Fender Blues DeVille OPT

[CJ]

rebuilding a Blues DeVille into a 5F6a Bassman,  12AY7A , Cathode Follower, the whole 9 yards,

why not,  got  a four 10" tweed cab  already,

the amp sounds great, but we want it to be a bit better as something seems to be just a tad off,

only culprit left is the OPT, 

you know what that means,

 

[CJ]

typical Shumacher,  4P -3S interleaving, and a wood shim, 

 

[CJ]

lead documentation starts with the breakout

 

[CJ]

let's check the splices

sec

wht = 2 ohm tap, 4 coming in,  4 going to the next tap (grn/wht)    but no,

look close, if there are four wires going to the white 2 ohm tap, and four leaving, should not all four wires go to the next tap (4 ohms-grn-yel) like they do on most transformers?

but there are two wires going to the grn/yel tap (4 ohms) and two wires going to the  green tap (8 ohms)

what gives? should not all the secondary winds be in series?  no because we always see something different with every transformer we unwind.

so the four wires split off the white tap, two go to the grn/yel finish and two go to the grn  finish.

this is why it is helpful to record as much data as you can get lest you have a question or your take-a-part notes do not add up,

in this case, we used the pics and the voltage measurements and a little experience with the magic 1.41 (root 2) number that all the engineers use to figure out how to double the impedance of the next tap (multiply turns by 1.41)

also, the Fender schematic that shows color codes on the OPT can help also.

 

[CJ]

pri side,  Fender always starts with brn and ends with blu, at least on most of the blackface and early silverface stuff,

pretty good work, nice chunk of pvc for the lead plate,

 

[CJ]

i see a lot of this lately, unsafe margins on the HV side, creepage can be an issue with that left margin being so close to the core, you don't here this at low volumes, but when you overdrive the amp, the trouble starts in the way of artifacts that can be heard, they show up on the scope as hi freq hash if you are lucky enough to spot it,

turn out the lights like they use to do at the transformer shop to see the blue corona,

 

[CJ]

sec winds look good,

 

[CJ]

in the middle of the coil you often find a bi fi section,

17 turns in this case, 

why do this? probably because that is where the least flux is, as this wind is further away from the steel core, but that is just a guess.

wires are soldered in parallel,

 

[CJ]

inner primary looks nice, usually they get squished around a bit by the heavy speaker winds,

 

[CJ]

here is the print, will measure wire tomorrow and add to print, feels like 32 AWG.

hitting this core pretty hard, 14 KG at 400 volts,

50 watts max power into 8 ohms = P=I^2R = 50 = I^2 x 8,  I = 2.5 amps,

2.5 / turns ratio (to get pri current) = 2.5/22.8 = 109 ma,

50 watts / .109 = 456 volts,  so we are gonna get distortion near full power so more turns are in order.

we get 14 kilo Gauss at just 400 volts, this steel is gone by 18 KG.

 

[CJ]

rewind done,  simple winding structure P-S, eliminate 2 and 4 ohm taps,  (no need for ext cab with four 10's)

this turns insulation space into copper,  capacitance goes up a bit but a random wind keeps it under control, leakage inductance goes down as there is more copper in the window,  the coil has some "heft" now, and there is a better balance between pri and sec wire (about half and half).

used 3000 turns instead of  1870 so pri ind goes up to  286 Henries at 40 Hz (used to be 190) and the B-max goes way down to a reasonable 12 KG at full power, (down from 19 KG) so distortion at the low end will go down.

we are also raising P to P Z up to 4.6K which means less plate current for the same power.

and this is bobbin wound so there will be less worry about creepage from the coil to the core.

pri DCR goes up to 222 ohms which means a little more sag which should sound more like a tweed amp.

frequency response comes down from 100 KC to about 20 KC before a slight rolloff, this will help keep the amp stable, 

we took a bit off the stack, down to 1" from 1  3/16", this will help Capacitance (less surface area) and puts us right between a Super Reverb and Pro Reverb as far as core weight goes, (the Pro Reverb uses 100 EI, this guy uses 125 EI)  Henries still go up with the 3000 turns as L=N^2.


will compare sq wave response before and after rewind next.

 

[tonton]

I want the fender super reverb output transformer winding. Super reverb opt and Blues deville  are the same? 4P -3S, 1870 turns pri.

 

[CJ]

rewind xfmr did not sound that great,  do not know why, maybe too many Henries, should have used a butt stacked core like the Triad tweed OPT's,

but we have another wind fresh off the bench with a 5-4 interleave which has a nice DCR balance and low leakage,

also uses a bigger core so lower turns but enough inductance.  thick gauge barn roof steel,

 

[CJ]

print

 

[CJ]

leakage

 

[CJ]

sec layers do not have a wind length that goes from flange to flange, so we stick it in the middle.

if you space the turns out to spread the layer evenly across the bobbin, you get an uneven platform for the next primary layer.

 

[CJ]

filling the cavities with some cardboard from an old album cover,

 

[CJ]

19 leads to splice, mark the bobbin with a sharpie,

 

[CJ]

lace up the core and test before bakeout,

smooth response,  light peak and a dip that goes almost to zero before returning to normal V, 

 

[CJ]

and a summertime moth, species unknown,