• 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.

Using Toroidal PTs as tube OTs - methods and sources

Hello all --

I am trying to get my arms around using Toroidal dual pri/dual sec type power transformers as OTs for tube amps. Here are the main questions I have:

1. How do you figure out which primary/secondary values you need? How do VA ratings for the xfmr translate to wattage? Are there formuli or decoder rings somewhere?

2. What topologies are best suited for these xfmrs? Any tricks to broaden frequency response?

3. What are some good USA toroid suppliers? Shipping costs for European suppliers are BRUTAL. I am aware of Antek - but they do not offer PTs (>25VA and <300VA) with the specs you most often seen mentioned as usable (115v/115v/6v/6v, 115v115v/18v/18v). Antek offers a full line of PTs with 12v secondaries -- could these be used?? which VA ratings are usable??

I have looked around in several threads but have not found definitive info - I am hoping to pull this info into a single thread so it is useful to the community at large. Thanks for the help!!!!
 
If I'm wrong, somebody please correct me, but as I see it secondary values are calculated as follows:
Find your target primary impedance from the tube data sheets, let's say 4500 ohms as an example. Now take your output impedance, let's say 8 ohms, and divide. 4500/8 = 562. The impedance ratio is the square of the voltage ratio, so take the square root of your previous result (562) and you get 23.7. If your primary is designed for 230 volts, then you would look for a transformer with 230/23.7 = 9.7v secondary.

Best topology for a toroid is parafeed, which completely eliminates DC. Next best is Push-pull, but the OPT has to be very well balanced to null DC or else it will saturate. Single Ended is out due to the lack of an air gap.

I'm in the same boat, I'm looking at toroids for an OPT. Looking at Antek specs, it seems thier bigger toroids have larger core losses, so bigger might not be better. Also, while toroids may be wide bandwidth, that bandwidth may not be linear, i.e. peaks and dips at some frequencies. I've heard of success stories with toroid OPTs, but I've also been warned off of them by experts like John Broskie. When I finish my o-scope I plan on testing the distortion of some toroids to see how suitable they are.
 
shoog has the bully on toroidal opts and the tricks. George (tubelab) and others are also fans of this approach. My next build is a parafeed SE which I want ot try toroids in. Calculations are the same as for any opt (square of this, ratio of that... you know the drill) but it does pay to Mcdonald the VA rating by a factor of around 50% to avoid saturation effects and ensure bandwidth. So, where you might think a 60W toroid will work fine, opt for the next size up...
 
Saturation depends on core area, core material, number of turns, and voltage. For most of these power toroids, it will occur at rated voltage and 40 Hz or so regardless of size. So if you want lows, you'll need to have less voltage across it, or a toroid with a greater rated voltage. These are all rated 120 / 240... and a push-pull amp with 300V power supply might put 1000V p-p (350VRMS) across the primary. It will saturate at full power around 60 Hz.

So low voltage tubes, twin-coupled amp, or maybe find a 240:240V transformer and add your own voice coil winding...
 
I'm somewhat stymied by this use of power transformers, toroids or otherwise, as output transformers. :eek: Are some DIYers really this frugal that they resort to this. Perhaps for voice only PA systems, but not for quality home audio. The output transformer is the most important part of a tube amplifier. (Not counting OTLs of course.) This is no place to be "el-cheapo". Why not just save up your shekels and buy the best "real thing" that you can?
 
the opt is ONE part of the tube amp. It can be the weakest link, but so can the driver circuitry, feedback, power supply etc etc.

There is a looooong history of repurposing components in diy, often successfully. F'rinstance, sweep tubes as outputs, G2 screen drive, even u/l connections in reality are an adaptation of a pre-existing component's design.

THe only abiding question should be does it work?
 
I have been thinking about building Crowhurst's twin-coupled amp using toroids.

This sound way cool to me - where could I find a schematic to look at?

Do both OTs connect to the same speaker???? In parallel ?? series??? In which case each OT needs to be half the impedence? So to drive a 4 ohm output you need two OTLs with 2 ohm secondaries -- is that right??

thx much for the post!!:D
 
Toroids work fine. I am using a pair of 34VA clunges in my workshop amp. 6080 push pull using CCS.
You really do need to be very careful about DC imbalance. Use some CCS. Also its not completely correct (it is in part) to use a bigger transformer. The smaller ones have more turns per volt, which equates to more inductance per unit volume (if that makes sense). Also they seem to have less leakage inductance and capacitance. This does make sense, since a larger toroid has way more windings in contact with each other. Also most run of the mill toroids are scatter wound, take a look at some, they really are bad. Crossed over windings are just asking for trouble.
I bought a whole box of these little trannys, I have not tested them as far as frequency response, leakage inductance etc. but I do have the kit to do it. So if any one is interested I could test them out as typical of the breed small toroids. I will have to repair my old HP 334A though (I have a Fluke true RMS meter but it gets sketchy above 15Khz)
Anyway I think the use of mains toroids as output trannys to be a great idea. The thing is to use valves/tubes with a low Anode R. Ie 6080/6532/6C33CB(the supply voltage is then quite low eg 180-200V). Then chose toroids with low secondary voltage like 6V or even 3V to get the required ratio. Or even use a cathode follower output stage(insane problems driving it though).
Nice.
Matt.
 
