Nickel lams on input TX

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

DAN_000

Well-known member
Joined
May 22, 2005
Messages
104
Location
Ganimedes
Im trying to get some info about nickel laminations for input mic/line transformers.
There are 49%, 80% nickel , .006 and .014 tickness

For example, why I found 49% nickel in vintage neve mic input transformers? 80% has more permability and maybe is more suitable to use there with less noise.

What about "the sound is in the iron"? 80% saturation point is lower than 49.

This vintage transformers like vintage St. Ives or Marinair sounds "warmer" and "sound is in the iron" effect with 49%.  Why not 80% was used  there ?

Maybe Im missing something...
 
line level means higher voltages than mic level,

so you need 49% for ine and 80% for mic

80% takes about 5 kilo gauss

49% takes about 12 to 14

and regular microsil, aka grain orient aka silicon steel takes the most at about 16 to 18 kilo gauss

thin lams means less eddy currents which means better freq response

but they are expensive and only used on a few transformers
this is because it is harder to stamp out the soft nickel lams compared to regular steel lams, they bend all over the place and bending lams takes away the heat treating that orients the grain,


80 % has less core loss at high frequencies
and since the permeability is a lot higher you can use a smaller core
which means less capacitance and thus better frequency response,



 
CJ said:
80% takes about 5 kilo gauss

Noob questions, sorry - but much transphormery is still foggy to me:

What does Gauss parameter mean/reflect in real-life? Voltage-pr-turn before overload? Transferred power before overload?

When winding transformers, is it important to split winding area evenly ½/½ between Pri's and Sec's, i.e. use thicker wire for lower-inductance section?

Very happy to see you here again..

Jakob E.
 
gyraf said:
What does Gauss parameter mean/reflect in real-life? Voltage-pr-turn before overload? Transferred power before overload?
Original permittivity is higher so you need less turns for a given inductance, but saturation comes earlier and in a less abrupt way so max level is less. As CJ says, iron is better at handling high levels and Ni is better at tiny levels.
When winding transformers, is it important to split winding area evenly ½/½ between Pri's and Sec's, i.e. use thicker wire for lower-inductance section?
That is a good practice, however it is often difficult, in particular when using sandwiched construction (the thicker wire may grind into the smaller one). And considering the overall performance, anything that's lost on undersizing the primary is more or less compensated by oversizing the secondary and vice-versa - up to a point, of course. More important is making sure the bobbin is fully used and the coupling optimized (well adjacent turns, primary and secondary windings of same width...)
 
Thanks. Beginning to make sense.

..but saturation comes earlier and in a less abrupt way..

Still not quite clear on what this core-size/saturation is related to? Voltage-per-turn? Transferred power?

What I mean is: If I have a core with e.g. 1000:10000 windings that will take 1V in before saturating (at given F) - wouldn't that same core take 10V at 100:1000 before saturating (at same F) because of less magnetization-per-volt?

I guess I'm looking for some perspective on voltage-capability vs. core size vs. number-of-turns.

I'm sorry for not being completely clear on my question!

Jakob E.

 
gyraf said:
..but saturation comes earlier and in a less abrupt way..

Still not quite clear on what this core-size/saturation is related to? Voltage-per-turn? Transferred power?
Both. The B-H curve defines the inductance; the higher the inductance, the less turns per volt. The typical Ni curve starts with higher permittivity than iron, but the curve inflexes gently as it saturates at an induction of about 0.5T. OTOH, iron would continue to about 1.5T and inflex very suddenly.
The consequence is that the Ni xfmr will allow less turns for a given impedance, but will not have the same handling than iron.
What I mean is: If I have a core with e.g. 1000:10000 windings that will take 1V in before saturating (at given F) - wouldn't that same core take 10V at 100:1000 before saturating (at same F) because of less magnetization-per-volt?
No, it's just the contrary. Let's say we use your starting point of 10V into a 100 turns winding, if you have a 1000 turns winding, you'll have 100V.
I guess I'm looking for some perspective on voltage-capability vs. core size vs. number-of-turns.
Voltage capability increases linearly with nr of turns. The larger the core section, the larger the voltage.
 
