Blocking dc-bias

farss wrote on 9/29/2011, 4:54 PM

"I should be able to make up a short jumper cable with some capacitors inserted, but I don't know the values or how to wire them."

All you need are capacitors in series with the signal to block the DC.
The value depends on the circuit impedance. If it is real 600ohms then you'd need reasonably large values to avoid too much bass roll off. 10uF gives a corner frequency of 50Hz which would be acceptable in your application I'd say.
They should be nonpolarised caps.
As it's very unlikely the circuit is 600ohms, probably in reality 10K ohms, then 1uF 100V caps would be just fine. A lower voltage rating would be just as good but most non polarised caps are rated at 100V or more. Just be aware these are physically large so you may need to makeup your DC blocker is a small plastic box.

Bob.

paul_w wrote on 9/29/2011, 5:54 PM

Just to add, you can use a standard polarised capacitor, as long as the positive lead connects to the positive supply, in your case thats the microphone pin from the Nikon. The negative side goes to the preamp, which should be about zero volts when idle. Its fine to do this because the cap is always charged the correct way and never reversed. Cap value, about 10uf 16v. Bigger the capacitance, lower the frequency response.
Paul.

musicvid10 wrote on 9/29/2011, 8:02 PM

I'm sorry, we are talking about audio coupling, which is a rather precise science. Plain old mylar or electrolytics will not do unless you don't care about big signal losses, phase and harmonic distortion (all capacitors phase shift, the bad ones do it unevenly).

They must be big enough to pass 10Hz at the nominal net circuit impedance, which you need to know. They must be top-grade polystyrene, polypropylene, or teflon (the latter being more expensive). And they must be in a properly shielded enclosure with RF shunt capacitors to keep from picking up your local C/W station.

Even in a passive untuned circuit, you should really have an engineer design this or you may be in for lots of trial and error and unneeded parts.

That being said, can the dc bias be turned off at the camera? Is the Beachtek designed to pass the dc voltage through to the mini jacks?

ChristoC wrote on 9/29/2011, 8:11 PM

I'm surprised that this problem has arisen, as compatibility with numerous brands of camera is claimed, and that Beachtek must have encountered or at least thought of such a problem as they are suggesting interfacing with MIC inputs; surely the best solution is a fix by Beachtek themselves achieved by modification to the unit, not an adaptor.

BTW, pity the user who plugs a standard dynamic MIC into that camera's input!

musicvid10 wrote on 9/29/2011, 8:25 PM

"BTW, pity the user who plugs a standard dynamic MIC into that camera's input! "
A balanced dynamic mic (think SM58) will not be affected by dc phantom voltage because the bias is also balanced across the legs. Running dynamics and condensors on the same 48V bias is common practice.

Laurence wrote on 9/29/2011, 9:42 PM

Maybe my diagnosis is wrong. What I'm getting is a huge pop every time I go into record on my camera. I am not monitoring through the camera. I am surprised that anything that would happen at the output of the beachtek would even affect what I am hearing over my headphones.

Between playing in bands and doing sound reinforcement for many years, my ears ring constantly and this popping really hurts my ears.

ChristoC wrote on 9/29/2011, 10:47 PM

I stated "BTW, pity the user who plugs a standard dynamic MIC into that camera's input! "
Musicvid: >A balanced dynamic mic (think SM58) will not be affected by dc phantom voltage because the bias is also balanced across the legs. Running dynamics and condensors on the same 48V bias is common practice.

Yes that is true for Balanced Inputs; however this D5100 camera has an Unbalanced input with a DC bias, therefore the coil of a connected dynamic mic will experience a DC voltage across it; the voltage does not null out in this case.

Of course the presence of a DC bias may be conjecture; have you managed to measure it, Laurence? ...... should be easy enough to connect a mini-jack and measure with multimeter.

musicvid10 wrote on 9/29/2011, 11:12 PM

The Nikon specs say it is a stereo input (not a single mono input), so since the bias is referenced to the sleeve, it also references to the sleeve (only) on a TRS adapter, thus pin 1 on the XLR, so still no problem. The two legs of the mic would merely record to the two input channels with no DC differential. Of course the audio gain would be so low as to be almost unusable on a recording, so someone going to all that trouble would still be disappointed . . .

ChristoC wrote on 9/29/2011, 11:32 PM

Now you are talking gibberish.....

musicvid10 wrote on 9/29/2011, 11:40 PM

"Now you are talking gibberish..... "

If you can't follow the logic, that's OK. You took the discussion OT, so I'll leave it at that.
;?)

http://dl.dropbox.com/u/20519276/dynamic.png

musicvid10 wrote on 9/30/2011, 2:42 AM

Laurence,
Apologies for getting so far off topic, but in doing so I came up with an idea for you.

How about a stereo pair of passive direct boxes between the camera and the Beachtek? They are really small, inexpensive, and enabling the ground-lift switches should make the DC bias go bye-bye!

ChristoC wrote on 9/30/2011, 3:15 AM

Why would I plug a dynamic mic into a stereo input as per your diagram; it simply would not work; where did I suggest using a TRS adaptor?

farss wrote on 9/30/2011, 4:19 AM

"I am surprised that anything that would happen at the output of the beachtek would even affect what I am hearing over my headphones."

So am I.
Firstly I would doubt the camera is switching the DV voltage on the mic input on and off as the camera goes in/out of record.
Secondly I'd think Beachtek would be aware that there's a DC voltage on what their unit connects to and they have some basic capacitor to block that.
Thirdly we have one of those Beachtek units and we've not had any complaints about any "pops" but it's only been used with Canon 5D and 7D cameras that I know of.

