• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Toroids VS EI Transfomers

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I have done a search and could not find a thread specific to this argument. There is certainly a lot of discussion regarding the topic in assorted threads. If there is some relevant discusions you know of, please post a link here in this thread.

I am starting to get back into tubes after a long haitus and building frenzy with FETs. I find toroids in power supply and choke application to be a God Send in solid state design. I am not happy with the rapidly escalating prices of toroids from top of the line manufacturers like Plitron and certainly not with the price of tube related toroids, output or supply.

Price aside, I would like to see hear some practical discussion on the differences between Toroid and EI style transformers in there applications with tube amps. In solid state it is generally accepted that a good toroid is the best thing going, quite the opposite with tube designs it would appear. Is this just romantiiscm for days past or real Science?

Regards

Anthony
 
My preference for EI's is due to the narrower bandwidth. Why let all the c*r*a*p on the mains in in the first place? Did some measurements in a test lab years back and the inclusion of a huge EI iso tx dropped the dirt from the mains significantly. I also always use electrostatic foil (not another winding) screens wherever possible. None of my system has a torroid PS.

Torroids are only better due to the smaller size and lower radiated field, neither of which are a particular issue in DIY where it doesn't get moved about as much, it can be designed to negate these effects nor usually have to fit in a specified form factor. The power tx's in the MI gear we build are torroids for these reasons, and about 15% or so better price for the torroids I have made from a specific supplier.
 
What Brett said...

Even my next chip amp will use split bobbin EI transformers in preference to toroids. The last thing you want in a power transformer is wide bandwidth.

Edit: the best thing about a toroid is that you can pull the secondaries off a cheap one and wind up your own centre-tapped filament transformers to the exact voltage you need. Good for those 2.5V and 4V valves.
 
Konnichiwa,

Coulomb said:
Price aside, I would like to see hear some practical discussion on the differences between Toroid and EI style transformers in there applications with tube amps.

The issues are identical to those in Solid State Amplifiers.

Coulomb said:
In solid state it is generally accepted that a good toroid is the best thing going,

It is? I must have missed that.

So you consider a transformer "a good thing" which has the following fundamental flaws:

1) no airgap (an EI Transformer invariably has a distributed airgap even if tiny) and very little iron in the core, leading to early saturation, waveform distortion and high sensitivity to DC offset in the mains and "flattopping" and which in the real world as a result run hot, often hum and are inefficient, unless extensive external circuitry is appended to offset these issues which do not exist in EI Transformers

2) low primary inductance which as a result stores little energy which could be used to "bulk up" the current flow on the rectification peaks (measure the primary current waveform of EI vs. Torroid to see what I mean)

3) Exteremely high coupling capacity between sections which makes sure all mains related junk sweeps right through the PSU and onto the supply rails.

4) Due to construction and the problems around high shunt capacitance and low leakage inductance makes snubbing out the ringing from the reverse switching all the more difficult

Given the drawbacks compared to more traditional shapes of mains transformers it strikes me that the only thing Torroids offer is more profit to the transformer maker and/or equipment maker as they are MUCH CHEAPER to make and in materials than EI Types.

I use torroids where I have no no real choice and fit the required circuits to reduce their drawbacks but would prefer to use EI transformers of suitable construction instead and to just connect them.

Sayonara
 
Kuei Yang Wang do not beat around the bush, tell us how you really feel! :)

So if Toroids are cheap to make why are they many times more expensive than conventional designs? Is it because a lot of us fall into the newer is better trap?

With conventional EI and R core transformers who has a global distribution base other than Hammond?

Does Hammond make the best EI transformers or do they have the best marketing?

BTW please post links to your favourite suppliers of transformers if you wish.

Regards

Anthony
 
Konnichiwa,

Coulomb said:
So if Toroids are cheap to make why are they many times more expensive than conventional designs?

They are not in Europe. It seems in the US the market supports a higher pricing for torroids than for EI Transformers, in Asia and Europe this is not the case. If cost/size/weight is a concern you fit a torroidal, if quality is more important R-Core, C-Core or EI types are used.

Coulomb said:
With conventional EI and R core transformers who has a global distribution base other than Hammond?

Hammond is largely unknown outside NA. Many manufcaturers for commercial grade EI transformers exist. Just look into the usual industrial electronics suppliers catalogues.

Coulomb said:
Does Hammond make the best EI transformers or do they have the best marketing?

I think neither. They are non-entities out here. As are Plitron BTW.

Coulomb said:
BTW please post links to your favourite suppliers of transformers if you wish.

