Choosing inductance and capacitance

Runem wrote:

Jalex2 wrote:It has only 9 turns on it and reads 102 Mh on my LC meter the same as the open air coil.

Yeah... thats typical for ferrite power chokes when you try to force them to operate at higher frequencies than what they are intended for. The Q of the core is so bad at frequencies in the MHz range that they will force the measurement oscillator to go waaay lower than normal and hence give you false readings. Use a scope and a siggen. instead, with a known capacitance in parallel to the choke, and tune the peak down in the range that Ivan mentioned (i'll guess below 200kHz will work just fine).

Thanks
That was helpful. I didn't know that. I am playing with a different project now and I think I was trying to run that oscillator at around 2 mh. I will have to go back to it and see if I can get it running at 200 kh instead.
Lately I have been playing with this circuit. I wanted to see if I could make this run with a single ended power supply but I have my droughts now. LOL
I didn't like having all those regulators but I guess for a circuit like this it's a necessity. I never had seen a 555 or power mosfet used in this way.
I am not really a fan of metal detectors, I just though it would be fun to play with his circuit ideaand see what I could do with it.
http://www3.telus.net/chemelec/Projects/Metal/Metal.htm

Ahhh seeeee? I knew I'd misunderstood you

I don't know, is the simple answer to your question - maybe it's different because Admin can set the depth of quotes, but probably not disable bold italics etc?

I mean that BB-code is a small text formating command set that is available anyway with the little editing buttons. So everyone can write on bold, italics, quotes, etc. So why does the BB-code need to be enabled? That was my question.

/Harry

Admin wrote:

DragonForce wrote:Look in your profile settings and enable BBcode - by default it's set too off.

test wrote:
blah blah blah

If you reply to a post then do you not have the bold/italic/strikethrough and justify, bulets, etc at the head of the text entry panel? I can see to attach pictures, and even quotes. It was my understanding that BBcode was just a means of manually programming the text formating. Or have I missed anything?

/Harry

Not sure what you mean, I can still use bold, italic underlined text ect if I want, whilst replying (and quoting you) - I still have them as an option whilst composing this message in the text entry panel. Maybe there's an advanced option setting hidden someplace? Or have I completely misunderstood you (about normal for me)

DragonForce wrote:Look in your profile settings and enable BBcode - by default it's set too off.

test wrote:
blah blah blah

If you reply to a post then do you not have the bold/italic/strikethrough and justify, bulets, etc at the head of the text entry panel? I can see to attach pictures, and even quotes. It was my understanding that BBcode was just a means of manually programming the text formating. Or have I missed anything?

/Harry

Right, I think it's sorted out now

Look in your profile settings and enable BBcode - by default it's set too off.

test wrote:
blah blah blah

Jalex2 wrote:It has only 9 turns on it and reads 102 Mh on my LC meter the same as the open air coil.

Yeah... thats typical for ferrite power chokes when you try to force them to operate at higher frequencies than what they are intended for. The Q of the core is so bad at frequencies in the MHz range that they will force the measurement oscillator to go waaay lower than normal and hence give you false readings. Use a scope and a siggen. instead, with a known capacitance in parallel to the choke, and tune the peak down in the range that Ivan mentioned (i'll guess below 200kHz will work just fine).

Jalex2 wrote:...I have an oscillator running with a 6 inch coil I made and it works fine. I made another both with exactly 102 uh and put it on the same circuit and it doesn't oscillate at all. The second coil was made using a toroid and the first was just wound on a 6 inch plastic form....

Hi Jalex,
the inverters in fluorescent lamps usually work between 40 and 120 kHz. The toroid is probably optimized for this range. It may lose its quality on higher frequencies, while air coils and coils on dielectric paramagnetic formers work OK up to at least 100 MHz.

