Telefunken V 72 Input Transformer

CJ · Nov 30, 2009

here are a couple of mic preamp input transformers, one newer than the other,

one has a popped secondary, that means the other one is ready to go also.

since we are done taking apart our back log of dissections, we are going to rebuild these guys.

okgb · Nov 30, 2009

No Cj , No .............
I mean Yes Cj , Yes

wtf, over

desol · Nov 30, 2009

omg. The V72 input. This transformer's been on my mind....like georgia.
This is awesome. oK, Cool CJ. ...and it's been a tear down fest lately?

CJ · Nov 30, 2009

we need to rebuild these 72's, so input-choke-output-. should be fun. the output is a lot easier.

CJ · Dec 1, 2009

here are the parts, 4 layers of mu metal with brass foil inserts, those silver pri. disk's are toxic mercury!

so we will have to break open a few thermo meters to wind this thing right.

the primaries are good for 14.5 Henries each, or 54 Henries in series, 2 K input Z, so, 200K sec Z.

88 ohms DCR vs about 4.4 K sec. had to multiply the 1.1 K disks, as Sec #4 is open.

the lams are close to 625EI, but are actually 48 mm wide, the tongue is 16 mm, this is a European lam.

but you could probably bobbin wind a 625EI core and only be a few mH short. 45 ea. in this sq. stack.

the lams have a burr, so you want to keep them exactly the same as you pulled them. lucky, no wax.

that black thing on the right is the foam pad for mechanical shock .

AndreasS · Dec 1, 2009

Nice work,

the Braunbuch documentation is on the IRT website:

http://www.irt.de/IRT/publikationen/braunbuch/V72_EUE.pdf

Regards Andreas

CJ · Dec 1, 2009

Awesome! the turns are 400 pri and 2,000 sec,

one of these v72 inputs has a pri that is wired different than the other, the terminals are swapped.

with those prints, i can figure out which one is right, so thanks on that!

there are a few tricks on this older one that are not on the print, i have some pics coming up...

the cardboard former is cut a bit longer than the overall coil length, then the ends are glued over>

here is the lam, as you can see, it is very close, just a hair bigger than the 625, It is called 48EI in Europe.

that speckeled "mother of bowling bowling ball" finish is a shot peened process with toxic chemicals.

the lams are allowed to expand and contact (magneto-restrictive force) by way of these Al clamps>

the mu metal housings are processed just like the lams if you notice the finish>

that rivet goes thru the core bracket and copper grounding foil, then the wires get soldered>

pri side>

so if you go by this print provided by Mr. Andreas, then maybe this xfmr is mis wired>

so either i'm on crack, or one of these guys is mis wired, the guy on the right looks right,

Michael Tibes · Dec 1, 2009

In the V72s (used in the beatles console) input transformers were wired somehow differently afaik, maybe that's a hint?

Michael

guavatone · Dec 1, 2009

Michael Tibes said:
In the V72s (used in the beatles console) input transformers were wired somehow differently afaik, maybe that's a hint?

Michael

I would think they are talking about how they are wired outside the xfmr, right?

Thanks for the picture shows CJ. Thoroughly enjoying em.

Biasrocks · Dec 1, 2009

I believe Oliver said that there are differences in the transformer itself.

Here's a post from Prosound on gain modding where he talked about the differences between the normal and "S" V72 versions.

even the V72 circuit looks a kind of simple, do not get fooled. It is a perfect harmony of all components, so you can not just change or add a part to get some more gain without throwing off the balance.

Please take a look at V72/77/78, all pres build around the same concept, with different max. gain and most of all sound....

The IRT V72 and 78 are identical chassis, same x-former with different hook up, so that means different impedance, response and sounds vary quite a lot...

The V72 input coils are hooked up in series, the output in parallel, +0/-0.5 dB 20/20 (the S version). AC/DC negative feedback, plus a positive feedback to tame the resonance, very stable at 16 to 10k Ohm load...

The so called EMI version the V72S, input is hooked up the same than the IRT but the coils are wound different and mechanical in phase, instead the out of phase of the IRT version. The circuit is nearly identical only two resistors are different and with a higher value plate inductor.

