RCA BA73 Transformers

equal time for the solid state iron, let us rip apart some vintage RCA stuff, what up?

a couple of BA-73 preamps  were thrown into my car as i was leaving Andys garage, plus another BA-71 that will get racked,

but these 73's have been worked, all the original transistors are gone, just a bunch of ECG stuff now, so you know what that means, torch time! 

been around the block with these soldered cans, you can spend a lifetime with a big soldering iron and it will get you nowhere fast, so we need to break out the power tools,

lid is drilled out, now we need to heat it up to loosen the grip of the potting compound,

lid off, knife comes out,

we have many transformer scalps tied to the saddle, but this guy is playing Tough!

however, with a little patience and a lot of heat, we can dig that corpse up,

finally got his sucker loose, they used some pretty heavy duty stuff, the more you heat it, the tougher it gets, however, if you chip away at it on flake at at time, you will succeed,

we have 70 ea, lams of 75 EI, regular silicon steel,

pretty hefty core, this is a cap driven xfmr so no DC,

time to unwrap the coil,

this thing is wound Sec-Pri Sec with 2 copper foil shields in between,

about 300T-81T-300T, so Z ratio is (600/81)^2 x 600 = 55.5, so Pri Z is \

600/55.5 = 10.8 ohms, so they want those big output transistors to see a small load,


we will run an inductance and freq response test on a second unit,

then we tear it down and use the cans for the Fairchild 670 as they resemble the Y-6432 cans, 

pics a little latter, going to I-Hop for some green eggs and ham, 

There are at least 3 different companies that made this one over about 15 years time. 

I hate to see the input torn down, I'm always trying to find those, but maybe you must. 

Good work on all the recent analysis. 

hmmm, after a search on evilbay i see nothing,

i guess i could sell it, but then it sits on some guys shelf for 50 years and nobody learns anything,

howabout a non destructive take apart, i open it up and use the PRR method to get the turns, see what core they use, then re solder the can?

only problem would be that the coil geometry might be lost,

looks like the BA 71 uses the UTC AR-535 and these BA 73's use the UTC LR 878,

do they have the same specs?

Take it apart, I'll control myself.  They made enough of them, I think!

Yes, they use the same input spec.  There are 2-4 variations of part # over the years. 

There's another input that externally looks the same, but is marked 1:1 (600), which is found on the BA-x4 series power amps and some cart machines. 

I hate to see the input torn down, I'm always trying to find those, but maybe you must. 

Click to expand...

What do you use these in? I have a pair of ba73s that have been sitting around for a few years- I've never gotten around to fixing / racking them, because from the specs they don't seem like they would be too usable. But maybe the transformers would be good with a different circuit?
It's a low ratio input input so I've thought of trying it in a 990 or neve 1290 circuit.

Why, BA-31/41/71/72 clones, of course.  Or the early General Electric SS circuits.  Such as the JENSAMP project here. 

http://www.groupdiy.com/index.php?topic=28546.0

Likewise, I think I have a post somewhere here about the BA-73.  I've hacked another 20 dB or so of NFB into them and gotten something much more useful.  They sound cool, bigger sounding than BA-71, but noise is high with germ and that much gain.  I'm sure you could futz around with silicon transistor mods in the front end to get noise down. 

> 10.8 ohms, so they want those big output transistors to see a small load

Suspect they stole a loudspeaker-amp plan. 8, 10, same difference.

+26dBm needs 11.2V p-p on the 10.9 ohm side. It's not a huge-NFB design so they need to stay well below clipping for good THD numbers. Plus losses in bias, drive, even in transistors? 24V-28V supply seems likely and was commonplace.

-30V PSU.

yes this xfmr looks like an output flipped around, wire gauges on the 600 ohm winding, which would be the push-pull winding were tweaked to get DCR balanced, as you would see in a speaker OPT.

"+26dBm needs 11.2V p-p on the 10.9 ohm side"

i think that 26 DBM is on the sec side as 11 volts on pri would equal 11 * 7.5 = 82.5 VAC on the Sec side?

maybe 2 volts pri = 15 volts sec = 26 DBM?

B max would be too large with 11 volts on the Pri as we only have 600 turns,

here are the inductance readings>

here is the freq plot, tight engineers, barley made the 20 hz spec,>

electrical, note input level DBM ref. = 11 ohms, low B max level means more perm, 29gaM6 can run at 18 K Gauss, so we are closer to u-initial, distortion will be lower down there, hefty 1 inch stack of 75 EI barely fits in the can,>

almost forgot the coil pics, DCR was carefully balanced, outer sec was one ga wire, inner sec of just 300 turns had 2 different gauges to get it perfect at 12.8 for both inner and outer windings,
you can tell by the first pic that i had a hard time with the black epoxy pot job,  :

pic 3 is a screen, then the big pri wire, then inner sec,

ok now we are on the input>

should be able to put this guy back together in the same can, we will straighten out the lead plate,

wax comes a bubblin like bubble up merle haggard style, good and bad,
good as it gets into the seam and makes getting the header plate off easier,
bad as it insulates the soldering iron from the metal so you have to clean the tip all the time to get heat transfer,

there is another can inside the can, input shielding.