Making Neve Style Input Transformer for 1073/1290 preamp

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

Spencerleehorton

Well-known member
Joined
May 12, 2012
Messages
4,031
Location
Felixstowe, Suffolk, UK
Hi,

Thought I'd Document my journey for making a Neve style input transformer.
So at least I have something to refer back to once I'm done, if I want to make more.
Initially I'm just going to make two but I do need to make about 10 of them for my pre amps.
I have bought some nickel lams and bobbins EI-35 size first of all and need to look at getting the correct winds, enough inductance and low resistance.
I have my own winder and was thinking to start winding the following, if it fits on the bobbin.

PRI winding - 84 turns X 2 , SEC windings - 168 turns X 2

This should hopefully give me the step up 1:2 ratio I will need for the mic pre amp.

I'm not sure whether it needs to be gapped or not, I'm pretty sure it's not, but would be better to confirm this.

I also have some mu-metal cases I can pot them in.

If anyone could chirp in with any info on bifilar or reverse winding any of the winds needed etc this would be fantastic.

Regards

Spence.

 
It definitley does not need to be gapped. I think you will need a lot more turns than you mentioned. What is the Al value of the core you are using?

Cheers

Ian
 
Hi Ian,

High Quality Nickel Permalloy Cores and Bobbins for Line Input&Output Transformers

Specification:
• Core Material : Nickel Permalloy Core(PB 49%) 
• High Saturation Magnetic Flux Density(15000 Gauss) and Low Distortion
• Thickness : 0.35mm
• Size : EI-35
• Bobbin : EI-3510(10mm Stack, 5pin + 3pin)
• 2 Sets( 2 10mm Stack Cores + 2 Bobbins)

Is that any help?

Like I said, my first numbers are complete guesses!!!!

I don't know how many turns of 0.1mm I would get on an EI-35?
Anyone know?

Regards

Spence.
 
I have asked CJ but no word back yet.

I have worked out the approx space for the wire and its volume is 14625mm2.
I've divided that by 1 turn which is 40mm which is then 366 turns.
Then divided that by 3 to then give 122 turns for the PRI then 244 for the SEC.
Then half each of the turns to make two different windings.

Does this sound on the right lines?

Regards

Spence.
 
Does this sound on the right lines?

No.

here's my take, I may have misread some specs somewhere but..._

EI-35 bobbin looks like something like 12mm x 15.2mm window - e.g. http://durgaelectronics.com/resources/pdf/ei35ver12pin-0123.pdf.pdf - totalling ca. 182mm2 space for wire

I don't have my calculator for wire-fill at hand, but say worst-case we'll be able to fill 10x10 (=100) 0.1mm-wires into a 1mmx1mm (=1mm2)window.

Which means that we could squeeze 18200 windings onto 182mm2. Now it looks like something.

You'll need to know your approximate Al-value: wind e.g. 100turns on a bobbin, mount, measure inductance, and deduct Al-value:  http://powermagnetics.co.uk/calculator

Once you have Al, you can start predicting how many turns you'll need for having what sort of bass in what sort of impedance.

And when turns and ratios are in place, it'll probably be time to start worrying about how to control inter-winding capacitance, which will try to get to roll of your highs, just when you have your lows where you want them....

I'm sure that CJ will chime in too... :)

Jakob E.
 
Hi Jakob,

thanks for this, i knew my guesses were completely wrong!!!! but didnt realise they were that far out!!! ;)

so it looks from that possibility that i will be able to get a good 3000 turns on each secondary and by my DCR calculations i get around the 120ohms which isnt too far off the 130ohms stated by the carnhill version.
and if the primary is 1500 turns then that is about 30 ohms not too far off the 24ohm stated by the carnhill, so i must be in the ballpark with these turns count. maybe the primary has a little thicker wire which will lower the DCR.
not that im worried about DCR but im interested in possibly how the other transformers are wired and would like to at least hit those parameter, hopefully get a little more out of it.

any thoughts CJ?

regards

Spence.
 
im understanding up to this point

Once you have Al, you can start predicting how many turns you'll need for having what sort of bass in what sort of impedance.

