I know. Monolith Magnetics does that as well (it's called IT-03) but FR is not as good (still very good nonetheless) and is more expensive.
The only potential drawback of the bifilar transformer is the max anode voltage. However wires have improved a lot over the last few years and unless one needs more than 300-400V anode voltage , depending on the specific transformer, I can't see any limitation. The IT-01 can safely work with P/S DC voltage up to 375V which is enough for driving the 211 biased at minus 57-60V. 300V anode voltage, -20V bias for about 16-17mA for the 10Y will be more than enough. If self bias is used for the 211 in class A1, then 350-360V anode voltage for the 10Y can be used without any trouble.
The only potential drawback of the bifilar transformer is the max anode voltage. However wires have improved a lot over the last few years and unless one needs more than 300-400V anode voltage , depending on the specific transformer, I can't see any limitation. The IT-01 can safely work with P/S DC voltage up to 375V which is enough for driving the 211 biased at minus 57-60V. 300V anode voltage, -20V bias for about 16-17mA for the 10Y will be more than enough. If self bias is used for the 211 in class A1, then 350-360V anode voltage for the 10Y can be used without any trouble.
Last edited:
Yes and no. Comparing the Monolith's that have certified performance, the non bilifar transformer will only reach 100KHz with lower source impedance than 10Y. I am referring to -3 dB point here. However inductance is lower at 75H and it costs 304 euros each vs 273 euros needed for the IT-01. That's 60 euros difference for the pair. Not really negligible...
I would not even start to make the amp with a transformer that bad!
Anyway it is not the case here. The Hashimoto will do -3dB @45-50KHz. The two transformers, the IT and OT, will add up. So why waste the time in getting a better FR from the OT if then such effort is wasted with the IT?
I only use IT's and input transformers that do not compormise FR at least up to 40-50KHz.
You can prefere it but 375V of the MM is a safe voltage which means is well under the real limit. There are third party standards test to certify it. If one doesn't do such tests then it's guessing. In the latter case I would agree with you but it's not the case of MM.
I am referring only to my own stuff.
With 100 KHz for the "normal" IT and 120 KHz for the bifilar IT, my choice is simple. Some 100H primary inductance (meant for 10Y).
Agree that with todays qualities in magnet wire there is little risk with bifilar, but as transformers are expensive parts I go for optimal reliability.
As for "bad" 211 OPT's: with not reaching 20 kHz flat I mean 20Hz-20kHz -1dB (15k primary impedance). A whole different league than -3dB at 45Hz for a 10k transformer which is not even HiFi spec to begin with....
With 100 KHz for the "normal" IT and 120 KHz for the bifilar IT, my choice is simple. Some 100H primary inductance (meant for 10Y).
Agree that with todays qualities in magnet wire there is little risk with bifilar, but as transformers are expensive parts I go for optimal reliability.
As for "bad" 211 OPT's: with not reaching 20 kHz flat I mean 20Hz-20kHz -1dB (15k primary impedance). A whole different league than -3dB at 45Hz for a 10k transformer which is not even HiFi spec to begin with....
I can't see and don't know that. So no comment.
If one asks an opinion I reply with what I know. You should reply to him not to me....
Again....regarding MM, the normal IT does about 80-90 KHz -3 dB with 5K while the bifilar does 110-120KHz. This is not a big deal ok but it is better nonetheless.
However the normal has 75H while the bifilar 100H. The bifilar is over 60 euros cheaper for the pair. There is no reason whatsoever for going for the normal one if there is no trouble with the one and only limitation of the bifilar. If I needed 400V plate voltage then the normal one would be necessary but this is not the case.
There is no risk if things are done properly. The only risk is going beyond recommended operation. It is the same as the capacitors. Why aren't you worried about ALL capacitors? If they fail they might also blow transformes.
The Hashimoto is flat to 20KHz and has the full bandwidth for modern digital sources up to 50 KHz. So it is HiFi.
No point in discussing this IT thing further 45; both approaches, well executed, will give by far good enough results.
I understand that modern digital sources start at 40 Hz??
The output transformer used by the TO is too small, and has not enough primary inductance for a 10k transformer.
Core saturation is very likely; it is not a grown up 211 output transformer.
I understand that modern digital sources start at 40 Hz??
The output transformer used by the TO is too small, and has not enough primary inductance for a 10k transformer.
Core saturation is very likely; it is not a grown up 211 output transformer.
Last edited:
Inductance has nothing to do with core saturation. The specified power handling is correct. Frequency response is more than good enough....
