DIY OPT info needed

Hi all
I'm glad to found this community with such a great know-how.

I hope you guys can help me with some questions

1) Do we want our OPT to saturate @ low frequency, as a part of soundforming? I’ve heard a lot about „good“ OPT saturation being important for guitar tone.
If yes – is there any knowing numbers, for example
xx kGauss @ yy Hertz @ zz Volt RMS = good tone ???
I wound 2 OPTs (not that much) and both were very oversized,  like 12KG @ 40Hz at full voltage swing. Both sound OK, but maybe they could be better with bit more saturation???

Is there any rules like: for modern high gain – stiff OPT with good low end, for vintage tone – more saturation, or others? Or is it like bigger = better?
Any help would be appreciated.

2)  Primary inductance
Is important for low end response, but is frequency and level dependent. And the measurement is quite tricky.
Do we have to take care of it at all? Let me explain:

From my understanding, it is almost impossible to have too low Lpr with decent design.
Small example: For high 8k Raa the impedance of Lpr should be at least 16k @ lowest frequency of interest (say 40Hz) and @ > ½ power (from turneraudio).
This is about 65H. (80H for 30Hz) 
Taking even small core with 28mm x 32mm and path length = 168mm and calculating the required core-mue for very low Np= 1000 (formula from turneraudio page) 
I’m getting mue = 11900,  and
Mue = 5290 with Np = 1500.

GOSS (M6) will have about 10000 with 1x1 lace, right? So even smallish 1000turns primary will be good enough for 40Hz response, from inductance-point of view

So flux density (core saturation) limits the low end performance on almost all designs (using M6 steel), not the inductance.
Please, correct me if I’m wrong.

Thanks

dimashek

transformers are math heavy so spread sheets are nice,

you will wear the buttons off a TI 30XA doing B-max calcs all day long,

here is a simple sheet right here with the cell formula at the top>

Hi CJ

Thanks, i will make a spreadsheet like this, good idea.

Anyway, designing a OPT from scratch, what would be a good starting point for a flux density (and as a result: needed core size/primary winding number)? Let's say for a decent M6 iron and 40Hz lowest frequency.

Or other way around: at which frequency I want my OPT(guitar amp) to saturate @ full level?

There are many examples for Hi-fi trannys, taking <10kGauss for 20Hz...  ??? ???
I would like to see some numbers for a typical guitar amp OPT

Best Regards
Dimashek

if you go to the search engine for this site and punch in OPT  and then cj you will get a lot of guitar output transformers

you have to expand the search window by clicking on the flower thing next to the members sign.

Ok will do so

Thank you, CJ

Wow, that's a lot stuff to study

Found some OPT data for:
-marshall 50W Drake, 5150 100W, Acoustic 164, few Fenders,  and! even one Partridge.

First conclusion : 
50W Marshall OPT will have 12.78kGauss @ 40Hz full power.
5150 100W - 13,5kGauss @40Hz full power  -  That's not much saturation.

This are good starting points for my designs

P.S. very strange to see non grain oriented high loss  0.19" steel in there. 

> you will wear the buttons off a TI 30XA doing B-max calcs all day long

They are tougher than they look.

SLIDE-RULE rules! Less finger bash. Faster data entry. And if you have a rough answer already, you can quickly explore nearby solutions without re-figuring your decimal point, so the main objection to the slide-rule is moot.

> TI 30XA

If you must use the lame cheat, be aware that TI keeps changing the 30Xa's details. For a while the decimal point was too small for over-30 eyes. In that period I resorted to scarfing all I could find at yard-sales and swapping batteries and buton-boards to keep one running. This one date-coded 0515 (2015?) is not so bad so I blew the $8 for it.

CJ said:

transformers are math heavy so spread sheets are nice,

you will wear the buttons off a TI 30XA doing B-max calcs all day long,

here is a simple sheet right here with the cell formula at the top>

Click to expand...

TI sucks, HP with RPN for the win!