|
Homebrew Equipment A place to show, design and discuss the weird and wonderful electronic creations from the hands of individual members. |
|
Thread Tools |
24th Oct 2006, 8:31 pm | #21 |
Dekatron
Join Date: Jan 2004
Location: Newcastle upon Tyne, Tyne & Wear, UK.
Posts: 8,171
|
Re: Toroids as O/P trannies
Hi Mike, I generally use an AVO for both volts and amps. They go down low enough for the current readings. It should ideally be a moving iron meter for the current measurement but the Avo is close enough. Is there a school/ colledge in your area? They may well have all the kit. The principals are usually described in "A" level physics books (but in !!**&&?? CGS units)
Proper engineering units (MKS) use the Tesla for flux density. Most silicon iron (old type mains transformer) would operate at about 1T. Modern laminations (GOSS, Unisil) will operate quite happily at 1.5T. There are special alloys for aerospace use that will operate at 2T. Ed |
26th Oct 2006, 11:02 pm | #22 |
Nonode
Join Date: Mar 2005
Location: Nuneaton, Warwickshire, UK.
Posts: 2,034
|
Re: Toroids as O/P trannies
Hi All,
I just spotted this:- http://www.rapidonline.com/productin...moduleno=77229 Looks like toroidal outpou transformers do exist. I'd not seen one before. Cheers Aub |
27th Oct 2006, 7:43 pm | #23 |
Dekatron
Join Date: Jan 2004
Location: Newcastle upon Tyne, Tyne & Wear, UK.
Posts: 8,171
|
Re: Toroids as O/P trannies
Hi Gents, there is a book published by Elektor that details the use of toroidals "Modern high end valve amplifiers" . It appears to come from The Netherlands and a specialist design company who sell these transformers.
E-mail is mennovdv"at" noord"dot"bart"dot"nl Ed |
20th Dec 2006, 9:46 am | #24 |
Pentode
Join Date: Jun 2006
Location: Malaga, Spain.
Posts: 235
|
Re: Toroids as O/P trannies
Hi, Now that my heater supply is sorted I can get back to the O/P tranny. I have removed the sec windings and tape etc and now have just the original primary winding on the toroid. Hidden under the tape was a centre tap connection. Bad thoughts here! How good is the insulation offered by the lacquer on the copper wire, probably not good enough but it would save me rewinding the primary? I have made a wooden shuttle anyway. Because the Amp will be experimental (what isn't) and I'm not sure about the final O/P configuration I thought that if i have to rewind it I will add taps for G2, 40% being the most quoted figure, but is it 40% from the centre tap or from the Anode?
Mike. |
21st Dec 2006, 11:25 pm | #25 |
Pentode
Join Date: Jun 2006
Location: Malaga, Spain.
Posts: 235
|
Re: Toroids as O/P trannies
Made some tests on the primary and it turned out that the centre tap wasn't a centre tap. I tried unwinding the toroid but after about 30 turns I ran into a problem with the wire passing underneath another turn that wasn't anywhere electrically close, ie the only thing to do was cut the wire. tried again but after a few turns ran into the same problem. Only solution out with the cutters. The winding was more than a bit strange with bifilar and trifilar winding methods, the afore mentioned CT was connected to the end of these windings. In order to know how much wire to get to rewind the tranny, I weighed the wire that I took off, 434 grms. I will make some plastic cheeks to fit on the toroid to separate the windings laterally. The transformer will have four primary windings, 2 X 40% for G2 and 2 X 60% for the Anodes. eg 120 turns and 180 turns. the two will be wound separately which should help re insulation and each layer will be insulated also. The 2 windings will then be connected in series. The alternative and better solution would be a 100% winding for the Anode and a separate 40% winding for G2, enabling a dedicated G2 supply, I may do this later but for now I will take the easy option. More to follow as it progresses.
