Custom Output Transformers

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Hello Ian

I just read both.

Thinking out loud, there is some design with small transformer including fixed pad before to lower the distortion.
This need extra make-up after, so trade off with the noise floor...

Best
Zam
 
zamproject said:
Hello Ian

I just read both.

Thinking out loud, there is some design with small transformer including fixed pad before to lower the distortion.
This need extra make-up after, so trade off with the noise floor...

Best
Zam

The line input of my tube mic pre designs uses a 34dB pad in front of the the mic transformer. The transformer has a gain of 20dB so overall the signal at the input of the tube amplifier is -14dB so you only need 14dB of make up gain so the noise floor is not badly affected.

heers

Ian
 
ruffrecords said:
The line input of my tube mic pre designs uses a 34dB pad in front of the the mic transformer. The transformer has a gain of 20dB so overall the signal at the input of the tube amplifier is -14dB so you only need 14dB of make up gain so the noise floor is not badly affected.

heers

Ian

Now I remember this  ;D

Best
Zam
 
volker said:
Nice read Ian! I'm glad it worked out with the manufacturer as well :).

Many many thanks for all your help Volker. The UI transformer would not have been possible without you.

We had a small hiccup with the first production prototype - they forgot to reverse the windings on core 2 but they soon fixed that and the second sample worked perfectly. There is enough material left over to make a dozen more which is just what I need for the first Mark 3 mixer.

Cheers

Ian
 
zamproject said:
Now I remember this  ;D

Best
Zam

The other interesting thing I came across when testing transformers if the effect of driving source impedance on the microphone transformer. If you drive it directly from a low source impedance generator you end up with a significant peak in the frequency response around 25KHz. It turns out this is quite normal. Most 1:10 mic transformers are designed to work with a 150 ohm source. As soon as you use the line input, where the pad has an output impedance of 150 ohms, the peak disappears. I got caught by this early on when I thought the peak was caused by the output transformer. So I always use the line input pad now when testing output transformers.

I should also mention that the frequency response figures given in the blog are the overall values and therefore include the effects of the input transformer, the tube amplifier stage and the output transformer. Just shows you how good tubes and transformers can be.

Cheers

Ian
 
I was half hoping Volker might write up the transformer design in glorius  detail  and permalink it for the good of humanity :D
With the dual bobbin symetry and the  low z out of Ians design  your not far off what the laws of physics dictate.
I think Ians  previous tests show a very gracefull overload , at what really amounts ,by modern standards to a worst case scenario ,ie 600 ohms , most likely than not 10 or 20 k will be what it see;s in a modern setting, that should in all likely hood mean the 3rd harmonic due to 'core' should fall off considerably.





 
I suspect Volker would be happy to publish the design. I certainly have no objection. One thing to bear in mind is that UI laminations and bobbins are not readily available in small quantities. I emailed the main UK suppliers of both and got no reply. It was only when Electro Mag, who are in the trade, contacted them that we were able to obtain any. Even then I ended up paying for enough parts for about 15 transformers just to send a enough to Volker to have a prototype made.

Cheers

Ian
 
Ive encountered that 'closed shop' mentality in relation to transformers also .
I guess back in the day rivalry kept proprietary stuff a closely guarded secret , at this stage the closed loop mentality really is only hastening the demise of the art of transformer design and winding . Luckily here in GDIY theres plenty of interest from a generation  ,like myself, who couldnt get the transformer knowledge anywhere else because it is/was considered out of date , unfashionable and uneconomic in main stream electronics .
 
Just to add to Volker's information that the laminations are UI39 in a 10,2mm high stack. The IEC code for the laminations is YUI 1-13.

Attached is the data sheet for the laminations.

Cheers

Ian
 

Attachments

  • UI_Laminations_Zg_Abmessung.pdf
    74.7 KB · Views: 26
Somehow I'd missed the updated posts on this topic ,

Many thanks Volker and Ian ,
I love the old fashioned style schematic like you see on Tele and vintage German gear , it just adds nicely to the abillity to visualise the winding scheme.


Finally now almost a year to the day later than planned  ,I have the CNC winder on the way  , price dropped from around the 550 to 475 euros which is good news. The guy at the winding place also has about 50 'end of '  spools of copper for me at maybe just above the scrapage rate , typically they contain from a few ounces up to a pound ,this happens because the end of the roll wont ever co-incide with the exact number of turns you need and your never going to splice in the middle of a motor coil or even use copper from two different batches in the one job normally.

Sincerest thanks to all once again.





 
Can't the series/parallel pri/sec coils  be rearranged to allow for other impedence ratios  ie 1:1 or 4:1 or  step up  ?
 
Tubetec said:
Can't the series/parallel pri/sec coils  be rearranged to allow for other impedance ratios  ie 1:1 or 4:1 or  step up  ?

Probably. The input has all four windings in series for a 2400 input impedance. If you wire them in series/parallel or all in parallel then this impedance drops appropriately. The secondaries are wired series/parallel for a nominal 600 ohms output load

Cheers

Ian
 
Its a nice side benefit of the 2x4 coil construction ,5 positions of source vs load depending on how you wire it , is each winding brought out to  its own set of pins , ie . 16 pins in total ?   
 
Tubetec said:
Its a nice side benefit of the 2x4 coil construction ,5 positions of source vs load depending on how you wire it , is each winding brought out to  its own set of pins , ie . 16 pins in total ? 

Yses, see Volker's post with winding details and pin numbers.

Cheers

Ian
 

Latest posts

Back
Top