Very low duty cycle oscillator circuit based on 7414 IC

  • 256 Views
  • Last Post 2 days ago
cd_sharp posted this 1 weeks ago

Friends, I'm going to try to make an oscillator with these requirements:

  • very low, adjustable duty cycle in range 0.01% - 0.1%
  • adjustable frequency above 50 Hz
  • use IC 7414

I will use as a starting point this circuit from Chris:

which translates to:

  • Liked by
  • Chris
  • Vidura
Order By: Standard | Newest | Votes
Chris posted this 1 weeks ago

Hey CD,

For Input Pins not used, good idea to Ground them. See Video:

 

I believe this is correct unless I have missed something simple:

 

 

Note, this is not enough to achieve your Goal, extra stages are needed.

See: 74HC14 Datasheet

I redrew the Circuit from Akula's Circuit:

 

 

When you replicate the Circuit, you will see some parts are redundant, not necessary. This is a handy circuit to have on hand for simple tests!

   Chris

  • Liked by
  • cd_sharp
  • Vidura
cd_sharp posted this 1 weeks ago

Thanks, Chris, but the circuit is a little bit different. I've put it directly in Circuit Wizard, exactly how I see it in the video, except the Pin13 connection to the ground, which the author neglected. It's more stable with it.

I prototyped it on a bread board:

It looks good, it's stable, goes from 200 Hz to about 90 KHz, and despite what's being said in the video, the variable resistor adjusts the frequency and the duty cycle at the same time.

Still, this is a step forward.

 

  • Liked by
  • Chris
Chris posted this 1 weeks ago

Hey CD,

Yes, you're right, the grounding is for stability.

Excellent work! I should say, the Diodes, they are acting as Directional Devices, the Resistance on one path is different from the other path, thus the Duty Cycle can be adjusted ever so slightly, but the video is not the best way to do it, as you noted!

The stages help, start with a Frequency Stage, push this on to another stage and start looking to adjust the Duty Cycle on another Stage, this way the Duty can be adjusted without affecting the Frequency. Because they are on different stages:

 

 

I hope this makes sense.

One of my Breadboard Prototypes: Notice no Grounded Inputs in my layout...

 

 

I started the same as you!

   Chris

  • Liked by
  • cd_sharp
Chris posted this 6 days ago

@CD

Re-reading my last post, this may have been a little confusing.

The simplest 74HC14 Circuit is as follows:

 

 

This uses One Hex Inverter, what I meant by a Stage..

 

 

The Diode is better if its a Signal Diode, they are faster than the 1N4001's faster the better. Replace the Diode with a Resistor of the same Value as R1 and this should change the Circuit dramatically.

I am sorry if I was confusing. Also the above Circuit, the Capacitor should be Ground one leg, Pin One the other leg. I will fix this and replace the image.

   Chris

  • Liked by
  • cd_sharp
cd_sharp posted this 6 days ago

Thanks buddy, it's a real pleasure to think how this stuff works.

This

is this:

Also, I found out and checked practically that this controls the duty cycle only:

These are the stages, but I didn't understand yet how they can work together. I'll do some more studying.

  • Liked by
  • Chris
Chris posted this 5 days ago

Hey CD,

The goal, Short Sharp Duty Cycle, I have tried for sometime and its not an easy goal.

I have used Microcontrollers to get some pretty good results. Thus the Microcontroller Category on the right hand side at the top under Hardware.

I have done some Arduino and also .NET Microfamework. 

About as low as I can get is 0.1 Duty Cycle, but its nice and sharp.

I can share some more data if you wish?

   Chris

cd_sharp posted this 5 days ago

Hi, Chris! Thanks for letting me know. 0.1 duty cycle is also the limit of my function generator. I'll stick to Akula's circuit, I'll try to reverse engineer it, although that is very tough considering the low quality video of lantern no 4. I can definitely see he has 2 transistors near the integrated circuit. A mistery to solve..

  • Liked by
  • Chris
Chris posted this 5 days ago

Hey CD,

The two Transistors are a Push Pull driver for the Coil Pulsing Transistor. You can see the Push Pull on some of Ruslan's replications:

 

 

The average Mosfet Driver is based on the same ideas:

 

Ewf: https://www.quora.com/What-is-the-purpose-of-a-MOSFET-gate-driver

 

 

Not all Mosfet Driver IC's are based on the same layout, but generally the Mosfet Gate is bought up High and bought don low as fast as the signal input requires.

