[BC] Re:111C repeat coils

Sat Apr 7 12:40:25 CDT 2007

The Tellabs equalizers can be made to sound GREAT. They have some major design
deficiencies that can easily be overcome. First off the output stage has
current booster transistors before the output transformer. Trouble is, these
were not put within the feedback loop of the output op amp! So you have a
class AB output stage with zero negative feedback. Moving one resistor fixes
this, cutting distortion by a factor of 20. Even better, connect the 5534
opamp directly to the output transformer through a 50 ohm resistor and use
these output transistors for another purpose (below).

The equalizer section of these use a 741 opamp, one of the worst opamps you
can use for audio!  Replacing this with a 5534 (with a 22 pf compensation
capacitor between pins one and 8) makes a major difference.

Finally, these units use a virtual ground set at 1/2 the power supply voltage.
This is derived by another 741. Problem is, a 741
doesn't have the output current capability to do this properly. This is where
the build out transistors mentioned above come into service. Put these in the
741's feedback loop and use them to drive the ground. Now there's decent
current capability there.

These mods literally transform the sound of these equalizers. The bass gets
*tight* and the highs very *smooth*, Fortunately, the mods can be done fairly
easily-only one clad cut is required. The rest of it involves lifting the
leads of a few parts and extending them with jumper wires. Of course, the 741
opamp in the equalizer also has to be replaced.
I have no idea who originally designed these puppies, but they had no idea how
to design good sounding circuits.

-D

------ Original Message ------
Received: Sat, 07 Apr 2007 10:59:40 AM EDT
From: Mike McCarthy <Towers at mre.com>
To: "Broadcasters' Mailing List" <broadcast at radiolists.net>
Subject: Re: [BC] Re:111C repeat coils

The 4008 was never designed for audiophile applications.  Somewhere I have 
the schematic for that board and it's not brain surgery.  I think the total 
dynamic range on the card is maybe 90 dB if set to lowest gain.  When you 
max out the gain, you are lucky to have 70dB dynamic range (noise floor to 
clipping).  But since tariff for a 15Khz circuit is only 68dB, (22dBrn), 
that's all they were designed for.

I have in the past tried pre-equalizing some circuits at the sending end 
where there is no practical limitation other than the hardware before the 
111C and have been able to get away from needing so much/any EQ at the 
other end. Just a simple gain block--20dB +/- is enough.  In one case, I 
was able to one man EQ a pair of circuits by simply sending the audio 
directly through the FM composite input and using the modulation monitor at 
the studio to set the amount of equalization.  Of course, I modeled the 
chain before hand using the digital path (which the dry pairs backed-up).

I then passively summed the two at the TX end in order to check phase and 
level match by driving the send end with inverted audio.  That worked 
remarkably well for years until the station moved and dry pairs to the TX 
were no longer an option.

MM

At 11:59 PM 4/6/2007 -0400, Xmitters at aol.com wrote

>In a message dated 4/6/07 11:39:29 AM Central Daylight Time,
>broadcast-request at radiolists.net writes:
>
><< The telco 111-C "repeat coils" are available on the surplus market. 
>They are
>  probably the finest audio-frequency transformers --ever. They have four 
> shield
>ed coils and the core is a toriod made of mu-metal strip. Very impressive 
>for s
>omething designed in the 1920s. The companion shunt equalizers are also 
>availab
>le although I don't remember the number. They consist of a tapped inductor 
>and
>tapped series resistors. Basically, you shunt-out >>
>
>OK, here is a dumb question. If telco copper loops are still around and 111C
>and equalizers are not considered Old Fart Engineering (see  my previous
>post/coinage of term) then why are there not companies still making the 
>111C?
>As
>ROTL, et. al.  pointed out, Jenesn makes some nice transformers. It appears
>that the choice is either Jensen or surplus 111C. To me,  it is odd if 
>equalize
>d
>copper is still not antique technology today, that 111C type performance 
>could
>not be had without resorting to the surplus market. Seems like you could
>outfit a specialty plant in China with some knowledgeable technicians, and 
>you
>could turn out high quality affordable product. A new 111C made in the
states
>would likely cost $500.00 I would guess. Just wondering.
>
>The last I looked, the companion repeat coil card for the other end of a Tel
>Labs 4008 equalized circuit, was a bit anemic at the low end as well as 
>anemic
>in the head room department.
>
>I'm sure that there are others out there besides me, that have restrictions
>about buying used stuff. What I do is, if used product is the only 
>alternative,
>
>I pay for it myself, then donate it to the station. I can only do that if
the
>item costs but a few bucks.
>
>
>Jeff Glass
>Northern Illinois University
>
>Dell CPi-D266 Win98SE AOL 5.0
>**************************************
>  See
>what's free at http://www.aol.com.
>_______________________________________________
>
>The BROADCAST [BC] list is sponsored by SystemsStore On-Line Sales
>Cable-Connectors-Blocks-Racks-Wire Management-Test Gear-Tools and More!
>www.SystemsStore.com       Tel: 407-656-3719    Sales at SystemsStore.com

_______________________________________________

The BROADCAST [BC] list is sponsored by SystemsStore On-Line Sales
Cable-Connectors-Blocks-Racks-Wire Management-Test Gear-Tools and More! 
www.SystemsStore.com       Tel: 407-656-3719    Sales at SystemsStore.com