In this topic, I would like to discuss impedance matching. For some time now, I have been having a hard time wrapping my head around it. I have read about impedance and watched a lot of videos and still couldn't figure out how it works. I just couldn't see how to match the input and output of a transformer, to match the input impedance of the Line signal, to the output impedance of the Load circuit. Until today. This is what I want to get into now, because I see how important it is, in what we are trying to do here.

What finally made me understand how it works was a video on YouTube. It showed the math and explained it to me in a way that I could understand it. Quite simple now.

I have seen the work that Webmug has been doing with his impedance test, and that will help us out here. I also seen some of the math the Ronnie has given us lately with matching the primary and secondary coils. This all comes into play here now that I have a better understanding on what we need to do in getting impedance matching.

After watching the video, it brings up the first question. What is the impedance of our line (source) and Load?
"Line" is a normal term for the input power or signal, and "Load" is the consumer of the input power/signal. I will refer to the Line as Source from here on out.

In this exercise the primary and secondary is what we will be trying to figure out here to match the impedance of our Source and Load. Therefore the impedance of the primary and secondary coils are not factored into these measurements, but are what we need to calculate to impedance match the Source to the Load.
 
In order to match the impedance of our set ups, we need to know the impedance of the Source first. Now the source is not the primary windings of our transformer, but the driving circuit we put into the primary coil. In this case it would be our frequency generator. OK so what part of the frequency generator do we use as the impedance? This is the first question we need to answer.
Since I am using a mosfet to drive the primary, is that where we need to measure the impedance? Is that the only part of the circuit that needs to be measured, or is there more? Again, we don't need the primary coils impedance here.

Once we have the Source impedance known, we then need the Load impedance. What constitutes the load impedance? Do we need to know the impedance of the choke coils and the cell, add them together for a total impedance of the load?  I would think that question is yes. Again, we don't need the secondary coil's impedance here.

Once we have the Source and Load impedance, we then have what we need to wind our transformers primary and secondary coils.
To figure the transformer primary to secondary ratio we need to figure the ratio of the Source to Load impedance ratio.
The formula for that is Source impedance divided by the Load impedance squared. Lets say we have a source impedance of 10ohms and a load impedance of 1000ohms. 10/1000 = .01 or 1/100. Now we square that and we get 10. So our turn ratio of the transformer need to be 10 to 1. If we have 200 turns on the primary, we will need 2000 turns on the secondary.

This is how you impedance match a transformer to Source and Load. So lets discuss how we get the impedance values for our setups. Do we need the impedance values taken at resonance of the choke coils and cell, or just the basic impedance. Meaning, an 8ohm speaker for the most part is just 8ohms, but at resonance it is much greater, but it is 8ohms most of the time. Only at a resonant frequency is it greater. Is this the impedance of the coils and or cell that we need to use and not the impedance at resonance? That's what I think we need, the basic impedance not at resonance, for impedance matching.

So this is where I plan on taking my research, any input is welcome.
Don

If the primary and secondary are of the same impedance it is beneficial because we know there are no reflections or standing waves of energy between them, the core material of the transformer also plays a part in this. The chokes at their own self resonant frequency are likely to be an impedance mismatch than the rest of the network and the secondary they are connected to unless you make the whole network run at the resonant frequency of the chokes. If the secondary is an impedance mismatch while the chokes are at resonance then it becomes just a short between them.

Interesting, so do we just put an impedance meter on our source when it is not powered?  This is where Stan's set up of one 2N3055 transistor on one side of the primary and a TIP120 transistor on the other side might be beneficial.  Both sides being controlled could be higher impedance?  The one side that most of us aren't controlling would be an impedance of Zero in theory?  Plus we have to figure in the 220 ohm resistor in parallel to the source, not the coil?

Quote from nav on January 3rd, 2015, 11:31 AM

If the primary and secondary are of the same impedance it is beneficial because we know there are no reflections or standing waves of energy between them, the core material of the transformer also plays a part in this. The chokes at their own self resonant frequency are likely to be an impedance mismatch than the rest of the network and the secondary they are connected to unless you make the whole network run at the resonant frequency of the chokes. If the secondary is an impedance mismatch while the chokes are at resonance then it becomes just a short between them.

If the primary and secondary are at the same impedance they are exactly the same size. Impedance matching the Source to the Load has nothing to do with the impedance of the primary and secondary coils. The impedance of the primary and secondary coils will have the same ratio as the Source to Load.

