Altec 9067b Passive Filter

[aldaraia]

Looking for a schematic for this puppy. Saw it on one of my favorite producer's gear lists and I've been in need of something like this in tracking and mixing situations. It seems pretty damn simple but I don't know anything about these things to really have an idea how it may work. Any input?

 

[ruffrecords]

All I can find on line are pics. These show that Altec thoughtfully pasted the outline schematic of the two passive filters on the bottom of each one. The hi cut is an LCL type and the low cut is a CLC type. Both are 600 ohm 'T' type filters. Thanks for bringing this product to my attention. I have a client who needs something very similar and, who knows, this might turn into yet another EQ project for me?

Cheers

Ian

 

[aldaraia]

Again, I do not know enough about electronics to successfully translate what you said. I can read and follow a schematic and I can source parts, but interpreting electronic concepts is a bit beyond my level. I can infer, though, that it appears the effect itself is controlled with rotary switches wired with the proper number and values of inductors and capacitors.

I have not found the photos you have seen of the filters with the schematics printed on, but I am very interested if you could track them down again and share.

 

[ruffrecords]

The pics were on photo bucket. There is a whole series of them here:

http://m764.photobucket.com/albums/lonepariah0/Altec%209067b/?src=www

Cheers

Ian

 

[emrr]

Same concept as the Cinema, Langevin, RCA, Pultec, UA, etc.  Slenderchap just promised us a version of the Pultec or UA as a new build kit, it goes lower and higher than the earlier types.  It is simple in concept, not in build.  Someone posted one of the variants here not too long ago, I don't remember which one or where it is. 

I own the Langevin and the Altec, honestly they aren't much use other than as HPF at the very lowest setting, which isn't low enough for most modern work IMO.  There's not a thing they do that I won't pick a software filter for first, other than usefulness while tracking, when appropriate, which is very seldom.  The first link contains another link of response plots.

http://www.groupdiy.com/index.php?topic=28452.msg343906#msg343906
http://www.groupdiy.com/index.php?topic=50848.0

 

[dandeurloo]

I think I have some of just the modules.  I have never racked them up.  Maybe I should look at them again.

 

[aldaraia]

Well, I'm not too big on "modern" production. I try and be creative, and the things I've heard from inductor based filters and EQs has definitely been creative, so I'm interested in attempting to make or acquire one of these. I assumed, perhaps for the worse, that it being merely high pass and low pass filters that it was relatively simple to make. From what I've seen of similar designs it looks like it operates along the same lines as a passive stepped attenuator: each steps increases the number of inductors and caps and this the filter level.

@danderuloo, if you have any modules you may be willing to part with, let me know and we can maybe work something out?

 

[emrr]

I think I have a spare set of Langevins in a box too.  Make me the right offer on my 3-banger and it could be yours.

As to reality, I'd suggest playing with the same points in software with 18 dB/oct filters and see how you feel about them. In my book there's a world of difference between LC EQs and filters. 

 

[aldaraia]

I think I have a spare set of Langevins in a box too.  Make me the right offer on my 3-banger and it could be yours.

As to reality, I'd suggest playing with the same points in software with 18 dB/oct filters and see how you feel about them. In my book there's a world of difference between LC EQs and filters.

Unfortunately, I do not at this moment have the knowledge to reasonably assume a decent price for those things, and thus making you an offer would be moot. I could lowball, offering the point that you don't use them and they're just sitting in the box, and it's good karma to offer unused things to people who will use them, but I also know this is the internet where people do not really know each other and those pleasantries are essentially worthless. In short, let me know what you want for them and, truth be told, what you're even offering because I haven't the faintest idea what "Langevins" are, and I will let you know if it is something I'd be able to swing.

Also, still looking for a schematic if this doesn't work out. While emrr has all but convinced me that this may be something totally useless to me, it still seems like an interesting project and I'd like to learn how it works and maybe try to build one just to see if I can and try and put it to use.

 

[emrr]

I can't find that schematic here, I thought it was the UA or Urei version, but it's gone.  Not seeing it in the tech docs section.  It was within the last year. 

As Ian noted, it's LCL and CLC type filters, pretty easy to calculate if you dig out the right formula.  K-derived filters come to mind, but I am quoting from memory, maybe wrong.    So two identical multi-tap inductors and a bunch of caps in one, twice as many caps and one multi-tap inductor in the other.  From scratch would be tough, maybe winding your own inductors being possibly the easy path.  Cap values not off the shelf either.  3 deck 11 position switch, better get a reliable one given all the parts you are hanging off of it. 

I'm not saying they are useless, I am saying from years of experience with them they don't get used much.  If software replicas feel good to you in terms of freq points, then it's a party.  In a modern one I'd want more steps lower in the HPF and more steps higher in the LPF. 

