Author Topic: DIY Simple Output Transformer  (Read 46715 times)

Offline Mars

  • Commercial Member
  • Trade Count: (+1)
  • *****
  • Posts: 1,034
  • Total likes: 1
  • “The path is the goal.” Mahatma Ghandi
    • Mars Tube Audio
Re: DIY Simple Output Transformer
« Reply #30 on: November 09, 2010, 08:20:36 PM »
Yip... long live the learning curve. I've just started (2 years of designing and rolling), and I'm enjoying the ride. BTW It is surprising how much muscle one has to apply to manifest a transformer. Some real health benefits there :D

Thanks for all the feedback. The fact that Ip0 should be half the value I used, is welcome news. Thinner copper can be used and more windings rolled :-*

A question: does the current density chosen have an effect on the total losses registered? My mind says no; the total losses are a function of wire thickness entered and current + iron losses.

OEM manufacture of Tube Audio
Power and Integrated Amplifiers
Phono Pre-amplifiers
DIY Kits
DIY Workshops

Offline Mars

  • Commercial Member
  • Trade Count: (+1)
  • *****
  • Posts: 1,034
  • Total likes: 1
  • “The path is the goal.” Mahatma Ghandi
    • Mars Tube Audio
Re: DIY Simple Output Transformer
« Reply #31 on: November 09, 2010, 09:01:56 PM »
From Turner (http://www.turneraudio.com.au/output-trans-pp-calc.htm) I get this:

39.   Is the leakage inductance low enough?
Calculate reactance of LL at 100 kHz.

ZLL at 100kHz = L in henrys x 2 x pye x F
                          = L x 6.28 x 100,000Hz, ohms

OPT1,
ZLL at 100kHz = 0.00435 x 6.28 x 100,000 = 2,731 ohms .................2,731 ohms

Is ZLL less than PRL at 100 kHz?.............................................................yes/no

Eg, we have RL = 5,000 ohms,  ZLL = 2,731 ohms at 100 kHz...................yes.

40. If answer to step 38 is yes, leakage inductance is low enough.

My mind is shutting down here... simple calculation to done later :)
OEM manufacture of Tube Audio
Power and Integrated Amplifiers
Phono Pre-amplifiers
DIY Kits
DIY Workshops

Offline Mars

  • Commercial Member
  • Trade Count: (+1)
  • *****
  • Posts: 1,034
  • Total likes: 1
  • “The path is the goal.” Mahatma Ghandi
    • Mars Tube Audio
Re: DIY Simple Output Transformer
« Reply #32 on: November 10, 2010, 05:26:52 AM »
Calculate ZLL at 100kHz

0.0205(see the Leak L in light blue block) x 6.28 x 100 000 = 12874

Is ZLL less than PRL at 100kHz? No (PRL s 5000 ohms which is much less than 12 874 ohms  :'( )

We should aim for a figure of 8mH Leakage inductance. Combined with a lower total loss figure, we might have a working transformer?

So let's see what the software recommends.
OEM manufacture of Tube Audio
Power and Integrated Amplifiers
Phono Pre-amplifiers
DIY Kits
DIY Workshops

Offline Mars

  • Commercial Member
  • Trade Count: (+1)
  • *****
  • Posts: 1,034
  • Total likes: 1
  • “The path is the goal.” Mahatma Ghandi
    • Mars Tube Audio
Re: DIY Simple Output Transformer
« Reply #33 on: November 10, 2010, 07:31:33 AM »
Thanks for the link to the software Yves! It is a bonus to have the Power Transformer designer in the same package. We'll have to do a DIY Simple Power Transformer sometime ::)

Tip for user of OPT-da: Drop the file called core.tbx from your old OPT installation into the new folder (MyPack) and your lamination data will be transferred. Does that make sense? The new installation lacks your entered EI96 info we calculated above.

Ok... I'm starting to realise that this design is coming to a head. Not wanting to sound negative, I do not see a way we can satisfy all our design criteria with this proposed core and layering arrangement. Let us see how close we can get. It would be a very workable transformer nevertheless.
OEM manufacture of Tube Audio
Power and Integrated Amplifiers
Phono Pre-amplifiers
DIY Kits
DIY Workshops

Offline Yves

Re: DIY Simple Output Transformer
« Reply #34 on: November 10, 2010, 06:13:55 PM »
Thanks for the link to the software Yves! It is a bonus to have the Power Transformer designer in the same package. We'll have to do a DIY Simple Power Transformer sometime ::)

Tip for user of OPT-da: Drop the file called core.tbx from your old OPT installation into the new folder (MyPack) and your lamination data will be transferred. Does that make sense? The new installation lacks your entered EI96 info we calculated above.
Yes, it makes sense  ;)
Quote
Ok... I'm starting to realise that this design is coming to a head. Not wanting to sound negative, I do not see a way we can satisfy all our design criteria with this proposed core and layering arrangement. Let us see how close we can get. It would be a very workable transformer nevertheless.
Good point !

