CBS EQ"S

[wilebee]

Does anyone here know anything about the modular passive eq sections from the CBS recording studios from the 60's I have 2 eq's and I'm trying to figure out how to wire them up w/out a pinout, schematic or even knowing what impeadence in or out that they are Help Wil

Wilebee

 

[wilebee]

anybody? Please

 

[Mark Burnley]

Hi Wilebee,

Passive EQ's are usually pretty easy to figure out- no extra PSU connections to get in the way!

Have you got any pics of the internals, especially the wiring to the connector.

If it's unbalanced there are in effect only three terminals- Input (hot), Output (hot) and the common ground. The common ground is usually pretty obvious because of the common ground conductor through the unit- look out for a buswire or larger trace on a PCB.

If it's balanced, there are only 5 connections- Input (hot and cold) Output (hot and cold) and Common Ground/shield.

Once you've found input and output just connect it up, set EQ to flat, measure dB loss and insert suitable buffer/gain stage.

:thumb:

Mark

 

[wilebee]

It has 5 pins and I was told it needs 15db of gain make up I'll see if i can get some pics put up Wil

 

[wilebee]

I poked my head back into these eq's and heres what I've got. 5 pins on the connector 7 total wires to the connectors 2 are doubled up. there really is no recognizable ground these babies are all handwired turret lug boards the back section is comprised of the torroidal inductors the mid section is the capacitor board (I've never seen a 5uf vitaminQ before)and the front is the ceramic switches. thewireing doesn't seem to follow any rhyme or reasin I've seen. I have pics if some one could host them. The build on these are baeautiful and I bet they sound great Wil

Wilebee

 

[fum]

Here's a pic of the EQ's that Wil forwarded to me to post. Nice flowers!! :grin:

http://www.shinybox.com/diy/cbseq.jpg

ju

 

[wilebee]

thanks Jon

 

[emrr]

I racked a pair of these for someone recently, they are interesting.  Controls don't say anything and don't do what you would expect either.  The pair seen above are totally stock, match the ones here.  I've seen a few that were recapped by some misguided fool.  The pair here matched very well.  I went transformer in, and a 990 based output amp with transformer.  There has to be a small bit of isolation between the EQ and the amp input or the low band cuts will shut the amp down, I did a 2dB 'T' pad and it became stable.  Simple series resistance didn't do it.  I took all the measurements multiple ways and drew it out as much as I could, there's an entire row of terminals buried under the wiring loom and you'd seriously risk a dead unit trying to get at it to decode the paths.  I dread the modern cost quote to remake these as-is, no one would ever do this again.  Party on....

 

[Bo Deadly]

Man I love stuff that's all transformers, inductors, film caps and switches. Noise performance should be pretty good, the parts generally don't burn out, wear out, dry out, etc and there's no power to worry about.

 

[emrr]

The downside is every different way you load them leads to a different response and range, just slightly.  Best to optimize an environment and leave them there, as they were in the consoles. 

 

[Bo Deadly]

The downside is every different way you load them leads to a different response and range, just slightly.  Best to optimize an environment and leave them there, as they were in the consoles.

But most stuff driving and loading is within a known impedance range. Specifically, you have a good chance that the source impedance driving it is between 30 and 100 ohms and the load impedance is almost certainly 10K. Just about every vaguely modern piece of gear there is is very low impedance with a series resistor of at most 100 ohms and 10K inputs.

So whatever the characteristic impedance of the gear is, just add resistors to match. For example, if the characteristic impedance is 600 ohms, then add a 500 ohms in series with the input and add a 638 ohm load on the output. Then most stuff will just work.

Of course if you have a lot of passive stuff like this, then it will not work without running it through something that buffers. But if you have a LOT of passive stuff like this you could make a unit that is just buffers. Like a patchbay that is just buffered mults.

 

[ombudsman]

Of course if you have a lot of passive stuff like this, then it will not work without running it through something that buffers. But if you have a LOT of passive stuff like this you could make a unit that is just buffers. Like a patchbay that is just buffered mults.

This is mostly over my head, but I do have 600 ohm passive EQs including some of these CBS 4 bands.

What are the parameters that determine whether some buffering is needed ? What I do with an EQ like this is patch it into a 600/600 Altec fader and then from there into my Metric Halo LIO 8 where I can give it some makeup gain. I thought it was working OK.

I just got some API 325 cards that I was going to have set up for makeup gain at some point with resistors to set the gain on the cards specifically for what each EQ needs.

 

[Bo Deadly]

This is mostly over my head, but I do have 600 ohm passive EQs including some of these CBS 4 bands.

What are the parameters that determine whether some buffering is needed ? What I do with an EQ like this is patch it into a 600/600 Altec fader and then from there into my Metric Halo LIO 8 where I can give it some makeup gain. I thought it was working OK.

I just got some API 325 cards that I was going to have set up for makeup gain at some point with resistors to set the gain on the cards specifically for what each EQ needs.

I'm not familiar with the specific circuit of these EQs but gear like that usually has a "characteristic impedance" of 600 ohms. This means that the source impedance must be near 600 ohms and the load impedance needs to be near 600 ohms. The filter is then said to be "critically damped". If the filter is not critically damped, you will get either resonance like a wah pedal or loss at some frequencies. Although in practice it's usually ok for the source impedance to be quite a bit lower and for the load impedance you want to be a little over rather than under. So it sounds like driving it with the 600 ohm altec fader would satisfy the first condition. But you would need to load it with 600 ohms. If the I/O is a terminal strip, you can just strap a resistor across the output. But I would use more like 680 ohms which is a common resistor value that when paralleled with a typical 10K input (what is the input impedance of the Metric Halo and API 325 cards?) gives you a load that is 637 ohms which probably great. Ultimately you want to feed some white noise through it and look at the spectrum in your DAW if you have something like that. Or use whatever FFT software. Then play with different source / load resistors until you get a nice low-loss flat response. Also bear in mind that the typical output impedance of a modern output is usually between 20 and 100 ohms so you might need to take that into account. But if the source is the altec fader then you don't need to worry so much about that.

