Transformer, 10K 1:1 - which one?

[Wavebourn]

Hi fellow smart people,
what transformer may be used for 1:1 symmetrical input from 16Hz to 20 KHz?
I tried one plugin from Altec Lansing, it is from 600 Ohm to 10K, but it is good from 16 Hz, but up to 10 KHz only, so I suppose their 10K:10K plugins perform the same or worse.

Any experiences?

[jrdmedford]

Good Lord. So many choices. Favourites:

Sowter L3575 (got mine from www.euphoniaaudio.com)
Lundahl 1544 (see www.kandkaudio.com)

Then there are additional offerings from Jensen Transformers,
Cinemag. There are also additional options in the Sowter and
Lundahl Catalog. The parts I mentioned are "budget parts" but
will run you about $120 for a pair of the suckers. The Sowter
is high-nickel cored and the Lundahl 1544 is amorphous strip core.

If you want to try cheap, there is the Edcor WSM10K:10K. They
are $11 ea and will go 20Hz - 30kHz. They are not shielded however, but in many applications, they work very nicely as input transformers. They are M6 cored.

-- Jim

[Wavebourn]

Thanks Good Lord, I've found a Guru!

Jim,

when you tried them what differences did you see?

I mean, THD and bandwidth. Loss is not critical, if it is in the reasonable range. Why $120 vs $11?
I forgot to tell, I need for +4 dB with some usual headroom...

Anatoliy

Originally posted by jrdmedford
Good Lord. So many choices. Favourites:

Sowter L3575 (got mine from www.euphoniaaudio.com)
Lundahl 1544 (see www.kandkaudio.com)

Then there are additional offerings from Jensen Transformers,
Cinemag. There are also additional options in the Sowter and
Lundahl Catalog. The parts I mentioned are "budget parts" but
will run you about $120 for a pair of the suckers. The Sowter
is high-nickel cored and the Lundahl 1544 is amorphous strip core.

If you want to try cheap, there is the Edcor WSM10K:10K. They
are $11 ea and will go 20Hz - 30kHz. They are not shielded however, but in many applications, they work very nicely as input transformers. They are M6 cored.

-- Jim

[SY]

I've used Altec plug-ins (worked better than yours, apparently), Jensen JT11P1, and the Cinemag (forgot the part number, 15:15, something like that). I preferred the latter two, easier to get the bandwidth right without ringing. Both the Jensen and Cinemag had distortion lower than I could measure from 100 Hz up, and distortion somewhere south of 0.05% at 25 Hz and 2.5VRMS. Sonically, they both performed the same, i.e., superbly.

[valveitude]

Howdy,

I contacted Edcor a little while back requesting some detailed info on their 10k/10k tranny (leakage inductance, winding capacitance, -3dB points ect.). I received an email a couple of days ago informing me that there is a scheduled run of a hundred units next week, and that they would measure a couple and let me know.

I will post the results when I get them

-Casey

[Wavebourn]

Originally posted by SY
I've used Altec plug-ins (worked better than yours, apparently),

I need to drive by a matched impedance for better power transmission (for maximum dynamic range). However, with typical OpAmp output to 600 Ohm taps they will work like a charm...

Originally posted by valveitude
Howdy,

I contacted Edcor a little while back requesting some detailed info on their 10k/10k tranny (leakage inductance, winding capacitance, -3dB points ect.). I received an email a couple of days ago informing me that there is a scheduled run of a hundred units next week, and that they would measure a couple and let me know.

I will post the results when I get them

-Casey

Thanks, I'll appreciate the info!

[cerrem]

What impedances are you going to and from????
ARe you going 500 to 500???
If i know what your source and load impedances are, as well as voltage levels...i can help spec out the transformer for the required bandwidth you prefer..

Chris

[jrdmedford]

OK, so you want to use these as input transformers to receive signal from
a 600 ohm balanced source and the use the phase split output to drive grids?
You can definitely get the Edcors to do that and do it well. In fact, if you need
stepup, you could use the 600:10K (1:4 stepup).

The Edcors work phenomenally awesome if you are driving the balanced line
with op-amps. I tried this using the balanced line driver chip from Analog Devices
(SSM2142) and the results were impressive. You wont have any problems driving
the line and the input trannies with opamps.

Level wise, you can get 10-13V out of the 10K side, so, +4dBu wont be a problem.

As far as the other options... You pay more for the nickel cores. However, nickel
will definitely sound more detailed and have better "fidelity" because it has lower distortion than iron. You will also pay more to have the nickel cored transformer put in a mu-metal shielding can. Mu-metal is not cheap. The result is a honestly well engineered device that will last a very long time and, as a result, comes with a fair price. You will also pay to get a premium transformer with symmetric windings - so that the impedance and phase of each leg is perfectly balanced. In the case of the Edcors, expect some variation due to winding geomtry.

The Edcors will pickup stray hum... So they need to be carefully kept far away
from power transformers, chokes and even filament supply wires. You can add
your own shielding (not as good as mu-metal), by putting the Edcor in a steel box.

The Edcors can be used as "Line In" for pro equipment and they can be used
as phase-splitters for DIY tube gear. Due to the 10-13V that you can get out
of them, they are excellent for driving grids of tubes like EL84, 6V6 and 7591s.

It's up to you if you want the Nickel. It's a better product, but you'll pay... I
think the Sowter L3575 is reasonably priced... About $130 for a pair of them.

-- Jim

[leadbelly]

Originally posted by valveitude
I contacted Edcor a little while back requesting some detailed info on their 10k/10k tranny (leakage inductance, winding capacitance, -3dB points ect.). I received an email a couple of days ago informing me that there is a scheduled run of a hundred units next week, and that they would measure a couple and let me know.

Which model, the WSM10K/10K, or the XSM10K/10K? I have both, I would love to see the results whatever they are.

[jrdmedford]

I've never measured stray C or L, however, I've watched their transformers on an oscilloscope. Cant measure distortion below 1/2 % with my FFT scope. OTOH, I looked for phase symmetry by putting
the scope on CH A - CH B mode and watching the difference between phases. I could only see phase imbalance at the 1mV/div setting up above 35 kHz. So, for phase splitting, these are likely doing an awesome job.

The freq response they quote is accurate, its 20 Hz to 30kHz.

They cant take any DC current imbalance in the primary except for maybe a milliamp or so. So, using them on the plates of tubes as drivers is not doable. Parafeed, of course, works just fine because there's no DC current on the primary. Even with PF, you need low Rp tubes, like 2000 ohms or less even when using a 10K:600 as a line out transformer. The Lpri is only like 4H or so. This is why opamps work so well because their source impedance is low. There's probably only a few tubes suited as parafeed drivers for Edcors - the 12B4 is one of them, possibly the 6BX7. SRPP would be the only option for using a tube driver with more common things like the 12xx7 series or 6SN7.

For input transformers you cant go wrong... I listened to a Edcors
for a week on a rebuilt XLR-out preamp and it was the bomb! The preamp was a Gemini PA-7000 refitted with AD826 buffer driving AD SSM2142 balanced driver. The amp was a custom 811A SET with an Edcor 10K:10K going balanced to SE as the line in. It was a really impressive test for a preamp with balanced out capability.

-- Jim