UTC M-1433 OPT

[CJ]

thanks guys!  i bet that GE/Triad uses a butt stacked core with all the lams going one way, this would drop the perm down to 400 to 500, or about 1/4 th the perm of the UTC,  inductance is a function of perm as a linear equation, so inductance would be reduced from around 100 to 25 H.

Since this is a gapped core, inductance will be fairly linear as a function of frequency.

note that the F/B wind has 83 turns on the inside and 84 on the outside, this is done to make up for the decreased linkage of the outside wind as it is further from the core,

i will change the print to 18 dbm, thanks for that!

 

[emrr]

I don't think it's a Triad, look at the posts.  Look like Thordarson, which is what the earlier GE is.  That's the first version too. 

 

[rackmonkey]

Yep, I think you’re right, Doug. Good catch.

Thanks!

 

[rackmonkey]

there is a lot of leakage inductance with coil from from the 4 screens and the core lacing, so the 25 K primary impedance is suspect. we get more like 20 K : 600, from a voltage test, the turns ratio would suggest 16K : 600, 

maybe Rackmonkey can do a voltage test on his UTC to see what he gets with the secondaries strapped for 600 ohms,

Sorry, I missed this earlier. I’ll do it tonight.

 

[PRR]

> we get more like 20 K : 600, from a voltage test, the turns ratio would suggest...

I noted a quite high DCR on the secondaries. I ran out of thumbs, and lost my matchbook, but if a 600r load has 300r of sec and primary copper loss then 16K based on turn-count comes out nearer 25K to the tube (of which only 66% of power gets to a true 600r load).

 

[CJ]

that sec DCR is high compared to the 10 to 20 ohms seen on most 600 winds,  they used small wire, probably to make room for all those screens to keep the F/B phase going good,

the smaller wire does extend hi freq response a bit,

i made the mistake of forgetting to load down the 600 wind with a resistor which would drop the voltage enough to make the ratio = 25 K.

hopefully Rackmonkey will try using a 1 K resistor across that 600 ohm wind,

 

[rackmonkey]

That’s a variable I didn’t think about, PRR.

FWIW, I did a voltage measurement and with a measured voltage of 1V rms out of the function generator at 1kHz into the 600 ohm secondary (pins 5-8, tie 6/7) I read 5.24V on the primary. Calculates to about 16.4 k and change primary impedance, which is what The Langevin version of this OPt calls nominal (16.5k).

But PRR’s chisanbop may explain GE’s/UTC’s reasoning in calling it 25k. In another FWIW, I measured impedance using my normal quick method, terminating the secondary with 600 ohms and measuring primary impedance with a Peak Atlas Imepdance meter configured for a 1kHz signal (results of which usually align very closely to nominal value) and I got 22.8k.

BT

 

[CJ]

ok so what PRR was saying is that the DCR gets reflected with the 600, so more like 777 x 5.22^2 = 21 K i get it,

 

[rackmonkey]

I might have missed or misunderstood what you wanted, CJ. Reading a couple of your posts back, it looks like you wanted me to terminate the 600 ohm winding with 1k and then do a voltage test? Let me know if I got that right. I’ll do it first thing in the morning.

 

[CJ]

yes, feed the primary with the sig gen and slap a 1K 1/2 watt on the 600 ohm wind,

 

[rackmonkey]

Tested with a 5V rms signal on primary with secondary terminated w/a 1k resistor. Voltage out on secondary was 0.767 V.

 

[PRR]

LINE-drivers feeding unknown lines/loads typically padded the output so that "wrong" impedance only made resistive loss of level, not distortion or reactance troubles.

Note that if this is an extra 300r feeding 600r (=900r), connection to a 150r load is still 450r (half design load, not 1/4) and a dead short is still 300r (1/3rd design load). So random 50-500r loads won't sound bad, just less strong, which can usually be recovered in the next box.

A cleverly cheap tranny designer could work the series resistance of such a pad IN to the transformer just be rounding-down a couple of gauges.

 

[CJ]

DCR forms legs of H pad,  wow, those Hammies were pretty smart ! and you have two winds so you could  increase that load all the way down below 150 when using a single wind, Thanks PRR!

5 V / .767 = 6.5, now we are getting somewhere, 6.5^2 = 42.37 x 600 = 25 K!

 

[emrr]

LINE-drivers feeding unknown lines/loads typically padded the output so that "wrong" impedance only made resistive loss of level, not distortion or reactance troubles.

Note that if this is an extra 300r feeding 600r (=900r), connection to a 150r load is still 450r (half design load, not 1/4) and a dead short is still 300r (1/3rd design load). So random 50-500r loads won't sound bad, just less strong, which can usually be recovered in the next box.

A cleverly cheap tranny designer could work the series resistance of such a pad IN to the transformer just be rounding-down a couple of gauges.

Yes, lots of examples of a 2-6dB output pad for isolation purposes. 

Y'all may or may not have noticed me bringing up this question over and over again over the last decade, and this is the only answer so far, and one that makes sense.  WE, Langevin, and a lot of RCA transformers run really high DCR on output secondaries (like up to 440Ω on a 600Ω spec winding).  I had speculated about controlling response or impedance curve under varying conditions, and isolation would help do that too I think. 

 

[emrr]

Just looked at the earliest round Thordarson version of this output, also pretty high secondary DCR.