Quote from firepinto on January 3rd, 2015, 12:25 PM

Interesting, so do we just put an impedance meter on our source when it is not powered?  This is where Stan's set up of one 2N3055 transistor on one side of the primary and a TIP120 transistor on the other side might be beneficial.  Both sides being controlled could be higher impedance?  The one side that most of us aren't controlling would be an impedance of Zero in theory?  Plus we have to figure in the 220 ohm resistor in parallel to the source, not the coil?

I would have to say yes to the first part where Stan used the 2N3055 and the Tip120, but no to the last line. The 220 ohm because it is in parallel, will be part of the primary winding and not the Source impedance.
The data sheet for one of my mosfets say it has only .032 ohms for RDS on. That's the drain to source resistance. So is that my impedance I need to match? That's what I'm looking to figure out.
Don

Quote

If the primary and secondary are at the same impedance they are exactly the same size

Not always, you can use different gauge wires. For example, you could have 600 turns at 28 gauge that has the same impedance as a coil of 500 winds of a difference gauge wire.
That's a rubbish statement you made.

Quote from nav on January 3rd, 2015, 01:13 PM

Not always, you can use different gauge wires. For example, you could have 600 turns at 28 gauge that has the same impedance as a coil of 500 winds of a difference gauge wire.
That's a rubbish statement you made.

Good point, didn't think of that, but then again Stan's coils were all wound with 29 gauge wire.
The primary and secondary coils aren't what we need to impedance match. They are the result of matching the Source and Load.

@Don,

Can you link the youtube video please.

Here's the video I mentioned

Dynodon, when talking about coil impedance it is sometimes better to imagine them as real world objects. So, if we want to relate dc pulsed frequency to impedance I do this:
Imagine the secondary coil of a transformer is a bucket that can be filled with water, the coil wire gauge is the diameter of an hose pipe you fill the bucket with and the pulse frequency is how long you turn a tap on and off for. Then imagine the load impedance as being how quickly you can empty the bucket with another hose pipe.
Now lets say you fill the bucket with an hose pipe and it takes 10 seconds to fill it, this represents V+ on a dc pulse width. After the 10 seconds have elapsed the bucket is 90% full and the load has an opportunity to empty the bucket. If the load only has a tiny hose pipe then during the 10 seconds the tap is switched off, it may only half empty the bucket. When the tap is turned back on, it will turn back on for 10 seconds but the bucket is half full already so the bucket will overflow. The tap compensates by turning down the pressure it needs to fill the bucket.
However, if the load is very quick at emptying the bucket and empties it in 5 seconds when the tap is switched off for 10 seconds, we are left with a gap because the tap is still switched off. To close the gap we have to reduce the rate at which the load empties the bucket or have a larger bucket.
This is what impedance matching is, making sure that we fill and empty buckets at the same time.
If we don't, then more energy is expended and water ends up everywhere.

Quote from nav on January 4th, 2015, 04:10 AM

Dynodon, when talking about coil impedance it is sometimes better to imagine them as real world objects. So, if we want to relate dc pulsed frequency to impedance I do this:
Imagine the secondary coil of a transformer is a bucket that can be filled with water, the coil wire gauge is the diameter of an hose pipe you fill the bucket with and the pulse frequency is how long you turn a tap on and off for. Then imagine the load impedance as being how quickly you can empty the bucket with another hose pipe.
Now lets say you fill the bucket with an hose pipe and it takes 10 seconds to fill it, this represents V+ on a dc pulse width. After the 10 seconds have elapsed the bucket is 90% full and the load has an opportunity to empty the bucket. If the load only has a tiny hose pipe then during the 10 seconds the tap is switched off, it may only half empty the bucket. When the tap is turned back on, it will turn back on for 10 seconds but the bucket is half full already so the bucket will overflow. The tap compensates by turning down the pressure it needs to fill the bucket.
However, if the load is very quick at emptying the bucket and empties it in 5 seconds when the tap is switched off for 10 seconds, we are left with a gap because the tap is still switched off. To close the gap we have to reduce the rate at which the load empties the bucket or have a larger bucket.
This is what impedance matching is, making sure that we fill and empty buckets at the same time.
If we don't, then more energy is expended and water ends up everywhere.