I have no preferred supplier per se. I tend to peruse the Catalogues of Farnell, RS-Components and others.

If I want something wound to my own spec I do have links to a number of suppliers, but they are industrial, so don't bother asking for a pair, minimum order on custom designs tends to be 10 - 20 pieces if you supply the full design, more if they design it.

If I need a transformer or two personally I can always drive over to Stevens & Billington on a saturday morning and wind my own, just pay for cores & wire.... ;-)

A company I never ended up dealing with (they quoted way too long delivery time and where impossible to pay when I asked - which was during their startup so should be better now) is Reinhoefer Electronics in (former East) Germany. The guy behind it really knows his ransformers and his designs are reputed very good, but as said no 1st hand experience.

http://www.roehrentechnik.de/

Sayonara
 
Hello all.
Well, if you read throughout this posts you can see why I make my transformers and inductors myself.
Cost? Who cares? I don't!
I'm not an industry I'm a DIY so why I should care for the cost of somenthing. Right the way not to use gold or platinum wires for windings. Let's think... I'm investing many hours a day for making something really unique, a beatifull object, a nice amp, with nice tubes, so, why to worry about few dollars or euros for a transformer? This will be somenthing that will last for decades! Even when I'll be dead it will be working, with the pleasure of the peple who remember the author of such a hard work fruit of patience. So I use the best components regardless off their cost.
What will be the real cost for making an amp? Time. Borrowed time from my life.
Further, I consider the magnetic components of the amp the key for good results from main transformers with shielded primary. Low output resistance high iron weight for low flux density, non magnetic field around and no hum etc... Who will be concerned about a couple or more pound of weight? Is this transportable?
Now, if you aren't sick yet, there cames the answer to toroid or not. I find very difficult to wind toroids myself, meanwhile, I enjoy to wind perfect bobbins, using polypropylene for insulation. I also enjoy to make brass covers bright polished or hand docored and then chrome plated, or even gold plated.
I could'nd do this with a toroid. Who will in turn show all the troubles you all stated in the above posts.
Nevertheless, fine hearing people warship MOS amplifier powered by toroids. But they are a class apart!
Have a nice day!
 
Hi there.....come on Larry.......burning the midnight oil winding EI trannies ain't so bad.....you can think of other things while you are doing it.....Re mains trannies......if you wind the primary so it is properly balanced to line common mode noise you shouldn't need an interwinding screen. Tried it ? Wind every alternate winding on the primary the opposite direction with the core the other way round......you know what I mean but it's alot of work. I did a tube amp tranny rewind for a friend who was having alot of problems on stage as so much lighting control uses phase controlled junk. The problem disappeared.
If you fit an interwindg cu/Al screen on audio parafeeds, it will cause ringing and overshoot.
I agree with you on large cross EI section core....but there are advantages with toroids......far lower winding resistance and Z for the same VA with an EI and the regulation factor is nearly halved. What many don't realise .....most toroids are pile wound and have a temp rise limitation whereas a well wound EI can gracefully stew.

rich
 
Konnichiwa,

planet10 said:

This article is correct on all counts but one. Torriodal transformers are not more difficult or expensive to wind and make than EI ones.

Once you have the right machinery torridal transformers of the generic kind are very quick and chap.

1) The core has less iron and what is there is tape wound into the doughnut shape from a strip of iron, no stamping out laminations etc. The core for a torroidal transformer contains much less labour and material than an EI one.

2) There much fewer turns on most torroidal transformers, this means less wire is used, one can even get away in many cases substituting cheaper aluminum for copper in the winding, as the DCR will still be lower than EI Types.

3) It may be possible to wind more EI Transformer bobbinss in one go than torroidal ones, but a torroidal transformer is fully assembled once it comes from the machine, which winds very fast btw, the EI Transformer invariably requires additional steps to fit the core and to clamp the assembly together.

Hence, at the same level of technology applied a Torroidal is invariably cheaper to make. Of course, if your EI Transformer winding machines are the same since 1960 and paid many times over and you don't have the capital for big, efficient torridal widing machines (which are arguably very different) then EI's are cheaper to make, but ONLY then....

