16:06 21/04/2005	OPT_da	Yves Monmagnon@wanadoo.fr

 ***************************************************************************
 *Version 2.1.1:                                                           *
 *Changed core database from Acces "dbm files" to "flat" text files.       *
 *Added core grade.                                                        *
 *Improved leakage inductance precision.                                   *   
 *Improved Hcu calculation now based on layers rather than total volume.   *
 ***************************************************************************						

						INSTALLATION
Create a new folder, says "OPT_da"
Unzip the uploaded file into.
Execute "Startup.exe" , follow intructions.


This is the "Standard Microshit" method. If you already use VB6 you may not need to load all those huge files.
Sources files are available too (so you can see how dirty there are) but you have to recompile.


 
********************************************  IMPORTANT ***************************************
*     THIS PROGRAM USES THE DOT (.) AS DECIMAL SEPARATOR, CHECK YOUR REGIONAL PARAMETERS      *
***********************************************************************************************	


						PURPOSE

This program doesn't pretend to compute the better output transformer, it simply helps you to design the one
which best fits your needs in assisting among near infinite iterations.

						DISCLAIMER
					(I even doubt of my incertitudes)
Great care was taken to make computation as precise as possible, however we don't live in a perfect world.
First at all, I'm less than perfect.

Predicting exactly how a transformer will work should (perhaps) be possible if it could be build exactly as
specified.
Some formulae are of the "cook receipt" type, specially for leakage inductance and parasitc capacitance.
Steel and copper specifications are subject to manufacturer's tolerance, winding execution and core assembly
are very prone to deviations.

	Will the insulation thickness between copper layers be exactly respected ? 
	Will the exact gap height respected ?
	Will the expected quantity of steel be inserted in the bobbin ?
	What kind of hardware wil be used in final assembly ?
	Is bell (or chassis) magnetic material ? Does it short out the gap ?
	Are you sure the sound that reach your ears through your loudspeakers is not affected by atmospheric pressure
	nor air moisture ?
						So long !

Consider results from this program for what there are: a reasonnable starting point and don't worry too much about subdecimals.


						APOLOGIZE

Sorry, my brain is unable to think in square mils, wire gauge nor cubic inches.

						STARTING

Launch programm as usual by double clicking on its icon or, from the file manager, by double clicking on "OPT_da.exe"

The worksheet shows five frames:
	The Green one is where you enter main specifications.
	The Orange one give results about winding and it's where you may alter wire size and turns.
	The Grey one give infos about core, and allow to select another one.
	The Blue one is where you enter data about insulation.
	The Purple one gives some useful additional infos about real load.
	The White one may receive free text.

If it is the first launch, default specs (in the green window) are as follow:
	20 hz lowest frequency
	10W 	
	700 Ohms tube's Rp
	Single Ended configuration checked
	3800 Ohms primary in five sections
	A DC plate current of 60 mA
	6 Ohms secondary with 6 paralleled wires
	a 0.2 mm gap (for SE operation)

From this specs, a solution is suggested, core data being taken from a small database (more about that latter).


					And now, YOU play !

Of course, you can change all default specs, but you can also alter:

	Standard Wire gauge
		Click in the 'Diameter' text box, replace by a new value and 'Enter'
		Recalculations are immediate.

		**********************************************************************************************		
	       	* BE CONSERVATIVE: The real quantity of copper that fits depends heavily of winding quality. *
	       	* Do some tests before real work, nothing more frustrating than an over filled bobbin !!     *
		**********************************************************************************************
		The value for 'Insulation Thickness Max allowed' takes care of the number of sandwich slices which is the 			consequence of the number of sections in each winding.
		It indicates how thick insulation can be between each slice, use a lower "Actual" value.
		Losses in copper (at specified power) are shown at the bottom center of the orange frame.
		The 'AC Only' value ignore the DC component.
		Increasing wire size reduces losses but quickly overfill the copper window.

