DIY Telefunken V76 Build attempt

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

Landins

Well-known member
Joined
Mar 3, 2013
Messages
156
Location
Umeå - Sweden
New to the Group I post som pics om my current ongoing V76 build attempt. This is by no means near the original spec but hopefully I end upp with a brutal preamp. Tubes: ef804 + e83f, 300:80kohm input, 40kohm:600 output, 3x150H chokes, 0,6H, tricky 4p12pos-switch yes.
 

Attachments

  • bild.JPG
    bild.JPG
    378.6 KB · Views: 696
Has anyone in this forum sucessfully diyed something that resembles the original? I have not yet been able to try my unit but it would be great to hear from someone who have embarked on this adventurous journey. Great forum btw!
 
Hi,

some members tried them, most of them were not real copies. Search a little, there are some good threads about V76s like this one about Pi filters: http://www.groupdiy.com/index.php?topic=18836.20
I'm a bit confused about choke values. Sowter only sells 500H for V76, while on schematics i see 250H, 400H and two other values in filter. Maybe it doesn't matter so much, except for the filter. I'm sure CJ knows this things well, i remember he used resistor instead of choke in his V72. Check his threads about V stuff.
Btw, are all your transformers Hammonds? Can you give model # for output? I don't see input transformer, it's probably encased in mu-metal can as it should be.
 
After some searching on the forum I found my question here quite obsolete. I excuse that. Interesting reads on the previous builds from CJ. Concerning chokes I went for the highest values availible  that I could find without having to pay a soccerteams salary. Or football as we call it in Sweden. And as you say, maybe the anode chokes can be replaced by plate resistors but I will try these out for smothing dc. If resistors will be equivalent in function I also will lose weight from this 10 kilo piece. In the filter section I will test caps to find best LC-conditions. Two options for input: beyer 1:30 or a vintage stancor 1:20. The later is not screened but will be. Output is a Triad 1682. I see this project as a skeleton that can be enhanced if the basic functionality proves good! More high end I/O-trx:s etc.
 
You may want to consider the placement of the power transformer - imho best not to have it inside the enclosure, rather than hanging off the back panel. And the xlr connectors are also quite close to it. I prefer to keep the transformer as far from anything as possible!

May result in more mains hum ending up in the audio. Especially with many-turns inductors nearby.
Maybe not.

Anyway - something to consider.

Good luck with your build.
 
Thanks for your comments! My main (!) concern has been the placing of iron but it has never crossed my mind that I could hang the mains trx off the back of the rack! Doh! It would also increase the space in this soon very overpopulated box. Its a mystery how the germans managed to squeeze so much in to the original casette without crosstalk. Again thanks for commensts. Having built mostly low voltage stuff like pedals, a modular synth, the mnats 1176 etc this is my first endeavour in the tube world.
 
Surely they can be inside and be fine - usually, it takes some clever steel plate as sheeting.
And nice steel chassis shell also helps.

Basically, comes down to proper engineering - something deutsch manufacturing has a habit of doing well.

And over doing even well-er :)


But no doubt that EM transfer happens from a power transformer at mains freq in particular, and harmonics, usually thru other  transferance means.  Then of course there is psu ripple and all that. I forget the V72 but I guess it has a good low-ripple psu filter.

Best defence is distance. Especially from low level circuits, such as nearby a 1:20 mic traffo! And from many-turns-inductors especially where they are not shielded.

I would try to find a placement for that transformer very well away from the input and then as far from the output transformer as well. Certainly the input traffo should be shielded. Preferably the output as  well.

And yes, sometimes hanging out back is the best solution. It does also depend on what other equipment you plan to run above and/or below your preamp. Having another big tube box with it's psu traffo hanging out back but unfortunately near to this one's sensitive areas can be just as much problem.

I try to keep my psu on one side consistently within a rack for that reason. You can't really tell the real-world noise performance until the unit is in  it's 'natural habitat'. You sometimes get weird interactions.

Sometimes I leave the final transformer mounting to the end in order to try some different locations. You could be amazed by the differences in noise performance you get by moving transformers around a bit.

On the other hand, I have a 'poorman 670' build with a big E-i core psu inside the enclosure full of unshielded signal traffos.
I had to place everything very carefully, testing as I go, but the unit ended up being one of my quietest, with with measured mains harmonics well below -95 dBu. (which is something like -105dBFS or so)

Anyway - carry on.

