I'm moving bravely onwards with my DC Darling project. Now that I have the voltages in a better shape (270V across the 1626, 180V across 6SL7), I'm looking into replacing the output transformers. Currently I have two Hammond 125DSEs working at around 27mA per channel. They're not so bad if you keep well below half of the max current rating.
I'm looking into James transformers. The looks certainly catched my eyes, and I've read much good on them on the forums.
Due to budget restrictions my choices seem to be the following:
JS-6112HS 70mA 2.5K / 3.5K 10W 29Hz~50KHz-2dB
JS-6113HS 60mA 3.5K / 5 K 10W 27Hz~60KHz-2dB
JS-6115HS 50mA 5K / 7 K 10W 25Hz~50KHz-2dB
I'm guessing the bandwidths are measured at max current; I will be running my Darling at 25 - 35 mA per channel, so well below saturation that bass response should be much better than the numbers indicate.
Which one would you recommend? Mainly which primary impedance would be best?
Any general comments on James transformers? Any experiences with SE James applications?
I'm looking into James transformers. The looks certainly catched my eyes, and I've read much good on them on the forums.
Due to budget restrictions my choices seem to be the following:
JS-6112HS 70mA 2.5K / 3.5K 10W 29Hz~50KHz-2dB
JS-6113HS 60mA 3.5K / 5 K 10W 27Hz~60KHz-2dB
JS-6115HS 50mA 5K / 7 K 10W 25Hz~50KHz-2dB
I'm guessing the bandwidths are measured at max current; I will be running my Darling at 25 - 35 mA per channel, so well below saturation that bass response should be much better than the numbers indicate.
Which one would you recommend? Mainly which primary impedance would be best?
Any general comments on James transformers? Any experiences with SE James applications?
I used James transformers with great success for years, and always found them to sound great.
A Malaysian Chinese friend was in their Taiwan factory and said he saw identical transformers coming down the assembly line, some getting James logos, the others getting Tamura logos. As they are way cheaper than Tamura's, way to go I say.
Regards, Allen
A Malaysian Chinese friend was in their Taiwan factory and said he saw identical transformers coming down the assembly line, some getting James logos, the others getting Tamura logos. As they are way cheaper than Tamura's, way to go I say.
Regards, Allen
I was pondering the same question.
In the end I ordered directly from Taiwan and bought the JS-6123HS. ( Don't forget to take the shipping costs in consideration. ) This one has 2.5K @ 90mA, 3.5K @ 80mA and 5K @ 70mA.
I know it is overkill (20Watt SE) for the 6L6GC SE (about 7 watts) but it offers the flexibility of changing/upgrading the design. Only mistake I made was ordering a choke which is not really big enough but I came up with a workaround.
AM
If the OP is in Finland, has he considered Lundahl in Sweden? A significant proportion of the cost of heavy OPT's is carriage. If one assumed James and Lundahl are of similar quality, the Lundahl would be better value landed in Finland.
As regards AW's comment, also mentioned in a previous thread, regarding rebadged James/Tamura OPT's, three interpretations come to mind:
Fortunately, with OPT's there's any easy way to assess what is what - measurement. I have never read or heard of a James transformer that shows the wide-range, flat frequency response of a typical Tamura OPT. Neither have I come across reports of a Tamura OPT measuring like a James OPT. Not even close. So, unless (1) or (3) are cropping up in China, is outcome (2) what was seen at the factory? It's a few years since I lived there. Perhaps some others might like to comment?
As regards AW's comment, also mentioned in a previous thread, regarding rebadged James/Tamura OPT's, three interpretations come to mind:
- James are legitimately producing OPT's for Tamura to the same specification as James OPT's.
- James are legitimately producing OPT's for Tamura to a different, higher Tamura specification.
- James are re-badging James spec OPT's as Tamura, illegitimately.
Fortunately, with OPT's there's any easy way to assess what is what - measurement. I have never read or heard of a James transformer that shows the wide-range, flat frequency response of a typical Tamura OPT. Neither have I come across reports of a Tamura OPT measuring like a James OPT. Not even close. So, unless (1) or (3) are cropping up in China, is outcome (2) what was seen at the factory? It's a few years since I lived there. Perhaps some others might like to comment?
