Hi Juan

first of all: it is important that the transformers in question fullfill the safety specifications that your end product must meet.

there will be 4 of those transformers between mains connected full bridge and the dc-link, and it makes the coupling capacitance 4 times bigger than single transformer alone.

I tried to find out something about the transformer from Ti website:

TMDSHV1PHINVKIT

http://www.ti.com/general/docs/litabsmultiplefilelist.tsp?literatureNumber=sprabr5

tells  that this transformer is used in driver boards for DC-DC converter having 15V in and +15/-12V out.

some info is in the schema Solar_HV_DC_AC_Sch.pdf but not for the driver board

from BOM one can find out that converter is controlled bu UC3845

http://www.ti.com/lit/ds/symlink/uc3845.pdf

there is a 1nF capacitor C10 it might be CT (pin 4 in soic8)

there are some resistors that one might be RT (also pin 4 in soic8)

if 15K resistor R1 is RT, frequency is 100KHz

then it should be straightforward to reverse-engineer the strange forward converter that does not have inductor in output.

I have not found further information about the driver boards, but it seems that there is a IGBT driver/optoisolator

and this transformer supplies power for the live side of the opto I cant tell from the picture which device is used.

BOM seems to list (Toshiba´s) TLP350 IGBT driver that does not need dual powersupplies, but single.

I looked at (Avago´s) ACNW3190 datasheet, and that does not need dual powersupplies, but single.

maybe the opto is used with +15V and -12V, that would still be 27V that is inside recommended range of 15 to 30V

for TLP350.

as you have the kit of  TI, you could measure the current running into one of the driver modules 15V supply

and find out the average power input, also you can measure transformers inductance and look from your favourite inductives supplier for such ready made transformer.

frequency can be easily measured by having small isolated coil touching the EE10 transformer when it is running,

and of course you could relatively safely measure input waveform to verify if it runs at 100KHz or some other frequency.

(such as http://www.coilcraft.com/prod_forward.cfm )

also for rather small powerlevel (below 4W ) flyback might be considered:

http://www.coilcraft.com/lpd5030v.cfm

sorry that I could not shed more light to this but without schematic it is difficult to tell.

I hope someone from Ti would respond too.