Sayonara
 
Hy there....
It looks that's somenthing I'm missing.....
Toroids are cheaper to build than EI with modern machines....
Uhm.... right!
This is obviously true, becouse we have a toroidal transformer maker here in Tuscany and is quite cheap.
But you don't ask them for a strange one or it is better for you to buy one from NASA suppliers. And a strange one means a one out of 120/220/380 to 12/24/48/110/115/120/220/380 volts.
So no way to use them for a tube amplifier.
Lets say that is the same for EI transformers. The price is about 50cents of euro per watt.
Commercial EI are a real bummer. Since it is supposed you are going to use it under load the wind that stuf wit a 1.6T of induction in the core, or even worse. So a variable load will turn those transx red hot. Not to speak about the em field radiated all around.
No way this is trash.
What I feel I'm missing, I think Is that around North America you should have better transformers makers than here, do you?

Yes, Richwalters, there are a lot of thinks to do wile winding bobbins, like hearing you wife's complains :dead: , your mother complains :cannotbe: , a friends tales :xeye: , friends talking about a soccher game :bawling:
It is a real good excuse to not to waist your time getting slippery bored. When they say: "Hey, are you there? What we where talking about?" you can answer: "I'm sorry I was counting the spires number, I bag your pardon?" :devilr:

Sincerely Larry
 
Konnichiwa,

Larry Lomax said:
This is obviously true, becouse we have a toroidal transformer maker here in Tuscany and is quite cheap.
But you don't ask them for a strange one or it is better for you to buy one from NASA suppliers.

Yes, that is true for any commodity item. If you wantn one-offs with different specs you pay the setup charge. For small scale items EI works out cheaper to make.

Larry Lomax said:
And a strange one means a one out of 120/220/380 to 12/24/48/110/115/120/220/380 volts.
So no way to use them for a tube amplifier.

Yes, there actually is a way.

Here is how.

Let us say you get a cheap inductrial 115V+115V:55V+55V Transformer with 500 - 1000VA. You get them quite cheap, I have several enclosed versions of these to make 110V balanced power for US Spec mains gear.

Take 2pcs, say 500VA.

Connect one to the mains as usual.

Connect the other in reverse to the first transformers secondaries. If you accept de-rating to 50% of the power you just connect one 55V secondary each.

You now have windings as follows which you can series to taste to get you a number of secondary voltages:

115V/0.7A
115V/0.7A
55V/0.7A
55V/0.7A
55V/0.7A

So you can have nominal AC of:

395V/0.7A
340V/0.7A
285V/0.7A
280V/0.7A
230V/0.7A
225V/1.4A
170V/1.4A

With that you can use Hybrid Graetz Bridges and Voltage doublers to cover almost ANY HT Voltage you want, up to around 1,200V/0.2A.

Add handwound secondaries for the heaters onto the first transformer (which is used to only 50% if hooked up as discussed above) up to around 250VA and you have a very cheap supply for Valve Amp's for little work (most torrid torroids work out at around 3 Turns/Volt).

However, if you do that, remember to filter the Primary side input for the first torroid transformer for common mode and differential mode (RF) noise plus maisn born DC and then you find that unless you can get the big torroids as surplus or equipment at ten cent on the Euro you might as well get someone to wind you a custom EI or LL device with balanced primaries and secondaries the primary windings placed on the outside, electrostatic screens plus a little airgap and a flux limited to maybe 1.2T.....

Sayonara
 
to wind you a custom EI or LL device with balanced primaries and secondaries the primary windings placed on the outside, electrostatic screens plus a little airgap and a flux limited to maybe 1.2T.....


Kuei

I know this has been discussed before but all the same it would be great to update the list of requirements for best sounding PS transformer. I have tried some of your recommendations like running at half flux, using electrostatic screens and removing dc from primaries. All have led to welcome but minor improvements. In comparison, a badly wound transformer on very old used and abused laminations sounds abslolutely amazing although hums (it hasn't been soaked in oil) and gets dangerously hot. It may have lower dc coils though.


What is the importance of primaries being on top? What are balanced windings?

Is it worth pursuing double C cores? R-cores?


Many thanks
 
Konnichiwa,

analog_sa said:
I know this has been discussed before but all the same it would be great to update the list of requirements for best sounding PS transformer. I have tried some of your recommendations like running at half flux, using electrostatic screens and removing dc from primaries. All have led to welcome but minor improvements. In comparison, a badly wound transformer on very old used and abused laminations sounds abslolutely amazing although hums (it hasn't been soaked in oil) and gets dangerously hot. It may have lower dc coils though.

Well, if we take a clean sheet of paper and design ourselves the "perfect" supply, mine would actually be fully electronic.

I would start with a PFC correcting input switcher, properly de-noised, so we get a well regulated DC to feed our main switcher, disregardless of mains condition and regardless of mains voltage.

