Altec M11 mic system impedance/termination (21B with "coke bottle" 150 base)

[soapfoot]

We have an Altec M11 mic system (often known as the "Coke Bottle"-- the 21B omni capsule with the 150 base and P518a power supply unit)

This is a really cool microphone, but when connecting it to a channel strip on our Neve console (31102 input module) we notice a peaky resonance in the high end that doesn't seem normal.

Looking through original documentation, I found this note:

The output of the M11 System is designed to connect directly into either a resistive load equal to the nominal output impedance or an input transformer whose secondary side is terminated in a resistance of proper value. When the microphone system is connected to an input transformer whose secondary is not terminated in a resistor, it is necessary to either place a resistor of proper value across the input transformer secondary or to terminate the microphone system output as shown in Figure B.

Failure to provide proper termination will usually cause peaks in the frequency response at both high and low frequencies. The degree of this distortion is deter- mined by the characteristics of the input transformer.

I checked, and our Altec's output transformer is strapped for 250 ohm impedance (presumably as it was when it left the factory).

Our 31102s have switches on the back for "Hi" (1200Ω ) and "Lo" (300 Ω) input impedance. They are all set for "Hi," and I presume this is the cause of our issue, as (per the note above) this mic wants a matched, and not bridged, impedance (correct?)

All of our other mics perform better with the 31102s in "Hi," and you have to remove the input module to access the impedance switch (not practical to do on a channel-by-channel basis).

Is there a simple modification I can make in the Altec's P518a PSU itself to terminate it properly?

Could it be as simple as placing a 250 ohm resistor across pin 2 and 3 of the output XLR connector?

Thanks in advance for any insight!

 

[emrr]

Yes, it wants a matched load.  Put 300Ω on it permanently at the PSU.  At least the Altec 428 preamps were matching inputs, at odds with most US theory. 

 

[gridcurrent]

frequency response on 2 Altec 518a microphone system power supplies

test conditions
600 ohm source Z
100 k analyzer load



first unit
(with chassis mounted XLR connector)

+10 dBu into T1 pin 7
no secondary termination
output in phase on XLR pin 3

20 Hz         +4.2 dB
100 Hz +0.4 dB
1000 Hz         0 dB (-24 dBu)
10 kHz 0 dB
20 kHz +0.1 dB
30 kHz +0.2 dB



second unit
(with captive cable XLR connector)

+10 dBu into T1 pin 7
no secondary termination
output in phase on XLR pin 3

20 Hz         -6.1 dB
100 Hz -2.2 dB
1000 Hz         0 dB (-23.2 dBu)
10 kHz 0 dB
20 kHz +0.1 dB
30 kHz +0.2 dB

 

[emrr]

frequency response on 2 Altec 518a microphone system power supplies

Huge difference in the 20Hz unless that’s a typo. 

 

[soapfoot]

Thanks to you both!

gridcurrent--

Very interesting measurements. Do you take them to mean that the captive XLR cable performs poorly and should be replaced with an XLR connector on the power supply and a modern XLR cable?

 

[gridcurrent]

Very interesting measurements. Do you take them to mean that the captive XLR cable performs poorly and should be replaced with an XLR connector on the power supply and a modern XLR cable?

no.
the measurements have nothing to do with the cable.
just a modification done by a well meaning technician decades ago.

 

[soapfoot]

Thank you!

So your tests, then, illustrate variance in individual examples? Or something else?

 

[mjrippe]

Ignoring the cable, you are testing just the transformer.  I would think the one with poor low end response is either magnetized or damaged.

 

[emrr]

Though the +4 doesn't make a lot of sense either. 

 

[mjrippe]

Though the +4 doesn't make a lot of sense either.

True.  Would like to see tests with termination.

 

[PRR]

Neither does the "+10dBu" test level.

 

[mjrippe]

The transformer is a big stepdown, so perhaps that what they needed to get 0dBu out at 1k.  Not really the best measurement though as the TBB 103 was designed as an input transformer for low level signals to tube grids.  Altec used it "backwards" in this design.

 

[soapfoot]

Okay, so I terminated it in the XLR connector itself, using two resistors as illustrated in Figure B in the document below.

I’ve lost a LOT of output level and the mic is very dark now. Does that seem right?

Or should I have done something else?

Thanks again for all of your help.

 

[emrr]

I've no idea where that comes from, but no you don't want to do that.  1 resistor.  300ish ohms should do it. 

Matching load may lose 6dB.  Comes with the territory, if you need it. 

 

[soapfoot]

Ah cool. Thanks. Simply between pins 2 and 3 xlr?

Thanks again!

 

[soapfoot]

Okay, I used just a single resistor across pins 2 and 3 XLR.

All is right in the world now! It sounds really cool.

Quite dark in comparison to a KM84 or a Schoeps CMC6. But a very haunting, familiar, and silky-yet-aggressive midrange. Slightly "hard" presence quality, not much "air" on the super top end... but a very characterful midrange and gorgeous low end.

Thanks again for the help!!

 

[mjrippe]

Glad you sorted it.  Yes, they are famous for their low end.  Used on a drum kit the slightly rolled off top coupled with the substantial bottom can be very pleasing indeed.

I would still like to hear from gridcurrent about some measurements with termination and lower input levels.

 

[soapfoot]

I ran some numbers with our most-used preamps in-house, and ultimately chose to use a 330Ω resistor instead of the recommended 300Ω.

Bogey value is 250Ω Z(in) total.

The 330Ω in parallel with the following preamps generates these numbers (if my calculations are correct):

Neve 31102 in "hi" (1200Ω)	258.8Ω (nearly perfect)
API 512c (1500Ω)	270.5Ω (close enough)
RCA BA-31b (600Ω)	212.19Ω (close enough)
In-house-built REDD.47 (600Ω)	212.19Ω (close enough)

The 31102 is what's in the consoles in both of our tracking rooms, so that's the one that was my first priority. But all of the above-listed preamps have been brought much closer to the target 250Ω value by adding the 330Ω in parallel.

Listening tests confirm that the mic performs better on all of the above. If all of our preamps had been 1k5 input impedance (like the API), 300Ω obviously would've been the ideal choice. But under the circumstances, 330Ω seemed like the best compromise.

Thanks again to all!

 

[emrr]

None of those preamps have those impedances, they are all bridging. 

The matching resistor can be many things above 250, pick whatever tames the resonance sufficiently to the ear. Could be 500.  Or anywhere in between

 

[soapfoot]

None of those preamps have those impedances, they are all bridging. 

The matching resistor can be many things above 250, pick whatever tames the resonance sufficiently to the ear. Could be 500.  Or anywhere in between

Interesting!

Those are the values listed under "input impedance" in factory documentation. I suppose this is a classic case of "knowing just enough to be dangerous," because I'm clearly missing something.

At any rate, it sounds good now!

Cheers