Hi:
I used the following transistor clamp in this link:
https://www.eevblog.com/forum/blog/eevblog-1157-transistor-zener-clamp-circuit/.I attached it to the output of the boost converter. Below is my circuit snapshot, and attached is the LTSpice .asc file.
Let me tell you what I want to achieve, and I want to do this without any ICs, without any microcontrollers. I want a solution that is discrete components only with transistors, resistors, diodes, capacitors, and inductors. I have a 9VDC battery. As the battery goes to 0VDC, it goes through a range of voltages from 9VDC to 0VDC. I would like the output of the boost converter to have a constant 6VDC at 100mA. The boost converter is fed from a stable oscillator that is made from transistors, resistors, and capacitors. So that the boost converter does not run off, I used a transistor clamp as shown in the link above to clamp the voltage to 6VDC. However, the output does NOT get clamped to 6VDC? Why? If I can't use this, then what circuits made from discrete components only can I use to achieve this goal? Also, is there any discrete component feedback loop circuit that stops the boost converter to run away?
Is this homework? Doesn't sound like something any school or university would ask.
Why the restriction of only discrete components ?
no this is not homework. Its my personal project. I like to reason and build from discrete components. I really don't like ICs, dont like anything except discrete components because I don't want things ready made. I like to understand it, and that knowledge would be helpful elsewhere where i am building personnel projects. I think everything can be built from discrete components and thru engineering.
Did you try just using a single zener diode alone? i.e. delete Q5, Q6 & D4 Whilst this may address the clamping issue, it will bring its own issues as well.
Ideally, you want a feedback network to directly adjust the astable duty cycle according to output voltage.
What you have designed does not have any feedback. Is this what you actually want to know?
looks like you are trying to clamp the 9V battery down to 6V through L1, D1. a booster is simply not able to generate voltages below input voltage (unless it uses a transformer or is of something like SEPIC topology, of course).
I have a 9VDC battery. As the battery goes to 0VDC, it goes through a range of voltages from 9VDC to 0VDC. I would like the output of the boost converter to have a constant 6VDC at 100mA.
Sounds like you want a buck converter. A buck converter has the same components of a boost converter but arranged differently, i.e. see:
https://en.wikipedia.org/wiki/Buck_converterA 9V battery is usually considered dead when it drops below 7V.
I have a 9VDC battery. As the battery goes to 0VDC, it goes through a range of voltages from 9VDC to 0VDC. I would like the output of the boost converter to have a constant 6VDC at 100mA.
9V to 6V is
not a
boost converter, so you have a fundamental flaw in concept here.
9V to 18V is
a boost converter.
How much does your load current vary, and what efficiency are you looking for?
No I want to do boost converter alright not buck converter. You realize I have a battery of 9VDC that can go to 0VDC. So range of voltages from 0VDC to 9VDC. I realize boost converter can't go below 6VDC so then what I am looking for is that below 6VDC I get 6VDC output and above 6VDC I get battery output. Do you know of feedback circuit that is made from discrete components that can control the duty cycle of the astable oscillator and do what I just said?
Did you try just using a single zener diode alone? i.e. delete Q5, Q6 & D4 Whilst this may address the clamping issue, it will bring its own issues as well.
Ideally, you want a feedback network to directly adjust the astable duty cycle according to output voltage.
What you have designed does not have any feedback. Is this what you actually want to know?
Yes I tried zener diode alone and deleted Q5, Q6 and D4. Same thing see attached. I realize this is boost converter so below 6VDC of the battery I like vout of 6VDC and above or equal to I like battery voltage no boosting. Yes I do want to know how I can design feedback to this circuit using discrete components only no ICs, no op amp, no controller only transistors, resistors, capacitors, inductors, diodes so that I can control the duty cycle. Can you help with that?
No I want to do boost converter alright not buck converter. You realize I have a battery of 9VDC that can go to 0VDC. So range of voltages from 0VDC to 9VDC. I realize boost converter can't go below 6VDC so then what I am looking for is that below 6VDC I get 6VDC output and above 6VDC I get battery output. Do you know of feedback circuit that is made from discrete components that can control the duty cycle of the astable oscillator and do what I just said?
If you are prepared to have the converter simply pass-through above 6V (via the diode and inductor paths) you have no regulation in that area.
All that requires is to never turn on the flyback FET, in that zone.
The circuit is also not going to work all the way to 0V and you need to decide on some sane minimum battery level to avoid over discharge.
