EDCOR XSE, GXSE and CXSE basic specs and measurements

[jamesdb]

I was at the Edcor site and found this post on the topic of using a higher rated opt, thought i would pass it along.
http://edcorusa.com/boards/t/149/best-option-for-4k-input-impedance-for-gxse.aspx
the idea of using a GXSE15-16-8K would not work for a GXSE15-8-4K.
The designs are logarithmic and you would not get the right voltages.

[Ty_Bower]

The idea of using a GXSE15-16-8K would not work for a GXSE15-8-4K.
The designs are logarithmic and you would not get the right voltages.

Huh? There is a known ratio of primary turns to secondary turns. The impedance ratio is the square of the turns ratio. For example, a 10:1 turns ratio gives an impedance ratio of 100:1. The voltage ratio is the turns ratio. If you put 120VAC into a 10:1 transformer, you'll get 12VAC out.

I think Brian must have made some mistake when he wrote that post. He must have been thinking of something else. You might not get the expected frequency response or power handling, but the voltages and impedances certainly should work out.

I'll be waiting for someone to educate me on this one. I just know I'm going to feel like that day in college, where the professor said "You remember that stuff they taught you in high school physics/chemistry/biology? Yeah, well that was all lies."

[EWBrown]

The designs are logarithmic and you would not get the right voltages.

I , too saw that and had a similar response

It's all a matter of primary to secondary turns ratio. "Logarithmic" just doesn't "compute" here...

A 10K:16 ratio is the same as 5K:8 or 2.5K:4 ohms. In all cases the impedance ratio is 625:1 and the turns ratio would be 25:1. The only differences might be the actual primary inductance (and its actual number of turns) and its DC resistance.

One thing that I have noted over the past few years, is that Brian is very adept at NOT directly answering questions posed on the "support" forum. He has the the polished skills of a master politician when it comes to "dancing" around answering specific technical questions. Maybe he is just overly "sensitive" about giving away "trade secrets"

I won't let that take away from my respect for his technical knowledge, and abilities, just he doesn't seem to be able (or to want to) give a straightforward answer a directed question.

/ed B

[EWBrown]

I wonder if there is also another consideration on the current rating to be made. If you slide down the ratio ladder, would you have to also derate the wattage because the current will be twice as high? Will the wire gauge handle the load, especially while in Class A?

Actually the primary current rating would be the same for all of OPTs of the same wattage ratings, IE, approx 75-80 mA for the 15 watters, 50-60 mA for the 10 watters, and 35-40 mA for the 5 watters. The higher impedance primaries apparently have higher inductance (noted during actual measurements and testing) , which means more turns, and an accordingly higher primary DC resistance. WHich also means better LF response. At no extra cosgt

EDCOR doesn't currently (pun intended) post the actual current ratings of their OPTs' primaries, but I do recall that the original website did, and I noted them at the time. Like I said earlier, maybe they're overly sensitive about divulging trade secrets... If all else fails, just use the published specs for the Hammond SE series, as a guideline. As they say "close enough for government work"

/ed B

[jamesdb]

Thanks for clearing that up. When you use this trick , using a 10K:16 as an 5K:8 equivalent, you hook it up the same way, you don't have to do anything differently?

[EWBrown]

So when you use this trick , using a 10K:16 as an 5K:8 equivalent, do you hook it up the same way or do anything differently?

Just hook it up the same way, no special "ticks" are necessary. It is simply a matter of ratios, and not absolute impedances.

I have been using technique this on an old build EL34 SE that I built about seven years ago. Its OPT is an One Electron UBT-3 3K: 4, 8, 16 ohm secondary.

I've been using it with the 8K sec taps into 8 ohm speakers (as one would normally do) but I decided to try connecting the 8 ohm speakers to the 4 ohm secondary, and the sound quality improved significantly, though the max power output is a bit lower..

Even though the original circuit design called for 2.5 to 3K OPT primary, it sounds just a bit too "raw" at times, but with the 4 ohm output into the 8 ohm speaker, the OPT primary now "reflects" as 6K, which now develops soemwhat lower maximum output power, but also MUCH lower THD, so it is a good trade-off.

I have found with this example, and other builds, that single-ended EL34s and 6550s just sound a lot better into 5K or 6K than they do into 2.5 or 3K.

/ed B

[Shannon Parks]

Great thread, guys.

Shannon

[Shannon Parks]

I'm looking at doing a GSG mod using all Edcor iron, and getting their specs is a bit frustrating.

For example:
http://www.edcorusa.com/products/509-gxse10-8-5k.aspx

This is one of the few SE outputs with inductance specs (and one that Ed has!). I assume that we are 1dB down at 40Hz. I will also assume a plate resistance of a triode like a 2A3. We'll use 800 ohms.

Bass -1dB point, ref. RDH3 page 211 and following:
(2 * pi * 40Hz * L) / 800 = 1.94
2 * pi * 40Hz * L = 1.94 * 800
2 * pi * 40Hz * L = 1552
L = 1552 / (2 * pi * 40Hz)
L = 6.18H

This is actually close to their own inductance spec of 5H. Thoughts?

Shannon

[Ty_Bower]

I'm looking at doing a GSG mod using all Edcor iron, and getting their specs is a bit frustrating.

For example:
http://www.edcorusa.com/products/509-gxse10-8-5k.aspx

This is one of the few SE outputs with inductance specs (and one that Ed has!). I assume that we are 1dB down at 40Hz.