Lower inductance will only cause lower impedance at low frequency and thus more distortion (mainly 2nd harmonic) from the tube. However the difference between 57H and 80H is not big at all. The effective impedance that the tube will see at 20Hz is 5.9K and 7.1K respectively for 8 ohm load.
Lower inductance will only cause lower impedance at low frequency and thus more distortion (mainly 2nd harmonic) from the tube. However the difference between 57H and 80H is not big at all. The effective impedance that the tube will see at 20Hz is 5.9K and 7.1K respectively for 8 ohm load.
57H is not enough inductance for a full range 10k output transformer; it should be 80H at least (2xPix20Hzx80H=10k), otherwise LF extension will be compromised. This is clearly visible in the measurement sheet in post #13 (20Hz -5dB).
The fact that the inductance is lower is caused by a limited number of primary windings; maybe there was not enough winding space for more windings, or for more windings the wire might have become too thin.
The limited number of primary windings causes higher core excitation (in T); core excitation can be lowered by more primary windings, also giving higher primary inductance.
More primary windings will cause higher DC resistance / copper loss and more winding capacitance. These factors are all design considerations.
The limited number of primary windings / low inductance situation brings the transformer closer to the saturation treshold; winding capacitance is minimized giving good HF extension. This seems to be the direction taken by the transformer designer.
When I look at the dimensions of the transformer and it's specifications, I would love to see at which voltage the core starts to saturate at 30 Hz.
You claim 17W @ 30 Hz undistorted power; that is 11,6 VRMS @ 8 ohm, or some 410 VRMS at the primary.
I am pretty sure that at those levels you will see considerable core saturation.
My opinion is based on extensive prototyping of output transformers for the 211 with 15k primary impedance, using massive HiB and Finemet c-cores with Afe of 36 cm².
Last edited:
You are confused. Those are 5 kohms not - 5 dB! The frequency response is the upper picture. The one you talk about is the reflected impedance (looking into the primary and secondary loaded with 8R) from which also inductance can be assessed. I am pretty sure about that!!!
Because the measurement is done with low signal inductance is bit less than 57H and the impedance at 20Hz is bit less than 5.9K but still more than 5K. That's the minimum. At full power will reach the target....
Hashimoto are a reputable company and don't cheat about power handling and other specs. I have Hashimoto transformers and they are fully in spec.
Last edited:
Ah you are right about the impedance/frequency graph; it shows that the transformer is actually not 10k primary impedance until 300 Hz.
Until I see measurements at 30 Hz however I am not convinced wrt power.
There will be considerable LF harmonic distortion because of the lowish load impedance (211 prefers to see load impedance nearer 15k).
Manufacturers seem to offer 10k transformers for 845 / 211. OK for 845; 211 is totally different tube and a difficult beast for output transformers. Good 15k transformers are rare.
Until I see measurements at 30 Hz however I am not convinced wrt power.
There will be considerable LF harmonic distortion because of the lowish load impedance (211 prefers to see load impedance nearer 15k).
Manufacturers seem to offer 10k transformers for 845 / 211. OK for 845; 211 is totally different tube and a difficult beast for output transformers. Good 15k transformers are rare.
You are right that is your opinion. Facts can be different because in reality the cross-section is only one parameter. They use Hi-B anyway.
That transformer can be done easily onto a HWR 70/32 Hi-B core 0.27mm that only has 15.4 cm^2 effective core area.
For example, using a split bobbin and winding a total of 4032 turns with 0.3mm wire with 7 primaries and 6 secondaries per each half, 0.25 mm spacer for a total air gap of 0.5 mm, primary inductance will be 57H and power handling even better than 17W @30Hz. In fact the transformer will handle up to 20W @30Hz with 80 mA DC working at a total induction of 1.1T ONLY which is perfectly fine for Hi-B core that sutarates at almost twice that value! Rdc of the primary will be about 260R that is also fine for a 10K transformer. The 13+13 sections and other tricks to redistribute capacitance will allow even better high frequency response. However cost will be higher and that's why they don't do it....still they make and sell a good transformer nonetheless at lower price and use more complex geometry on demand only....
Last edited:
It won't get much better with 80H. That's the SE transformer....
Metropolitan legend. 10-11K max are fine. If one wants more inductance can use a bigger core like the HWR 90/44 and get 80-100H, more power handling, more DC current capability and even lower copper loss at expense of complexity. End.