Yves.

Offline Ampdog

Re: DIY Simple Output Transformer
« Reply #35 on: November 11, 2010, 04:12:14 AM »
Mmmmm ....

This might fetch me the remark to rather go fishing. But, since I have tried from youth and have thusfar never caught a single fish  :'(  I might have more success with OPTs.

Two things:

That wonderful h.f. goal of 100kHz - it might make matters easier to take some variation on that. That was chosen on 'let's say ...', so why not 'say' something else? [It does have rather more to do with NFB stability than audibility (as long as it is above 20kHz), and NFB stability is governed by other factors also.] With this remark I mean that I am not going to chase after that figure studiously, bending the transformer configuration out of kilter to achieve it. It will be going OT to go further on that here; just mentioning.

Secondly - and again rather an allied subject: The chosen characteristics and choices governing them will only hold for class A push-pull. For AB, at least some of the time only one power tube will be working. During that part of the cycle one can discard the cut-off tube - and also its part of the OPT! How 'balanced' is the operation and what the leakage factor now? So one really sees two different leakages over any signal cycle. It does make sense to at least separate the two conditions and examine the two leakage factors. Especially in OPTs just glibly using part of the secondary windings for a lower loudspeaker impedance instead of some reconnection(s) to involve all of the windings all of the time, the wheels can come off (and does!).

But I did not closely examine Karel's design above, thus not saying he is at fault. I'm just mentioning this rather generally overlooked factor in some commercial designs. And then we come to the distributed load (UL) designs..... The screen taps are not simply going to another (high-impedance) grid. This time that grid contributes to the output, rather like a second little anode.

.... and then we go to the UL configuration with cathode NFB, like in the Quad II ... Ya OK! I will go fishing then!
Audio must be the only branch of engineering where lack of basics' knowledge is considered a superior form of wisdom. (Anon)

Offline Mars

  • Commercial Member
  • Trade Count: (+1)
  • *****
  • Posts: 1,034
  • Total likes: 1
  • “The path is the goal.” Mahatma Ghandi
    • Mars Tube Audio
Re: DIY Simple Output Transformer
« Reply #36 on: November 12, 2010, 06:34:27 AM »
I have downloaded Yve's latest OPT da from the link he kindly gave above, and updated my data files.

Let's recap:

We want the: Leakage L < 8mH
                   Primary inductance > 100H
                   Total Cu losses < 10%
                   Total Flux (B total) < 1.1 Tesla .........(laminations are Non Grain orientated)
                   Power 16.5 Watts
                   Bandwidth 30Hz to 100kHz (?)
                   Layering arrangement 2 Primary sections interleaved with 2 Secondary sections
                   Secondary limited to a single layer filling the whole bobbin width
                   Using EI96 laminations stacked 32mm high

I also took note that the current (Ip0) should be halved.



We come close to our target:  Leakage = 12.1mH
                                         Cu losses < 10%
                                         Primary inductance = 92.9H
                                         B Total = 1.09

Could this be the the one?

There is still a bit of room to fill in the window: Only 8.5mm of the available 14mm height of the bobbin is filled. Real world winding typically exceeds the predicted height, so lets come back to that later.

OEM manufacture of Tube Audio
Power and Integrated Amplifiers
Phono Pre-amplifiers
DIY Kits
DIY Workshops

Offline Yves

Re: DIY Simple Output Transformer
« Reply #37 on: November 12, 2010, 09:02:09 AM »
. . .
Could this be the the one?

There is still a bit of room to fill in the window: Only 8.5mm of the available 14mm height of the bobbin is filled. Real world winding typically exceeds the predicted height, so lets come back to that later.

Yes, you're in the ballpark to have a very decent OPT !
Using the "central" primary for one anode and the two "outside" ones in serie for the other will balance their DC resistance.
Beware about "phasing" !

Go rolling  ;D

Yves.

Offline Ampdog

Re: DIY Simple Output Transformer
« Reply #38 on: November 18, 2010, 05:46:01 AM »
Karel, Yves,

If I might add something more (adding, not criticising!) ...

I have always tried to fill the winding area - it is there after all. If the space factor [(actual winding space)/(available winding space)] is too small it means that one may have used thicker wire, thus improving efficiency, etc. Transformer winders frown at me, but I have wound (and had stuff wound) up to 93% space factor in OPTs*. One then has to compress the windings where the core goes over, but the space is there as indicated by the 'rounds'; no damage need result if carefully done. There is also the point (though again small) of having the outside layer as close to the core as the inside. Finally, the calculated leakage reactance shown is for exact winding. (Karel, I apologise if you have allowed for that earlier - I did not reread your design steps just now.)  Here there can be a measurable difference between a slightly oval structure and a compressed one - the distance between layers is a factor in the leakage reactance.

