UTC A-10 and HA100-X Blueprint

UTC input transformer  1:10  mic/line to grid

these two transformers are exactly alike only different, 

main difference is HA100-X has a bigger case and 3 nested cans for better shielding,

plus a few extra lams,

winding structure is the same,

Black vintage has the Starts at the top, slightly less turns and lower perm lams,


here are some pics of a HA100-X coming apart>

where's the print? 

more random pics starting with drilling the mtn holes to pop the plate,

a few more, these are collected from an older post,

and the last  of the pics,

those dual pie sec's  are tough to wind, probably took 3 or 4 hours per coil even with an experienced winder, thin insl between every sec layer, however maybe 10 or more coils were wound on the same stick and then sawed apart with a band saw,

you could do it easier with a dual chamber nylon bobbin,

I love you in a very big, manly way, CJ!!!

Let's explore the lam difference.  Only 84% in the A-10 compared to the HA-100-X.  Practical effect?

There's another obscure set of differences.  Let's look at one particular year, to keep it a little more simple. 

A-10: +15dBm, +/-2dB 20-20K, 500-50K, no DC in pri
HA-100: +18dBm, +/-1dB 30-20K, 500-60K, 0.5mA DC in pri
HA-100X: +16dBm, +/-1dB 30-20K, 500-60K, 0.5mA DC in pri


THEN YET ANOTHER DIFFERENCE!  This is what happens when you make something for 40+ years. 

1934 catalog:

HA-100: uniform in response from 40-12K, 500-90K  $217.31, adjusted for inflation. 

By 1940 the HA-100 lists like the topmost reference. 

so many different specs... Crazy! I guess the last one is wrong at 90k. Someone got that 6 upside down!

Nope, it's correct.  The LS-10 is a different ratio spec too. 

how did Manley get dragged into this? 

good to see SSL Tech back at the table,

let us see how the 31 UI lam deals with the flux,

B=100,000,000 E / 4.44 K1 Turns Freq Area(cm^2) K1=0.90 for 1 x 1 stack,

B=25,000,000 E /Turns freq Area,  use 20 Hz,

B=1,250,000 E / Turns Area, use 888 Turns for pri on A-10, HA100-X,

B=1,408 E / Area  , now we can get flux for different stack heights,

these lams have a tongue width of 5/16", convert to cm,

0.3125 * 2.54 = 0.79375, square that for area>  0.63 cm^2

B-max  for sq stack is 1,408 E/0.63 = 2,235 E

A-10 has 31 to 34 lams depending on vintage and thickness, which = 7/16",

convert to metric = 0.882 cm^2 so B-max = 1,596 E

and HA100-X has about 37 lams which = 1/2" stack =  1.008 cm^2

so B-max = 1,397 E

B-max A-10 = 1,596 E  B-max HA100-X = 1,397 E,

so HA100-X will take, in theory,  1,596/1,397 = 14% more voltage, before saturating,

15 dbm is about 4.4 volts, 16 dbm is about 4.9 volts, = 11% difference,

so at 20 Hz the propeller heads at UTC are running these  2 nickel cores at

A-10: 1,596 *4.4 = 7,022 Gauss, HA100-X: 1,397 * 4.9 = 6,845 Guass,

a bit high for 80 Ni, which saturates at 5 K Gauss,

but like PRR mentioned, most program material is not at 20 Hz, so i guess we will let them slide on this catalog stuff,

flux increases as frequency goes down, so the HA100-X will have less distortion and roll off at 20 Hz, which is shown in the catalogs as 20 vs 30 Hz,

what is not in the catalog pages?  the effect of the 3 extra cans inside the HA-100X,

this can change capacitance to ground, which means you short out a little more high end,

it can also increase leakage inductance a bit since the thin metal can absorb some flux,

here are some catalog pages concerning these 2 guys,

where did Emrr get that 30's catalog?  i need a copy,

these two transformer models were very popular and thus sold a lot of units,

maybe this is why turns counts varied so much,

i seldom counted the same pri and sec turns even for the same,

so the turns listed on the print are an average from about 5 or 6 take-a-parts,