 

[ombudsman]

thanks squarewave. I think I wasn't clear, I meant that I go from the CBS eq (which I believe you are correct about being 600 ohms characteristic impedance) into the 600 ohm fader which I thought would be correctly loading the EQ, and then into the interface.

(As to the input going to the EQs I am mostly using UA 1108 or Cinema Engineering or Gates old tube preamps, then into a 600 ohm balanced fader or rotary attenuator in each case,  so I think I am OK there.) As you see , while tracking I am more or less trying to run a vintage style mostly 600 ohm rig (with almost no technical understanding on my part) until the signals hit the digital interface.

An old API document lists the input of a 325 as 12K ohms. The Metric Halo line inputs are 10K.

I can show your post to my tech. I was mostly wondering if putting a 600/600 fader or attenuator in between a vintage pre or EQ and the higher input impedance of the APIs or the MH inputs would suffice for correctly loading the outputs of those pres and EQs as I had assumed (I thought it was working OK with respect to the MH but I could be missing something), or if it is more complicated than that and requires specific testing and more of a fixed arrangement.

thank you for your comments.

 

[Bo Deadly]

Same difference. If you put the 600 ohm fader after the EQ, then that is the correct load for the EQ. If everything is 600 ohm, then it should all work without modification just well as it did in the 1950's.

But these days modern gear is between 20 and 100 ohms output and almost always 10K input. So if you are trying to connect a 100 ohm source to the 600 ohm EQ, then you technically need to add 500 ohms in series to makeup the difference. If you are trying to connect the 600 ohm EQ to a 10K load (or 12K of the 325), then you need to add a 600 ohm resistor in parallel to load the EQ properly. Although again, the 500 ohms in series can be a lot less in practice. Some people probably skip it altogether but that will probably cause a bit of a ripple in the frequency response. And the 600 ohm load should be more like 680 because it's in parallel with the 10k.

Loading a passive 600 ohm source with 600 ohms is usually very important. If you don't, it will sound nasal. So when you're running a passive 600 ohm source into a typical 10K input like that of the MH or 325, you MUST strap a 680 resistor across the output terminals. Then it will be "critically damped".

 

[emrr]

These, like many, are preloaded.  You can load it again and lose more level if you want.  I put a 6db pad before the input transformer and then the 2dB T and 990 output makeup amp isolates the output.  Those curves look the best of any I took.

 

[ombudsman]

Doug does this generally look like it tracks with what you were seeing on these ?

8.5dB insertion loss

    High Frequency boost is a shelf starting around 7.5kHz: +2.5dB (position 4) and +4.0dB (position 5) measured at 12.5kHz
    High Frequency cut is a shelf starting around 7.5kHz: -3dB (position 2) and a steep low pass filter (position 1)
    High Mid Frequency boost is a wide bell centered at 4kHz: +2dB (position 4) +4dB (position 5) +5.5dB (position 6)
    High Mid Frequency cut is a wide bell centered at 4kHz: -3dB (position 2) and -5.5dB (position 1)
    Low Mid Frequency boost is a shelf starting around 200Hz: +2.5 (position 5) and +5dB (position 6) measured at 100Hz
    Low Mid Frequency cut is a shelf starting around 200Hz: -1.5dB (position 3) -3dB (position 2) -6dB (position 1) measured at 100Hz
    Low Frequency boost is a shelf starting around 100Hz: +3dB (position 4) and +4.5 (position 5) measured at 50Hz
    Low Frequency cut is a shelf starting around 100Hz: -2dB (position 2) and a steep high pass (position 1) measured at 50Hz

 

[emrr]

8.5dB  sounds right into 10K, the standalone measurements I have had a 620R tacked on the output when I took it, and accounting for that difference, seems right.  My later measurements were all within my assemblage, tuned for unity. 

Those all sound reasonably close.  If you try various various loads while measuring max settings, you'll see some movement.  You'll also see some change in the absolute shapes of the low and high cuts, my final setup they looked the cleanest/smoothest/most well behaved. 

Doug does this generally look like it tracks with what you were seeing on these ?

8.5dB insertion loss

    High Frequency boost is a shelf starting around 7.5kHz: +2.5dB (position 4) and +4.0dB (position 5) measured at 12.5kHz
    High Frequency cut is a shelf starting around 7.5kHz: -3dB (position 2) and a steep low pass filter (position 1)
    High Mid Frequency boost is a wide bell centered at 4kHz: +2dB (position 4) +4dB (position 5) +5.5dB (position 6)
    High Mid Frequency cut is a wide bell centered at 4kHz: -3dB (position 2) and -5.5dB (position 1)
    Low Mid Frequency boost is a shelf starting around 200Hz: +2.5 (position 5) and +5dB (position 6) measured at 100Hz
    Low Mid Frequency cut is a shelf starting around 200Hz: -1.5dB (position 3) -3dB (position 2) -6dB (position 1) measured at 100Hz
    Low Frequency boost is a shelf starting around 100Hz: +3dB (position 4) and +4.5 (position 5) measured at 50Hz
    Low Frequency cut is a shelf starting around 100Hz: -2dB (position 2) and a steep high pass (position 1) measured at 50Hz