Langevin is a company.  Altec, Cinema, and Langevin all made a lot of virtually identical stuff, Art Davis is the connecting name across the brands. If you looked at my link you see two Langevin sets and one Altec set.   

Price history, last I saw, several years back, they were selling untested in the $100-200 range regularly as individual filters.  So raw parts pairs in the $200-400 range.  That was consistent value for 5 years or so.  More when on the factory panels.  I haven't watched for one recently.   

The other thing you can look at is the Krohn-Hite and Spencer-Kennedy active bandpass filter sets.  There are both tube and SS versions, and they are continually variable.  Watch out that some are subsonic, and range from almost DC to only maybe 1kHz, designed for vibration testing.  They sound cool, and are sweepable in real time.  Many are scratchy sounding at various points, since they use 4 gear linked carbon pots in each filter.  They cost a bazillion times more than the stepped passive types, and only showed up in labs. 

 

[ruffrecords]

It is not too hard to work to the values of inductance and capacitance needed. Attached is a screenshot form LTspice showing the two basic circuits driven and loaded by 600 ohms. The formulae for calculation of the L and C values are given above each circuit and the values for 1KHz are shown on the circuit. Z0 in this case is 600 ohms.

Cheers

Ian

 

[emrr]

You will note there's a clear double/half relationship with values relative to direct octave moves. 

 

[aldaraia]

Much obliged, gents. This has been very enlightening and I will be looking at this and attempting to work it out as best I can in the near future. I just caught up with the slenderchap thread about the Pultec "pod" and stated my interest in one of his filter PCBs so provided they become or remain available, I might jump on that instead of attempting to make one of these myself. Probably cheaper in the short run.

 

[rickc]

the design of this is classic passive, and it follows what is shown in the outline schematic on the individual filter cans. There's a single L and two Cs for the highpass unit, and two Ls and one C for the lowpass unit.

The inductors were torroidal, with taps that were switched along with the appropriate capacitors to change the frequency.
The design equations can be found in the Audio Cyclopedia, or in Motion Picture Sound Engineering ((c) 1938 by the forerunner of SMPTE). MPSE can be found online. I think the actual filter type is Constant-K Tee section.

If you *really* can't find them, then you need to ask, nicely, and I'll post them here. But really, you ought to find MPSE because its section on filter and equalizer design needs to be on your bookshelf.  Once I found this book, I was able to successfully reverse engineer the Altec 9013  filter modules, used for their Acousta-Voice process (1/3-octave cut-only bridged-tee filters). The Audio Cyclopedia had *most* of the information, but was missing the critical parts about going from a bandwidth value to a real design.

40-something years ago, I was in a surplus place near Anaheim and ran across several bins of surplus torroid inductors. I had already started my reverse engineering of the Altec filters, and recognized the inductors for what they were. In this bin, I found several of the tapped torroids used in the 9067. How do I know they're for the 9067? Because I measured the inductance and then plugged those values into the formulas for the filters and they match what's on the panel of a 9067 that I own.  It's unlikely that I'll use these ever (haven't used them in 40-years...) so I'll entertain offers to sell them. No, these aren't VF-14s, but I think they're worth at least $25 each. I don't know how many I have; I need to go find the box and count them. 

aldaria gets first call on what I have since he posted the orignal question.

I don't think (99.9%) I have the inductors for the highpass section.  Chris Preston at Vintage Windings has the capability of winding these inductors. It might be worth sending him one of the ones that I have in the interests of science.

--rick chinn

 

[aldaraia]

Well, I am at least 90% certain that the design and build of one of these things is out of my league if I'm doing it entirely on my own with only a schematic and some equations, so as nice as it would be to get my hands on some things to take a crack at it, I'm afraid the components would be wasted on me, so if you're looking to sell them, someone with a bit more experience and know-how would be the person to sell them to. For the time being I've filed this under the "would be nice but not practical" folder and I'll be keeping my eye on slenderchaps thread.

 

[ruffrecords]

The design equations can be found in the Audio Cyclopedia, or in Motion Picture Sound Engineering ((c) 1938 by the forerunner of SMPTE). MPSE can be found online. I think the actual filter type is Constant-K Tee section.

If you *really* can't find them, then you need to ask, nicely, and I'll post them here. But really, you ought to find MPSE because its section on filter and equalizer design needs to be on your bookshelf. 

+1

And if you cannot afford to buy it, you can find a pdf version on line.

Cheers

Ian

 

[rickc]

Reply to Aldaraia,

the circuit is simplicity in itself, totally passive, so no active components. It's toroids, capacitors, and switches. This would be a good first DIY project.

 

[aldaraia]

Reply to Aldaraia,

the circuit is simplicity in itself, totally passive, so no active components. It's toroids, capacitors, and switches. This would be a good first DIY project.

I'd like to attempt it and I plan on searching for the schematic a bit more but failing that I haven't the know how to figure out the equations, much less know what particular components to order, etc. When it comes to building gear I am more of a "give me a PCB and components and I'll work it out" kind of person, much less a "give me nothing and watch me flail around" person.