First tube box is always kind of special. I still use mine all the time - an La-2a with a fender preamp added. :)

Enjoy and stay hv safe :)
 
I belive I have to condut in a trial-and-test-manner. In my design a flip of the mains should be easy. Just mounting it on the outside and drill a hole fort the wiring with my new love: the stepdrill. If the 50 cycle (swedish fq) hum is present i will introduce shielding with mu-sheets to bring down the noise. And now i belive that I will go for the well-shielded beyer input.

Filterissues: My experience in LC-circuits is somewhat limited. The original v76 is renowned for the "huge bass" and such. Has the filters anything do do with this? As I see it the LC can be tuned to different peak resonances depending on the H and C-values/relation. Will the filter then act as a boost for the respective fq:s dialed in by the four way C-switch grunded via inductor to ground?

Tip: The e-tools pro app provide an LC-calculator where I can give the 150H for the inductor, name frequencies desired and desired C-values are calculated. Am I out in the deep. Will I end up with a selfocsillating Buchla  :) ??
 
Germans used a lot of mu-metal to squeeze all this material inside cassette. You could do the same, but mu-metal sheets are very expensive and shouldn't be cut or bent. Alexc's solution with power tx outside of case should be sufficient without mu-metal. I found out toroids emit less EMI but of course don't look as vintage.
Beyer inputs are known for distorting at very small signals. You might experiment here as you also have Stancor.
I'm pretty sure filters here don't have anything to do with "huge bass" you mention, they just cut at certain frequencies. See response chart and some good info here: http://gwagner.50webs.com/v76.htm
At least one tube preamp i built have really huge bass and doesn't have any filters. When i use output tx with smaller core on same preamp bass gets a little smaller. I never heard bass like this with SS preamps....
Tubes are very forgiving and i'm sure you can build something close to V76 with material you have. I wouldn't care about filters because you need 20-20kHz response. People seem to take them away which makes building easier too.
 
I meant psu HV  rail filters in my posting. Smoothing choke+R+C+R+C type of thing after the rectifier.

Called a psu HV pi-filter.

Not sure how sensitive the v76 is to psu ripple - some circuits more, some less.


Generally tube boxes have such a filter or in more elaborate cases a HV voltage regulator.
 
My setup is -rectifier squre, 33uf, R/B+outtake/33uf, 33uf, R/B+ outtake, 2X150 H plate chokes. Hopefulle this will give a smooth anode VDC.

Concerning filters I intend to keep the four-way filter and 3khz-switch but remove the filter between the primary windings on the input trafo. In the fq-chart from gwagner i cannot see wich filter is bypassed. I belive ints the input one. Althou schematics of designs with no filters at all have been around on the itnernt. The turretboard design will give advantages in bypassing and testing out. The switch itself also have the Gerade/linear position.

In the gwagner fq-chart i cannot see wich filter is bypassed. Is it totaly bypassed? Or is is it the input filter that is bypassed?
 
I don't see psu as a problem. Original should work well and could be improved with quite simple regulators we use for other preamps (HT, heaters and phantom on one pcb). So after psu like this we have almost no ripple. Anode chokes add some gain, etc., i don't think they are there for killing ripple.
I found a thread where someone wrote a scheme without filters, it looks good and pretty easy to diy:
http://www.groupdiy.com/index.php?topic=43515.20
With turrets it wouldn't be hard to try both versions and decide if it's worth leaving the filters. Abechap024 might reveal you more about it, i'm not sure if he built it at the end. Wagner only shows part of the picture, i can't say what he did with filters.
Those Hammond chokes don't look like they could be used instead of something like Sowter V76 choke, although they might "work". I was wondering why Sowter only has 500H choke for V76 when originals were 400H and 250H; at different current levels inductance changes and this is probably why they make only 500H model.
I hope we are getting closer to good copy of this preamp. I'm also looking into it because i already have transformers and only need chokes. If this schematic gets us close it could be nice new project that more people would build.
 

Attachments

  • V76m without filters.jpg
    V76m without filters.jpg
    790.7 KB · Views: 381
Thats the one! Ive been trying to find it!. Clean one-knob-build. I wonder if it possible to hook up a 4 pole gain pot and attach the hot/cold wires from the input to it to get the original co-variance of input impedance and first stage feedback.

It dawns on me that my inductance knowledge is to bad. I´ll go for these Hammonds anywas and see what happens. I know they have the plate resistace requiered according to the ef:s datasheet. If the burn up/fail/melt/saturate or whatever i replace them with resistors. I got hold of beatuiful low-noise resistors from a closed down radio-repairs shop here in Sweden! The store-man hand picked three sets and marked them individually with schem-numbers :)

IF everything turns out to my favour I can present a turretboard layout.