The V78 input coils are in parallel, the output in series, you gain 18 dB by trading off the stable frequency response to a more home stereo spec of +/- 3dB 20/20. Slight modified feedback control (adjustable), no positive feedback, very unstable (you mentioned sluggish) at different loads, (needs about 1.5k output load)

The V77 has a different input x-former (similar winding style, higher ratio and cross coupled) but again hooked up in series, the output in parallel, a slightly different feedback, but stable at 50 to 5K loads....

95% of all V72 units are IRT speced and have a frequency limiting x-former, both high and low shuts off or better roll off quite fast... due to the way the coils are hooked up.
The first step is to change the x-former to the "S" spec version in order to play with the gain...

Most (home) made gain mods, are changing the DC controlled feedback, or just cutting out the DC (inserting a cap) while trying to control the feedback with an potentiometer. This will pass signal, but has nothing to do with the very high quality sound the V units are known for. Mostly you end up with a high end boost of 6 to 8 dB an octave, reducing the S/N ratio by 20 to 30dB.

Answering your question is quite tricky, and even I got help from one of the members of the historic V72/75/76 design team, it was quite troubling coming up with something that works, does not change the sound and is easy to implement.

In order to keep the unit stable the DC feedback (negative and positive) needs to be intact, you can decouple the AC via a cap, and with two buffer resistor control the AC feedback.
Still this only gives you a max. of 50dB, but at the same time you have to compensate the frequency rise (+8dB an octave) if you go over 40dB...
Changing the x-former strapping is out of question otherwise you change the very stable frequency platform plus the in/output impedance.
Going further into detail is proprietary info, but putting it in an easy way, in order to change the ratio of the output x-former to 600 Ohm(you gain the missing 10 dB) you need to change the inner impedance of the 2nd stage tube....

Best regards,

Oliver Archut

Thank you, that is very interesting.
I am especially curious about thedifferences between the v72 and the v72s input transformer.

can a V72s transformer be placed in a stock v72?

Does the V76s differ in this respect? Isn't the input transformer the same on the s version as on the broadcast version?

The actual x-former is the same, it is the way the coils are hooked up, V72>v72S and V76>V76S is identical how this was done.
In order to change the set up, you have to take the x-former apart, place the coils in the right order and hook them up to S specs...

Do not try this at home...

Best regards,

Oliver Archut

Mark

CJ · Dec 1, 2009

thanks for that!

when he says "the v72 input coils are hooked up in series", he means the sec coils.

there are 4 terminals for the input pri, so you could run the pri in parallel for a 1:20 ratio.

you can see from the pri pics that one of the pri windings is reverse wound,

easy, just a matter of flipping the pre wound disks. you need a steel mandrel with steel sides to wind.

if you use a bobbin with pre fabbed chambers, the wire dams up and bends the sides, hydraulic jack like,

the shield leads from the 2 pri windings are tied to the finish of the sec winding, all 3 see ground.

the bass roll off is caused by that cap on the pri, not the transformer. that huge core has mega bass.

desol · Dec 3, 2009

Uh..this may be some extraterrestrial influence here....

Michael Tibes · Dec 4, 2009

My transformer technology english is very limited, but I'll give it a try for the remaining bits in 'street language':

Gap:0, lams stacked overlapping (so there is no gap)

Wickelart - Type of transformer (Disk type)

Zweite Seitenwand...: Apply / affix the second side panel after the coil disk is mounted

Wicklungsschirmung...: Coil screening: Primary disks are screened with 0,03mm foil.The coil screen must not form a closed ring. Connection of the screen with 0,18 isolated (enameled?) copper wire. The screening foil has to be overlapping at the cut-off point (somewhat contradicting, how can it be overlapping, but not form a ring?)

Verbindlich...: Binding (relevant) is the geometric size of the disk coil, not the wire strenght (I don't understand what they mean by that??)

