What would be my output transformer imp?

[DeathRex]

I have a UCL11. The specs are:
Va = 200V
Vg2 = 200V
Vg1 = -8.5V
Ia = 45ma
Ig2 = 6ma
S = 9ma/V
Ri = 18KOhm
Ra = 4.5KOhm
Wo = 4W
Wa = 9W

The Ri seems low enough to use a 2.5K to 3.3KOhm transformer. Is that right?

[Geek]

Hi,

You need plate curves to dertermine impedance.

Without it, the hand way is a THD and power meter and various loads on a transformer of known primary impedance. You will find an anode load of lowest THD and one of maximum power. Most commercial designers set the load for the point between the two.

Cheers!

[DeathRex]

UCL11

[DeathRex]

I guess with 100 volts on G2 it's really not linear.

[dcriner]

I'm a little confused by the question. But here is a typical way to select an audio output xfmr:

Look in a tube manual for the plate resistance of the tube in question. You will want the load impedance reflected back to the xfmr input to be about twice that plate impedance. That reflected load impedance on the plate is the speaker impedance multiplied by the square of the xmfr turns ratio.

If you know the speaker and plate impedances, you can compute the appropriate turns ratio -- plus or minus 50% should be fine in most cases.

The xfmr turns ratio is the important thing. Audio xfmrs in catalogs often have specified input and ouput impedances, but that is somewhat meaningless except for one particular pair of impedance ratios.

Theoretically, the maximum power transfer occurs when the plate impedance and the plate load are equal. But because of distortion considerations, etc., it's customary to have the plate load impedance about twice the tube's plate impedance. The data in your original post didn't seem to include the speaker impedance - without which, we can't select the xfmr turns ratio.

[Geek]

Here ya go!




(diagram corrected 26-5-14, 1200h PDT)

Both will need NFB to linearize them for SE use. Personally, I'd go with the lower Z and deal with less power for more linearity (IF the tube can handle 50mA continuous bias).

[DeathRex]

Here ya go!




(diagram corrected 26-5-14, 1200h PDT)

Thanks Gregg. I need to learn how to do that.

Personally, I'd go with the lower Z and deal with less power for more linearity (IF the tube can handle 50mA continuous bias).

Wouldn't lower Z give me more power than a higher impedance?

[Geek]

Thanks Gregg. I need to learn how to do that.

No problem!

SE curves are easy and the tutorial given in the beginning chapters of RCA tube handbooks are a great place to learn.

Wouldn't lower Z give me more power than a higher impedance?

To a point. The science behind it is voltage swing across the load. If the tube can source enough current to swing lotsa volts into a low load, then yes.

Check out a detailed 6L6 or EL34 datasheet for example and it shows tube THD and power levels over a variety of loads. Below optimum load, power will give a peak and then drop far more rapidly than going to high in load. It actually looks like the response curve of a bass-reflex enclosure thet's too small.

Triodes and pentodes are a little different... triodes are more forgiving on the low-Z end of things (a 2A3 for example will be happy with anywhere from 2K to 5K load) and pentodes more forgiving on the high-Z end of things (an EL34 will crank out fine from 4.2K to 11K pentode or UL).

Cheers!

[dcriner]

The tube characteristic curves will define the plate impedance of the tube. The maximum power transfer occurs when the source impedance (the tube's plate resistance) is the same as the load impedance. The load impedance is the speaker impedance (which hasn't yet been mentioned) multiplied by the square of the xfmr turns ratio.

If the source and load impedances are different, no matter which is the larger or smaller, power transfer will be less than the maximum.

Fine point: the maximum power transfer actually occurs when the load impedance is equal to the complex conjugate of the source impedance. But, for practical purposes, we can compare only the impedance magnitudes. And, considering distortion, it's best to have the load impedance higher, perhaps two times the source impedance.

[DeathRex]

The load impedance is the speaker impedance (which hasn't yet been mentioned) multiplied by the square of the xfmr turns ratio.

The speakers are about 6-8 ohms. I'm looking at an Edcor XSE15-8-3.5K or GXSE10-8-3.5K or GXSE10-8-5K.
At 45ma the Ra = 4.5KOhm. How does internal resistance figure into this? The UCL11 is 18KOhms, very close to the EL34's 15KOhms.

[dcriner]

The speakers are about 6-8 ohms. I'm looking at an Edcor XSE15-8-3.5K or GXSE10-8-3.5K or GXSE10-8-5K.
At 45ma the Ra = 4.5KOhm. How does internal resistance figure into this? The UCL11 is 18KOhms, very close to the EL34's 15KOhms.

How does what internal resistance figure into what? In the case of speakers, impedances are normally stated at 400 Hz, and you can forget the DC resistance. If you are talking about the DC resistance of the xfmr windings, you can neglect that, too.

Taking the Edcor as an example, the nominal ratio of impedances is 3.5K/8 = 4,375 So, with, say, 8-ohm speakers, the reflected impedance to the input of the xfmr, is, indeed 3.5K. But, for an EL34 as a Class A1 amplifier, the plate resistance is 15K and the recommended load is 2K.

In push-pull, Class AB1, the recommended plate-to-plate load impedance is 6500 ohms.

This data is from the RCA Receiving Tube Manual, RC-30 edition. No need to dive into tube characteristic curves.

[DeathRex]

How does what internal resistance figure into what?

The internal resistance of the tube figure into what load it will drive.
UCL11:
Va = 200V
Vg2 = 200V
Vg1 = -8.5V
Ia = 45ma
Ig2 = 6ma
S = 9ma/V
Ri = 18KOhm
Ra = 4.5KOhm
Wo = 4W
Wa = 9W

[kheper]

Transcendar has a number 3K Single Ended outputs.

http://www.transcendar.com/2.html

Generally - in the datasheets - given the same plate voltage, the higher the plate impedance, the higher the recommended primary output impedance, but look at an 807 on page 4:

http://nekhbet.com/STC807.pdf

22.5k plate impedance and a 2.5k output impedance - about the same recommended for an an EL-34.

Using an output with a higher impedance (or placing your 6-8ohm speakers on the 4ohm tap) will maximize tube longevity. If you find a 2.5k output, I'd place the 6-8ohm speakers on the 4ohm tap. Maybe the Edcor 5k to 8 output would be the best compromise for this tube and your speakers?

[dcriner]

I think the plate resistance, as listed in the tube manual, etc., includes the whole enchilada. It is the dynamically measured a.c. resistance between the cathode and plate - the ratio of an incremental change in plate voltage divided by the resultant change in plate current.

[EWBrown]

http://frank.pocnet.net/sheets/030/u/UCL11.pdf

Looks like 4 watts output, with a 4.5K OPT. The more commonly available 4.3K or 5K primary impedance ,should be good with this particular kind of tube.

Definitely will need to use fixed bias for both triode (-2v) and tetrode (-8.5v) grids, since it has a shared cathode for both sections.

Interesting tube, 1942 WWII era. And, it takes a really unusual socket...

There is also the ECL11, 6.3V, 1Amp filament:

http://frank.pocnet.net/sheets/030/e/ECL11.pdf

/ed B