Peerless S-217-D DIY

[CJ]

we need a Pultec output,

so we build one,

here are the paper formers from a Peerless 16402, which is a very similar output,

it has no tert windings, it has more pri tuns than the 217 for more inductance,

Peerless 16402 is used in the Altec 436 tape machine i believe

S 217 D goes like this:

sec-tert-pri  times 2 since we have 2 coils,

sec - 352 turns of #29
tert - 71 turns of #36
pri - 1575 turns of #36

no E shield, core stacked 5 x 5

12 L lams, no longer available,

mu metal can, 1.5 x 2 x 2.75 , copper shield inside,

i do not know if the lams are the same alloy as the 217,  the stack height is the same,

so this thing should turn out very close to the original as we have orig formers and lams ( i hope!)

we can get the dcr and inductance and compare it to a real unit if drpat is still alive,

just sand a block of yor favorite hardwood into a mandrel,

add a drill bit up the center, then JB Weld it overnight,

mandrel should be somewhat convex so that there is a slight arch to the core tube, this keeps it from collapsing under the incremental tension of each turn,

 

[CJ]

here is the old can and the lams,

jb weld is dry, we are ready to wind the #29 sec,

notice the holes, Peerless E 204 D repeat coil has similar lams,

 

[ed rees]

Cool! I've wanted to start winding my own audio trannys. This will be a good kick in the pants.

 

[CJ]

first sec layer is 103 turns, don't ask me how i know,

you do not need a fancy machine to wind these gauges,

this layer was hand wound in about 5 minutes,

this is a very popular transformer, probably the #1 score on evilbay,

there is a trick to this sec, can you see it?

hint: the finish will become the start,

i put a layer of 0.002 mil KP (Kraft Paper) over the old rough waxy former to provide a smooth winding bed,

any rough surfaces will get magnified as bulges as you move along,

plenty of room for margins on the 1 9/16 inch long former,

there are taps on the pri and sec but we do not need them, less hassle,

and the leakage goes down because there is less tape and paper,

any time you bring out a tap, you lose valuable space for copper,

the dead guys probably had formulas for  leakage cased by taps for various cores,

 

[CJ]

4 th layer is 43 turns,

DCR is about 9 ohms,
stock is 8.9 ohms, so looking good so far,

next is a 73 turn single layer winding for the Pultec push pull feedback loops,

spread it out to reduce stray C as feedback windings are allergic to phase shift,

if the hi freq phase gets shifted too much, you gots trouble,

 

[CJ]

here is the tertiary windng, 71T #36,

stock fill for this partial layer is 28.5%,

but we can wind it any way we want, like make tiny lumped C's of different value,

is turn to turn stray C added in series or parallel?

if parallel then it will do no good, if series, then it might do something,

it could be quasi-linear, 

next is the 7 layer primary of 225 turns per layer , 90% fill factor

this will complete coil 1, coil 2 is a clone of coil 1,

you want to tape a partial layer good on an output as there is magnetic force.

go to Hoover Dam if you want to hear mechanical force in a transformer,

it is cased by current and creates mag field H, which is a force field,

at high frequencies, those wires will rattle around on the slick paper layer,

that is why we use wax,

here is 15 seconds of transformer noise>
http://www.youtube.com/watch?v=KmUKb1omkVM

guitar pickups display this mechanical feedback phenom, some people use it,

but i want to make sure no wires break during the star spangled banner,

here is an exploding transformer at 1:22 for all the sicko's who would rather see me destroy a transformer rather than build one,>  ;D
http://www.youtube.com/watch?v=rN4z8prp2VM


DCR on the tert winding is 8.4 ohms, a bit below the stock 8.7 ohms,

wire is a bit thicker

 

[MatthisD]

You mentioned doing this in the 16402 thread, amazing to see it being put together.

So the first primary layer of each coil is going to be connected together for B+ centre tap?...with one coil reverse wound.

There's probably more than one reason but is the primary wound last for higher resistance to better cope with any unbalanced DC?

 

[CJ]

bigger wire for the first layers makes a better foundation for the rest of the coil,

think wire tension, yo se less for smaller wire,

you can pull pretty hard on #29, but not #36,

if nyoy wind the #36 pri first, you can mess up the smaller windings beneath due to the increased wire tension

 

[MatthisD]

Ok, so if you were to compare with higher gauges such as a mic output transformer then it becomes less of a problem to wind thicker over thinner.

 

[CJ]

yes, as the wire becomes thinner, it makes less difference who comes first,

like the Triad HS29 inner-stage transformer, very thin wire for both pri and sec,

a big three phase transformer might use #4 rectangular wire which is about 3/8 th wide and 1/8 high, it makes a nice flat bed for the smaller round sec wires, so you really have no choice at that extreme end of wire mismatch,

 

[CJ]

if you look up the winding data for #36 wire in RDH4 (around page 1400) you will see that there is not enough room for 225  turns, but after checking DCR for the winding length of about 3.25 inches times 1725 tuns, = about 467 feet, divide by 1000

467/1000 = .467 times 415 ohms per 1000 feet (from wire table) = 193 ohms,

little higher than the 159 ohms listed but still ballpark.

if you crunch the numbers for #37 or #38, you get a very high resistance figure,
so we know that #36 is the right gauge,

DCR was 190 ohms for the first coil so that looks about right,

wire back in 1952 was a thin coat of enamel and a lot of copper,

now days we have 10 different coatings to choose from and thus the wire thickness will vary for the same gauge,

you can save tape by using the first turn to hold the insl while you tape the outside only>

 

[CJ]

for paper wound coils you have to watch out for the copper falling off the edges, no wall there like a nylon bobbin, easier t5o access the former while winding, taping is easier,

so add some border tspe to keep the windings within the margins>

 

[CJ]

using .005 nomex instead of .005 KP so we can have better hi-pot to core protection,

be sure to mark the orientation, especially on paper wound coils,

i use A for the left, most starts are on the A side,

for reverse winding, start at the B side, wind the section, then use the finish for the start,

most blueprints have A and B written on the bobbin,

 

[CJ]

ok this is coil #2, hand wound pri taking shape,

wire will change gauges on you,

sometimes you get 210 turns per layer sometimes 230,

so if you take apart a coil and find that the layers have different turns,

it is probably because the wire gauge was changing,

every spool of wire will vary as you wind, happens every 500 feet or so,

 

[CJ]

leave the coil on the mandrel when soldering hookup leads,

these cloth leads were to big so i switched them to some #22 gauge PVC.