Mike. |
22nd Dec 2006, 8:10 pm | #26 |
Dekatron
Join Date: Jan 2004
Location: Newcastle upon Tyne, Tyne & Wear, UK.
Posts: 8,171
|
Re: Toroids as O/P trannies
Hi Mike, have a lok at the Dutch book, he has lots of suggestions about G2 supplies, both independant and cross connected. There are some interesting claims for "crossing" the G2 taps on the output tranny.
Larger diameter wires (above 0.2mm), if double coated will withstand greater than 1Kv before breakdown. They should not normally be run with more than about 300V across them. Ed Ed |
23rd Dec 2006, 5:20 pm | #27 |
Pentode
Join Date: Jun 2006
Location: Malaga, Spain.
Posts: 235
|
Re: Toroids as O/P trannies
Hi Ed, Is there a site? By crossing the taps do you mean from one valve winding to the othervalves G2?
Mike. |
23rd Dec 2006, 8:17 pm | #28 |
Dekatron
Join Date: Jan 2004
Location: Newcastle upon Tyne, Tyne & Wear, UK.
Posts: 8,171
|
Re: Toroids as O/P trannies
Hi Mike, the book was published in 1998 by Elektor and written by Menno van der Veen. His company e-mail is mennovdv at noord dot bart dot nl.
Yes that is the idea proposed for the G2's Ed |
24th Dec 2006, 9:38 am | #29 |
Pentode
Join Date: Jun 2006
Location: Malaga, Spain.
Posts: 235
|
Re: Toroids as O/P trannies
Hi Ed, Thanks, I will send him a mail. I presume that the Amp with the crossed over G2s was working in class A, but I can't see a reason to do it.
I read something about optimising Ultra linear O/P stages where not only the G2 feedback ratio but also the G2 supply should be set at 40%. Every circuit that I have seen so far uses either the full HT( or a small R) or a separate G2 Supply. In my case I would have to drop around 500V. A big R with a big parallel cap from the 40%? winding? Mike. |
24th Dec 2006, 5:52 pm | #30 |
Dekatron
Join Date: Jan 2004
Location: Newcastle upon Tyne, Tyne & Wear, UK.
Posts: 8,171
|
Re: Toroids as O/P trannies
Hi Mike, no this was clas AB push-pull/ UL, from memory.
I believe the idea was that the side with the valve not conducting had a "stiffer" G2 supply. The Mullard book "Circuits for Audio Amplifiers" gives details of the various connections for UL use and the advantages of various tapping ratios from 20% to 43%. Ed |
24th Dec 2006, 7:03 pm | #31 |
Pentode
Join Date: Jun 2006
Location: Malaga, Spain.
Posts: 235
|
Re: Toroids as O/P trannies
Hi Ed, Most odd, The whole point of the tap (winding) is to provide negative feedback via G2 and to reduce the Ra to near triode operation. Connecting G2 to the other phase in class AB would have some strange effects. As for the % tap etc it would seem that this has optimum effect depending on the physical location of the G2 assembly wrt to the cathode and anode, eg for the KT88 it is 43% and for the 813 it is 40%. The same would appear to apply to the G2 potential. in my case with around 900 V on the anode i should have 360 V on the screen. This means either a separate supply or running G2 at 900 V as well with lots of bias (G1 or K). I wondered if I could feed G2 from the 40% winding via 22K (20W) paralled with a cap. The only resistors I have seen in series with G2 have been of the order of 100 Ohms, otherwise G2 is basically at the same potential as the anode even in single ended UL class A amps.
Mike. |
24th Dec 2006, 10:02 pm | #32 |
Dekatron
Join Date: Jan 2004
Location: Newcastle upon Tyne, Tyne & Wear, UK.
Posts: 8,171
|
Re: Toroids as O/P trannies
Hi Mike, yes 100R resistors are often used as "stoppers" in sewries with many electrodes.