CJ said:
line level means higher voltages than mic level,

so you need 49% for ine and 80% for mic

80% takes about 5 kilo gauss

49% takes about 12 to 14

and regular microsil, aka grain orient aka silicon steel takes the most at about 16 to 18 kilo gauss

thin lams means less eddy currents which means better freq response

but they are expensive and only used on a few transformers
this is because it is harder to stamp out the soft nickel lams compared to regular steel lams, they bend all over the place and bending lams takes away the heat treating that orients the grain,


80 % has less core loss at high frequencies
and since the permeability is a lot higher you can use a smaller core
which means less capacitance and thus better frequency response,

Because that I ask me why I found 49% in vintage mic input, like T-1454. Maybe newer 10468 uses 80% with less turns ?
thanks!
 
abbey road d enfer said:
More Ni = better low-level performance but less headroom, and more cost

Also, it gives lower DCR, as for given inductance (low end performance) it requires less turns, which also can be made with thicker wire. At the same time, with the same winding technique it improves the top end performance, because of lower capacitance.

Likewise, with the same turns count the higher Ni will have better low-end performance because of higher permeability of the core and thus higher inductance.

Best, M
 
you can also look at transformer design from a turns per volt ratio,

instead of a volt turn ratio,

a certain core with cross section xy will need a certain amount of turns to handle x amount of volts,

so if you need 1000 turns to handle 1 volt, then for the same core, you need 10,000 turns to handle 10 volts.

if a lamination maker gives you a spec for gauss per volt on a square stack of core X, then you can convert this to turns per volt by picking where you want to run your core at as far as max gauss at a certain frequency,

say you have a super 80 core, super 80 only handles 5 k gauss max before saturation (and distortion),

pick 5 k gauss  at  60 hz , this will distort at 20 hz, but this is just an example,

now you can develop a math  relationship between turns and volts for that core.

lets say you are using some 75 EI lams,

mag metals says on page 14 of their catalog that your flux equation for 75 EI is

B max =100 kilo gauss per volt/Turns at 60 hertz.

so plug in 5 kilo gauss for B, and you have

5=100 E/Turns,  where E is rms volts,  so

Turns = 100 E/5        or

Turns = 20 E.

we want turns per volt, so we re arrange to equation to read

Turns/E = 20

so if E is 1 volt,

Turns/1volt  = 20

or

we need 20 turns per volt for 75 EI in 80 super..

what if we need to handle 10 volts?

well, from our 20 turns per volt constant that we developed,

we see that we need 200 turns for 10 volts.


so a Neve output wound on a nickel core at 60 hertz will need about 20 turns per volt on the primary side.

figure out the voltage going into the pri from the circuit and you can design both sides of the transformer.

every lamination will have a gauss per volt spec, which means that

every lam will have a turns per volt spec , square stack, fixed frequency.

if you cut your min freq in half, you need to double the stack height,


here is a good link on the history of transformer metal,

see that development in this field, as well as most scientific fields, was used for self defense purposes, (ie, war)

http://www.sowter.co.uk/pdf/GAVS.pdf










 
CJ said:
a certain core with cross section xy will need a certain amount of turns to handle x amount of volts,
so if you need 1000 turns to handle 1 volt, then for the same core, you need 10,000 turns to handle 10 volts.

Spot on - that did it for me! Thanks, CJ!

Jakob E.
 
I know I'm pushing it, but another - somewhat related - question:

What is the word on amorph and grain-oriented metal core materials..

And possibly on high-AL ferrites like T38?

data here: http://info.ee.surrey.ac.uk/Workshop/advice/coils/mu/

and here: http://www.epcos.com/web/generator/Web/Sections/ProductCatalog/Ferrites/Materials/PDF/PDF__T38,property=Data__en.pdf;/PDF_T38.pdf

- I just can't interpret the data to make real-life transformer sense..

Jakob E.
 
gyraf said:
What is the word on amorph and grain-oriented metal core materials..

And possibly on high-AL ferrites like T38?
You're opening a big can o' worms... I won't try to answer in an organized way, just floating some ideas.
To my knowledge, most magnetic laminations used in electronic construction are grain-oriented. Unless someone says different...
Amorphous is the contrary of grain-oriented, because it has no "grains". It has unique properties, in particular requires less energy to initially magnetize, so it can be used in fast-switching power circuits with reduced losses. It has the same max induction and permeability as iron, but the same HF capabilities as ferrite. I don't think it is particularly useful in audio xformers.
Ferrites are not particularly good for audio transformers, because their permeability is less than iron or Ni, even the high Al types, so a ferrite-core transformer will need more turns, hence more DCR and stray capacitance, leading to poor HF response and noise due to losses.
High Al ferrite cores have no gap, so rely on perfect surface finish for flux transmission. As you know, nothing is perfect, so Al is generally guaranteed within poor tolerances, such as -20/+30%, so you have to build a safety factor if you want to maintain LF response, so, again, more turns.
 