I recently built a P-P 15W amp with toroidal transformers.. a few things contraditory to my original expectations..

1. Toroidal transformers are easily to saturated by imbalanced DC --my initial build had a DC 5ma out of balnace.. the amp was still playing music flawlessly for nearly one full week.

2. Smaller VA transformer has higher "inductance", hence should deliver better bass -- I used 150VA ones, I was shocked to hear tight and deep down bass I ever had.

3. An amp with toroidal transformers needs complex circuit and hence expensive to build --I read thru Shoog's related threads and added minor mods (due to parts avaliability)..the final circuit was not that complicate and cost me around $400+

Here it goes, my amp in action ..(hope pic comes up ok..)
 

Attachments

  • DSC_0047.JPG
    DSC_0047.JPG
    224.2 KB · Views: 500
Last edited:
I am perpetually amazed at the criticisms levelled at power transformers as OT by people who have obviously never used them and so have absolutely no practical experience to pass on. In a word shut up until you have some real experience to pass on. Give them a try and make an informed response.
I would contend that power toroidals will outperform EI PP output transformers in almost all situations. If i had a choice between most commercial EI outputs and power toroidals I would generally choose the superior toroidals. If I could afford them I would buy toroidals specifically designed for output duty - and would contend that they would better all other OT's.

I have built a Parafeed amp using 807's and toroidals, and the only limitation was the parafeed cap.
I have built three PP power amps using 6080's, 807's and PL84's using power toroidals. I have built a EL82 headphone amp (which I am listening to now- very nice:D) which uses power toroidals. I have used power toroidals as step down input transformers. The least successful experiment I conducted was as interstage transformers - but they worked adequately if you used enough current and ,low enough drive to overcome the interwinding capacitance.


How would I characterize the results. Extremely good bass down to 10hz, and treble extension out to 50Khz before roll off. This has been consistent across all designs. Others have listened and compared and found the amps very pleasing.

My practical observations are;
- smaller, within reason is better, because of increased inductance. However Core size will effect the ability to sustain bass notes below 50hz (the designed low limit). I therefore advocate something around 100VA. The batch of Telcam transformers I have used were about 200VA, and I haven't noticed any significant penalty for this gross overspec.
-I have used them to PP impedences of 6K with 4ohm loads, with plate voltages of 320V. Beyond this you cannot go because there are not enough turns left on the secondary. Because of this I have used low rp tubes such as the 6080 and EL86. I have also used Schade feedback to further lower output impedance which gives better bite on the primary impedance. I advocate this approach as the best path.
-Interwinding capacitance is the limiting variable, but is not a major issue with the step down ratios used - expect response up to 50khz min. Leakage inductance is a none issue as the core is fully enclosed. Wiring the primaries in one direction will make interwinding capacitance worse, and in the other it will make it better.
-Interwinding capacitance will lead to a resonant tank which will cause ringing out beyond 50khz - this sounds nasty, and can lead to core saturation and fall off in bass response.
-I like to get my DC imbalance down to less than 1mA, and the secondary benefit is hum cancellation. This is easily achieved with simple CCSinks (LM317's work very well here), and also by Blumfields Garter Bias - though this doesn't allow differential operation of the output stage.
-I have only ever used them in class A designs, and always use differential outputs.
-There main advantage is lack of hysteresis which means that unlike EI's there is no loss of detail at critical low level listening and crossover. This gives them micro-detail comparable to SE amps.
-If ringing is addressed I found frequency response largley flat across the audio range.
-I have used both multicorp and Telcam and both work well.

Within their acknowledged limitations I would say they cannot be beaten - anyone with doubts go ahead and build with them and compare the circuit to one using EI's, a quick substitution should settle all doubts.

Shoog
 
Last edited:
My practical observations are;

-Interwinding capacitance will lead to a resonant tank which will cause ringing out beyond 50khz - this sounds nasty, and can lead to core saturation and fall off in bass response.

-If ringing is addressed I found frequency response largley flat across the audio range.

Shoog

How have you managed to get rid of the ringing? A Zobel network?
 
When using them as input transformers I have found that loading the secondary with an appropriate resistor works.
With output transformers I have found that simply reversing the polarity of the primary eliminates most of the capacitance. I noticed the response falls gently away from 50Khz, and the bass response is sustained down another 30hz or so.

Shoog
 
Ferrite EI/EE core for OT ?

I am just wandering that the Ferrite EI/EE core is ok for OT or not.

I think the Troidal and Ferrite EI/EE transformer has the same core.
And The EI core has air gap and easy to wind by hand.

Is there any reason to against using the core for OT?
I need your feed back before try out the core for OT.
 
A scope will certainly help with fine tuning a circuit - but I found that the difference between having it wired and ringing, and wired and not ringing is sufficiently audible that tuning by ear is possible.

Toroidals are not possible to wind by hand (because of the large number of primary turns), and has the limitations of DC intolerance. Also the number of primary turns on a power toroidal is not optimised to get large ratios. EI's are a proven technology which can be done DIY, and if you have the time and resources to master the technique will produce an excellent transformer.

Ultimately either will do a very good job, I think that in absolute terms a toroidal will always have better performance - given due care about its own particular requirements. Also bare in mind that designed OT Toroidals cost more than conventional EI transformers.

Shoog