Telefunken V100 Tube mic pre frequency respons problem

[lars78]

Hi,

i have one V100 and one V101 they different age but fully recapped.
The V100 is like the V101 without the pickup (TA) input .

The V100 looks and listen like as if one high shelf. It starts at 1000Hz and @16k its up to 4db more.

Any suggestions where I can find the error?


specs:
http://starfishcoffee.de/v101_specs.jpg

pics V101
http://starfishcoffee.de/V101_01.JPG
http://starfishcoffee.de/V101_02.JPG
http://starfishcoffee.de/V101_03.JPG
http://starfishcoffee.de/V101_04.JPG

 

[mjrippe]

I would check the 5 mfd capacitor and 600k resistor on the first tube stage (top left corner of  schematic).

 

[lars78]

thanks a lot !

the frequency response is more equal after matching the 5pf caps. ( i´ve overlooked )

now the FR looks like this

guys, do you think it's possible getting the frequency response more equal... without changing all components (-:

both amps are fully recapped



Lars

 

[mjrippe]

Ah yes, 5 pf (not mfd) would make more sense!

As far as matching, try the easy stuff first - swap tubes.  Also check that voltages are the same throughout the circuit.  You might disconnect the TA input circuitry on the 101 if you are not using it.

 

[emrr]

Disconnect the 5 pfd and measure, I expect larger high boost.  Keep track of the one with the larger high frequency boost, it will probably need more than 5 pfd to match the other.  I think we are seeing tolerance differences in the transformers.  The wiggle in the red plot is a transformer resonance fingerprint, fairly normal. 

 

[lars78]

all tubes are new and matched pairs

the red plot is V101, and the one that sounds better and seems to be OK.

both caps are new 5.6pf

for testing the V100 ( blue plot ) i have added a 2.2pf  in parallel to the 5.6pf , and also replace the 5.6 with 6.6pf (3x 2.2pf parallel).
It makes no big difference....

how much voltage difference is insignificant?

Lars

 

[emrr]

Put a lot more, whatever amount makes the high end go away.  Then back it off until they match better.  Might be 20pfd, or more. 

Voltages might have to be 25-30% off to make any visible difference. 

 

[ruffrecords]

What load do you have at the output? Have you tried it with a 600 ohm load?

Cheers

Ian

 

[lars78]

the interface / measurement input is 1k7, adding a 1k2 parallel to the output transformer brings the load near 600 make no difference

"Z" measurement with NTI MR-PRO

adding a 2.2PF to the 4.7PF affects a low cut

blue plot V101 - 4.7pf - OK
purple plot - V100 4.7pf
green plot V100 - 7.0pf
yellow plot V100 - 13pf

 

[emrr]

I would be sure to redo all the plots at the same time.  I suspect your low difference is not the cap change, but possibly that it is a different testing time, and some unknown factor has changed.  I experience this frequently.  If that observation is correct, then the cap tailoring appears to be doing the right thing. 

 

[lars78]

now i put a DI Box ( Cinemag CM-DBX ) between lineout from interface / test signal and preamp input....

averaging 20 Measurements for all plots !

green plot - V100
yellow plot - V101
red plot - V100 with V101 input transformer 

i think the input transformer is the bad guy.... are you agree with me?


....any suggestion for a good pair of 1:20 input transformer ?

 

[emrr]

I'm probably confused, does one of these not have an input transformer?  Or did you just swap the input transformer to see?

I don't understand the large low frequency boost in the yellow and red plots.

The top may be tamed with a zobel, it's out of spec but doesn't look terrible compared to other vintage preamps. 

If you don't like it with simple modifications, I would sell them to a collector and build from scratch.  You destroy it's historical value by changing transformers. 

 

[lars78]

i swap the transformers for testing.
I would like to use them as a stereo set, so i played a little with different capacitors ( 5PF)  and resistor ( the 600K ), but theres no was to match them close enough.
if the high end is nearly ok, the low end is far away from that.
tomorrow i´ll test both amps with the same transformer, if that´s the way, maybe i´ll buy a pair of nice transformers and stock the old ones.

i´ll checked the input impedance
V100 Z=3.5 kohm  DC resistans  ≈ 40 Ohm
V101 Z = 5 kohm  DC resistans ≈ 80 ohm

by the way, the V100 was in very bad condition. the  wire pots for hum compensation were very oxidizes, so the wire was broken. I've tried to restore it with dive them in WD40 (OIL) over night and clean it with alcohol. After that i trie to solder the broke wire at the pots, but no way. All electrolyte caps are bad to. The previous owner had changed the lamp to LED, also he solder to long at the V100 input transformer, so the tranny was damaged (shorting primary). I could repair the primary, but the tranny wasn´t a beauty anymore. http://www.nrgrecording.de/nrg/viewtopic.php?t=2067&highlight=v100
I think, the V100 is not interesting for a collector.

 

[ruffrecords]

I think the very low frequency boost is due to the first stage 0.05uF coupling capacitor which is inside the NFB loop. With the 600K feedback resistor this turns over at 5Hz which is just about where there is a 3dB rise in response. This may have been deliberate to compensate for input transformer low frequency response. The transformer low frequency response depends a lot on the test generator source impedance and if this is very low then the low frequency response may well appear much better than it is with a 'normal' source like a microphone.

Cheers

Ian

 

[emrr]

But if I'm understanding correctly, red and yellow are with transformer, and green is without.  I'm still not certain I know what is going on with the lows in the previous plot having four traces. 

 

[ruffrecords]

But if I'm understanding correctly, red and yellow are with transformer, and green is without.  I'm still not certain I know what is going on with the lows in the previous plot having four traces.

Agreed. What we need is an accurate description of the test set up.

Cheers

Ian

 

[lars78]

a smart man once said about DIYGROUP.COM: the Lab where the knowledge is at

thank you guys for the angagemant to solve my problems !



test set up:

green plot - V100 - with the v100 input transformer
yellow plot - V101 - with the v101 input transformer
red plot - V100 - with V101 input transformer 


I think about whether it would be better to make the level and impedance matching with a resistor network to put the DI box

 

[emrr]

OK.  That is more clear. 

I would prefer to use a resistive U pad between the RME and the input transformer, one that matches the expected impedance with the shunt element, and gives a good series impedance to load the RME correctly while giving a satisfactory level drop (if needed).  A typical 20 dB microphone U pad would be sufficient, with the test signal from the RME measuring at least -30 dBFS, to avoid V100/V101 overload.