The Motown eq

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Orphan Audio has been preserving Langevin factory documents and NOS parts archives for several decades, (amongst other tenuous and pehaps ill-advised pursuits in the pro-audio industry) with Quad-Eight, Electrodyne, Sphere, Helios and others occupying equally massive space in the file drawers and servers for quite a bit longer.
Although I am sure other mentions have been made on GDIY, this is the first in depth discussion I have seen on these legendary passive graphic eqs, built in various forms by Cinema Engineering, Altec, Langevin and Electrodyne and of course potentially the most famous and impeccably fabricated "clones" ever built, the Motown rotary graphic eq.
This does not even include the many later (and equally impressive sounding) active reciprocal-inductor graphics from Quad-Eight, Electrodyne & Sphere.

I pulled out a few archived docs to help answer at least a few of the lingering questions posed in this thread.
(BTW: If I missed or mis-quoted something, please let me know).




Model numbers for well known 6 & 7 band Passive Graphics:

Cinema Engineering:
7080
As far as I know this is the only multi-band graphic that was supplied as a complete "plug& play" system with integral makeup gain amp and featured an easily remoted slider panel, since the actual audio electronics with seven modular canned filter nets and tube makeup gain amp was not a small object.

Altec:
7-band graphic: 9062A (aka 62A), 9072A (aka: 72A)
6-Band graphic: 9073A, (aka 73A), 9074A (aka:74A)
The only difference between odd and even numbered models is panel finish, Black Anodize or Altec Green.

Langevin:
EQ252: 7-band graphic, eq module only, intended for mounting on a console face.
EQ270: 7-band graphic mounted on a full width 19" panel.
The only difference between the two model numbers is the mounting arrangement.

FUN NOTES:
-Altec and Langevin 7-band graphics have cases and mechanical dimensions so identical, that their precision machined/engraved faceplates and panel mounting brackets can be freely swapped back & forth.
-Functional vs. Mechanical:
-Although the internal mechanical design of Altec graphics is superior in every way to Langevin (aside from the odd, no solder connections of band edge connectors to motherboard) they are electronically identical in every way, up to and including the size and winding of the toroidal inductors.




EQ Curves:
I attached a scan of the original Langevin 252 response curves and an Altec 9062A Ap sweep from final test for one of many restoration/rack jobs. The Ap sweep is low res (only 30 frequency steps) and typically used only to confirm performance before shipping, so you will have to "draw in" the actual smoothed curves in your head.



Topology / Architecture:
Whoops (or anyone!) do you have an idea about the topology or configuration of each filter here? how they are supposely conected (series, parallel...)?
The drawing near the start of the post showing dual paths may have been a Motown improvement, however Altec and Langevin graphics were built with all filter sections in single series pathed configuration.
Motown & Barry Gordy were known for relentless excellence, hiring (& mentoring) some of the most talented and innovative technicians and engineers on the planet, so improvement on prior designs/art seemed to drive that entire organization.

Although the schematic "looks" quite simple, each band employs a marvelously clever seven-pole switch array to allow a minimum parts count, swapping & reconfiguring the positions of one parallel and one series LCR net and associated precision-stepped resistor strings for boost, cut and a separate band bypass (loss net) in each band cassette.
That said, each 7-band eq requires no less than fourteen sets of large precision inductors and capacitors (two L,s, two C,s per band) and equally precise resistor strings.
I have seen/heard attempts at passive and active graphics using small inexpensive switching supply quality inductors (yes, sometimes size does matter), but the bh curves of their cores are very steep & optimized for square waves, so they tend to saturate quickly and harshly which makes their usable headroom extremely limited and sonics far less than pleasant. There are no cheap shortcuts to a good sounding passive eq.

I attached some theory pages from one of my favorite reference books, (Passive network Design I think) that cover both Equalize and Attenuate configurations of the BP filter elements in each band.




LCR values: (see attached parts spec sheets).
I don't have any knowledge or skills about tube, so I think it will be a transistor circuit.
I haven't picked up the circuit yet, but I'll upload a photo that isn't on the blog.
The lowest and highest frequencies.
Precise inductors and capacitors are required to make a copy.
Even though the R-PCA BOM calls out 5% resistors, careful examination reveals the carbon comp resistors were either hand selected (Altec) or hand-ground to value and sealed with clear varnish (Langevin).
To achieve precise and matching frequency centers for both Equalize & Attenuate, I have found that all parts values actually measure much closer to 1-2% accuracy.
Considering the low values in series/parallel resistor strings in each band, even a small variation of a few ohms would cause significant gain and frequency errors.
I have had the opportunity to test this in the real world while replacing damaged resistors on eq,s that arrived with mis-handled frequency modules and R-string errors do have a very measurable effect.



