Midwest P-5417-1 OPT

Used in 1962 Lowrey Organs,

output transformer for a pair of 7591 power tubes,

about 3.8 K to 8 Ohms,

never taken apart a Midwest xfmr before, as usual, we found something unusual with the coil structure,

luckily this xfmr was processed with wax only, so the lams come apart real easy with no tugging with needle nose pliers which is nice as we can rewind with un-damage lams,

cool insulator pad, never seen it done like that , saves taping ,

no speaker taps, so just plate-ct-plate and + and - spk leads,

coil leads usually hang out the right side as the motor is usually on the left,  so that is the way we mount it for unwinding,  if we guessed right, then the counter will run backwards.

why unwind it like this? so we can document the starts and stops the way they were on the print,

this can be important if there is a single layer wind,  or a reverse wind.

outer insulation being removed,

the lead plate has staples which is pre-fabbed before soldering,

here is where things get weird see those small wires attached to the speaker leads? 

what are they?  possibly wire screens? maybe, let's find out,

first section is one of the secondaries, wound with two pieces of #25 AWG, which is the sign of a good transformer. the wire lays nice and flat which reduces leakage inductance and limits damage to underlaying primary wires which are only .007" thick,

next section is a primary which is weird, as we expect the wire shields to be here. shields usually go between the primary and secondary, and are rarely seen on power output transformers, 

wire size is #34 AWG  which is a bit light for 7591 tubes,

nice winding job winding length is about 1 3/16" inch,  old school wire with the dark red enamel,

about 150 turns per layer , 1,440 turns for each primary half so 2,880 T total.

DCR is about 300 ohms.

ok so the weird windings that get hooked up to the 8 ohm secondary are located between the two primary sections, not the primary and secondary. 

there are two of them,  1 layer each, 150 turns, and both ends get hooked up to the secondary, the starts go to the black speaker lead, and the finishes go to the blk/yel speaker lead.

so they are not shields, they are lime a middle secondary, only  the wire gauge will not be passing much current.

the speaker wind also has 150 turns, so these #34 winds are in phase with the secondary, have the same amount of turns, and thus carry the same voltage as the speaker wind.  to see what effect this has on the response of the transformer, we are going to rewind the coil with leads for these two winds so we can connect and disconnect them to the

the inner primary wire is pretty depleted so after 57 years it is  time for a rewind.

first section of coil is a secondary, bi-fi #25 AWG so the final secondary, when the leads are connected will be a quad set of wires,  DC R will be different, we use a DC power supply to force about 1/2 an amp thru the wires and measure the milli-volts across them to get an accurate measurement which is faster and cheaper than a four wire ohm meter.

inner sec DCR was 0.937 ohms and outer was 0.937, so in parallel we have 0.555 ohms total sec DCR.

here is the print which shows what is going on with the weird winds,>

here is primary inductance vs frequency at low voltage, about 25 V-rms  from the HP 200CD

and primary inductance vs 60 Hz variac at various voltages,

here are some sq wave shots, sq wave into primary, scope on unloaded secondary,

ok we did some tests with the mystery winding, here are the results,

first pic is sq wave at 10 K before and after winding connected, no load on secondary,

not much change except the over ring spike is reduced>

unloaded 100 K before and after>

same thing 1 M Hz