Toroidal Transformers for Parafeed?

TriodeLuvr

TriodeLuvr

Well, not parafeed exactly, but the transformation would function in a similar manner (no DC current). The idea is to start with a SE tube amplifier that has an OPT designed to drive an 8Ω load. Then, in order to drive other loads, such as headphones, a toroidal power transformer would be connected across the output to step up the impedance. I've been told some of the toroidal transformers have very wide BW, despite being intended to operate on 50Hz or 60Hz. Any thoughts on this?

Jack
 
They have a rather hefty amount of parasitic capacitance, and are (IME) a bit more susceptible to picking up hum from nearby power transformers. It certainly doesn't hurt to give it a shot!
 
That should work Triode, At least I cannot see any problems except the losses involved connecting two transformers back to back.
Paul also mentions a particular fault with toroids, and that is the capacitance, as the primaries and secondaries are essentially wound over each other.
Also, BOTH transformers, your OPT and the following toroid have significant phase shift, but that shouldnt matter in your case.

Joe
 
That's interesting. I remember seeing 60Hz pickup in small RF toroids a few years back. It was very low level, but we had to shield against it with mu-metal. I was thinking that the capacitance wouldn't be such a significant factor at these low impedances. The concept is to attach a 2:1 toroid backwards to the 8Ω output. That would create a 32Ω equivalent output. A 6:1 ratio could transform to roughly 300Ω. I used to design bifilar and trifilar broadband RF transformers at similar impedances, such as 50:200Ω. Audio is different though, and I'm not familiar with the mix used in power transformer cores. Guess I'll just have to give it a try.

Jack
 
What are you planning to attach these to? For 300 Ohm headphones, the 8 ohm output of an amp that puts out a few watts is already a great place to connect headphones ;)
 
I don't have watts, more like milliwatts. The higher Z is not a priority, so I'll see about ordering a 2:1 transformer before going further with this idea.
 
Most toroids are 5% SiGO, so thats 5% silicon, grain oriented transformer iron. It doesnt saturate easily but when it does its an avalanche.
Not a problem in your case. Apart from that, a decent toroid ( from a good manufacturer ) will easily cover 100 Khz.
If you only have milliwatts, I would reckon a 20 watt core size, as anything smaller gets expensive. I dont know whats available in USA, but a power transformer from 120 volts to two times 50 volts will be close enough, and should be off shelf. Parallel the secondaries for your primary. Although that sounds like a big mismatch, a good toroid will really just work like a gearbox, BUT in this case, plenty of primary impedance.
I am waiting to hear your results.

Joe
 
You could also connect the toroid windings as an autotransformer for a step up. The Anticables brand toroids are autoformers.
 
Good ideas! I was thinking about a 120V : 60VCT for a 2:1, but an autoformer would also work well for this ratio.

I'm also thinking about the inductance required to reach 20Hz or 30Hz and how this compares to the use of the transformer at 50/60Hz. Ignoring losses, a 10VA transformer powered at 120VAC would be working at an operating Z of 1440Ω at full power (120V ÷ .0833A). This would seem to mean that it exhibits sufficient inductance to prevent excessive internal power consumption at 50Hz. As the operating impedance goes down (300Ω or less in this case), less inductance is required for a given frequency. There's the issue of core flux, which varies inversely with frequency, but this application will be way below the rating of the core.

I need to take a look later today and see if any of the low voltage toroids in my stash have dual primaries. That would allow testing as an autoformer.

Jack
 
I wouldnt even worry about flux density in your application. Most commercial toroids are wound at about 2 Teslas, or in my case, being an olde fart, 129,000 lines per square inch. If you are really going to stress about it, a commercial 30 watt or 40 watt toroid is only a few bucks more expensive and I would put money on it that it will near reach DC, in any case way below what your phones will get to.

Joe
 
It looks like you're exactly right. I found a 45VA toroid on the shelf and tested it a few minutes ago. It's a single 120V primary with a 16V single secondary. I drove the 16V winding from my 50Ω generator and terminated the 120V winding with 2.2K in order to load the generator close to its characteristic impedance. With 10V RMS output from the 120V winding, the -1dB points are 8Hz and 140kHz. Holy cow! This is much wider than the 5K:8 transformers planned for the amplifier proper. If this can be maintained at lower operating impedances, the transformation will be almost totally transparent. I'll order a couple 10VA units with a more suitable winding configuration and give it a go. 😁

Jack
 
There is one big drawback with toroids, You cannot have them single ended unless you drive them the way you intend to.
The performance of a toroid exists because there is absolutely zero airgap anywhere in the iron toroid bit, and no square corners for leakage to occur.
Which is the reason you can run such high flux densities, AND the reason the bandwidth is so wide.
I have tried cutting a gap into a core and totally ruined one toroid.
There are push/pull toroids available, mostly by Menno Van Der Veen, which are horribly expensive, and I have never heard or read any feedback apart from his own opinion on performance. (His company is Plitron ).