Since you are ok with quite unregulated output, you could design a somewhat fixed 1.5x boost, which will get you down to about 4V for 6V out.
However, most 9V batteries have only a tiny fraction of energy left below 6V anyway, so maybe you are smarter to make better use of the energy you have.
If your system can work at 6V, a buck converter can reduce the initial drain on the 9V battery to give you much more time extension than chasing the dregs.
@Le_Bassiste: Would you mind going into detail about how SEPIC topology can assist in my case? Also, can you assist me with feedback control of duty cycle using only discrete components with SEPIC topology and the one I showed post #1?
No I want to do boost converter alright not buck converter. You realize I have a battery of 9VDC that can go to 0VDC. So range of voltages from 0VDC to 9VDC. I realize boost converter can't go below 6VDC so then what I am looking for is that below 6VDC I get 6VDC output and above 6VDC I get battery output. Do you know of feedback circuit that is made from discrete components that can control the duty cycle of the astable oscillator and do what I just said?
If you are prepared to have the converter simply pass-through above 6V (via the diode and inductor paths) you have no regulation in that area.
All that requires is to never turn on the flyback FET, in that zone.
The circuit is also not going to work all the way to 0V and you need to decide on some sane minimum battery level to avoid over discharge.
Since you are ok with quite unregulated output, you could design a somewhat fixed 1.5x boost, which will get you down to about 4V for 6V out.
However, most 9V batteries have only a tiny fraction of energy left below 6V anyway, so maybe you are smarter to make better use of the energy you have.
If your system can work at 6V, a buck converter can reduce the initial drain on the 9V battery to give you much more time extension than chasing the dregs.
My minimum would be 3V to 6VDC. My load would need 6VDC at 100mA. I want a boost converter with feedback. The feedback to control the duty cycle in my circuit I want with discrete components only no ICs, no controller, no op amp, only transistors, resistors, capacitors, inductors.
Some other person also mentioned SEPIC topology, and I am also open to that with feedback control of duty cycle with discrete components only.
Main concern is how to design feedback using discrete components only that can control the duty cycle in my circuit and in the SEPIC topology.
Main concern is how to design feedback using discrete components only that can control the duty cycle in my circuit
That's easy enough to do.
Just vary the current pulling up one or both of the timing caps in your cross coupled NPN oscillator.
You can have parallel resistors to set the lowest current, and can vary one or both charging paths, typically with a PNP current source derived from your feedback error voltage.
If your input voltage is 3-9V and your desired output voltage is 6V, then you want either a SEPIC converter or a two-switch* buck-boost converter (the single switch buck-boost generates a reverse polarity output compared to the input, probably not what you want i.e. 3-9V input, -6V output). The two-switch solution might be a liability at low voltages due to the voltage drop of two switches. SEPIC might be tricky due to the large ripple current in the largish transfer capacitor (which in most applications is a large ceramic capacitor, not a tantalum or electrolytic).
A boost converter followed by a zener clamp, is IMO, the incorrect way to approach the problem.
*the two switches turn on & off simultaneously, not opposite phases.
As for feedback, that might depends on what kind of operation you want out of your regulator. PWM might get quite complex with only using discrete parts, I'm estimating at least a half dozen more transistors. However, bang-bang control should be relatively straightforward.
When designing, you need to start with realistically set goals. Yours aren't realistic; you've focused on optimizing the circuit to the point where the battery no longer works (3-6V), while neglecting the range where the battery works well (6-9V), wasting most of the usable energy in a Zener diode.
Look at the battery characteristics at high currents, below 6V, there's nothing.
Main concern is how to design feedback using discrete components only that can control the duty cycle in my circuit
That's easy enough to do.
Just vary the current pulling up one or both of the timing caps in your cross coupled NPN oscillator.
You can have parallel resistors to set the lowest current, and can vary one or both charging paths, typically with a PNP current source derived from your feedback error voltage.
Its not that easy. I don't understand what you mean by the sentence(s) you wrote. Show me, circuit-wise, what you mean and what your design would look like with component values? I just did it, and my oscillation just stopped. I have attached .asc file for LTspice as well for your convenience so that you can teach me how.
Main concern is how to design feedback using discrete components only that can control the duty cycle in my circuit
That's easy enough to do.
Just vary the current pulling up one or both of the timing caps in your cross coupled NPN oscillator.
You can have parallel resistors to set the lowest current, and can vary one or both charging paths, typically with a PNP current source derived from your feedback error voltage.