BR from Ivan

Thanks Harry
I haven't gotten back to that yet. It could be a bios problem on the circuit I used as I could see the signal on the base but not on the emitter. It has a 1k ohm to ground. The toroid I used was one of those from a fluorescent energy lights They might not be suitable for an oscillator.. It has only 9 turns on it and reads 102 Mh on my LC meter the same as the open air coil.

Jalex2 wrote:Hi Harry
I have an oscillator running with a 6 inch coil I made and it works fine. I made another both with exactly 102 uh and put it on the same circuit and it doesn't oscillate at all. The second coil was made using a toroid and the first was just wound on a 6 inch plastic form.

What would cause that?

Take another closer look at the two coils. There must be SOME difference. Are you sure that they are exactly the same with the tappings at the same point?

If there are no tappings to the coil then there must be tappings to the capacitors. You may have to adjust them a little since there will be differences in Q when you use different materials for the coil former. Ferrite has the effect of increasing inductance ramaticall for a given number of turns, which means that wire will be shorter and have a lower loss. But there could be eddy currents in the ferrite to have a loading effect and cause a reduction of the Q value. Re-adjust your feedback a little to make it oscillate.

BR Harry

(PS - Had my abdomen operation a week ago - now recovering fine :-)

Jalex2 wrote:Hi Harry
Thanks a lot. I will study what you said there a lot more and I think that will help. I see what you are doing here but where did you get the index numbers?

You take the basic frequency formula and transpose it so that you get LC = ....

Put in your frequency value in the right hand side then you will find a single index value. Simple!!

Hi Dragonforce
No I don't think it's the atmosphere. LOL
I think it's the idiot behind the soldering iron.In thin case I don't really care what frequency it is but some where around 1 mhz would be just fine but I will need to run the mixer and around 1000 hz from the incoming.
I had planned to use my new NE602 gilbert cells for this but then I cam up with this idea. I like to play with oscillators and have made many good ones but not at lower frequencies.
By the way I did get one to work with AM mod from a converted V5 so I am learning a little.

Hi Harry
Thanks a lot. I will study what you said there a lot more and I think that will help. I see what you are doing here but where did you get the index numbers?
Now I have another one for you. I have an oscillator running with a 6 inch coil I made and it works fine. I made another both with exactly 102 uh and put it on the same circuit and it doesn't oscillate at all. The second coil was made using a toroid and the first was just wound on a 6 inch plastic form.


What would cause that?

The circuit has a 330pf from the collector to emitter and a 330pf from emitter to ground and a 100pf across the coil.
I think I should use your simple oscillator in your building block page but was not sure if I could still mix the signal from the first oscillator on it's base.
I think Something is really wrong here as it looks like an ok wave form but it's on 538khz now.

http://www.sm0vpo.com/data/tune_01.htm

What interesting responses

I now use my above link for coils and caps all the time and these frequency indexes are very versatile. Simply divide the index value by a capacitance and you have got the inductance you need. It also works for superhet tuning a specific range.

For example, if you wanted to tune 1MHz to 2MHz then you have the index values 25000 and 6250. So if I wanted a coil to tune that range with 25 - 250pf tuning cap, then 25000 / 250pf = 100uH. Now at 2MHz you need 6250 / 100 = 62.5pf, so you need to add 38pf to the tuning cap to get resonance from 1MHz to 2MHz. It is approximate but it works. You can put the 40pf back into the first calculation to reduce the error.

Another method is to subtract one value from the other to work with difference (delta) values and get precise results: ie. use the tuning range and the capacitance tange to get the inductor you need. For example, repeat the above example:

The tuning range is 1MHz (1MHz to 2MHz) and the tuning capacitor you have is 250pf to 25pf. The capacitor range (variation) is 225pf.