Comments please gentlemen?
____________________________________________________________
*Not every winder will do this. I then accept the transformer parts as a kit and do my own assembly - no guarantee in that case, however, but I never had a 'blown' OPT. The windings in such a case must be interleaved layer wound to specs, as I believe Karel does. 
Audio must be the only branch of engineering where lack of basics' knowledge is considered a superior form of wisdom. (Anon)

Offline Mars

  • Commercial Member
  • Trade Count: (+1)
  • *****
  • Posts: 1,034
  • Total likes: 1
  • “The path is the goal.” Mahatma Ghandi
    • Mars Tube Audio
Re: DIY Simple Output Transformer
« Reply #39 on: November 19, 2010, 12:37:08 PM »
Ampdog,

I agree. I prefer to fill the window as much as possible. I'll do some real world tests on the bobbin to see how close we can get to 100% on this one. I have a suspicion that it is going to be a tight one ;D

My hand winder arrived today... (I have been borrowing one from a friend till now)




Now I need to prepare the wooden blocks to keep the bobbin centre and square.
OEM manufacture of Tube Audio
Power and Integrated Amplifiers
Phono Pre-amplifiers
DIY Kits
DIY Workshops

Offline ngvtanhb

Re: DIY Simple Output Transformer
« Reply #40 on: March 30, 2014, 04:18:22 AM »
Dear Mr Mars
Thank you so much for your software which you have presented
But i inputed the parameter into  the green block ??? I got a wrong result for Push-Pull of OPT with EL34 tube
Please help me and see the below my picture
« Last Edit: March 30, 2014, 04:22:30 AM by ngvtanhb »

Offline ngvtanhb

Re: DIY Simple Output Transformer
« Reply #41 on: March 30, 2014, 04:22:52 AM »
I have downloaded Yve's latest OPT da from the link he kindly gave above, and updated my data files.

Let's recap:

We want the: Leakage L < 8mH
                   Primary inductance > 100H
                   Total Cu losses < 10%
                   Total Flux (B total) < 1.1 Tesla .........(laminations are Non Grain orientated)
                   Power 16.5 Watts
                   Bandwidth 30Hz to 100kHz (?)
                   Layering arrangement 2 Primary sections interleaved with 2 Secondary sections
                   Secondary limited to a single layer filling the whole bobbin width
                   Using EI96 laminations stacked 32mm high

I also took note that the current (Ip0) should be halved.



We come close to our target:  Leakage = 12.1mH
                                         Cu losses < 10%
                                         Primary inductance = 92.9H
                                         B Total = 1.09

Could this be the the one?

There is still a bit of room to fill in the window: Only 8.5mm of the available 14mm height of the bobbin is filled. Real world winding typically exceeds the predicted height, so lets come back to that later.

Dear Mr Mars
Thank you so much for your software which you have presented
But i inputed the parameter into  the green block ??? I got a wrong result for Push-Pull of OPT with EL34 tube
Please help me and see the below my picture

Offline Mars

  • Commercial Member
  • Trade Count: (+1)
  • *****
  • Posts: 1,034
  • Total likes: 1
  • “The path is the goal.” Mahatma Ghandi
    • Mars Tube Audio
Re: DIY Simple Output Transformer
« Reply #42 on: May 14, 2015, 06:10:21 PM »
Well it is 5 years later, and time to update this thread. I would also like to add photos of the actual transformer winding process.

I have wound a few transformers and experience has taught me a few things. Hopefully you'll benefit from my remarks :blah:

Have a look at the screenshot of the transformer design as I would do it now:


You can see that I increased the diameter of the 6mm copper wire to accommodate the enamel. If I enter 0.62mm, I get the correct amount of windings per layer. To confirm, I rolled one layer of secondary on the bobbin, and it came to exactly 65 windings.
OEM manufacture of Tube Audio
Power and Integrated Amplifiers
Phono Pre-amplifiers
DIY Kits
DIY Workshops

Offline Mars

  • Commercial Member
  • Trade Count: (+1)
  • *****
  • Posts: 1,034
  • Total likes: 1
  • “The path is the goal.” Mahatma Ghandi
    • Mars Tube Audio
Re: DIY Simple Output Transformer
« Reply #43 on: May 14, 2015, 06:20:40 PM »
Start from 0



65 windings later




So now to proceed with the actual winding.



OEM manufacture of Tube Audio
Power and Integrated Amplifiers
Phono Pre-amplifiers
DIY Kits
DIY Workshops

Offline Steerpike

Re: DIY Simple Output Transformer
« Reply #44 on: May 15, 2015, 01:04:13 AM »
I am lead to believe that OT winding is a bit of a black art.

In a way, it used to be.
Before computer programs existed to do it, the maths was quite unpleasant.
The equations are not straightforward plug-in formulae. You had to make a few educated guesses, do the calculations and then see if the results gave the right specs. If not, change a few things, redo the calculations & re-evaluate.
By hand & brain, it is an iterative process - and you get so bored!
A computer can do the iteration a million times in a second.