I believe its a 50VA because thats all I could get from Farnell in that ratio. However since I now have it configured as a center tap choke with the secondary unused, if I was sourcing again I would use a 6V secondary at the lowest VA available (ie 20VA). I had it explained to me once that the smaller the transformer the higher the primary inductance and hence the better the bass.
In an output transformer there are other important considerations to take into account, such as voltage capability and low frequency response - below the designed 50/60hz. To achieve a good bass response I would use a VA rating 4x the expected output wattage.
Shoog
Although toroids appear to be the answer intuitively, I find one matter not raised here thusfar. That is the ability to design for particular values of intersection capacitance, coupling and leakage inductance. In a toroid this is problematic without predictable layer behaviour.
It will be likely that whatever figures emerge with toriods would be low enough, to yield little problems to get incorporated into stable design. But I suspect that due to a degree of unpredictable random winding, no two transformers would be alike.
I thus prefer C-core, but some care in assembly is required as they are capable of humming, and it can be difficult in partice to eliminate that.
It will be likely that whatever figures emerge with toriods would be low enough, to yield little problems to get incorporated into stable design. But I suspect that due to a degree of unpredictable random winding, no two transformers would be alike.
I thus prefer C-core, but some care in assembly is required as they are capable of humming, and it can be difficult in partice to eliminate that.
Shoog,
Regarding your amp circuit, and knowing by experience the huge loss of sonics if a toroidal OPT has any imbalanced DC in the windings, I would suggest you put in some sort of balance circuit to ensure both halves of the OPT's primary have EXACTLY the same current.
The 1.2V "reference" in LM317's are poorly matched in my experience, and unless you spent some time finding matched pairs I suspect you have less than optimum current balance in the OPT.
I would be looking for less than 500 uA imbalance - from cold to full hot operation, and much better than that if possible.
Regards, Allen
Regarding your amp circuit, and knowing by experience the huge loss of sonics if a toroidal OPT has any imbalanced DC in the windings, I would suggest you put in some sort of balance circuit to ensure both halves of the OPT's primary have EXACTLY the same current.
The 1.2V "reference" in LM317's are poorly matched in my experience, and unless you spent some time finding matched pairs I suspect you have less than optimum current balance in the OPT.
I would be looking for less than 500 uA imbalance - from cold to full hot operation, and much better than that if possible.
Regards, Allen
Johan Potgieter said:
This is certainly true for mains toroids as they are not optimized in this respect.
That being said, a winder with a will to do it, can wind a toroid exactly to spec regarding layers, although in toroids the concept is somewhat different than in a standard EI or C-core transformer. What they can do, depends on how much they are willing to tinker with the winding machine, the actual choice for the core, and how sophisticated the machine is.
There are in general 3 types of toroidal cores, the so called standard, the flat and the tall version. The difference is in the inner to outer diameter and, of course, to get the same cross-section, height - that's how you get tall and flat.
The usual way of winding a toroid core is to wind 'in one layer', although depending on the size of the core, this can be several turns of the whole core. For a standard core, this will normally be 2-3 turns, resulting in the wires being one over the other around the inner diameter, and one next to the other around the outer diameter. For a tall core, it will be 1-2 turns, for a flat core 4 to 5. In terms of standard layers, this would look like 2-3 layers for the standard core, but wound so that the 2-3 layers always go in the same direction on the bobbin, left to right. This way the layers' capacitances are 'bootstrapped' wrt to one another, but on a toroid, they all have fairly equal capacitance to the core itself.
The alternative way to wind a toroid is available on some machines, which do not rotate the core during winding in a continious fashion, but rather zig-zag. In this way the winding is truly made in one single turn of the core propper, and never overlaps except for the 2-3 successive turns. Instead of filling out the outer diameter of the core by going round it several times, the machine is set up to move the core back-forth in a rythm of 3 turns of the winding, this results in wire for the successive 3 turns being one on top of the othe on the inner diemater, and interlieved one next to the other on the outer diameter (turns 1-3-2-4-6-5-7-9-8 etc). This technique is a bit difficult as it requires manual intervention (holding the windings) at the start, to prevent the windings on the inner diameter from getting all jumbled up instead of stacking one on top of the other properly. It is also used with especially tall cores where inner and outer diameters are such that less thatn two wire diameters can fit on the outer diameter of the core. In this case the windings are wound one on top of the other, and what you get if you strip the wire enamel from the outermost 'layer' is a variac core.
I came across an interesting series of articles comparing PP to SE.