 

[rickc]

ok. a trip to my lab shows that I have 3 of these, and they're not quite what I remember. There may be a 4th toroid; I usually buy things in pairs.

This is a largeish torroid with 7 wires coming off of it. Here are the measured inductances (measured with a yellow Peak/Atlas component tester), values in mHy

blu                0                    96.17
grn              9.828              47.26
yel              25.11              24.34
org            35.88              15.91
red            44.88                9.626
brn            52.96                7.286
red/wht      96.17                0

Then, for the highpass the configuration is input -> series C, shunt L to ground, series C -> output
and the formulas are  C = 2*(1/(4 * pi * f * z))  and L = z/(4 * pi * f)

Then rearranging the L formula to pull F out, F = z /(4 * pi * L)


then using the L values that use red/wht as the common, we get 497, 1011, 1963, 3003, 4963, 6557
if you use the L values that use blu as the common, you get: 4861, 1902, 1331, 1064, 902, 497... I don't think this is the set to use

The last frequency isn't one on the 9067, and these are pretty close, and you can probably chalk the error up to the tolerance of my tester, and altec's winding tolerances. Since there are only 6 inductance values, and the calc'd frequencies are in the end of the 9067's range, it says that there needed to be another inductor. That is doubtful. I think that Altec would have just wound one coil, with 10 inductance values.

Bottom line here is that these inductors MAY be from a 9067, but probably not. I'm not sure if I want to take the cover off of mine to look at what Altec did. Maybe I'll decide I have time to do this.
It would probably be easiest to calculate the L values, and have Chris Preston at Vintage Windings wind them.

 

[rickc]

ok. I took mine apart, at least the highpass can, so I could see what was going on. It's going to be massively fiddly to put it back together. I'll have to disconnect the terminal board from all of the wiring.

There are TWO coils, one like the one that I described previously (I haven't confirmed that it's THAT part), I'm 99.9% sure it is, with another one, looking very similar, but with 5 wires coming out.

There is a custom switch, looks like 3-poles.

There is a black can, with a whole bunch of wires coming out, and this must hold all of the capacitors.

I haven't tried to draw a schematic, but the two inductors have red-white wires coming out of them, and these wires (2) go to terminals 2 and 4 of the terminal board. These terminals are jumpered together and the two red-white wires both connect to it. That says that rather than putting the inductors in series, Altec wound one to have the one set of values, and another to have the next set of values, and they did this so the two coils do not depend on each other.

Circuit wise, I think the input connects to one big bunch of capacitors, the other sides of those capacitors goes to a throw of the switch, and the opposite throw goes to another (duplicate) set of capacitors, with their right-hand connections all tied together and connected to the output. In the middle, are the wipers of all of the switches, and they're tied together, 3 of them. The first wiper goes to the left hand switch that selects one capacitor out of a big bunch of capacitors (one per frequency), and the 2nd wiper goes to the right hand switch that sends the signal to one capacitor (same value as on the left side) out of the 2nd big bunch of capacitors (one per frequency), and the 3rd wiper selects the proper inductance, which goes to ground/low-side. It is a classic circuit, classically simple, with no switch funny business going on.

Switch talk: we talk about switches in terms of poles and throws. A switch pole represents one electrical circuit. A switch throw represents a direction that you can send the electricity. A simple on/off switch, which just opens and closes one circuit, is the simplest switch. It has two terminals, and is called Single Pole, Single Throw. One circuit, that can go one direction (or off). abbreviated SPST.  The ground switch on a guitar amp connects a capacitor from chassis to one side or the other of the power line. The capacitor goes to the wiper (common terminal), and one throw goes to the high-side of the power line, and the other throw goes to the low-side of the power line (neutral). This is a single-pole, double-throw switch. 3 terminals. abbreviated SPDT. The bypass switch in a pedal is a double pole, double throw switch (DPDT). Input goes to one wiper, output goes to the other wiper. two throws are wired together, the other two go to input and output of the circuit.  In the 9067, there are 10 frequencies, and a 11th that is the OUT for that filter, so you would need a 3-pole, 12-throw, or 3P11T switch. The switch should be a SHORTING type switch (make before break), but those may be difficult to get. If you have to use a non-shorting switch, then it may not be totally silent if you operate it with signal going thru and speakers on. It won't go BANG, but there may be a soft click.

This should be easy to DIY, assuming that you can get the coils (i am working on that), and the capacitors. Altec was able to have custom capacitors made, and if you want one of these filters to work correctly, ALL of the values must match what you calculate. Ok, you might be able to round it to a decimal place, but that's about it. For the capacitors then, you'll have to build the values up from the stock values that you can get off the shelf. Of course, if you own a capacitor factory, then hey... the world is your oyster!