Sounds like a very delicate procedure scanning and maintaining impedances then, atleast with regards to a Meyer WFC setup, what would be the indicators of having acheived and continously accurately maintaining such a successful impedance matching process?
Or is that job done only once, after the cell has been finished the windings can then be ballpark estimated, or maybe even scanned using a low voltage ac signal and watching/comparing the impedances on a scope?
I guess what I'm really asking is if that's what the pickup coil is for or not?

You can use a pickup coil for several things. We can use it to indicate when the core is shut down during self resonance then we can shut the primary off to save power. Then the period of self resonance is equal to the gate. Or we can use it during normal charging to measure the impedance of the charging network in the core. That is how large the magnetic field is compared with value X. We can then pll value X

In the circuit, based on the components connection, it looks more like detecting the frequency of pulses coming from the core. From this, how do you detect, these 2 things, 1. core shutdown and 2, measure impedance of the charging network in the core

Quote from nav on January 4th, 2015, 05:50 AM

You can use a pickup coil for several things. We can use it to indicate when the core is shut down during self resonance then we can shut the primary off to save power. Then the period of self resonance is equal to the gate.

That would be the ideal way of gating, that's for sure :thumbsup:

Quote from nav on January 4th, 2015, 05:50 AM

Or we can use it during normal charging to measure the impedance of the charging network in the core. That is how large the magnetic field is compared with value X. We can then pll value X

What would the voltage on the pickup coil be then with regards to the impedance of the charging network in the core at ideal conditions?
As the voltage and the current feeding the primary are known, what would the pickup coil show us indicating that we're pulsing in the correct neckar of woods at all?
I know we're aiming for the lowest possible primary current, but at the same time we would want to be looking for it during a fairly narrow band and if we missed the train we're automatically going for all the lower current the higher the primary pulsing frequency gets as, well, that's just in the very nature of inductors to show increased impedance the higher the frequency is that which you feed them with.

Sorry for the noob questions, but I've only just recently got interested in all this, in no small part thanks to your very posts here Nav, many thanks for that.

Quote from Lynx on January 4th, 2015, 06:38 AM

That would be the ideal way of gating, that's for sure :thumbsup:

What would the voltage on the pickup coil be then with regards to the impedance of the charging network in the core at ideal conditions?
As the voltage and the current feeding the primary are known, what would the pickup coil show us indicating that we're pulsing in the correct neckar of woods at all?
I know we're aiming for the lowest possible primary current, but at the same time we would want to be looking for it during a fairly narrow band and if we missed the train we're automatically going for all the lower current the higher the primary pulsing frequency gets as, well, that's just in the very nature of inductors to show increased impedance the higher the frequency is that which you feed them with.

Sorry for the noob questions, but I've only just recently got interested in all this, in no small part thanks to your very posts here Nav, many thanks for that.

These values are simple Cmos logic sequences.
If the impedance of the load is a set value that never changes then there is no problem, we can set the impedance on the primary and the secondary to match the load. But in this case the load is reactive for reasons out of our control.
So we set the system up initially so that the core produces a magnetic field of the value of X when the coils are operating with matched impedance with a static load. The static load represents the exact center of another load that is reactive in either direction of center.
So the first Cmos logic question is 'is the core open for business yes or no? a normal pulse train from the primary begins. If the pickup can't see a response because the primary pulse isn't received then the Cmos answer is no and Pv=0 that is primary voltage = 0. No more pulses are sent for a period of time.
If the C'mos reply is yes then the Cmos moves to the next logic question. Is the value of Y (the magnetic field in the core) higher than X (value of magnetic field when coils operate impedance matched), yes or no? If its yes, then the C'mos reply is to lower the system impedance and keep asking that question until Y=X. If the answer to that original question was no then it goes to the third logic question 'is the value of Y lower than X? If the answer is yes then it highers the system impedance and keeps asking the same question until Y=X.
When Y=X then the Cmos will allow the primary to operate normally and while ever the pick up keeps sending the Cmos the correct answers to its questions the primary will remain open.
Finally, when the chokes begin to self resonate and the core shuts down the C'mos will get an answer that the core is shut and it will not allow the primary to operate until the core is open. This is why you only need one NEC 555 and a couple of Cmos chips.
It is a binary system like any IC, of questions with yes or no answers. Lynx, you said you went through my board schematic, well its just the same inside a Cmos. One transistor closing another transistor and opening another transistor that closes an SCR that opens a jfet until Q3 is closed etc etc etc. You can make transistors do anything in logic sequences.