I have never tried them as choke for parafeed, because you only need a couple of milliamps DC on any one of the windings and the toroid will saturate. No free lunch, AGAIN !!.

Joe
 
Well, it's almost a free lunch in this case - if it works. The alternative would be an expensive multi-tap SE headphone OPT like the one Magnequest used to sell. I wanted a pair of those for years, but just couldn't justify the cost. If it's possible to combine a toroid with a good quality conventional SE transformer instead, maybe the cost will be more reasonable without sacrificing sound quality. That's the plan, anyway.
 
It will work just fine, largely depending on how well they match together.
Its a marriage after all, some work, some dont.
BUT a toroid without DC is a fine little animal.

With ANY SE traffo there is that airgap, with that you reduce all the good bits. YES one can maike a fair SE traffo, but its expensive, AND its lossy, and I dont care who makes it. Physics is physics and doesnt tell fibs.
This world is round remember, NOT flat.

Joe
 
Specs on the Antek 5K are poor. They claim 0dB at 10Hz (relative to 1kHz), but it dips to -9dB at 20kHz. The 2.5K model is almost identical, and some of their PP transformers also share this trait.
 
I have been making transformers for about 55 years. I dont know all that much about SE traffos as I have NEVER had any real success with them. They are inherently lossy and with the core unbalanced it will NEVER be linear. I suppose one could reduce that somewhat by uses a stupidly large core, so that the DC bias would not be completely overcome the linearity.
SE does work, and it works well for kitchen radios, and portables because its lightweight, and its cheap to make. I would NOT say its Hi-Fi though, and I use Williamsons origional specification ( he DID coin the phrase after all ) 20 Hz to 25 Khz with less than 1 dB variation over the bandwidth at rated full power. Remember too that he used 807's in the origional amplifier, strapped as triodes, and running a big 15 watts to keep everything working at best linearity.

I know I will get arguments/discussion about this but read ANY decent electronics theory on transformers and trying to make them work with DC bias on them. I have already said that toroids dont do well ( MY experience only ) as they saturate very easily if the BH curve is upset.
Apparently they do OK in push/pull operation, but I have never tried it myself. Coupling would be excellent and capacitance pretty large due to the excellent coupling.
As far as DC bias goes, microwave ovens are half wave rectified and the power transformer has all sorts of "gimmicks " to prevent it saturating completely. Namely a big wedge of leakage designed into the traffo itself to prevent it saturating. PLUS, so far I have never seen a toroid in a microwave oven, always EI and always quite poor by Hi-Fi standards, the iron used. Its mostly 3/4's of a MM thick and even 1 mm thick in large commercial types. I use .35mm or 14 thou in my transformers, and if you want to pay extra, and put up with the wait time I will wind with 7 thou or .178 mm for really low leakage and wide response.

Even chokes designed to smooth HT and reduce ripple have the same specs. The BEST chokes are made with 7 thousands of an inch thick laminations.

All the above is just my experience.

Joe
 
SE can be a useful topology, and there are decent transformers out there for power levels up to a few watts or so. From what I can tell, anything more than that introduces serious compromises. In fact, data sheets for the higher power transformers appear to be cleverly camouflaged to hide the downsides. I'll never build another SE amp for more than one or two watts.

211 Monoblocks

I'm anxious to try Class A triodes in push-pull, but none of my vintage OPTs are sell-suited to that. They're all designed for AB1 pentodes, and the impedance is too low for optimal performance. Maybe I'll just sell everything off and buy the iron to build one last excellent push-pull amplifier.

Jack
 
You must log in or register to reply here.

Similar threads

TriodeLuvr
Output Transformer Impedance Specifications
Replies
6
Views
691
joebog1
J
M17XR2B
DIY The Oily Boy headphone amp
Replies
10
Views
1K
paul_b
P
TriodeLuvr
Distortion Test Adapter
Replies
8
Views
778
TriodeLuvr
TriodeLuvr
M
Sold Neumann BV33 stereo moving coil SUT
Replies
0
Views
718
manfredas370
M
R
Sold Neumann SMB2 valve phono stage with BV33 input transformers (WV2 / EMT)
Replies
7
Views
1K
airdronian
airdronian
Back
Top