How to calculate??????

[Geo]

Hi.
I attach a schematic from a oscillator. I want a help how to calculate the input resistance of transistor so after it to calculate the ratio N of the L1 and L2 and the AL of the pot core. L2 is connected to a one turn loop as the TM808.....
With L2 i want to mix the signal of L3 loop with the signal from oscillator.
Please your help

Regards

[Davor]

L1:L2 is a transformer with coupling ~1. In effect you are transforming the inductance of a L3 by (N2:N1)^2 ratio. The only quirk is that the leak inductance will be in series with the equivalent inductance. With good coupling that is negligible. Core calculation is done as usual with primary (L1) magnetisation current etc.
Unfortunately this can turn to be a bit problematic from the core saturation point of view. Make sure you are well within the recommended values. Good luck

[Geo]

L1:L2 is a transformer with coupling ~1. In effect you are transforming the inductance of a L3 by (N2:N1)^2 ratio. The only quirk is that the leak inductance will be in series with the equivalent inductance. With good coupling that is negligible. Core calculation is done as usual with primary (L1) magnetisation current etc.
Unfortunately this can turn to be a bit problematic from the core saturation point of view. Make sure you are well within the recommended values. Good luck

Hi.
Some values are fix (as the inductance of L1 (5mH) and the L3 who is a one turn loop). So for best adaptation of the input signal (from L3) i must adapt the Z of L3 with Z of transistor and after it to calculate the AL of potcore. This is the reason that i ask help to calculate the input resistance (Z) of the transistor.
Any way thanks for your reply.

[Davor]

This oscillator observes a transistor emitter in a switching mode of operation, and on the plus a coil is within a pi network. You may freely assume to have over 8V pp accross the coil, and that's where you start building your calculation.

In this case a classical 25mV KT/q with emitter current does not apply.

[WM6]

Hi.
I attach a schematic from a oscillator. I want a help how to calculate the input resistance of transistor so after it to calculate the ratio N of the L1 and L2 and the AL of the pot core. L2 is connected to a one turn loop as the TM808.....
With L2 i want to mix the signal of L3 loop with the signal from oscillator.
Please your help

Hi.
Some values are fix (as the inductance of L1 (5mH) and the L3 who is a one turn loop). So for best adaptation of the input signal (from L3) i must adapt the Z of L3 with Z of transistor and after it to calculate the AL of potcore. This is the reason that i ask help to calculate the input resistance (Z) of the transistor.
Any way thanks for your reply.

Regards

Geo, you decided here for Clapp oscillator.

L3 is part of oscillator here, so what sort of signal you wish to mix by L3 over L2 here?

Also, you do not need to adapt L3 to input impedance of transistor.

If value of L1 is fixed to 5mH, then L2 have to be about 50uH (better less, till say 15uH, than more).

Total sum of conductor cross section of L2 turns should be not less than the total sum of conductor cross section of L1 turns.

Impedance (practically resistance) of L2 should be no more than L3 resistance (practically L3 of about 1.5 Ohm + - 1 Ohm, so you are limited to use Alu conductor).

Ferrite pot should be out of low frequency ferrite with magnetic permeability of 2000 or more.

Self resonant frequency of L3 (coupled over L2) should be above your working frequency (you can measure this using an old HAM GDM, or signal generator in connection with scope to observe amplitude, even better if you have sweep generator and spectrum analyser).

[Geo]

Hi.
The schematic is from a commercial metal detector and i like to construct it but i have not more details from the turns of L1 and the AL of pot core. So i try to calculate all these!!!
All that i know is L3 = 5mH, turns =???, L2 = 1 turn, L3= 1 turn loop, AL=????? and frequency is near to 100 Khz.
I have a pot core with AL2800 but i am not sure if i wind it for 5mH if the ratio L1/L2 will be near to original

Regards

[Geo]

This oscillator observes a transistor emitter in a switching mode of operation, and on the plus a coil is within a pi network. You may freely assume to have over 8V pp accross the coil, and that's where you start building your calculation.

In this case a classical 25mV KT/q with emitter current does not apply.

The signal accross the coilL1 is near to 70 mv p-p only.....

[WM6]

Hi.

All that i know is L3 = 5mH, turns =???,

You need to wind 10 (or better 20 or 30) testing turns on coil body of L1, put your ferrite pot together and press it (gently) and measure inductance of your testing windings. If you get 500uH for 10 turn, then you need for 5mH: 5mH/500uH = 5000uH/500uH =10, so you need in total for L1: 10turns x10=100turns. So for L1 you need in this case about 100 turns.