 

Quote from nav on January 4th, 2015, 09:22 AM

These values are simple Cmos logic sequences.
If the impedance of the load is a set value that never changes then there is no problem, we can set the impedance on the primary and the secondary to match the load. But in this case the load is reactive for reasons out of our control.
So we set the system up initially so that the core produces a magnetic field of the value of X when the coils are operating with matched impedance with a static load. The static load represents the exact center of another load that is reactive in either direction of center.
So the first Cmos logic question is 'is the core open for business yes or no? a normal pulse train from the primary begins. If the pickup can't see a response because the primary pulse isn't received then the Cmos answer is no and Pv=0 that is primary voltage = 0. No more pulses are sent for a period of time.
If the C'mos reply is yes then the Cmos moves to the next logic question. Is the value of Y (the magnetic field in the core) higher than X (value of magnetic field when coils operate impedance matched), yes or no? If its yes, then the C'mos reply is to lower the system impedance and keep asking that question until Y=X. If the answer to that original question was no then it goes to the third logic question 'is the value of Y lower than X? If the answer is yes then it highers the system impedance and keeps asking the same question until Y=X.
When Y=X then the Cmos will allow the primary to operate normally and while ever the pick up keeps sending the Cmos the correct answers to its questions the primary will remain open.
Finally, when the chokes begin to self resonate and the core shuts down the C'mos will get an answer that the core is shut and it will not allow the primary to operate until the core is open. This is why you only need one NEC 555 and a couple of Cmos chips.
It is a binary system like any IC, of questions with yes or no answers. Lynx, you said you went through my board schematic, well its just the same inside a Cmos. One transistor closing another transistor and opening another transistor that closes an SCR that opens a jfet until Q3 is closed etc etc etc. You can make transistors do anything in logic sequences.

This would explain the modus operandi in operating/switching the primary once the "correct" frequency is known/found, right?
In that case that makes perfect sense, that should be the logic way to operate/switch the primary.
What I'm wondering about though is during, say a scanning phase, which purpose would be to subject the primary to a slowly increasing pulsing frequency in order to find the "sweet spot" frequency, how would that manifest itself in the pickup coil, would it's voltage all the sudden just rise, or sink for that matter, at the "correct" frequency, or would it show itself in quite another way?
I'm just trying to get the gist of it all, to learn the basics about what you could expect to find in the pickup coil that which let's you know that you're using the correct primary pulsing frequency.
Or maybe it's a combination of things that let's you know that you're using the correct primary frequency, duty cycle, voltage, current, etc etc.

Just to remind everyone here, this topic is for impedance matching the source to load. Try to keep it that way please.  Rav already has a topic started on the operation part of the circuits.

What I want to do is to figure out how we can get the impedance of our set up right. Lynx was asking about how to keep impedance matching in place when scanning for resonance. For the most part, the impedance doesn't change much. It will peak at a point of resonance at a given frequency, but most of the time it is pretty constant. Take a look at the graphs Webmug made of his coils. You can clearly see the point at resonance, but most of the time it's pretty consistent.

What I want to figure out is, what part of our VIC and WFC is considered the load. Is it the coils and cell, or just the coils or cell? That's what we need to figure out. I am getting ready to start building a new cell with ten tube sets. Then I will need to find out what the impedance of it is, as well as the choke coils. This then needs to be matched up to the source.

So this is what I want to get input on here in this topic.
Don

Quote from Dynodon on January 4th, 2015, 04:32 PM

Just to remind everyone here, this topic is for impedance matching the source to load. Try to keep it that way please.  Rav already has a topic started on the operation part of the circuits.

What I want to do is to figure out how we can get the impedance of our set up right. Lynx was asking about how to keep impedance matching in place when scanning for resonance. For the most part, the impedance doesn't change much. It will peak at a point of resonance at a given frequency, but most of the time it is pretty constant. Take a look at the graphs Webmug made of his coils. You can clearly see the point at resonance, but most of the time it's pretty consistent.

What I want to figure out is, what part of our VIC and WFC is considered the load. Is it the coils and cell, or just the coils or cell? That's what we need to figure out. I am getting ready to start building a new cell with ten tube sets. Then I will need to find out what the impedance of it is, as well as the choke coils. This then needs to be matched up to the source.

So this is what I want to get input on here in this topic.
Don

Sorry Don, didn't mean to hijack your thread.