It pointed out that one of the main deficiencies of a PP transformer over an SE was in low level listening and micro-detail. It then went on to explain why that was the case, and put it down to hysteresis and crossover magnetisation blurring the signal at the critical microwatt level.
For the very reasons that a toroidal will saturate with minute amounts of DC current in the core, they have very small hysteresis and so are the most capable of the PP transformers at faithfully reproducing micro-detail, and low volume listening.
I have found this to be true.
Shoog
It pointed out that one of the main deficiencies of a PP transformer over an SE was in low level listening and micro-detail. It then went on to explain why that was the case, and put it down to hysteresis and crossover magnetisation blurring the signal at the critical microwatt level.
For the very reasons that a toroidal will saturate with minute amounts of DC current in the core, they have very small hysteresis and so are the most capable of the PP transformers at faithfully reproducing micro-detail, and low volume listening.
I have found this to be true.
Shoog
winding a toroid core
Hi!
What could be the best way to winding a toroid core for PP OTP?
-the first layers are the primary than isolation layer and after come the secondary windings. (It is the usual way for power toroid)
-trifilar winding
-sectional winding: in one part of the core is the primary and in another 2 parts are the secondary.
Greets:
Tyimo
Hi!
What could be the best way to winding a toroid core for PP OTP?
-the first layers are the primary than isolation layer and after come the secondary windings. (It is the usual way for power toroid)
-trifilar winding
-sectional winding: in one part of the core is the primary and in another 2 parts are the secondary.
Greets:
Tyimo
AndrewT said:
I have also come across suggestions about this, and my question is always: How micro-level - in real life there is such a thing as threshold-of-audibility, after all, even allowing for some uncertainty about signals-buried-in-noise (which had been researched). What is the purpose in analysing distortion below this?
AndrewT said:
Also my concern. One can of course do a non-destructive leakage test, but in production for every transformer? I have recently lost a 500VA power transformer (almost catastrophically) because a winding slipped onto another with a few 100V between them. It is not uncommon in large tube amplifiers to have >1KV differences somewhere in the output transformer.
Toroid interstage
Shoog!
Could you suggest me a toroid size, ratio, voltage values for an interstage trnsformer between my SE tube preamp and balanced Mosfet amp??
The tubepre has ca. Zout 100R and the power amp has Zin 90K/side.
I would use this toroid as a SE/balanced converter.
Greets:
Tyimo
Shoog!
Could you suggest me a toroid size, ratio, voltage values for an interstage trnsformer between my SE tube preamp and balanced Mosfet amp??
The tubepre has ca. Zout 100R and the power amp has Zin 90K/side.
I would use this toroid as a SE/balanced converter.
Greets:
Tyimo
Tyimo,
I'm not certain I would recommend it for this application.
Where I have used a toroidal as an interstage transformer there were always issues with high frequency extension. What I was using was a 110+110 : 55+55 with the secondaries in series. It was rated at 50VA, but apparently primary inductance goes up as VA goes down, so smaller is better. I had to push 40mA of current through the driver to get satisfactory top end response. I cannot say that it didn't work, and that it didn't sound excellent - but it always troubled me the need for such high drive current just to overcome the interwinding capacitance. With the likely high Mosfet gate capacitance you could face similar issues.
What I did about two weeks ago was convert this to a center tapped choke by simply ignoring the secondaries. If you did this then it would be possible to go down to a 10VA transformer with say 6V secondaries. This has produced a noticable improvement in top end extension though I haven't scoped it out. It may have introduced a bit of inbalance - but its difficult to say and I can't say that the sound has suffered (though slightly rolled off top end is appealing).
I have another suggestion for you - drop me an email.
Shoog
I'm not certain I would recommend it for this application.
Where I have used a toroidal as an interstage transformer there were always issues with high frequency extension. What I was using was a 110+110 : 55+55 with the secondaries in series. It was rated at 50VA, but apparently primary inductance goes up as VA goes down, so smaller is better. I had to push 40mA of current through the driver to get satisfactory top end response. I cannot say that it didn't work, and that it didn't sound excellent - but it always troubled me the need for such high drive current just to overcome the interwinding capacitance. With the likely high Mosfet gate capacitance you could face similar issues.
What I did about two weeks ago was convert this to a center tapped choke by simply ignoring the secondaries. If you did this then it would be possible to go down to a 10VA transformer with say 6V secondaries. This has produced a noticable improvement in top end extension though I haven't scoped it out. It may have introduced a bit of inbalance - but its difficult to say and I can't say that the sound has suffered (though slightly rolled off top end is appealing).
I have another suggestion for you - drop me an email.
Shoog
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.