? impedance wise i think i need about 140 Henries, think i read that the carnhill is about that?

I don't understand what you mean about what sort of bass?

impedance wise it just relates to the ratio right? so what ever the total figure of winds i divide by 3 then i have the 2:1 ratio right?

regards

Spence.
 
output impedance of the mic and primary inductance of the transformer are forming an high-pass filter. The higher the primary inductance, the lower the corner frequency of this filter, thus the "bass" response.

140H is a LOT for a mic preamp transformer primary inductance, this maybe the total transformer inductance (with all windings in series) which means the primary inductance is around 15H.
 
I don't know whether anyone else has noticed but most schematics, PDFs, jpgs etc that were available are not available from old posts now?
There were a lot of winding sheets that I had on my other iPad and sadly don't have now as the links are dead.
Has anyone else noticed this?

Regards

Spence.
 
Spence, dir you see this one? It adresses some of the questions you might have asked yourself.
http://www.jlmaudio.com/Neve%20transformer%20info.htm
 
that Neve might have some humbucking qualities to  it with the dual coils, the jury is still out on that,

if you use EI lams you will have a single coil, so your start and finish leads will not be the same,

that core will have plenty of inductance if you use the same amount of turns as the Neve,

i bet you can get Neve lams and bobbins from Chameleon Labs or i would go for the K-241-D coil if using EI.
 
Hi CJ,

What order should I wind on a EI lams?

Also I reckon I could get more winds on than 3000 turns total, should I try and calculate it up with thicker wire or stick to the thickness used?
Also should I use 2 earth screens, if so between which layers?

Regards

Spence.
 

Attachments

  • IMG_0145.PNG
    IMG_0145.PNG
    292.7 KB · Views: 54
Hi There!! I am working on similar stuff since June  ;)

I am using 10x10 coil bobbin with a window of 12.8x3.9mm with a wire of 0.012mm (AWG37) and i can't fit more than 2200 turns with insulation.
I use GO lamination, and i have recreated both 1166 and an input transformer with 1:4 ratio.
I also did two different pcb for test them. The output stage is the BA340 with the 1166 style OT; the input is the BA338 with the Input transformer. The first test are really good, around 22dBu headroom for the output @ 30VDC (the BA340 is a bit modified to fit 30VDC) and good linearity from 20Hz to 18kHz.

 

Attachments

  • IMG_8223.jpg
    IMG_8223.jpg
    128.1 KB · Views: 75
This is the output board with the OT 1166 style and the BA340.
 

Attachments

  • IMG_8224.jpg
    IMG_8224.jpg
    122.6 KB · Views: 85
start simple and see if you can get a good freq response with 1000 turns primary then a 2000 turn secondary with a copper shield in between,
 
or tape over the start of the copper shield and overlap the finish,

solder the shield lead wire to the middle of the shield to balance capacitance,

here is a dimensional comparison between the EI 35 metric lam and the EI 375 standard lam,

first we converted the metric dimensions  to English, then  compared them with the English dimensions of the EI 375. The English dims are shown in parenthesis,

the tongue is about the same but the window is different enough to cause problems if you were to try and use an English bobbin on the metric core, however, a good  mill bastard could be used to knock off some nylon along the flange,

700 turns on the primary will give you the same inductance as the original Neve mic input and will add 3 db input level, 1.4 v-rms means 2700 Gauss on a core that will handle 5000. That core you bought off evilbay has two errors on the spec page, 1)-49 Ni will handle 12 KGauss and 2)-if you do an inductance test, you will find that what he sold you is actually 80 Ni with a max level of 5 KG, however, you want 80 Ni for a mic input and 50/50 for line input,

Original Neve inputs generate 2700 Gauss at 1 V-rms input level.  but there might be a pad on the console so this point is mute, or moot, depending on which dictionary you use, Webster or AES.
 

Attachments

  • s-l400.jpg
    s-l400.jpg
    68.3 KB · Views: 94
Back
Top