It's not bad but it isn't better at all.
It has lower permeability than high grade GOSS. On top of that the real cores have issues because the material is brittle and it is not easy to work with. Often parts do not match perfectly as GOSS, making leakage and minimum air gap worse. So actual permeability can be even worse.
If you take two Lundahl transformers that are identical but one with GOSS core and the other amorphous you will see that the the GOSS transformer is clearly better, including subjective sound impressions. The first impression I have always got is harshness (more overall distortion, I think) and this seems to be quite common with other users. The fact that such harsness "disappears" after a while only means that one gets used to it but in reality it is still there....I do not like it at all. It's just more expensive and not as good as Hi-B.
Reading the datasheet you cite, I find: There is one specific topology that does not suffer from the aforementioned limitations: a 1:1 interstage transformer. Typically, a bifilar wound transformer is used .....
Bifilar sure can reduce leakage L. 250KHz is not absurd.
To about 9KHz. BUT this IT would be used into "open grid", a nearly UN-loaded condition. No-load (or 100K as specced) will not run into leakage L until much higher frequency. If you want to run output grids positive.... ah, this thread is about push-pull output, and the IT-03 25 appears to be single-ended.
As for "saturation".... 140V at 20Hz is badly bent, yes. I would run 3X lower, near 50V, to stay below obvious saturation. Then we can up-rate with F. 20Hz is very rare in music. 50Hz is often the lowest strong tone. We can up-rate 50/20 to 116Vrms or 160V peak. Very few domestic power tubes need 160V peak drive. 2A3, 300B, run 40V-70V peak drive.
As for "hype"-- they have to sell the stuff, and sadly that has always meant hype.
The particular transformer that I cited (IT-03 25) has "non-bifilar winding scheme".
Bifilar winding is not a magic bullet. It improves one characteristic (leakage inductance) at the expense of the other (winding capacitance). At high frequencies, a bifilar SE to SE transformer, especially one working into a high impedance load, will act not as transformer, but as coupling capacitor. For a coupling capacitor, 250 kHz is of course not absurd.
Yes, if we regard high frequency rolloff due to leakage inductance. But leakage inductance poses the additional problem of transformer high frequency resonance, especially if the secondary is unloaded.
This is a guess that one wouldn't have to make if the manufacturer provided signal level in dBm for the rated frequency range.
All transformers have resonances so the question is just how to use it properly. Its behaviour at resonance depends on the source impedance, leakage inductance, shunt capacitance and the load. No transformer is unloaded. The power tube is quite a load! Testing is done with a passive network but it's just a test to show what it's capable of....
Having no grid resistors across the secondary is a good thing! For sure it means maximum efficiency and no waste of power no matter how small it can be. If one needs resistors a priori it means that is the only way to avoid undamped resonances.
dBm is not a unit created for and used in HiFi. NEVER. If the impedance changes the dBm value changes. It's useless for HiFi! At worst it could be used for low impedance line transformers such as 600R or 150R. But even there the common unit is dBu, I think.
There is no specific impedance for interstage transformers. It can be defined only if the secondary load is fixed. If an interstage transformer can only work with a specific fixed value resistor across the secondary to get 20KHz @-1dB or worse then it is not a good transformer in my book.
Last edited:
If 140V means saturation one can happily get at least 65-70V clean with standard cores, including EI (i.e. with 1.8T saturation induction they can run without troubles up to about 0.85T). High quality C cores are better than that and can run close to the onset of saturation without troubles.
Because magnetic saturation is asymptotic, it is impossible to tell at what exact H value saturation occurs, unless the B value is specified, for example 90%, 99%, or 99.9% of Bsat. This is why saturating signal voltage alone is meaningless in characterizing low frequency saturation distortion behavior of a transformer.
Yes-- I now see their writing is indirect and I mis-read and mis-spoke.
Perhaps Voltage (dBm is fraught with misunderstanding; has no meaning in a no-power coupling). But also THD @ that level, or we are still in the dark.
Also let's not forget that here is a comparison between a SE-SE interstage vs a PP-PP.
In the SE interstage, depending on the operative conditions, saturation @20Hz might never occur because there is also another limit on the other side which imposes that Bac cannot be bigger than Bdc.
So those 140Vrms of the MM include the DC bias caused by 25 mA anode current and if anode current is lower, say 15mA, one might never reach saturation before zero crossing.....
The other point is that one can ask the manufacturer because details are not known. With vintage stuff it's either guessing or finding out AFTER buying, in addition to the limited frequency response.
In the SE interstage, depending on the operative conditions, saturation @20Hz might never occur because there is also another limit on the other side which imposes that Bac cannot be bigger than Bdc.
So those 140Vrms of the MM include the DC bias caused by 25 mA anode current and if anode current is lower, say 15mA, one might never reach saturation before zero crossing.....
The other point is that one can ask the manufacturer because details are not known. With vintage stuff it's either guessing or finding out AFTER buying, in addition to the limited frequency response.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.