My GSG is all Edcor iron. Well, all Edcor except the Triad choke. I've probably got the most posts in this thread:
http://www.diytube.com/phpBB2/viewtopic.php?t=3421

I used the GSXE15-2.5-6K. The "odd" 6K secondary rating is because I believe my nominally 8 ohm rated speakers were really more of a 6 ohm load over most of their range. I will note that this amp isn't quite as authoritative in the bottom octave as some of my other amps, at least compared to other single ended ones. In particular I'll compare to an amp using a set of Transcendar 10 watt 5K-8 outputs (EL34/6L6GC/KT88 power tubes).

Anyway, the 2.5K designed primary doesn't seem to have quite enough inductance for my liking. At least, this amp (with the Edcor outputs) doesn't have the same cojones as my other amp with the Transcendars. If I had the chance to do it all over again I would have sprung for something like a 5K-12 output. If they had such a thing. Which they don't.

[EWBrown]

Their published specs for the GXSE10-8-5K differ quite a bit from my own earlier measurements, and if anything theirs are more "conservative", especially the 5H inductance, which is about 30% of what I measured at 100 Hz. 5H seems unusually low, and even the humble Hammond 125SE series has a stated 9.58 H primary inductance.

My basic measurements: (from first page of this topic, with corrections)

These were taken with one of the older style open-frame OPTs, and not the newer style with bell ends.

GXSE10-8-5K:
Maximum Power: 10 Watts
Primary DCR 244 ohms,
Primary inductance 17H,
Max Current: 60 mA DC, 40 mA AC.


Theirs:

Wattage 10W
Primary (input) Impedance 5K Ohms
Primary (input) DCR 187 Ohms
Primary (input) Inductance 5H
Screen/Grid Tap 40%
Secondary (output) Impedance 8 Ohms
Secondary (output) DCR <1 Ohms
Frequency Response 40~18K Hz., <1dBu

They still do not specify the primary current rating, mine were taken from their earlier website, when some of these specs were actually published.

I will be able to make new GXSE10-8-5K and GXSE15-8-5K primary inductance and DCR measurements, in another 2 or 3 weeks, when my EDCOR order should arrive, in early July.

/ed B

[Shannon Parks]

Yeah, it could be under a certain level of DC saturation, or a different design, or something else? Working on a test jig out on the bench, straight out of RDH3, page 217.

Shannon

[EWBrown]

I can put a small amount of DC voltage (up to 12V or so) on the inductor / OPT under test, my HP 4276A LCZ bridge does allow selecting the measurement frequency, starting at 100Hz, and can apply a modest DC "bias" across the primary / inductor under measurement.

Dependin g on the actual device under test, I usually see only a little difference between "biased" and "unbiased" measurements.

I can get stable inductance readings up to 1 to 2 KHz, above that, the readings go "crazy", with increasing frequency.

It allows for setting the DC "bias" voltage, but I do not know what the actual current is - I suppose I can temporarily insert a DC milliameter in series with the OPT ' primary.

The RDH3 circuit looks reasonably easy to build, it would be interesting to see how it compares with an LCR / LCZ meter, which can test at 60 Hz..

Aother approach that I was thinking about, is to build up a simple low frequency audio oscillator, and use a simple L-C circuit consisting of the OPT primary, or choke, and capacitors of known value, measure the output frequency, and then do the math to determine the inductance value.

Kinda kludgey, but that should also work.

/ed B

[msmpe]

60mA sure seems light for 10W. Im putting together my 7695 SE-TUL and the 7695 draws 100mA at 140V for 4.5W as a SE pentode (per tube manual). I'm using the OEM OPT, but I would consider a pair of Ecors for a later upgrade - 'cept 60mA don't cut it!! Guess I'd have to go with the 15W GSX.

[EWBrown]

Definitely the GXSE15 would be the better choice, the price isn't all that much greater, and the current handling capability is at least 25% better, and the GXSE15 has a significantly higher primary inductance than the GXSE10. (see my next posting of July 1, 2011)

I wouldn't try to push the OPT to 10 wats, probably 2 or 3 is the most power before things start getting a bit strange Use the GXSE for greater than 3 watts, up to 5 or 6. Above that, some more "serious" and pricey iron would be necessary.

I'd SWAG that 60 and 75 mA are fairly conservative ratings for the GXSE 10 and 15 series, though how far these ratings can be exceeded without core saturation has yet to be determined by actual in-circuit testing.

If more plate current is needed, then the CXSE series (25W) should handle the job, though at a significantly higher price.

/ed B

[EWBrown]

My EDCOR OPT order just arrived (29 days friom initial ordering to delivery). Four GXSE10-8-5K and four GXSE15-8-5K.

So I decided to do some quick primary measurements, DCR and L @ 100 Hz.

GXSE15-8-5K:

Pri DCR: 161 Ohms Red - Blue
Pri L @ 100Hz: 17.5H
Pri L @ 1KHz: 12.5 H

GXSE10-8-5K:

Pri DCR: 188 Ohms Red - Blue
Pri L @ 100Hz: 9.6H (yes, these have a lower primary inductance value)
Pri L @ 1KHz: 5.8 H (close to their published value).

Primary (input) DCR 187 Ohms
Primary (input) Inductance 5H

This is also significantly different from my earlier measurements of the older style, open-frame GXSE10-8-5K:

GXSE10-8-5K:

Primary DCR 244 ohms,
Primary inductance 17H,
color]

So, it appears that the newest version of the GXSE10-8-5K is not the same, as the original version, (larger core) GXSE10-8-5K, which, IMHO was a far "better" ttransformer, overall.


And a single GXSE15-16-10K - from Pre-Bay:

Pri DCR: 301 Ohms [color=red]Red - Blue
Pri L @ 100Hz: 29.0 H

I had tested the GXSE15-16-10K previously, and the readings were idenrtical (at least my LCR meter is still on cue). These could be a good alternative for a 5K:8 ohm SE OPT, with higher DCR and inductance values.

HTH

/ed B