AEA active mics aren't playing well with API 212Ls. SOLVED

[sircletus]

So, I've had multiple AEA active ribbon mics crap out - very noisy, crackly - over the past few years.  But there's a twist: they've ONLY died when using them with API 212Ls: the ones built in to our console AND the ones in a 200 rack in my home studio.

I'm wondering if anyone else has had issues with this particular combination.  AEA have been great about fixing these at no cost, but I have a hard time believing the crapping-out of three active ribbons  - ONLY when used with these particular mic pres - is a coincidence.

Would love to get to the bottom of this.  Is it related to the L200 power supply design?  Anyone out there have a similar experience?

 

[abbey road d enfer]

So, I've had multiple AEA active ribbon mics crap out - very noisy, crackly - over the past few years.  But there's a twist: they've ONLY died when using them with API 212Ls: the ones built in to our console AND the ones in a 200 rack in my home studio.

I'm wondering if anyone else has had issues with this particular combination.  AEA have been great about fixing these at no cost, but I have a hard time believing the crapping-out of three active ribbons  - ONLY when used with these particular mic pres - is a coincidence.

Would love to get to the bottom of this.  Is it related to the L200 power supply design?  Anyone out there have a similar experience?

API consoles have restricted phantom power, due to the 10k resistor in series with the 48V, before the two 6.8k resistors. This is a mistake!, which, for some reason has been perpetuated. It should be a 100 ohms.
It doesn't explain why your mics get damaged, though.
Did AEA give you a fault report?

 

[Khron]

Even if the phantom power ended up being too low, at the mic... How in the hell COULD that cause any damage, really? 

Maybe unless somehow the signal out from the ribbon transformer ended up overly-reverse-biasing whatever input transistor the circuit might have, but... Even that's quite a stretch  ???

 

[sircletus]

Even if the phantom power ended up being too low, at the mic... How in the hell COULD that cause any damage, really? 

Maybe unless somehow the signal out from the ribbon transformer ended up overly-reverse-biasing whatever input transistor the circuit might have, but... Even that's quite a stretch  ???

It's strange, for sure.  I did notice something a bit strange both on the Vision's 212Ls and my 212s in the L200R rack.  For sharts and gargles I stuck a TT cable in the patchbay and measured 1.5V just...hanging out on the 212 inputs.  This can persist for several minutes after phantom power has been switched off.  I've even measured this upon starting the console, without ever having engaged phantom on any channel.  So I'm wondering if the process of pushing/pulling a TT cable into/out of jacks with DC on them is subjecting the AEA FET buffer to little spikes that are juuuuust enough to damage a FET.  But even then, I can't imagine a mere 1.5V being enough to do any significant damage.

AEA are going to send me a demo mic to "beat on" under controlled circumstances to see if the issue is consistently repeatable.  Should be interesting!

 

[abbey road d enfer]

I did notice something a bit strange both on the Vision's 212Ls and my 212s in the L200R rack.  For sharts and gargles I stuck a TT cable in the patchbay and measured 1.5V just...hanging out on the 212 inputs.  This can persist for several minutes after phantom power has been switched off. 

that is normal. There is a 10uF capacitor that discharges in a 10k resistor and an LED. The 1st part of the discharge is rather quick, since 10uF and 10k result in a time-constant of 0.1 second. It means that after about 1 second, the voltage drops down to 1.6V, at which time the LED ceases to conduct and the capacitor has no discharge path other than its own leakage, so the voltage stays nearly constant for a long time.

 

[sircletus]

that is normal. There is a 10uF capacitor that discharges in a 10k resistor and an LED. The 1st part of the discharge is rather quick, since 10uF and 10k result in a time-constant of 0.1 second. It means that after about 1 second, the voltage drops down to 1.6V, at which time the LED ceases to conduct and the capacitor has no discharge path other than its own leakage, so the voltage stays nearly constant for a long time.

Excellent point, makes perfect sense.  Thanks for the insight!

 

[TwentyTrees]

Yeah, that's actually really useful in helping me grok some similar behaviour in a couple of 500-series preamps. Thanks!

 

[sircletus]

Sure enough, lingering phantom is shorting the FETs when pulling TT patch cables.  So for those of you out there using AEA active ribbons and any API mic pres, disconnect your mics AT THE MIC (after killing phantom of course) before pulling TT/TRS patch cables!

 

[ubxf]

i've seen U87 die a sudden death due to phantom and tt/trs mingling. But i never thought residual voltage after phantom is turned off could be deadly. Thanks for the heads up

 

[Khron]

I'd be curious to see the actual circuitry in one of those mics. "Lingering phantom shorting the fets", while perhaps plausible, still leaves the "how" unanswered ???

 

[emrr]

Sure enough, lingering phantom is shorting the FETs when pulling TT patch cables.  So for those of you out there using AEA active ribbons and any API mic pres, disconnect your mics AT THE MIC (after killing phantom of course) before pulling TT/TRS patch cables!

http://www.thatcorp.com/datashts/AES127-000183.pdf

 

[abbey road d enfer]

http://www.thatcorp.com/datashts/AES127-000183.pdf

his describes how a transformerless preamp could see its input devices destroyed be its coupling capacitors discharging when the input is accidentally shorted, but it does not explain how the same cause could damage the mic's electronics.
I don't see how the classic transformer output circuit or the xfmr-less Schoeps-type with PNP outputs could be damaged this way.
That suggests AEA uses a different topology, which we'll probaly never be enlightened with...

 

[Khron]

I don't see how the classic transformer output circuit or the xfmr-less Schoeps-type with PNP outputs could be damaged this way.
That suggests AEA uses a different topology, which we'll probaly never be enlightened with...

Well, not until a brave soul might be adventurous and determined enough to take a peek under the petticoats ;D

In case the OP would mind offloading at least one of the dead ones (in case they haven't all been RMA'd yet), I'd be willing to be quite such a soul  :

 

[sircletus]

Well, not until a brave soul might be adventurous and determined enough to take a peek under the petticoats ;D

In case the OP would mind offloading at least one of the dead ones (in case they haven't all been RMA'd yet), I'd be willing to be quite such a soul  :

I've had some interesting conversations with AEA folks over the past week.  It's been enlightening. They're very good people and I don't want to air their dirty laundry in a public forum like this.  From what I understand, now that the discovery has been made, they are revising the circuit.  In the meantime, we'll all have to be extra careful with our active AEAs and TT patchbays, I suppose.

 

[abbey road d enfer]

In the meantime, we'll all have to be extra careful with our active AEAs and TT patchbays, I suppose.

Patching mics on TT is a bad idea from the start. Dangerous for both mics and preamps.

 

[Matt Nolan]

Patching mics on TT is a bad idea from the start. Dangerous for both mics and preamps.

Exactly. There's a reason that XLR connectors are used on mics

 

[pucho812]

Patching mics on TT is a bad idea from the start. Dangerous for both mics and preamps.

why I recomend all mic lines are on an XLR patchbay.  a little cumbersome, but it won't have you patching on a tt where one side shorts to ground on the way in and out