 

[CJ]

never laced up L lams before, pretty simple once yo get a few lams going,

why the holes in the lams?

1:  lets wax into the core, where it then seeps in between all the lams,
2: flux diversion, kind of a miniature air gap, lowers perm to prevent saturation,
3: allows drive pins to be inserted to keep the core together,

this is regular grain orient, GOSS, if you want a xfmr buzzword,

stands for grain orient silicon steel,

somehow i had the idea that this was a nickel core, who started that rumor? 

if you squint real hard you can see the grain oriented  sideways to the lam,

we went with a lace 1 X 1 core for max henries, orig is probably like the 16402,

which uses like a 5 x 5 for a little air gap action,

Peerless catalog says that this thing can be used for SE of P-P,

i am guessing the 5 x 5 lace up is for SE operation,

the Pultedc is Push Pull so we can skip the 5  x 5,

 

[CJ]

core is laced up, pins installed,

one coil is mounted upside down,

why?

and who told you?

Winston O Boogie?

who told him?

so a polarity chart showing starts and finishes is a real help,

we have 14 leads coming off this beauty, a flipped coil, and a rev wind,

what a mess! there are a million ways to go wrong so you have to work slow,

for a dual coil xfmr, you go S-F-F-S for hooking up coils, this is a humbucker,

notice that this is opposite EI cores,

why?

picture a C core, flux goes around in a circle, up one coil and down the 2nd coil,

so we need to flip the leads on the second coil to get the phase right,

what is weird is that if you goof up, the transformer will still work,

but there will be no low or high end and excitation current will be 10 times higher,

this is why you need a schematic for this transformer,

read the S 217 D data sheet in the Peerless catalog,

http://www.bunkerofdoom.com/xfm/PEERLESS1953/Peerless_1953_08_0001.jpg

is says xfmr may be used backwards because the windings are balanced,

i believe they use the buzz word Astatic, which means balanced in electronics,

so there were clues as to winding structure in the data sheet,

now the pri has no reverse wind, but half of it got flipped,

so it goes  S-F S-F instead of S-F-F-S because of the flipped coil,

and flipping means upside down, not rotated on the core 180 degrees,

rotating the coil would not change the polarity, so you will never see rotation,

just inversion.

use a dot to signify a ref for current flow,

tell yourself that current moves into the coil where the dot is,

the dot can be on a start or a finish, usually it is a start,

if the dot is on the finish, this could be a reverse wind,

i verified the flipped coil by noticing the lead lengths for the hookup wires,

leads from the bottom of the can will be longer in order to reach the header,

why the flipped coil instead of rev wind?

easier.

to reverse wind you have to flip the mandrel, which is a pain because of taped leads,

so there are two ways to get a rev wind, one is to flip the coil,

the other is to flip the mandrel.

they both do the same thing as far as polarity,

but a true rev wind has the start closer to the core than  finish,

a flipped coil will have the start further from the core than the finish,

how is that possible? we are using the S for the F, remember?

ok, so is there a diff between the two? 

yes!  think stray C. leakage, that is. it will be different for the two methods,

are transformers complex? compared to what? the Hubble telescope? no,

but understanding everything  can be a PITA sometimes,

lab work will verify which is best and easiest,

you can use brass or aluminum pins if you want to get retentive about lams shorting,

but talk to a good xfmr engineer and they will tell you it is no biggy,

you might lose 0.001 henries to leakage, who cares? is that important?

no, the feeling you put into the recording is a million times more important than brass,

but i used brass anyway, 

 

[CJ]

ok, we have the smaller wires attached so we can pot this thing,

this should sound pretty close, orig core, orig coil formers, orig can, 

we will wrap some 005 Nomex around this thing, better than fish paper,

this insulates the hi voltage pri windings from the can, as it is wound last,

after that, we are going to heat up some polycrystaline (sp) and fill er up,

wax was courtesy of Doc Hoyer of Audio Transformers (Marantz and Mac)

 

[CJ]

ready to install this thing,

it will be joining it's cousins the Triad HS-29 brand new from Andy's Garage,

and a Triad HS 56 just for the hell of it,

we have an external supply for less noise,

a custom toroid with better Q at the hi end due to sectioned windings,

(ever hear people say they love the low end of the Pultec but the high is just OK?)

we also have relays running off the rotary switches for longer life,

add in caps selected by band and we are ready to race,

freq plot for the S-217-D matched the specs, flat from 5 to 60 k hertz,

rez peaks at 367 k hertz and 512 k hertz befor final roll off,

inductance peaked at 220 Henries at 30 hz,

minimum excitation current was 67 microamps@ 4 k hertz,

what does that mean? i do not know, but it sounds cool so wtf, over? 

this will be the first time i have heard a 217 D, so excited i could spurt, 

i had a Canadian tease me about my hockey puck toroid, but we get hi end with this,

 

[jackies]

sounds cool so wtf, over? 

Great job CJ! Especially the reverse windings and connections part - trying to wrap my head around that!
Do you plan to make an HS-52?