Its not that easy. I don't understand what you mean by the sentence(s) you wrote. Show me, circuit-wise, what you mean and what your design would look like with component values? I just did it, and my oscillation just stopped. I have attached .asc file for LTspice as well for your convenience so that you can teach me how.
Simulation of self oscillating circuits is fraught with challenges. You've picked something that is very hard to do
and not explained why it is necessary or what the goals are (so others are unlikely to put much effort into working on it).
If you don't understand that sentence (which is plain and clear) that suggests you need to learn more about the language and practicalities of electronics before attempting this project. That suggestion by PCB.Wiz is beautifully complete with many important details that you have glossed right past.
I explained it well what I need why I need many times here. I dont want stuff with ICs, stuff with module and any stuff that doesnt involve discrete component. If I wanted that I am electrical engineer so I can just easily built ready made. I dont want those recommendations. Just built thru discrete stuff. I am electrical engineer so I know. Street stuff pull resistor down etc i dont get it. Besides that I attempted circuit as well. So its not like I put in my effort. Shame none of you all want to help design it or give recommendation of your circuit actual circuit not words with discrete components so astable oscillators duty can be controlled thru discrete components. Thats all I am after here.
Its not easy to get transistor in actual saturation, active region where pcb wise suggested in location because its oscillating. It wont drain current properly and control duty cycle if its not at properly biased and cant get it to work because oscillation at the point pcb wise suggested . If you know how then help and show the circuit not just words
Anyway can write words oh put transistor and have it as switch on/off but hey try giving me circuit with actual values and see if that works!
When designing, you need to start with realistically set goals. Yours aren't realistic; you've focused on optimizing the circuit to the point where the battery no longer works (3-6V), while neglecting the range where the battery works well (6-9V), wasting most of the usable energy in a Zener diode.
Look at the battery characteristics at high currents, below 6V, there's nothing.
Thanks. I will be using SEPIC topology or one that does either buck and boost convert both as necessary. I get it. However what I am after is my astable oscillator in post #1 an actual circuit with discrete components only that can control astable oscillators duty cycle. That way it can lower or upsize the converter voltage as necessary. Thats all I am after here. Duty cycle control circuit thru discrete components of my astable oscillator.
Check the Roman Black website. He has some extremely ingenious switching regulators using only transistors.
... I have a 9VDC battery. As the battery goes to 0VDC, it goes through a range of voltages from 9VDC to 0VDC. I would like the output of the boost converter to have a constant 6VDC at 100mA. ...
When a 9V battery is dead (like when the voltage is 6V) the internal resistance is high. You won't be able to get 100mA out of it.
Thanks. I will be using SEPIC topology or one that does either buck and boost convert both as necessary.
You might also be interested in single-transistor flyback and inverting topologies.
Both can increase and decrease voltage.
Thanks. I will be using SEPIC topology or one that does either buck and boost convert both as necessary. I get it. However what I am after is my astable oscillator in post #1 an actual circuit with discrete components only that can control astable oscillators duty cycle. That way it can lower or upsize the converter voltage as necessary. Thats all I am after here. Duty cycle control circuit thru discrete components of my astable oscillator.
Controlling the current through R1 & R4 are the key to adjusting the duty cycle.
The problem is that you will need to control both of them for full range duty cycle, because R1 only alters duty cycle between 0-50%, and R4 only alters the duty cycle between 50-100%. I'm not sure how to transition between the two control schemes smoothly (which is normally required for power converter oscillators).
Hi:
If I wanted to get feedback loop using discrete components only then for a SEPIC converter what I need to control to change duty cycle and output voltage? What about for a two-switch* buck-boost converter?
...
However, the output does NOT get clamped to 6VDC? Why?
...
That is because the clamping circuit is based on the reverse breakdown voltage of the BE junction of the transistors. Often, that is about 6V or so, not always exactly 6V but around that value, between 5..10V or so, but anyway much lower than the BC or than the CE breakdown voltages.
Usually, the transistor models in SPICE do not simulate the breakdown voltage (or breakdown currents) at all, so that is why nothing happens at 6V. The casual BJT model does not include that behavior because usually one must never reverse-polarize the BE to its breakdown voltage.
However, there are SPICE models that can simulate the breakdown voltage, an example in this post:
https://www.eevblog.com/forum/beginners/help-choosing-a-simulator-ltspice-qspice-or-tina-ti/msg5329655/#msg5329655
Try to simulate your circuit using the transistor model attached there (use 2N3904_BD instead of 2N3904), and it should work (but I didn't try).