333 and 0.667 x Vcc, the output swings to GND + 0.05 V and Vcc - 1.5 V (working from memory). When the output is pulling down, the sink voltage is 0.05 V, but when it is pulling up the source voltage is 4.5 V. To be symmetrical it should be 5.95 V. This difference in driving voltage changes the charging current to/from the timing capacitor, and makes for an asymmetrical output waveform.

Figure 4.4.8 on your link shows how to modify the circuit to compensate for this. You probably do not need R1 and R2.

ak
 

Alec_t

Joined Sep 17, 2013
11,387
The output stage can swing pretty close to GND, but only to within 1 V (or more) of Vcc
.... and that will limit your choice of FET. The gate voltage will be only about 4.7V, which is insufficienet to turn many FETs on fully. You would need to use a 'logic-level' FET.
 

Thread Starter

dllu

Joined Aug 3, 2017
10
Why do you need 50% duty-cycle?
Your fly-back circuit doesn't need it.
Hi, is it possible that you could clarify? I'm quite new to circuits in general so I'm not sure I understand. Since I am trying to produce an AC waveform, I want the output to be symmetrical, right? So wouldn't a 50% duty cycle be necessary in this case? I guess I'm not familiar with fly-back circuits, so I'm not sure.
 

Thread Starter

dllu

Joined Aug 3, 2017
10
Is anyone able to suggest a design or create a rough schematic on how they would convert a 7.4 V DC into a 2000V AC at 100 kHz? In the end, I would like to implement this circuit on a small PCB to fit within the device I am making. I think if I see an alternative design, I may get better insight on how this circuit works.
 

Thread Starter

dllu

Joined Aug 3, 2017
10
There was no need for the R and C to be so small. You mentioned 100kHz, but the values you chose would have given you almost 2MHz; which a 555 timer can't do.

Choose some sane values, keeping in mind that the tolerance of capacitors will limit your frequency accuracy more than the resistor.

Also keep in mind that the configuration you're using isn't guaranteed to give you a 50% duty cycle. It depends on the output voltage swing and worst case for HIGH output is VCC-2.25V. For a 50% duty cycle, it's better to use @MrChips suggestion of doubling the frequency and using a flip flop.
Can I ask what equation you are using to find the frequency? The one that I am using (frequency = 1/(1.4*R*C)) with a resistance of 72 ohms and 0.1 microfarads gives me approximately 100 kHz.
 

bertus

Joined Apr 5, 2008
20,564
Hello,

Make the capacitor 1000 times smaller and the resistor 1000 times larger to have the same frequency.

Bertus
 

AnalogKid

Joined Aug 1, 2013
8,486
The 555 datasheet has recommended limits on the resistor values. As bertus indicated, once you have the basic values you can scale them up or down to find reasonable real-world values.

And:
convert a 7.4 V DC into a 2000V AC at 100 kHz.
Why? What is this for? That voltage and frequency can cause a severe electrical burn.

ak
 

Thread Starter

dllu

Joined Aug 3, 2017
10
The 555 datasheet has recommended limits on the resistor values. As bertus indicated, once you have the basic values you can scale them up or down to find reasonable real-world values.

And:

Why? What is this for? That voltage and frequency can cause a severe electrical burn.

ak
I am making a portable low-temperature helium plasma ionizer for my summer internship project. I generally stick to the mechanical parts, but I was struggling to make this circuit. In order to fully ionize the helium, I have to send an AC waveform of at least 1 kV to one of the electrodes. So this is what this circuit is for.
 
Top