Last edited:
Something to think about bifilar wound interstage transformers is that after a certain number of windings and frequency, they look like a coupling capacitor in parallel with the magnetic coupling. This is because of the high primary to secondary series capacitance. Is the coupling capacitor effect a good thing to the sound? Afterwards many of us are trying to escape from the coupling capacitor and replace it with an IT.
But they are so wideband. I once wound an experimental line level bifilar 1:1 transformer with 6 vertical sections that makes almost perfect 10MHz square waves.
But they are so wideband. I once wound an experimental line level bifilar 1:1 transformer with 6 vertical sections that makes almost perfect 10MHz square waves.
Last edited:
Yes it is if it's a good capacitor and other solutions are worse....
I am not a fanatic of transformer coupling and only use it when it brings real advantage or more than good enough performance at least.
For 211 output transformers I would not consider 10k primary impedance to begin with.
My customers ask for 15k transformers, and they do it for a reason.
To provide you with some (apparently) necessary information, check different load lines for the 211.
Among others, these guys did, and they conclude that 15-16k primary impedance for the 211 makes the tube more linear, and with this load it can even give more power:
J&K Audio Design: February 2016
So more power and audibly less distortion. That's why we want 211/845 don't we? Otherwise stick with 300B. Even the 845 would benefit with a higher than the "normal" 10k load impedance.
Your transformer design is based on textbook calculation, which by the way is a good approach. But don't come back before you actually wound your transformer, and back up your claims with real world measurements to prove being right.
End
Last edited:
As usual I should say.
You should not come back before you actually make a 211 amplifer. I have already made a 211 amplifier. I don't need your advice.
Audio Note made the 16K transformer only to show how good they were at making transformers but in reality one can get those 27W at 5% THD like the Onkagu with 10K as well.
I have given the transformer details to compare with the smaller Hashimoto so that everyone can do it. I can design a transformer and get the specs right because I know how it works. 50AE here can do the same!
Don't need to make 200 prototypes based on wrong assuptions to get something barely usable and weights a ton.....You have no idea sorry! You might make transformers for work but you do not actually know how they really work. You still make fundamental mistakes like confusing saturation with inductance. The only one making claims it's you. I have never seen measurements of transformers of yours and yet you discredit other manucturer's transformers that show real and actual data. End.
You should not come back before you actually make a 211 amplifer. I have already made a 211 amplifier. I don't need your advice.
Audio Note made the 16K transformer only to show how good they were at making transformers but in reality one can get those 27W at 5% THD like the Onkagu with 10K as well.
I have given the transformer details to compare with the smaller Hashimoto so that everyone can do it. I can design a transformer and get the specs right because I know how it works. 50AE here can do the same!
Don't need to make 200 prototypes based on wrong assuptions to get something barely usable and weights a ton.....You have no idea sorry! You might make transformers for work but you do not actually know how they really work. You still make fundamental mistakes like confusing saturation with inductance. The only one making claims it's you. I have never seen measurements of transformers of yours and yet you discredit other manucturer's transformers that show real and actual data. End.
Last edited:
I wanted to be specific - the capacitance coupling effect between the primary and secondary that mimics a coupling capacitor.
I'm not implying fanaticism either. Just wondering how would that influence the sounding (or performance) as compared to a classic sandwich-interleaved transformer with lesser pri-sec capacitive coupling.
First, 16K with 77H means that at 20Hz actual impedance will be 8.3K. Worse than the small Hashimoto. A bit bigger 10K transformer with 80H will present 7.1K.
Second, what's the point of targeting 40W at 20Hz if the DC current is only 60mA and inductance 77H??? It will never do it! The max power @20Hz into 16K cannot be better than 10.5W at the primary for the simple reason that Bac cannot be bigger than Bdc. At 10.5W Bac = Bdc and it cannot go any further. No zero crossing for SE operation. One should run at no less than 120mA to get 40W at 20Hz into 16K with 77H inductance which is not possible on that transformer and is impossible in general with the 211 if the plate voltage has to be no less than 1000V.....the other option is 154H primary inductance with 60 mA! I can only imagine how bad at high frequency or otrageously complicated would be such transformer....16K, 154H, 40W @20Hz...pretty much dreaming....
Last edited:
You could wind a tightly coupled layer to layer interleaved transformer to compare with the bifilar one for sound.
The former will not reach the same bandwidth (but more than enough).
The difference in sound might be a subjective thing; I know people who don't like bifilar wound and prefer the classic approach, and the other way around. I don't hear enough difference to make a choice.
Bud Purvine has built his reputation on the capacitive coupling sound at much lower frequency than bifilar IT's. And it's a rather good reputation....
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.