	Primary Inductance
		Is shown in Henrys above the scroll bar at the upper center of the orange frame.
		Use scroll bar to change it as needed, recalculations are immediate.
		
	Core size
		When no other compromise is possible to fit all necessary copper, you may choose
		a larger core by double clicking on the desired core name in the list (Grey frame).
		Core name now appears in red, click on it to return at the suggested.
		You may choose to accept an higher lo frequency cut off (EG. 80 Hz for a guitar amp) that will
		allow for a smaller core.
		Another compromise is to permit higher induction, resulting in risk of saturation at lower
		frequency. 
		The induction value is given for the rated power at lowest frequency, this condition
		seldom occurs in real world. It's up to you !

	Steel grade
		You may choice between (actually) two steel grade: M6X and std (for standard).
		See "STEEL GRADE DATABASE" below to create your own.

	Interleaving
		In the green frame, both primary and secondary winding may be divided in "Sections" and "Wires".
		One section means this winding is not fractionned, two sections means the winding is divided in two
		halves, and so on.
		More than one wire means that the winding is in fact constitued by more than one windind to be
		tied in parallel.
		This affects the "Shunt Cap", the "Leak L" as well as the "Fo" values shown in the blue frame 
		according to the insulation thickness and dielectric constant.
		
	Actual load
		The table in purple frame shows the actual load reflected at the primary according to the
		frequency and the real secondary load incuding wire resistance, primary inductance and
		leakage inductance. It should be fine to separate the real and imaginary components but
		this is well over my math competences. Help needed !
		
	You may iterate as long as you are not satified.
	Beep warns you if something goes too wrong.
	Design may be saved at any time with same or different name.
	More than one worksheet can be launched at a time to easily compare solutions.
	When you launch a new worksheet, it spawns with the more recently saved values.

	You may simulate a single choke by leaving secondary impedance blank.
	Information about parasitic capacitance and leakage inductance are then meaningless

					SIMULATION

	You may want to see what happen at different frequencies or power level for the actual winding.
	To do that, click on "Turns per Volts" text to lock the actual winding so it will not be recalculated.
	
	Click again to return to normal processing.
	

					CORE DATABASE

	Core data are stored in a file named "core.tbx"
	Existing samples may be updated or deleted as well as new core may be created.
	Use the "Edit HighLighted core" button to open the edit window.
	You may then alter existing values.

	To create a new entry, type a new name in place of the current one and strike TAB key.
	If it not allready exists you will be prompted to accept saving current values under
	this new name.
	At this time you have just created an 'alias', now alter values as needed and click on
	the "Save" button.
	You may also alter values first and then change name. 
	Note that if you enter an existing name, values for this core are displayed.

	Cores are named according european convention, eg. EI96 is an 96 mm width EI stack having
	a (96/3 =) 23 mm tongue width.
	Indices, (A,B,C) when present, denotes various stack height.

					STELL GRADE DATABASE

	Each steel grade is described in a file named "xyz.dat". 
	
	They appear in a list in the grey frame.
	You may create your own data file using NotePad.
	The format is as follow and somewhat self explanatory:

		******************************************************			

			; Core data for standard lamination

			[GENERAL]
			; Max allowed induction (in Tesla)
			MaxB=1.2
			[PERM]
			; B=MU
			0=500
			0.27=3600
			0.55=5100
			0.82=5400
			1.1=4600
			1.37=3500
			1.5=1000
			2=100

		******************************************************			
	
		A line begenning with a ";" is a comment
		The section [GENERAL] contains only one key:
			MaxB=X
			where X is the value of induction (in Tesla) above wich you will be warned
		The section [PERM] contains as many keys you want, each indcating the permeability
		for a a value of induction:
			T=P
			where T is the induction (in Tesla) and P is the permeability.
		Be sure to always have a value for 0 induction.
		The program will interpolate a permeability value or the actual induction, so,the
		more value you give, the more precise is the interpolation.

				And a splendid time is garanteed for all !