The Mullard book only defines the tap ration in terms of valve behaviour abd effect on power and distortion. As you say, this will vary for different valves depending on their electrode spacings. (How will this work for different manufacturer's product ?). A great bag of worms here for the Audiophools! |
24th Dec 2006, 10:15 pm | #33 |
Retired Dormant Member
Join Date: Dec 2003
Location: North London, UK.
Posts: 6,168
|
Re: Toroids as O/P trannies
Thinking about a couple of points that have been raised.
The voltages between a pair of series connected primary windings is still high when used in the original mains tx application. Obviously an audio amp may use even higher voltages so the insulation might still not be adequate. As Ed Dinning rightly says, output tx are usually run at rather lower flux densities than mains tx. From what I've seen of mains tx they are often run right up to the limit set by excess heating and to hell with linearity. Just means you need to keep the power down and use more turns per volt. My understanding of ultralinear operation is that it runs the valves in a sort of halfway house between triodes and tetrodes/pentodes. The 40% tap seems to have been chosen as a good compromise between power and distortion, espeically when the output stage is run with little or no negative feedback. I feel that Ed is wrong about the audiophool can of worms. No self respecting audio phhol would contemplate anything other than directly heated triodes |
25th Dec 2006, 10:02 pm | #34 |
Dekatron
Join Date: Jan 2004
Location: Newcastle upon Tyne, Tyne & Wear, UK.
Posts: 8,171
|
Re: Toroids as O/P trannies
Hi Jeffery, an apt reminder ref Audiophools, perhaps the more impoverished ones would consider pentodes even if semi-triode strapped!
From memory there was a tantalisingly small amount of info on the "why " of UL tappings in the Mullard book. I believe there may be more data in the Philips technical library series that was published from the late 30's through to the early 50's. These really gave the low down on valve design and applications, even going back to the basic fundementals of valve design. Has anyone out there got the volume that does UL? Ed |
25th Dec 2006, 10:18 pm | #35 |
Retired Dormant Member
Join Date: Dec 2003
Location: North London, UK.
Posts: 6,168
|
Re: Toroids as O/P trannies
Just found this reference which suggests that UL was invented by Blumlein. However, the reference he quotes is Alexander's biography of Blumlein. When I reveiewed that book in 2000 I found that it was not as accurate as it might have been.
http://web.telia.com/~u43200663/blocks/ulaudioamps.htm If anyone can access the UK patent (496,883 dated 5 June 1937) this should give more info. I can't find a public on-line database of patents and my friend at the Patent Office is away until the new year. |
27th Dec 2006, 9:53 am | #36 |
Pentode
Join Date: Jun 2006
Location: Malaga, Spain.
Posts: 235
|
Re: Toroids as O/P trannies
There's a lot of info out there split between UL and distributed load, but it's mainly just statements without anything to back them up with a few references to the already quoted sources + Mullard, Quad and a couple of other manufacturers from the 60s.
IMO the distributed load versions don't really stand up due to the fact that the current source Z is much lower than the G2 Z. The winding carries the G2 current but the effect of the winding overall is to Modify Ia and Ra and not to add G2 derived signal current. I found a site which rattled on about optimised UL which made a lot of sense and which brought me back to the DC G2 conditions which seem to be ignored in most amp designs apart from those in the n x 1000$ class. There things get even more complicated with separate G2 and K windings on the O/P transformer. My scope is just to make a reasonable amplifier with quite a high power O/P without spending lots of dosh and without spending a week trying to wind a transformer with dubious results. The toroid for a tenner is still the way frorward until proven otherwise. Splitting the windings on the core will mean that the PDs between layers will be reduced and should simplify the winding. Later this week I will get the wire and wind it and then put some test signals on it from another amp and see what comes out. Mike. |
19th Jan 2007, 11:29 pm | #37 |
Pentode
Join Date: Jun 2006
Location: Malaga, Spain.
