Marshall JCM 900 OPT 100 Watt Dagnall C3070

we need to rewind this, so that means we need to get the winding blueprint,

Marshall gets off cheap by using only one end bell, but they have to chop a big square in the chassis for the  side mount wiring scheme,

not the most fun in the world, de-lam time,

burned by the heat gun, varnish fumes leading to an early death,

fortunately there was only partial impreg on the processing, but the lams were laced 1 x 1 so more work than a 3 x 3 and the lams bend easier, have to worry about messing up the grain orient as we don't want to lose perm,

but we can add a few turns to compensate,

heat gun is resting on a core from the infamous W.E 111c transformer,

finally over,  did not need to bend up many lams,

photo doc the breakout,

unbelievable, they used masking tape for insulation on this thing,

first time i have seen this,

outer pri has big margins, this could be a capacitance tweak or a high voltage tweak, or both,

masking tape used in between pri and sec, not good, low budge,

too much moisture that won't dry out in a 220 oven for 2 hours,

plus the sec wire is big and you need to protect the pri layers underneath from all the torque needed to wind the #14.

getting into a sec section,

have to be careful when all three wires go out at the same place, some go to one lug, some to another,

thought there was a wire shield going on here as it appeared that a layer had the start and finish connected to the same place, but this turned out to be false upon further investigation,

tap for the feedback resistors to come off off, seems like a lot of work to do this, must be a reason, Partridge did some great work on Marshall transformers in the early days, this design might be one of his,

middle pri is wound all the way to the bobbin flange, no margins like the outer layer,

random winding, wonder if this is written  out in the original print,

inner sec, had a crossover on the right side,  seems like you would want this heavy wire to lay flat,

don't like to see crossovers on the first layer, even in a random wound coil, heat can expand the core and put pressure on that first layer, so we like to wind it flat and loose, as this is where  90% of the transformer and choke failures occur,

ok so we have a pri-sec-pri-sec-pri  job with 820T:80T as the ratio,
10:1 into 16 ohms, 20:1 into 4 ohms,

did you see how big the stack is? (look at the bobbin) so not many turns needed on this 125 EI OPT, but we might add a few to get some bottom end going on, and use Nomex 410 instead of masking tape, sheesh! 

oh wow, Neil Young Bridge School concert, sound garden, pearl jam, brian wilson, neil, and.........................

Tom Jones!  ???


print coming up...

here is the print for this particular transformer,

another spec sheet on this model # calls for 310 turns and 620 turns for pri sections instead of 205 and 410, and 120 turns for the sec winds,

let's rewind this contraption,

1'st layer going on, we are going to use #26 instead of #27, no need to play with wire sizes to exactly match DCR in a git OPT,

plus, since the first layer is the most likely to fail,  we can get a little insurance with a thicker wire and lay it down flat with no crossovers for even more insurance,

rag keeps corrosive sweat out of the coil,

re-wrapping the heavy #14 sec wire was a no go, don;t want to pay and wait for new copper which will be almost as hard to wind, this heavy wire needs a ton of tension to wind right, stresses out the machinery and stresses out the thinner pri wire underneath,

what to do.....

borrow idea from infamous Peerless 16431 Williamson OPT, use two sets of quad-filar #26 for the sec speaker wind, 16431 is flat to 100,000 cps, 

lot of work splicing, kind of a pain to wind, but the winds lie down nice and flat so we won't run out of winding room as we build this thing up,

quad wind, using #26 for the whole transformer, nice to have red and green, just de-reel wire onto a couple of small spools, enuff for one transformer and you have 4 spools, take a small piece of plywood and hack it in two, drill for wood screws and put the four leads in between the wood, this keeps the wire flat so the leads don't cross over while winding. massage the wire around the corners to get a tight coil,

going to keep the margins the same on the outside pri like original just in case this is a trick to limit surface area (capacitance) next to the sec wind, pri and sec winds form a capacitor which is excess baggage that we don't want to claim,

last layers, looks like we have just enough room to finish this thing, hate to have and tear it down again to reduce turns and do more quad winding,