I thought that I read somewhere that James used to make the Tango, not Tamura, transformers. That is what this guy on eBay is saying, though I do realise not to trust everything said on that site! James Transformers JS-6123HS pair! Tube amplifier Tango - eBay (item 270594635118 end time Aug-15-10 06:31:26 PDT)
Good question. You've indicated an idle point at 270V, 27mA. Have you drawn the load lines on the plate curves?
http://i69.photobucket.com/albums/i43/Ty_Bower/Tubes/1626.gif
If I sounded political, sorry, I didn't mean that way. My point was, James Transformers are made in Taiwan. There are products made in Taiwan and made in China. None was manufactured in Taiwan but stamped China and vice versa. Anyhow, let's go back to what the poster asked.
I had a pair of JS-6115HS in a SimpleSE and provided more than adequate performance. It compared well to the Transcendar TT-018-OT which is also very good but doesn't look as good. I think the TT-018-OT's provide better grunt in the bass department in higher output; however, I have no scope to measure. So, it can be just illusion from my not so trusty ears.
Hey Fred, guess i was in one of those "pedantic moments"!!Good question. You've indicated an idle point at 270V, 27mA. Have you drawn the load lines on the plate curves?
http://i69.photobucket.com/albums/i43/Ty_Bower/Tubes/1626.gif
Ty, I'm curious as to your thoughts regarding these load lines as this subject has been the focus of my latest reading, although I admit to confusion on the subject. One criterion is said to be the need not to drive the grid positive. Please give us your analysis of which slope (primary impedance) is the best choice and why. Thanks
Well, I originally started trying to compose my thoughts when I realized I had no conclusion. At least not anything coherent, so I chose to simply post the loadlines and leave the rest as an exercise to the reader.
There are many considerations. Believe that the impedance curve of the speaker itself is not flat, and you will know that the load reflected by the OPT won't be exactly 2.5K, or 3.5K, or 5K across the frequency spectrum. In other words, your 5K:8 OPT is only 5K when your speaker is always 8 ohm. It isn't.
Another factor is the idle point. Regrettably, the plate curves I used don't have the convenient dissipation limit envelope sketched on them. Often you will find that a different load might push the plate dissipation over the limit, causing you to want to choose a lower idle point for that load. You might also find that one load offers less "swing" around the idle point, at least symmetrically. Ideally a + change in the grid potential will cause a movement along the line that is the same distance as an equivalent - change. Perhaps moving the idle point would make a different load perform better. It is almost certainly easier to change the idle current, rather than change B+.
The general rule of thumb is that a higher load will offer lower distortion at the expense of lower maximum power output. Today's trend seems to be that people are willing to make that trade-off, preferring the 5K load over the 3K. In this particular case the 1626 is already power challenged, making perhaps three quarter watt with a tailwind. The designer might not be willing to sacrifice some power in exchange for better distortion figures.
Yet another thought that popped into my head was the specifics of the particular transformers under consideration. I'd always rather use the full winding whenever possible and avoid using taps partway along the winding. I've read the issues are worse when dealing with multi-tapped secondaries, and are fairly benign when you've got multi-tapped primaries. Essentially, you're going to have less than maximal coupling between the primary and secondary. You might also have additional stray capacitance or leakage inductance caused by the unused section of winding "flapping in the breeze". I'm won't argue it's a deal breaker, since many high quality transformers have multi-tapped windings. I'm just stating my preference. Anyway, after that long winded preface I'd observe that all of the James transformers in question have multiple impedance options available. I'd look closely at the specs and determine which tapping uses the entire winding, and prefer that option over the other.
Not knowing much about the 1626, I'd also think that the textbook answer for the 45 (another relatively low powered tube) is 5K. The 2A3, which is an awful lot like a pair of 45 in parallel, generally takes a 2.5K load. I'd guess the 1626 is more like a 45 than a 2A3, and probably lean towards a transformer than can provide a 5K load. Completely unscientific... sorry about that. I'm also sorry for the rambling... I did warn you about the lack of coherency.