 

Note: Akula used a KT805 Transistor to drive the Coils.

 

 

Its worth noting: Class A and B Amplifiers use a similar NPN PNP Transistor pair called a Push Pull Transistor Array.

 

 

I am no EE Guru, I know enough to be dangerous, but this is what I have learnt.

I would not use a Push Pull Transistor pair as in the Akula Circuit, I would replace this with a proper Ultra High Speed Driver. Way easier to get a good result!

   Chris

 

High Def is available, but only one person has it, how did they get the High De Version?

 

 

Chris posted this 5 days ago

Hey CD,

A little research on the fastest Hex Inverter, fastest I can find is: 2.5 ns

This is the: 74LVC04

Most are around 6 to 12 ns. 2.5 ns is a big difference!

   Chris

 

Attached Files

Vidura posted this 5 days ago

Hi Chris,

This super-fast IC are very good indeed, but we have to consider if it is really necessary , for building a nanopulser it might be a good choice, but i don't think that it's really necessary for this circuits, and if it makes much sense. A fast mosfet driver have also a rise-fall-time of 5 to 10 ns, then the mosfet and all other components AND wires -traces have an inductance,which has a notable effect when we are dealing with nanoseconds. I would guess that better improvements can be made relatively easy and cheap making a good ,clean layout and wiring, the important components close together, use the capacitors nearby the mosfet, eventually use hi frequency electrolytic caps, so you can achieve a quite fast response. I hope that this helps a little.

  • Liked by
  • Chris
Chris posted this 4 days ago

Hey Vidura.

I agree, your' e correct!

We know Akula's Frequency was: 256.360Hz

 

 

As you know, the Period of the Frequency is: 1 / f so the Period is: 1 / 256.360Hz = 0.003900 Seconds

The On Time period is approximately: 19μs

So, we can calculate: 19μs /  3900μs = 0.004871μs * 100.00 = 0.4871% Duty Cycle.

That is approximately if my bad math is correct. 0.5% Duty Cycle is pretty easily achievable.

The only reason I bought this up was CD's specification of sub 0.1% Duty Cycles. I hope I have not confused anyone!

   Chris

cd_sharp posted this 4 days ago

Guys, awesome info, lots of stuff to learn. I understand the push-pull circuit.

But, if it's using a NPN and a PNP transistor then why are they oriented one in reverse of the other?

I think it's most likely that the 2 transistors are identical and judging from the connections, they are NPN. I'll be posting revision 3 soon.

Again, thanks for all the help.

  • Liked by
  • Chris
cd_sharp posted this 4 days ago

Guys, all the pots I know have 3 pins or 6 if they are stereo. The one in the image above has a leg connected to the zero voltage rail (the one closest to the integrated circuit). On the other side, the pot looks like it has 3 pins aligned on the same column (3rd column counting from the left) and one on the 5th column (closest to the IC). So, that makes 4 pins and this confuses me.

Anyone has any idea if this type of pot has 4 pins indeed and why would Akula use this type of pot?

  • Liked by
  • Chris
Chris posted this 4 days ago

Hey CD,

A good question:

But, if it's using a NPN and a PNP transistor then why are they oriented one in reverse of the other?

 

This is because the Pins are different from NPN to PNP:

 

Guys, all the pots I know have 3 pins or 6 if they are stereo. The one in the image above has a leg connected to the zero voltage rail (the one closest to the integrated circuit). On the other side, the pot looks like it has 3 pins aligned on the same column (3rd column counting from the left) and one on the 5th column (closest to the IC). So, that makes 4 pins and this confuses me.

Anyone has any idea if this type of pot has 4 pins indeed and why would Akula use this type of pot?

 

Some, not all, some Pots ground the case, this helps stop noise and this is also to stabilize the Pot on the PCB, helps stop the breaking of connections to the pins already connected.

   Chris

  • Liked by
  • cd_sharp
cd_sharp posted this 3 days ago

Yes, it makes a lot of sense. Thanks!

  • Liked by
  • Chris
cd_sharp posted this 3 days ago

Friends, this is my latest progress. with Akula lantern no 4 circuit reverse engineering:

Here I tried the driver circuit as I "read" it from the video. Very big effort:

The signal that supposedly is applied to the power transistor base is similar to this.

We know that the signal must be very sharp square wave. So, I probably "read" something wrong.

If you have any idea how to modify the above circuit to output a sharp square wave please let me know.

  • Liked by
  • Chris
Chris posted this 3 days ago

Hey CD, whats your signal before the NPN / PNP Transistor Array Driver?