1MHz index = 25000
2MHz index = 6250
The tuning range is therefore 25000 - 6250 = 18750.
18750 / capacitance range (250 - 25 = 225pf) = 18750 / 225 = 83.33uH

So calculate the capacitance you need for 1MHz = 25000 / 83.33uH = 300pf (your 250pf max plus a 50pf fixed capacitor).
At 2MHz you need 6250 / 83.33uH = 75pf (your 25pf min + 50pf). Problem solved

This was shown to me by Stuart Revell a few years ago (last century) and I have been using it ever since

Regards Harry - SM0VPO

Do you think it might be all that lead in the atmosphere due to leaded fuels? I remember during the days of the CB radio, having to get a can of "SWR spray". Basically, it was a petroleum based liquid, much like WD40 - one simply sprayed it onto the aluminium elements of a half wave dipole, and then wiped off the surplus liquid and metallic lead.

I think it was called Aprilla No1 or something similar - you can buy it at most CB radio shops. Maplin used to sell it but they ran out of humour

Seriously though, anything to do with coils is a black magic art, although I have found Harry's turns/picofarad rule tends to work quite well - an extra turn here, one or so less etc. Maybe it may help to take a few general notes - for example, when I was experimenting with home made IF transformers, 220 turns and 220pF worked pretty well at 455KHz, although I also used a ferrite tuning slug to trim them out. My point is, if I ever need to make another, I have notes that will guide me - very much like I know 80 - 100 turns or so will cover the MW broadcast band with a160 pF variable - 65 or so with a 365pF.

Hi,
"Perhaps Swedish air has a different permeability?" No ! The iron ore burried in the ground must make the difference ! Ha ha

VBR from Ivan

Jalex2 wrote:Hi Harry
Isn't it possible to make a frequency calculator like your's that would tell when the capacitance it not optimal for the inductance used and vice versa. I use your calculator a lot to get me in the ball park but sometimes I see the oscillators don't want to work well and I am sure that is the reason. Maybe there is a rule of thumb that I can use to make sure I will get a nice uniform wave form them. I think you talked about that once in one of your pages but I did understand it well.

One of the reasons is that the "ballpark figure" takes no account of impedance. For example you could use 25pf an 1000uH to make a tuned circuit for 1MHz, but all the wire on the coil would addcapacitance, more than 25pf. The trick is to use something a little more practical, such as 160pf and 160uH (Z=1000). In most applications I try to aim for a tuned circuit impedance of about 150Ω to 400Ω. Any higher than 1000 and the tuning becomes quite touchy.

You may want to take a look at http://www.sm0vpo.com/data/tune_01.htm which will give you a single calculation digit for all frequencies.

If you want a more broad "rule of thumb" (sorry!! I hate violence but I had to use the expression) then you can take a 6mm air-cored former and use the number of turns = number of PF = wavelength. For example, at 14MHz use 20 turns and 20pf. Now you are in the right area. A quick check of capacitance tells you that the impedance is 500Ω so use 10 turns and 40pf. Just adjust it to get the impedance you want.

I was surprised to learn that you are always "out" and that the formulas are not so accurate. Perhaps you and I are using different formers? Perhaps Swedish air has a different permeability?

As regards oscillators, you should really examine the circuit impedances and make some form of estimate when matching. The bipolar transistor base (with emitter decoupled) has a very low impedance so your tuned circuit should not have so high that it is difficult to match. I used to shed a lot of tears trying to get oscillators to "hoot" but now it seems second nature to me. 1/2 the time I know not what I have done right: I just know it is right without taking off my socks to calculate

BR Harry

Hi
Here is what I have been playing with just for fun. What if I took two oscillators tuned to the same frequency like the V5 has and feed the base of the second one with the out put of the first.. Then shift the frequency of the first like a metal detector does. Wouldn't I get a beet frequency equal to the difference at the emitter of the second transistor? This all sounds pretty crude but I think it might work.

Hi Harry
Isn't it possible to make a frequency calculator like your's that would tell when the capacitance it not optimal for the inductance used and vice versa. I use your calculator a lot to get me in the ball park but sometimes I see the oscillators don't want to work well and I am sure that is the reason. Maybe there is a rule of thumb that I can use to make sure I will get a nice uniform wave form them. I think you talked about that once in one of your pages but I did understand it well.