There are many considerations. Believe that the impedance curve of the speaker itself is not flat, and you will know that the load reflected by the OPT won't be exactly 2.5K, or 3.5K, or 5K across the frequency spectrum. In other words, your 5K:8 OPT is only 5K when your speaker is always 8 ohm. It isn't.
Another factor is the idle point. Regrettably, the plate curves I used don't have the convenient dissipation limit envelope sketched on them. Often you will find that a different load might push the plate dissipation over the limit, causing you to want to choose a lower idle point for that load. You might also find that one load offers less "swing" around the idle point, at least symmetrically. Ideally a + change in the grid potential will cause a movement along the line that is the same distance as an equivalent - change. Perhaps moving the idle point would make a different load perform better. It is almost certainly easier to change the idle current, rather than change B+.
The general rule of thumb is that a higher load will offer lower distortion at the expense of lower maximum power output. Today's trend seems to be that people are willing to make that trade-off, preferring the 5K load over the 3K. In this particular case the 1626 is already power challenged, making perhaps three quarter watt with a tailwind. The designer might not be willing to sacrifice some power in exchange for better distortion figures.
Yet another thought that popped into my head was the specifics of the particular transformers under consideration. I'd always rather use the full winding whenever possible and avoid using taps partway along the winding. I've read the issues are worse when dealing with multi-tapped secondaries, and are fairly benign when you've got multi-tapped primaries. Essentially, you're going to have less than maximal coupling between the primary and secondary. You might also have additional stray capacitance or leakage inductance caused by the unused section of winding "flapping in the breeze". I'm won't argue it's a deal breaker, since many high quality transformers have multi-tapped windings. I'm just stating my preference. Anyway, after that long winded preface I'd observe that all of the James transformers in question have multiple impedance options available. I'd look closely at the specs and determine which tapping uses the entire winding, and prefer that option over the other.
Not knowing much about the 1626, I'd also think that the textbook answer for the 45 (another relatively low powered tube) is 5K. The 2A3, which is an awful lot like a pair of 45 in parallel, generally takes a 2.5K load. I'd guess the 1626 is more like a 45 than a 2A3, and probably lean towards a transformer than can provide a 5K load. Completely unscientific... sorry about that. I'm also sorry for the rambling... I did warn you about the lack of coherency.
Last edited:
some considerations :
lundahl make is own double c core in amorphous material (it's the price).Tamura use only amorphous or permalloy core. And they claim it.
Yes, amorphous core is the top quality for OPT.
James says EI core. In what material ? Response : EI core
Have a look at this test, James 6123 "is far behind" the lowest price tango transformer (U-808)
SAC Thailand.
Thanks to excuse me
Philippe
lundahl make is own double c core in amorphous material (it's the price).Tamura use only amorphous or permalloy core. And they claim it.
Yes, amorphous core is the top quality for OPT.
James says EI core. In what material ? Response : EI core
Have a look at this test, James 6123 "is far behind" the lowest price tango transformer (U-808)
SAC Thailand.
Thanks to excuse me
Philippe
Not incoherent at all, Ty. I get the need to stay within the dissipation limit and I see your point about the swing. Thanks for tour of thoughts. Like most complex things, there is no simple answer. Embrace the ramble!!!
As far as I remember James used Hi-B core for PP and SE transformers. Choice between high-quality GOSS steel and amorphous is (at large degree) question of taste and and not quality of finished product. Its possible to make both excellent products and total crapware with both.
I'm sorry to say but square wave tests mean basically nothing. Testing audio output transformers with square wave is useless, it will be never used to do so (it makes sense for high-frequency transformers used in switch mode power supplies). Instead, you should look for frequency response curve 10Hz - 100KHz, sinewave THD % measured between primary/secondary, leakage inductance and stray/inter-winding capacitance, primary/secondary DC resistance.
According to tests published on web site you mention, James has unusually high leakage inductance 167 mH, a tenfold value of other units in bench. I suspect it is a typo and they mean 16.7 mH, not 167 mH. As for the dips/peaks behind 20 KHz, it is not audible anyway.
Last edited:
- Status
- Not open for further replies.