Posts: 235
|
Re: Toroids as O/P trannies
A quick update. The workshop which is my wire source has been closed but reopens nxt week.
Anyway over the last few weeks I have been reading up on transformers, especially in regard to leakage inductance. In regard to toroids and leakage inductance I have found only statements and that includes papers (Universities and institutes) on toroids and company blurb. there are a lot of real bullxxxx explanations about limiting flux paths etc but nothing that is really backed up. There are many docs that deal with flux distribution within the core and flux polarities across the core. Most of the arguments that are used within the traditional E I core regarding leakage inductance are contradictory when applied to a toroid. An example in one instance that was quite laughable consisted of a pseudo toroid made up out of 4x I sections!! The flux path in an E I Trannie has windings over only around 30% of the flux path and in the case of a C core will not exceed around 50% of the flux path. Commonly "stated" figures for leakage inductance claim that toroid trannies are at least 10 times better than E I cored trannies without resorting to interleaved windings etc. Your thoughts please and any real info on winding layouts which is backed up technically. Cheers, Mike. |
20th Jan 2007, 10:47 pm | #38 |
Dekatron
Join Date: Jan 2004
Location: Newcastle upon Tyne, Tyne & Wear, UK.
Posts: 8,171
|
Re: Toroids as O/P trannies
Hi Mike, leakage inductance can generally be thought of as not all of the primary flux coupling to the secondary.
In general terms this can be due to gaps in the core, laminations not butting together correctly and having air gaps etc. These effects are very much reduced by the use of a toroidal core. BUT The primary and secondary windings cannot occupy the same space so this also results in a loss of coupling. Hence the practice of interleaving and sectionalising windings to give better coupling. This can give problems in output transformers, as the interleaving will result in increased capacitance between windings; this in turn can effect the feedback. On way round this is to fit screening layers over the sections so that the capacitance is to ground, not to the other winding, and hence feedback. This is fine but makes for a bigger transformer. Something to think about! All good engineering is a matter or inteligent compromise. Ed |
20th Jan 2007, 11:53 pm | #39 |
Dekatron
Join Date: Aug 2003
Location: near Reading (and sometimes Torquay)
Posts: 3,086
|
Re: Toroids as O/P trannies
Has anyone checked the linearity of using a solid cored transformer for audio? I was told that audio transformers have an air gap (and I have seen some that certainly do) so as to linearise the magnetic path.
|
21st Jan 2007, 10:38 am | #40 | |
Retired Dormant Member
Join Date: Dec 2003
Location: Solingen, Germany
Posts: 727
|
Re: Toroids as O/P trannies
Quote:
the air gap is neccessary if you have AC+DC. In push pull the magnetic field of the DC components is in opposite directions and you don't need an air gap. An importand aspect you should know is that only the low and mid frequencies goes over the lams. The high sound frequencies are coubled from coil to coil. To verify this, put 15KHz to the amp and disassamble the output transformer. Nothing changes! The interleaving and sectionalising Ed is talking about (#39) is neccessary for the high audio frequencies say above 10KHz and does not depend in the material or design of the core. In a pentode amp (not the ultralinear (UL)with tappings for g2!) the stray inductivity is not that importand. The pentode is a current scource and generates automaticly more voltage at high frequencies to make sure that there is a constant voltage over the whole frequency range at the load resistor. This is why a pentode works well with a not interleaved and sectionated mains transformer. The problem here at the low frequencies current caused by the impedance of the transformer reduces the output voltage. ( https://www.vintage-radio.net/forum/...ad.php?t=13650 see #1 point 3) ) I don't like UL because the voltage ratio between the tappings is not constant over the frequency. Attached a pic of an output transformer with grain oriented lams and interleaved and sectionalised windings. (See triodemultiplier thread https://www.vintage-radio.net/forum/...ad.php?t=13624 ) Kind regards, Darius Last edited by oldeurope; 21st Jan 2007 at 10:56 am. Reason: adding links |
|