It is not so critical, one turn more or less, you need to consider design rule I posted in previous post for other transformer and one-turn coil data..

AL (nominal inductance) of pot core you used is depending of ferrite pot producer. Better to follow this data LF (magnetic) ferrite type "FerroxCube" u=>2000 and then test it for turns needed as described.

[WM6]

I have a pot core with AL2800 but i am not sure if i wind it for 5mH if the ratio L1/L2 will be near to original

You first need to know if your AL values are given in uH or in mH (consult factory specifications).


For powdered iron cores AL values are usually given in uH per 100 turns.

For ferrite cores AL values are usually given in mH per 1000 turns.


Equation in first case to calculate turns you need is:

N (number of turns) = 100 x square root of the (L (inductance of your windings) / AL (of your powdered iron core)) or: N=100x ⱱL/AL


Equation in second case to calculate turns you need is:

N (number of turns) = 1000 x square root of the (L (inductance of your windings) / AL (of your ferrite core)) or: N=1000x ⱱL/AL

[Geo]

You first need to know if your AL values are given in uH or in mH (consult factory specifications).


For powdered iron cores AL values are usually given in uH per 100 turns.

For ferrite cores AL values are usually given in mH per 1000 turns.


Equation in first case to calculate turns you need is:

N (number of turns) = 100 x square root of the (L (inductance of your windings) / AL (of your powdered iron core)) or: N=100x ⱱL/AL


Equation in second case to calculate turns you need is:

N (number of turns) = 1000 x square root of the (L (inductance of your windings) / AL (of your ferrite core)) or: N=1000x ⱱL/AL

Maybe i don't give you to understand!!!!
I know how to wind a 5mH coil. The problem is if the turns that needing for 5 mH are adaptating with the input resistance of transistor .
For example from your formula....
If the AL= 1000 then N=1000xsqr root(5/1000)=2250 turns
If Al =9000 then N= 750 turns.
Who turns are the better adaptation???

Regards

[Davor]

The signal accross the coilL1 is near to 70 mv p-p only.....

Ah! I missed the 15k collector resistor. It limits the transistor current to below 200uA. With ffe of well above 400, it gives below 500nA base current. The famous formula containing KT/q says:

base ... KT/q/Ib=25mV/500nA=50Mohm
emitter ... KT/q/Ie=25mV/200uA=125ohm

In this configuration I expect well over 70mV with any reasonable tank Q, say half a volt or something. But in any case this seem to be in a butter zone for most of the ferrites, especially if you are winding that many turns. If you are making it from a toroid core, I'd suggest you to break it to facilitate winding, forget about the L2 and glue the whole thing together with L3 going through a centre of the toroid core as a single turn. This configuration is used in Harris MW broadcast transmitters. If you assume the coupling to be near 1, the only thing to calculate is the N1: N2 ratio. Since N1 is exactly 1, you only need to find N2. Your goal inductance is ~3mH, while a 10cm diameter single turn has ~150nH ... it gives you near 5000 windings for the coil... you can use just about any ferrite material at hand.

Good luck.

Edit:

I realised that 1:5000 just can't be right, and that my fingers must have been faster than my brain, and sure enough - it isn't 5000 turns.
The required inductance ratio is 1:20 000 and square root of that gives us the turns number, which is 141. This narrows a choice of ferrites to those that will exceed a target 3mH by a factor of at least 2 with 141 turns. This narrows a choice of ferrite materials a bit, but you can use the likes of FT50-H or FT50-J or FT50-W, or similar size Ferroxcube 3E25, even 3F3.

[Geo]

Davor, sorry for late answer..
I am afraid that 50 Mohm is very high resistance for a bipolar transistor....
Anyway i winded one coil on a new potcore with AL=315 and it needed only 125 turns for 5mH. Also i will make another coil in a pot core with AL=3800. I"ll connect both and i"ll see who is better.

Thanks for your help.

Regards

[golfnut]

I think you want to know what is the inprut impedance of the transistor at the frequency of operation - so you can match the Tank and transformer to the transistor.


One of the values you require is hie the transistor base input impedance - but datasheet does not list.

A BC108 is about 4kohm at 1kHz so this is a guide for you.

S