Thanks for all the interesting posts and links. I really enjoyed Dr G.A.V. Sowter's paper (CJ's post above). It is well worth a read. I have also recently started my own transformer winding journey. To date I have only played with traditional laminations, so don't have a lot to offer in terms of direct experience with glass or amorphous cores.

However, when I saw Abbey's thoughts on amorphous cores, I was reminded of a reply that I received when I asked Josephson about the differences between some of their very high end microphones (C720 and C715). I'm sure they won't mind if I share the some of their helpful reply here.

"The Circuits are very similar (adjustments made to fit each capsule) with the major exception being
the C720 had a metglass core transformer and the C715 has a nickel core transformer. The metglass
transformer is notable for having close to no "vibe" or "character of its own" as far as it can given it is still
a transformer. The Lundahl nickel core transformer has a "smooth but detailed" character or vibe."

I have not yet heard the C715, but the C720 with the "metglass" core is an outstanding microphone.

(PS, my background is physical chemistry, where we use 'glass' and 'amorphous' to describe non-crystalline materials. I would not be surprised if the metallurgists and electrical engineers have found different uses for these terms.)

Stewart
 
Amorphous cores have been around for a while, from the Lundahl site>

"Old News July 2004
Amorphous core OPTs.
The LL1620, LL1623, LL1627, LL9202 and LL1679 are now available with Amorphous C-cores. (Sorry, no data sheet is yet available. I expect most data to be unchanged, except max signal/power level)
If these transformers are received well, we plan to get cores also for the LL1660 size transformers.

LL1688
is a new BIG tube output transformer designed for 845 applications.
The transformer weights 4 kg and is the biggest transformer we can presently manufacture.

LL1689
is a new tube preamp output transformer based on the LL1660, but with turns ratio 9+9:1+1+1+1.

LL1922
is a high level line input stepup transformer (turns ration 1+1 : 4+4) similar to the UTC-LS10. "

these cores have to be bigger than the steel cores because there is a bigger air gap between the layers, so for the same volume, you have less core with the amorphous stuff.

but some audiophile people like the sound, probably because of the increased frequency response of the core which can mean less distortion and phase shift at the high end.

Kevin at K and K audio , back in N. Carolina can tell you about the differences, he has actually done listening tests on a lot of these fancy transformers,

here is their website with some cool projects,

http://www.kandkaudio.com/

if you go to this link and enter "amorphous" into the serch text box, you will get many threads having to do with amorphous cores.

http://www.audioasylum.com/forums/KandK/bbs.html

it looks like the power companies are starting to build big amorphous power transformers, >

"More than ever, electric utilities and industries today are searching for technologies that will reduce their operating costs and improve energy savings throughout their systems. New transmission and distribution (T&D) technologies are now available to help utilities meet these goals.

With a new generation of Metglas® amorphous metal distribution transformers (AMDTs) -- with up to 80% lower core loss than conventional transformers -- Metglas, Inc. is helping utilities worldwide to achieve their efficiency objectives. When you consider that 10% of all electricity generated by utilities is lost in the transmission & distribution process, the potential savings through reductions in core loss can be significant."

looks like tons of companies are starting to use amorphous, global warming being one of the motivating factors,

some more on low level stuff:
"Amorphous magnetic cores allow smaller, lighter and more energy efficient designs in many high frequency applications for Invertors, UPS, ASD (Adjustable speed drives), and Power supplies (SMPS). Amorphous metals are produced in using a rapid solidification technology where molten metal is cast into thin solid ribbons by cooling at a rate of one million°C/second. Amorphous magnetic metal has high permeability due to no crystalline magnetic anisotropy.

Amorphous magnetic cores have superior magnetic characteristics, such as lower core loss, when compared with conventional crystalline magnetic materials."

ok, now on the grain question,

not all cool audio transformers use grain oriented lams,

there are some old western electric models that use regular "barn roof" non grain oriented lams,

but most non grain lams are used for low cost power transformers or motors, etc

there are two main advantages to using non grain,

the first is that it is dirt cheap, about 1/2 to 1/3 the price of microsil.

and the second is that there is very little magnetic aging with the non grain stuff.

when you align the grains in a lamination, they eventually become disordered again,
this is due to heat and stress and maybe a few other things, so the property of the grain orient tends toward the non grain,
where as the non grain is already where it is going to be in thirty years, which is non grain.

so you might lose a little perm and thus bass with a grain orient that is really old.
 
Back
Top