 

 
If you look at this schematic 4 pos. of the switch act as a pad ("same" like Jensen's pad we all use, it just has 4 positions) at the input, the other 8 pos. are changing feedback and consequently gain. So after 4 pos. it has nothing to do with input (except passing the signal of course), it changes feedback resistors as one pole switch.
If you are asking if four pole 12 pos. switch could do what original does the answer is yes. First 4 pos. use 3 poles, other 8 pos. use only 1 pole. It could also be done with simple 3 pole 4 pos. swtich and one potentiometer. This is what Abechap did although without the input pad.
I don't understand what you mean by attaching wires to get original co-variance, maybe you didn't really understood what switch does.
My knowledge about chokes as anode loads is very bad too and it will take some time before i get it well. Thanks to Abechap i'm starting to understand what each filter does. Would you please explain how you know required choke resistances for this two tubes if this is what you meant? I think you won't damage tubes (certaintly not chokes) if you use your Hammonds in the circuit. Maybe test each stage separately; add I meter and watch for excessive current from the tube. This way you won't damage anything even if it's very wrong.
This is such a great learning experience for me, i'm just starting one of RCA's preamp, but V76 with it's filters and "internet magic" is challenging me for a long time. When i'm done i will probably build 76 or 72. It is nice that people like you post their work and ask question because we are loosing so much with all this kits all the time. It's really annoying.
 

Attachments

  • V76 swtich positions.jpg
    V76 swtich positions.jpg
    180 KB · Views: 270
Yeah! My professional experience is so far away from this field of activity as you could get so its a lot to learn: thus questions on forum. But i like to learn. Thanks for the recap of the switch. My thought (that i see will not work) was to jack up a 4-pole non-step potentiometer to conduct in the same manner as the 12-way switch. But as you say: the two input poles are only "active" the first steps as a pad. If all poles had a linear function i could have been done. I have soldered up the switch, see below! I found cheap and sturdy 4p12pos switches from Vilnius, Latvia. One dollar each.  Old military stuff from Russia. Concerning dc-resistande of chokes i misread the datasheet for the ef:s. "Typical operation" section gives anode-resistance of 300k but the hammond choke has only 3,3k:s. I have to place a resistor i series to get a higher value. The sowter has a dcr of 14k.

Work is progressing and I have jacked up the heaters and the old tubes now glow. Luminus! The HV- supply is givig me a hedache. I jacked up the first to tubes, said a prayer, flicked the switch. And got a good reading on screen and anode voltages. But suddenly polarity changed in the circuit! Coffebreak. As i see it now i underdimesioned the rectifier bridge! The rms-rating is to close to limit so they cannot handle the peak voltages.

No retreat - no surrender
 

Attachments

  • image.jpg
    image.jpg
    152.8 KB · Views: 166
One think i didn't catch; at pos. 4 fourth pole switches on to nfb resistors, while pad resistors stay at the same position as at 3rd position. I hope you didn't fall into the same trap, Germans knew how to place them effectively :)
I can't think of another preamp that is as complex as this one. RCA OP6 comes close, but the problem there is not so much complexity of the circuit as are unknown values of feedback resistors at the input. Output stage is also choke loaded.
Radiotron Designer's Handbook is classic about tube designs, you might check third and fourth editions for choke loading:
http://www.tubebooks.org/technical_books_online.htm#Vacuum%20Tube%20theory%20&%20circuit%20design
 
Yup! It took me some time to figure the configuratikn out. One should not forget to bridge the remaining posts on 1st and second poles. I generally find these multipole switches confusing but i think i got it right. The total resistance of the plate feedback to input tube should be around 30 k. Todays business is the filterswitch. Phew.

Thanks for the link! Havent looked into it yet but it will certianly give som insight on plate load. What ive seen"regular"tube amps load the output tube via the output-trx  and use its inductive properties. Thats not the case in our cases. Certianly the germans have an argument for this that a humble diyer would understand :).

Complex yes. But the v76 is quite well covered. Good luck with your build!
 
The choke drops minimal voltage VS a resistor doing the same work. Then you can squeeze every last bit of voltage swing out of the output tube, rather than dropping it across a resistor. In CJ's last V72 choke adventures, he noted a 7 volt drop across the choke, that leaves a LOT of plate voltage to work with.
 
wtf? I never sought out this schematic before; there is a whole lot going on here...some comments:

the switched feedback arrangement and input biasing is similar to that used in  Knifs Therapy pre.

the way the input xfmr is tuned with an RC network between primary coils is interesting.

thanks for turning me on to this

 
Back
Top