Seems there is a lot of capacitance in there?

   Chris

Vidura posted this 3 days ago

Hey cd The transistors npn-pnp are reversed in your schematic, check it out.both emittors should go to the output(mosfet gate)

  • Liked by
  • Chris
Vidura posted this 3 days ago

The HD version of the video is most certainly copyrighted by akula and steho energy, there seems to be a comercial relationship with OU.com also.Recently they showed a video disassembling a TPU device made by akula too.

  • Liked by
  • Chris
Chris posted this 3 days ago

Hey CD,

I have used the MCP1403 and a few other Driver Chips. They are very easy. I have had problems with Transistor Array, getting tidy clean switching:

 

 

In the Akula Circuit, the Transistors have to be matched and biased correctly. I have just found an easier way, no fiddly messing around.

Yes, perhaps I am a little lazy when it comes to this, but the way I look at it, better things to spend my time on!

   Chris

cd_sharp posted this 2 days ago

Hey CD, whats your signal before the NPN / PNP Transistor Array Driver?

Seems there is a lot of capacitance in there?

The signal is almost constant voltage. It's clear that I got something wrong near the IC. The capacitors values can be found only be experimenting.. Lots of work.

The transistors npn-pnp are reversed in your schematic, check it out.both emittors should go to the output(mosfet gate)

You're right.

I have just found an easier way, no fiddly messing around.

Please show me!

  • Liked by
  • Chris
Chris posted this 2 days ago

Hey CD,

My Thread: Reliable and Flexible Switching System is where I have tried to share information on these topics.

I will tell you, these tools have made my work way easier!

I have been using Microcontrollers for some time to get to the end goal. Its a huge time saver and a very clean accurate way to take note of input settings!

   Chris

cd_sharp posted this 2 days ago

Hey, man! Quadratron seems complex with those 4 stages. For now I just need one. I have no experience with microcontrollers.

Perhaps a basic, simple intro would be better. I'm experienced in .net, so I would prefer .NET Microframework.

I understand I need to buy some microcontrollers and a board that helps interfacing and programming the microcontrollers. That's all I know about microcontrollers.

Thanks for you guidance.

  • Liked by
  • Chris
Vidura posted this 2 days ago

Hey CD!

I am not familiar with the .net framework, i only used microchip tools and software. 

If you want to continue with one single switch i would recommend you to use a mosfet with a mosfetdriver , the MCP1403 which suggested Chris is fine, I use  MCP1407 with a single input, these drivers are simple and will enhance the switching a lot. They can operate upto 18V supply voltage, But 12V is OK. Just connect the signal directly to the input, connect also a 100nF decoupling cap together with the supply pins of the IC , and the output pins directly to the mosfetgate, normally no other parts are required for a single mosfet.

If you which to employ it in the circuit from the video replace  R2 , C6, R5, Q1 and Q4  with the driver, the supply on C5+ and 0V.

Input IC1 pin 12 , output directly to the mosfetgate, check if you have a clean waveform on the input also.

PD: For practice with microcontrollers it would be best to purchase a board like the arduino or similar experimental tool ,some have integrated programming interface, ready to connect to your PC; and start with simple projects to become familiar with programming and so.

Good Luck!

  • Liked by
  • Chris
Chris posted this 2 days ago

Hey Guys,

I am happy to help and share what I have learned. You don't need all 4 stages, you don't need the whole circuit, just the MCP1403 / 7 or similar Driver IC and a few resistors and Caps would be fine!

I believe Netduino is the only current .NETMF Controller on the market right now. GHI Have one of two but I don't know anywhere selling them.

It is possible to port the .NETMF to another Microcontroller with GHI Electronics .NET TinyCLR OS, but this is a bit much for now.

A very simple PWM would be:

Microsoft.SPOT.Hardware.PWM MyPWM = new PWM(Cpu.PWMChannel.PWM_4, 2, 0.01, false);

All that's required is to import the Microsoft dll: 

C:\Program Files (x86)\Microsoft .NET Micro Framework\v4.3\Assemblies\le\Microsoft.SPOT.Hardware.PWM.dll

 

 

Of course this is a very simple layout, took me no more than 15 minutes. The Driver IC is very easy to use!

The HUGE benefit, I can easily program the Micro to a huge range of parameters and easily save the settings for my records. I can share software also if you decide to go down this path.

This is really easy. I am more than happy to help, sharing is where we have a huge edge!

   Chris

Close