emud: Audio Output transformer Pri Open. Replacement needed.

Gentlemen,

I need an audio output transformer for this model Rekord Junior 196 (Export USA).  Emud part number BV70725 (ECL82 output tube). Primary 5300 ohms Tap @ 6.25%. Sec 4-6 ohms. 

Where might I find one?

Let me know if this is the wrong place for this request. 

Thank you,

Paul Pinyot.

To thank the Author because you find the post helpful or well done.

Exact replacement will be tough. Another Rekord Junior 196 or a similar model from the same factory and era. Or take it apart, estimate winding gauge and turns, re-wind it.

A "functional" replacement is easy. Any one-power-tube output transformer, preferably for 4 ohm load.

The "gimmick" is that tap. It isn't (as in Hi-Fidelity) for feedback. It is "ripple cancellation" for lower hum/buzz (without increased capacitor cost).

I don't think you really need it. R27 and C1B give lots of ripple rejection for low-level nodes, and the plate is not so sensitive. That the tap may just have been handy from some other model.

See attached.

I use Hammond DSE125 for example, but the "BSE125" is probably better. Another option is "Fender Champ" output transformers: they are a little smaller, are competitively priced, and the slight impedance discrepancy (7K vs 5K3) will not be a problem unless you must hit the *exact* output spec of the original.

On the original: GRN is plate. Blue is raw B+, RED goes to filtered B+. Tie Blue and Red together and call it B+. Connect your secondary and speaker.

If there's a low buzz, and you have checked C1A C1B for health, add the 100r 40uFd B+ filter to clean the plate supply. Maximum power drops <10%.

Attachments:

• Rekord Junior 196 repair (23 KB)

To thank the Author because you find the post helpful or well done.

Paul,

I am glad to see your solution.  This is just about what I did on the bench yesterday.  I have a Hammond universal in a project box that I hooked up in SE.  I used a 50 ohm resistor in in series with R27 place of the 6.25% tap.  It sounded just fine with no hum for the hour I ran the radio.  Vanode was right on the money. 

I think C1b has the hum problem covered.  And with modern new 47uf 450v Nichon caps it should perform just fine (I hope).

The Fender Champ is a great little transformer and it is a great price.  It looks like it will fit too.  I may have to strip off the outter bells.     Magnetic Components makes this transformer. 

The specs sheet show a 50ma primary current.  The ECL82/Emud is 45ma (Icathode).  There is a 8000 Z and a 5000 Z tap on the primary.  And a 4 or 8 ohm secondary.

Thanks again for the confirmation.  I'll probably order one Monday unless a used Emud Xfmr pops up.

Paul Pinyot.

To thank the Author because you find the post helpful or well done.

I'm glad it works good.

Another 50 ohms in series with 1K R27 makes no real difference to the ripple. It may get the B+ on-the-mark, but that's not too critical. The "wasted money" is of course totally moot for DIY and restoration work. (When Muntz was making lots of sets, he'd snip every part to see if it could work without it... a penny saved is pure profit.)

Since several radio makers used this tapped OT trick it may be worth discussing.

AC-to-DC rectifiers make ripple-y DC. Larger caps reduce ripple but we can't get absurd with costly caps. A 2-stage filter (such as I sketched) helps, but adds cost.

One reason pentodes got popular is that, when transformer-coupled, they tend to reject ripple. The plate resistance tends to be 10 times higher than the good-power load impedance. This means only 1/10th of the DC B+ ripple appears across the load. (With power triodes, more like 2/3rd.)

Still, 1/10th of big ripple may be audible.

The plate is only part of the total DC load. The tuner-tubes draw significant current.

If power for the tuner tubes is pulled "through" the OT, but in opposite direction to the plate current, the two ripple currents tend to cancel. They can null, but this is very sensitive to small variations in the power pentode plate resistance. It appears that most designers hoped for 1/2 to 1/3rd the ripple, 6db-10db hum/buzz reduction, 1/2 to 1/3 the cost of capacitors for "acceptable" ripple.

The "6.25% tap" is suspiciously precise. In real production, the null-ratio could vary from 5% to 10% depending on pentode grid-plate alignment, also wall-voltage variations from city to city. I bet what it really is, is "1/16". That the OT is wound with 16 (or 32) primary layers, and tapped at a whole-layer point so the tap wire does not have to run across windings.

RCA had a patent similar to this. I suspect it was not valid (obvious), but who could fight RCA? It may be quite old, may have expired by the time AM/FM radios were common.

Yes, the generic Champ OT is great for nearly any 200V-400V SE output. (110V sets like a lower impedance.) That 5K/8K 4/8 part is even more useful: with 4 ohms on the "8" terminal it acts like 2.5K/4K 50mA, which covers most 110V sets, near-enough. It is bigger than a penny-pinched 110V set would have, but it is getting hard to find the puny little OTs these days, except by over-paying for a rusty cobwebbed donor radio.

> specs sheet show a 50ma primary current.  The ECL82/Emud is 45ma (Icathode).

You probably have 40mA cathode. This OT is 25% bigger than you need. OTOH being gapped, it may go 10% lower than the rated 100Hz. It is a good replacement when exact-looks are not required (or not affordable).

To thank the Author because you find the post helpful or well done.