Ampeg Portaflex Power Transformer

today we are going to have some fun with a transformer,

full props to a GDIY member for taking the time and money to ship this out just for us,

here is the transformer after the toaster oven>

it just oozes out from it's cage, about a millimeter an hour,

that's because of the sticky icky,

Houston,  we have separation of the lunar module,

enuff toaster oven heat for extraction And delamination,

finally down to the last of the Mohicans, now the rest is cake.

fingers lookin like we just chained smoked a carton of Lucky Strike non filters,

LSMFT-Lucky Strike Means Fine Tobaco

now we put the coil on a mandrel and unwind it while counting the turns,

we will save the wire so we can get the DCR as this transformer had some problems with shorted turns,

unwinding it will un-short the wire and prevent meter error from the core messing with the digital chip,

after we get the turns, we can calculate flux from the lam catalog, 138 EI in particular,

Mag Metals specs this core at

Bmax=30.8^3 /K1N Gausses per volt at 60 Hz, which is where we run this thing,

K1 stacking is 0.90 - interleaved one per layer

this is what we have so far: 

here is a print, still need wire ga will measure today,

weird structure, HT wind starts with  the CT, ends at the plate terminal, then we start with the other plate terminal and finish with the CT which gets spliced to the first CT start, buries the hot ends, maybe done to limit radiation into audio,  nice, but you have to watch insl as you have 780 VAC right next to each other,

heater winds on top of everything, risky business as pressure can damage small sec wire underneath which is probably what happened to this guy,  oh yeah, PRR says bigger radius = easier winding,

notice Ampeg says 117 pri on orig schematics, so 390 VAC becomes 400 at today's 120 VAC pwr line,

10 volt increase which means 14 volts extra plate voltage after filtering, if you already run tube at the limit then you risk failure as 470 becomes 484 volts plate

so we can wind 1007 turns on each sec instead of 1033 to compensate, or lower if we want a solid state bridge,

pri wire is heavy, only 2.3 ohms, good deal, steel is barn roof .018, so low Henries,  means we have maybe 30 ma excitation current plus tube power,

nice powder coat can with this xfmr,

Wow CJ, that was fast!  And yes, it was a short in the HT which caused this one to become an organ donor  ;D

What are you going to use to re-pot this?

Dreams said:

What are you going to use to re-pot this?

Click to expand...

The replacement from fliptops.com looked like it was potted in beeswax.  As long as it stays cool that should be fine.

gonna wait for a roofing company to pour some black tar and sneak up behind the truck,

ever get stuck behind one of those trucks in rush hour? sheesh!

ok measured the wire, it goes 321 pri #29 HT sec and #16 for the heaters, (AWG)

> notice Ampeg says 117 pri on orig schematics, .... today's 120 VAC pwr line
> so we can wind 1007 turns on each sec instead of 1033 to compensate

You "should" re-figure the 6V and 5V turns also. Which gets into part-turn numbers, no can do.

You are still running higher flux than when first designed.

If you want a 117V design in a 120V world, you perhaps should take the correction on the "117V" side. 310 turns, times 120/117, is 317.95 turns. Try 318 turns: "117V" becomes 120.02V. Since "120V" is still a rough approximation (125V to 115V are likely), 318 turns is a best-bet primary. Only you can say if another 8 turns will fit with #21, or you must go to #22.

To really future-proof it, throw on 340 turns with a couple of taps about 11 turns apart. 318, 329, 340 gives 120V 125V 130V.

0.377 Volts Per Turn is a fairly large lump.

that's a good idea, you see so many transformers running heaters to high,

you could try different heater wire sizes, using DCR to tweak voltage but copper is 15 bucks a pound thanks to hybrid cars,

plenty of extra room for primary wire when switching from layered paper to nylon bobbins as you do not have to have those margins, just bobbin thickness which is .04"


here is the primary,  thin poly wrapper, Nomex would be good here,

here is the secondary,

this looks like the reason for failure, heavy heater wire was pulled down too tight over the secondary,

probably started out as a shorted turn then went open,

staple plate for the pri/sec leads which sit under the heater layers,

good strain relief if you don't wind too tight,

5 volt wind, remember that the 5 volt wind will have high voltage superimposed on it from the rectifier tube, if you use a directly heated cathode,

that 6.3 wind sits right next door and has a path to ground, those 100 ohm virtual CT resistors for hum cancel,

so leave some room in between and don't spare the cloth tape,

no lead connections, they just use the actual winding wire as the leads, more dependable, saves labor and does not add to coil build which might hit the core window,

measure your turn length so you can order just enough of this heavy copper to do the job,

thicker wire is cheaper as far as price per pound as the bigger the wire, the easier to make,

different anchor style for heavy wire, usual method for medium gauge wire is to strap 2 or 3 turns together,

but this stuff is so heavy that it has a lot of spring to it, so you strap it around the coil,

a sideways anchor might loosen up around the last few turns,

and here is the final wind, 6.3 volts, 18 turns, same size wire as the 5 volt wind,