Marinair LO1166 vs Carnhill LO1166 - VTB9049

[KraftWerk]

I made the first transformer according to CJ notes.
I compared that transformer with a modern Carnhill VTB2569 and that transformer sounded ´´Vintage´´ ( what a surprise ).
I thought, more bass and more air sounded nicer than the transformer I made.
CJ told me that old Marinairs do not track sam new Carnhils.
I decided to buy an old Marinair (599 euros) to assure myself and adjust one I made to old Marinair specs.


This is the result.



Aligned




CJ have the right.
They are noticeably different, not only have different graphs, they sound different.
On listening tests, the old one is a little quieter, a little less bass, more in the middle, and highs is softer than Carnhills.

 

[daristiguieta1]

Very cool, putting to rest the stigma that various users seem to place on the Carnhill...They seem close enough, would you agree?

 

[KraftWerk]

They are night and day, Marinair have bigger bottom end, softer highs,and Carnhilll is tinner in the low end and way open from 10 to 30 khz..when you swap these two transformers in device, you can easy notice difference...region from 70 Hz at Carnhill easy start to drop, and around 20hz have -0,3 dB down, on graph looks so small, but is much hearable..I'm wondering, is it same with new Neve transformers exclusively made for AMS Neve products?
I saw transformers with sticker AMS Neve...maybe they are closer to Marinair Lo1167...

 

[ruffrecords]

You cannot just "swap" these transformers and expect them both to work optimally. You need to adjust the output stage bias current for each one to the correct value.

Cheers

Ian

 

[KraftWerk]

You cannot just "swap" these transformers and expect them both to work optimally. You need to adjust the output stage bias current for each one to the correct value.

Cheers

Ian

I didn't know that,. I expect, If I have 2 same type transformer, and swap them, in my AML EZ73 they don't need bias re adjusting They must be same, same turn ratio,same core,same size wire, outputs is always in range +- 0,5 dB.
If Carnhill is direct replacement,why need bias re adjusting?In any case, I will try to do what you saying.
This weekend I'm at home and have time...

 

[SIXTYNINER]

You cannot just "swap" these transformers and expect them both to work optimally. You need to adjust the output stage bias current for each one to the correct value.

Cheers

Ian

Also a check about something happen on the "Harmonics" ?
cheers.

 

[ruffrecords]

I didn't know that,. I expect, If I have 2 same type transformer, and swap them, in my AML EZ73 they don't need bias re adjusting

They do, that is exactly why the bias adjust pot is there

They must be same, same turn ratio,same core,same size wire, outputs is always in range +- 0,5 dB.

They are very similar. The turns ration will be identical and the amplifier ac gain is set by negative feedback so you expect the output to be within 0.5dB. The bias is a different story entirely.

If Carnhill is direct replacement,why need bias re adjusting?In any case, I will try to do what you saying.
This weekend I'm at home and have time...

Carnhill may use a different wire size and a different tension. Both will affect the dc resistance of the the primary and hence the bias.

Cheers

Ian

 

[KraftWerk]

Today I did some tests because I wanted to see how the 3 transformers behaved when changing the bias.
In my microphone preamp aml EZ 73 is a standard Bias 90 mV.
I was able to change this bias from 50 to 130 mV.

I compared the original Marinair Lo1166a with modern replacement Carnhill VTB 9049 transformer and the transformer I made named Kraftwerk VTX.

I also compared the original transformer in the Aml ez 73 is marked VTB 4049, which turned out to be thinner than the others, but also with extended high frequencies, and I was not impressed.
This transformer is made for the 500 series and its coil is reduced compared to the standard coil for ei57 lamination.
Ruffrecords was right when he said that the bias affects the transformer spec and it is better to adjust the bias again when the transformer is replaced.
However, combining all the transformers with different bias values, I could not adjust the Carnhill Vtb 9049 to have the same depth as the Marinair LO 1166.
I believe I know the reason why this is so, because by experimenting I got the same results as the stock Carnhill, and with additional experimentation, I surpassed those results at 20 hz, which was my goal, my transformer has a few shades of lower tone than Carnhill and one shade deeper tone than Marinair.View attachment 134381
After many hours of listening and comparing,I noticed that a seemingly small drop starting from 80 hz towards 20 hz has an impact on the depth of the sound and what in the measurements is 0,05 db or 0,1 decibel in hearing tests turns out to be audible.
For me, all these specifications are not as important as it is important for me that the transformer has a good sound.

this is how Carnhill behaves when changing Bias

this is the difference at 20 hz

this is how a Marinair behaves


this is how my Kraftwerk VTX transformer behaves



this is how the distortion changes with the change of bias for Carnhill




at a bias of 90 mv, this is what the distortions of all three transformers look like



There would probably be much bigger differences if you increased the input signal and then followed the distortion change, but tthat is much work, if anyone want so see, i can do...

 

[ruffrecords]

That is an excellent, comprehensive series of tests. Well done. The only thing I am not clear about is exactly what method you use to set the bias. Where are you measuring?

Cheers

Ian

 

[KraftWerk]

That is an excellent, comprehensive series of tests. Well done. The only thing I am not clear about is exactly what method you use to set the bias. Where are you measuring?

Cheers

Ian

AML ez 73 have points for adjusting bias

 

[ruffrecords]

AML ez 73 have points for adjusting bias

OK, and as I understand it those two points allow you to monitor the current draw of the output stage. That is not the "official" Neve way of setting the bias. The Neve way is to load the output transformer with 600 ohms, feed in a 1KHz signal such that the output is +26dBu and adjust the bias control for best symmetry of the 1KHz when viewed on an oscilloscope.

Cheers

Ian

 

[KraftWerk]

Latest version

 

[dreamix]

Latest version

Looks awesome, i will buy one when its available

 

[KraftWerk]

I

Looks awesome, i will buy one when its available

Thanks Dreamix, Now they sound amazing...

 

[thomasdf]

Good job ! Amazing to see all this work coming to completion !

 

[KraftWerk]

This one is for fun

 

[ToneShack]

Latest version

Absolutely gorgeous work!

 

[Whoops]

That is not the "official" Neve way of setting the bias. The Neve way is to load the output transformer with 600 ohms, feed in a 1KHz signal such that the output is +26dBu and adjust the bias control for best symmetry of the 1KHz when viewed on an oscilloscope.

That's the method I used to set the Bias in my DIY clones and also in original Neve modules that I serviced