output transformer impedence matching

[Randall]

When matching an OT to output tubes, how close is good enough? Example, the RCA Receiving Tube Manual shows the Effective Load Resistance for a pair of 6L6s in AB push-pull mode as 5.6K ohms. Looking at a couple of transformer manufacturers catalogs, I see 30W 5K and 6.6K OTs, but no 5.6K ohm. Why not, since the 6L6 is a very popular amp tube? I recently bought a 6.6K ohm 30W at the recommendation of an engineer at Mag Comps for a 6L6 project. Why not the 5K, since it's closer in value?

Or am I not doing this right?

[bbsailor]

When matching an OT to output tubes, how close is good enough? Example, the RCA Receiving Tube Manual shows the Effective Load Resistance for a pair of 6L6s in AB push-pull mode as 5.6K ohms. Looking at a couple of transformer manufacturers catalogs, I see 30W 5K and 6.6K OTs, but no 5.6K ohm. Why not, since the 6L6 is a very popular amp tube? I recently bought a 6.6K ohm 30W at the recommendation of an engineer at Mag Comps for a 6L6 project. Why not the 5K, since it's closer in value?

Or am I not doing this right?

Randall,

See this very good lesson on transformers from Jensen. http://www.jensen-transformers.com/a...%20Chapter.pdf

The tube-seen impedance of your transformer is only valid for a speaker load of nominally 8 ohms. Your transformer may have a variety of speaker loads such as 4 ohms or 16 ohms. Nominally means the rated impedance but this will vary over the audio spectrum. What you are seeking is for the 8 ohm speaker to look like a value close to what the tubes impedance likes to see for the maximum power transfer. At 6.6K the transformer sees a turns ratio of 28.72 (6600 divided by 8 equals 825; take the square root of 825 to be 28.72). Your speaker changes impedance at different frequencies and it will reflect back to the tube by the real speaker impedance, at a particular frequency, times the transformer turns ratio squared. Don't worry too much about the exact match; just get a transformer with enough iron to put out enough power at the lowest frequency you desire.

Try to obtain a speaker with a higher sensitivity (more effective magnetic structure) as it will put out more sound per watt of applied power.

I hope this helps.

Joseph Rogowski

[Randall]

I think I follow this, but is there a typo in parenthesis?

[Enzo]

Yes, of course he meant to say square root of 825, not 625.

[jazbo8]

Where did the Ra-a of 5.6k come from? The RCA datasheet I have shows Ra-a of 6.6k, which matches up well with the maximum output and minimum distortion chart I posted recently on another thread.



http://music-electronics-forum.com/a...on-vs-load.gif

[nickb]

Also worth mentioning, the optimum plate impedance varies with operating conditions i.e bias point, plate voltage and screen voltage. In general the actual value is not that critical.

[loudthud]

The data that Jazbo8 posted above is from the older RCA data sheet for the 6L6 {metal}, 6L6G and 6L6GB. The newer RCA data sheet for the 6L6GC does indeed show that for the GC version only you can get 55W with 5.6K Raa, B+ 450V, screen supply 400V. There was a change in the maximum ratings for the GC version, higher B+ and plate dissipation.

[ReadyTeddy]

Here's a dumb question. Is the rated input impedance of an OT across each coil (i.e., from the tube plate to B+), or across the two coils connected in series?

[Randall]

That's not a dumb question at all, and I await an answer as well.

[jazbo8]

The data that Jazbo8 posted above is from the older RCA data sheet for the 6L6 {metal}, 6L6G and 6L6GB. The newer RCA data sheet for the 6L6GC does indeed show that for the GC version only you can get 55W with 5.6K Raa, B+ 450V, screen supply 400V. There was a change in the maximum ratings for the GC version, higher B+ and plate dissipation.

Good catch, had to get a copy of the 6L6GC datasheet.

Here's a dumb question. Is the rated input impedance of an OT across each coil (i.e., from the tube plate to B+), or across the two coils connected in series?

Ra-a is plate-to-plate or the impedance across the whole primary winding.

Randall - since you already bought the 6.6k, 30W transformer the chart I posted is still applicable, the Ra-a=5.6k is intended for higher operating voltages (Plate=450V, Screen=400V) and higher power output (55W), which your transformer is not rated for.

[loudthud]

[Steve Conner]

It is indeed across the whole winding. And what's more, the impedance from either end to the centre tap is not half of the end-to-end value, it's one quarter. One of the great mysteries of transformers

[Randall]

Makes sense, thanks for all the good advice.