I would follow this with a simple, open loop, slewrate limited open loop switcher aimed at 50% Duty Cycle. This feeds a Transformers that provides us 12V or 24V eff. Squarewave AC at 76.5KHz, so basically the output is a safe HFAC voltage.

All this ideally in a simple sealed module with a flange to allow heatsinking, ideally IEC Input Socket and external insulated low voltage "on" switch. I'd like this available in a variety of power levels (50VA/250VA/500VA/1KVA). Options to allow several modules to be syncronised for switching appreciated.

The above will form a highly effective barrier to mains borne noise and other modulation below the switching frequency.

We then use small local EE, Pot or Torroid Core HF Transformers to take our safe 12V or 24V HFAC Voltage to whatever level we desire, locally directly next to our given load, internal distribution of the HFAC current via a nominal 100R terminated balanced line or something like that, with suitable shielded cable.

The very high switching frequency and the 50% duty cycle will make for a very solid and low ripple supply even with very small value reservoir capacitors, the remaining switching noise can be filtered quite effectively.

That would be what I would like to have.

If you want a traditional transformer, here my list:

Good tolerance to DC offset

Low DCR and other losses

High primary inductance

Narrow bandwidth

Ability to snub out primary/secondary ringing with a simple Zobel (series RC circuit)

analog_sa said:
What is the importance of primaries being on top?

It leads to less noise coupling to the core, especially if electrostatic screens are present.

analog_sa said:
What are balanced windings?

Same as in any other transformer. You make a winding that with respect to cores and secondaries is balanced, thus any common mode noise is rejected quite effectively, an electrostatic screen improves this further.

analog_sa said:
Is it worth pursuing double C cores? R-cores?

Maybe. They do not, from what I note have any intrinsic advantages over torroids (or EI), as usual it is a question of implementation.

You can make airgapped torroids wound balanced and with electrostaic screens quite easily, but they will drive the price up and likely more than an equivalent set of one-off EI cores would cost.

Double C-Cores are just torroids formed differently, the same can be said of R-Core transformers. A double C-Core and R-Core type is best likend to an LL Core Transformer (they where common in the eastern block) and has the main advantage that it allows EI production technology to be reused with more modern single piece (as opposed to multi lam) cores and a low profile.

The bottom line is that a transformer designed explicitly to meet the above "good audio" criteria (as opposed to meet highets power for lowest bucks/size/weight) will be a good transformer, regardless of construction.

Among commodity transformers quality EI types become scarce (in europe at least), otherwise EI and LL commodity types would be my choice with LL leading, as LL tends to be inherently balanced" and low leakage, moreso than EI types.

Sayonara
 
I would start with a PFC correcting input switcher, properly de-noised, so we get a well regulated DC to feed our main switcher, disregardless of mains condition and regardless of mains voltage.


Sorry to disappoint ... I've beaten you to it... A pfc boost smps on the business end is very viable with 0.99 power factor. A truely ++++ move with way better regulation....I chose this for my tube 150W+150W amp because it sucks rather alot of juice....however I get constant heckling from the pro audio loby because nobody can swallow switchmode in tube amps........my trumpet quality sounds just the same.....but far louder.....
In my locality I get billed for crummy power factor usage.....so that puts the curtain down on power toroids with cap input filtering......Bear in mind a tube amp using choke input B+ isn't so bad after all.

I also have a design that uses an EI mains tranny as a step down and heater supply while the HV windings go to the pfc......endless possibilities..to explore.


rich
 
Good morning!
I'm sorry to say, that I work everyday year round, on setting up industrial up-to-date, real modern hardware, plenty of digital electronics, AD/DA converters, supersensors, mega-amplifiers, ultra-regulators, fast-inverters. And I get sick of those allways-truble-generators. That real small SMD resistors is not well soldered so sometimes everithing crashes. That real big inverter 650KVA sometimes generates excessive harmonics.... everithing hung up....... Gentlemen, this is work.
When I like to get delighted with electronics, I love to use few very good uman-sized components disregarding the price.
I like to spend a 10$ on a mica capacitor, rather than in a switching power supply. I admit, anodic stabilizer could be a real good enhancement. But I don't like to start struggle with IGBTs MOSFETs controllers IC and ferrite transformers.
My hobby is a hobby. No compromises. How is the price for a component? Allways the right if this is the best component availlable.
I made computers just for fun, somebody recalls old AMPRO little board? I delighted with digital electronics. But.... Complaints where more than ..... how to say...... take a sit and listen!
:)
Bye!
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.