Don
How you will find out ( measure )  the impedance of your cell?
thank
andy

andy, that's one of the things I'm still thinking about. Webmug uses a special circuit board that is just for that, it cost @ $60 at mouser. It will also measure the coils.
Don

Quote

What I want to figure out is, what part of our VIC and WFC is considered the load. Is it the coils and cell, or just the coils or cell? That's what we need to figure out. I am getting ready to start building a new cell with ten tube sets. Then I will need to find out what the impedance of it is, as well as the choke coils. This then needs to be matched up to the source.

The tubes are the load but are reactive. That is what we have been talking about. It is not a fixed impedance as such. When the cell is full of water and the capacitance builds up, it builds up as a step charge process, as the capacitance increases across the load, the dielectric property of water changes and when it begins to produce gas, as bubbles of Hydrogen and Oxygen begin to fill the cell, its dielectric property changes again because there volume of water decreases as gas increases. Its a moving target and that is why we talk in terms of phase lock looping. This effect is felt right back to the primary in the network.
It is not as simple as finding the impedance of the cell and working with a fixed value. The cell's impedance at 100v is different from when it step charges at self resonance to the Kv ranges.
That is why I suggested that people either build a Gamma match or a beta match into the cell so that they have a variable impedance part of their cell. The electronics do the rest with pll controlled by Cmos integrated with an oscillator.

Nav
Is the impedance of bifilar coils dependant from frequency at which we pulse them?
thank
andy

Quote from nav on January 5th, 2015, 07:06 AM

The tubes are the load but are reactive. That is what we have been talking about. It is not a fixed impedance as such. When the cell is full of water and the capacitance builds up, it builds up as a step charge process, as the capacitance increases across the load, the dielectric property of water changes and when it begins to produce gas, as bubbles of Hydrogen and Oxygen begin to fill the cell, its dielectric property changes again because there volume of water decreases as gas increases. Its a moving target and that is why we talk in terms of phase lock looping. This effect is felt right back to the primary in the network.
It is not as simple as finding the impedance of the cell and working with a fixed value. The cell's impedance at 100v is different from when it step charges at self resonance to the Kv ranges.
That is why I suggested that people either build a Gamma match or a beta match into the cell so that they have a variable impedance part of their cell. The electronics do the rest with pll controlled by Cmos integrated with an oscillator.

I know the tubes are the load, but would the chokes be a part as well? After all, the LC resonance takes place in the positive choke. I agree with most of the first paragraph, but not total convinced with the second one. I don't know if voltage effects the impedance. If it does, than it makes the tuning part almost impossible to get to. Impedance is a byproduct of frequency, it changes with frequency and reaches peak at resonance. How voltage can change the impedance, I have not see that described before of even seen a formula to show that type of change. Impedance after all is the resistance to AC voltage. So when we are testing, does changing the applied voltage effect the impedance? I would need to see that somewhere in text to agree with that comment.

With that being said, then it would be almost impossible to tune a cell manually. I doubt that Ronnie is using PLL for tuning his set up. He claims he has his working, and I think it is tuned by hand.
Don

Quote from Dynodon on January 5th, 2015, 11:31 AM

I know the tubes are the load, but would the chokes be a part as well? After all, the LC resonance takes place in the positive choke. I agree with most of the first paragraph, but not total convinced with the second one. I don't know if voltage effects the impedance. If it does, than it makes the tuning part almost impossible to get to. Impedance is a byproduct of frequency, it changes with frequency and reaches peak at resonance. How voltage can change the impedance, I have not see that described before of even seen a formula to show that type of change. Impedance after all is the resistance to AC voltage. So when we are testing, does changing the applied voltage effect the impedance? I would need to see that somewhere in text to agree with that comment.

With that being said, then it would be almost impossible to tune a cell manually. I doubt that Ronnie is using PLL for tuning his set up. He claims he has his working, and I think it is tuned by hand.
Don

Voltage does not effect impedance but the reactive effect of voltage on the dielectric layer does.
Water is the dielectric layer, as you change water into gas, the ability of the water to act at one particular dielectric property changes because it is being replaced by gas all the time.
The more gas you produce between the plates of the capacitor the less water there is. Therefore the impedance at beginning of water production is different to when the cell is at full swing.
The cell is a reactive load.

Quote from andy on January 5th, 2015, 09:42 AM

Nav
Is the impedance of bifilar coils dependant from frequency at which we pulse them?
thank
andy

No, look at my signature.