Hope this helps filling some gaps

Michael

edanderson · Dec 4, 2009

the part about the electrostatic (foil) shield means that the ends have to overlap MECHANICALLY (so that all of the windings are covered) but must not form an ELECTRICAL closed circuit (which would act as a shorted turn, not something you want).

verbindlich in this context would be like "more important" -- indicating here that the finished dimensions of the pre-wound disc is more important than the exact wire size used; presumably there is a small range of wire sizes that could be used to get a disc of acceptable size with the required number of turns.

as was pointed out earlier, it is easy to convert one of these discs to a "reverse" winding by flipping it over. it seems that the difference between the S and non-S version is which discs are flipped. the braunbuch document shows that (as CJ noted) one of the two primary discs is reversed from the other (and the secondary windings). from oliver's description, the S version has no discs flipped, so that the primaries are mechanically in phase with each other.

ed

CJ · Dec 4, 2009

Awesome you guys! I will get those changes in there.

8)

hey, do they still use the "Eszett" nowadays? because that thing caused me some pain in the back,

for non German speakers, the capital looking "B" signifies two s', so it can throw you on a search.

there are a couple of good online German Technical Dictionaries that i ran into, i will link it,

you could have fun playing with the disks, i figure there are 6 factorial ways to stack them,

so 6 * 5 * 4 * 3 * 2 * 1 =720 different ways to wire that thing, maybe divide by half for reality,

scope out each combo and map the freq response , yikes, might take a while, ???

CJ · Dec 4, 2009

updated all the changes, everything is final except page 3, the funky wax thing.

oh, and the weird test on page 4, #3, leakage in ohms? ???

this other v72 input has a plastic bobbin with loose wire, i think, will find out shortly,

i heard of coils being wound on heated mandrels, that would explain the wax process, that is,

how they keep it viscous while the 2,000 turns are going on. i have heard of yoke coils like this,

there is even wire that you heat up after forming a coil and it forms itself solid with some type of wax.

ok, thanks again for the help,

bonus pic> 1 ea. - primary:

bonus pic #2> a start of a disk.>

CJ · Dec 5, 2009

photos of the TAB dissection, which is a bit different from the Siemens version.

coil disks are not the solid varnished type, they are wound and then tied off at a few locations,

then somebody has the wonderful task of assembling the nightmare disk onto the phenolic bobbin.

then they get spliced. ???

this coil was no different electrically from the Siemens version, same reversed Pri ,

i guess somebody either messed up, or the unit was supposed to be wired 180 out.

or the output transformer polarity is swapped also, we will find out after i take apart a choke.

ok then,

notice the lack of the foam shock absorber, that extra seperator is mentioned on the print, ahh! :-X

the 3 clear plates in the middle are thicker than the 2 on the ends, which are actually 2 thin plates ea.

showing primary clearance

showing secondary clearance

you got to keep em separated...

maxwall · Dec 5, 2009

6 coils all together , including two primaries ?

Do the other 4 coils make up the 2000 turn secondary , connected in series ?

I see two wrapped coils, are those the two 400 turn primaries, looks like they are wrapped in copper foil tape or cloth tape ?

I suppose the hot wax during the winding process binds the wires on those skinny coils and keeps them together during the final coil placement assembly process on the bobbin. Otherwise those wires would unravel all over the place during handling.

CJ · Dec 5, 2009

yes, two primaries, wrapped in the silver foil,

this TAB xfmr has no wax or anything, pretty amazing, i imagine they had a bin full of copper scraps,

the V76 has 6 secondaries, i believe. :

CJ · Mar 17, 2013

just a bump since we are doing all the V72 iron,

both primary windings are brought out to the terminal board so that this input trans can be wired for 500 ohms input or 2K input,

2K was used at Abbey Road studios?

there might also be 2 versions that differ by having one of the pri discs put on backwards which would represent a reverse wind,

we will investigate all this reverse winding stuff in a coil experiment coming up soon,

Telefunken V 72 Input Transformer

Help Support GroupDIY Audio Forum:

Well-known member

Well-known member

Well-known member

Well-known member

Well-known member

Well-known member

Well-known member

Well-known member

Well-known member

Well-known member

Well-known member

Well-known member

Well-known member

Well-known member

Well-known member

Well-known member

Attachments

Well-known member

Well-known member

Well-known member

Well-known member

Similar threads

Join the conversation!

Join today and get all the highlights of this community direct to your inbox. It's FREE!