Gain makeup:
I typically use vintage Langevin AM4701 Class-A transistor amp cards for gain makeup and to try to keep the signal path as close to vintage as possible. With a simple variable gain mod these cards can provide up to 36db of incredibly clean gain and can drive up to 1watt into 600 ohms.
6/7-band passive graphics typically only need 16db of makeup (if they are wired within a 600ohm environment), but I prefer to place a bridging input xfrm at the front to isolate from modern low impedance sources so another few db is required to compensate. An output xfrm is also desirable for full isolation and proper balancing, since a raw eq is unbalanced in/out. Configured as mentioned they can be patched-in with no risk that any type/termination of source or destination gear can affect or load their passive eq nets in any way.
Since these eqs are capable of tremendous level handling, any associated I/O transformers should be selected to match.
I suggest ones that can handle up to +28dbm and offer the smallest or cleanest sonic footprint possible.
Just an opinion,.... but any transformer or amp "tone" permanently added to these eq,s signal path would be an interference, cloud and/or unnecessarily color what is already an amazing sonic experience and will frankly, just detract from what they are all about.




NOS parts in stock:
Yes I have inventory of NOS Langevin L & C parts to build these beasties, but stock is limited and I do not have every value of inductor.
Also available are panel mounting ears, a few remaining vintage engraved front panels, Nylon slider guides, motherboards, and for the truly obsessed few, complete manufacturing drawings if you want to go down that road. (a few pictures attached)

I threw in a few more pictures to illustrate the "be careful what you wish for" effect. One shows a typical "staging" and cleaning bench with one of many batches of passive graphics that come through the shop each year. This is just a collection of 7-bands, but just as many 2-band (Altec 9061, Electrodyne & Langevin 251) come across the benches in various conditions.
This leads to the pic with Langevin eq band modules disassembled to show the progression of restoration for several 252,s that came in completely scrubbed free of all Gold & Nickel on every contact surface by someone with an ink eraser or perhaps just some 4-0000 steel wool (ouch!).
This required three hours (per band!!) of buffing/polishing scratched & oxidized bare copper with jewelers rouge, hand re-plating all contact surfaces with two layers of hard Nickel, one heavy layer of 18k Gold, then re-assembly, alignment and re-timing of each bands slide contact arrays.
Beware of anyone who says they can "restore" the contacts on these beauties by applying aggressive cleaning techniques. If the Gold over Nickel plating has been compromised by wear or abuse, the only certain solution for noise and level-jump free operation at this point is to polish and re-plate all contact surfaces, and if possible, leave them completely dry of all greases or lubricating contact treatments. Grease or any lubricant no matter how little can eventually collect and trap airborne dust and grit, which over many years, morphs into a very efficient abrasive compound, accelerating wear of the conductive precious metal contact surfaces, exposing bare copper (which reacts badly with many vintage petroleum based greases) creating oxides and noisy contacts.

The last is just one of the many active rack jobs we have built, this one with all the bells/whistles (Hard bypass, Front panel +/-4db gain trim, Unity gain adjust).

Please take good care of these pieces of history if you are lucky enough to acquire them, since once they are gone there is no replacing them.
 

Attachments

  • EQ252-270_ResponseCurves.pdf
    1,021 KB
  • Altec_9062A_ApSweeps.jpg
    Altec_9062A_ApSweeps.jpg
    3 MB
  • EQ252-270PeakDipTheory.pdf
    285.9 KB
  • Langevin_EQ252_LC-PCA.pdf
    173.3 KB
  • Langevin_EQ252_R-PCA.pdf
    193.9 KB
  • EQ252capacitors.JPG
    EQ252capacitors.JPG
    5.8 MB
  • EQ252inductors.JPG
    EQ252inductors.JPG
    5.4 MB
  • EQ252pcb_hdw.JPG
    EQ252pcb_hdw.JPG
    5.3 MB
  • Langevin_252_ContactRestoreSequence.jpg
    Langevin_252_ContactRestoreSequence.jpg
    3 MB
  • BeCarefulWhatYouWishFor-PassiveEQmageddon.jpg
    BeCarefulWhatYouWishFor-PassiveEQmageddon.jpg
    5.7 MB
  • Altec9062aRackedViews.jpg
    Altec9062aRackedViews.jpg
    1.2 MB
Last edited:

Latest posts

Back
Top