Do you have anything connected to the RCA connectors on the unit that are labelled "Monitor In" ?

Do you have the AGC Disable feature On in the Beachtek?

Bob.

farss wrote on 9/30/2011, 4:35 AM

"I'm sorry, we are talking about audio coupling, which is a rather precise science"

Here we're not talking about 600 ohm impedance matched transmission lines.
All that is being dealt with is an input for electret microphones. Impedance and resistance is going to be so close to identical the difference can be ignored.

In reality a 10uF cheap electrolytic would do just fine, that's almost certianly all that's inside the camera itself, if that.

Plenty of information on the typical circuits involved here: http://www.epanorama.net/circuits/microphone_powering.html

Bob.

paul_w wrote on 9/30/2011, 4:47 AM

"I'm sorry, we are talking about audio coupling, which is a rather precise science"
Errm, musicvid, now i'm sorry because thats not 'always' the case. EMC shielding ? polystyrene caps? come on. :) We are talking about the most basic circuits possible next to a battery and a bulb. Its a simple DC block.
Some people try to make audio into a black art, a bit like radio. Yes of course you can dive into the frequency response maths - but in this case the data simply is not available and will change anyway depending on what you have plugged in.
Port impedance of both input and output ports - do you know that? does anyone?
But really, this is very basic electronics. Block the dc and maintain the low end response. 10uf 16v standard eletrolytic. That is a good starting point with experimentation. This is not rocket science - not in this case. EMC shielding, ok put it in a metal box and connect to shield. More than good enough for audio shielding.
Funny thing is, go look at the mic in socket on the nikon - im 99% certain it would be a simply electrolytic for audio decoupling and a resisitor for dc bias.

Paul.

paul_w wrote on 9/30/2011, 4:48 AM

ha, Bob, we posted at the same time :)

Paul.

Anthony J C wrote on 9/30/2011, 6:25 AM

What has Beachtek got to say about it???

Anthony

Laurence wrote on 9/30/2011, 6:38 AM

Beachtek hasn't answered my email yet. When I called, all the techs are at some sort of conference this week so I couldn't talk to anyone.

The levels match between the Beachtek and the D5100 is actually quite good. The D5100 has no metering, but it has four audio input settings: low, medium, high and auto. Auto is different than what I've seen in other AGC circuits in that it is like the medium setting gain wise, except that it limits anything that goes beyond that acceptable range. It does not bring up the low volumes, it just limits the high volumes. Quite nice actually and quite useable. With the camera input set to medium, the Beachtek input level LED is perfect. Flickering green is a nice healthy level and flickering red has a little bit of audio distortion, but not enough to give me that nasty clipped digital audio sound. I am happy with it aside from the pops at the start and stop of record.

As far as the DC bias voltage goes, aside from the pop in my ears over the Beachtek, I really like having it available. It means that I can plug in my stereo pair of electret clip on mics that I got from microphonemadness.com directly into the camera and get useable interview audio. A very nice thing to be able to do, but on my paid gigs I want to have a little more control and reaffirmation that I am getting good audio as I record.

As it is, it is quite useable. I just have a loud pop in my ears every time I go into record. I have been doing sound systems for years though and my ears ring all the time. My ears are so sensitive these days that the pop really bothers me.

PeterDuke wrote on 9/30/2011, 6:55 AM

"I have immediately run into a problem in that Nikon cameras (like Sony) have a dc-bias voltage to power electret condensors on their mic input."

Apart from the plop, what evidence do you have that your camera supplies DC out of its mic input?

I have tried to measure the DC voltage on the mic input of my camera (Nikon D7000) using a Fluke multimeter, and see no evidence of it. The manual is woefully inadequate in discussing mic input specifications.

Are you sure that the DXA-SLR is not providing the DC offset?

If you plug a passive stereo mic (or other device you know does not have DC) into your camera, do you still get the plop?

PeterDuke wrote on 9/30/2011, 7:21 AM

I have tried an Azden SMX-10 mic with no battery and it still records (as it does with a battery), so I guess that proves that it does have a DC output. However, I get no plop when I start recording.

paul_w wrote on 9/30/2011, 7:22 AM

This may help.
ME-1 Nikon D5100 external microphone:

ME-1 is powered through the camera, and also features a cable stop to keep the cable out of the way when using the camera, and reduce instances of noise from unintentional contact. Because it utilizes a standard 3.5mm stereo jack, the ME-1 is ideal for the D5100 and other Nikon HD movie capable D-SLR’s such as the D3s, D300s, D7000, as well as the COOLPIX P7000 and any other camera with a 3.5mm stereo input jack. ...

"ME-1 is powered through the camera" that being the point. I read that as if the power is coming from the camera. And i suspect its being switched high when you hit record. This voltage, i assume, is back feeding into the preamp output. And you hear the thump. Hence the dc-block cap needed. Well, there is guesswork going on here, thats true.

Paul.

Laurence wrote on 9/30/2011, 7:33 AM

Yes, you can also use the D5100 with other dc-bias powered mics. It is exactly the same bias voltage as is used in the Sony cameras, the Kodak PlayTouch and Zi8, the Zoom audio recorders, the Olympus audio recorders, etc.

> "ME-1 is powered through the camera" that being the point. I read that as if the power is coming from the camera. And i suspect its being switched high when you hit record. This voltage, i assume, is back feeding into the preamp output. And you hear the thump. Hence the dc-block cap needed. Well, there is guesswork going on here, thats true.

That was my train of thought as well.

Think of it. In a zoom recorder, when you press the record button, the headphone monitoring begins, the meters become active and the bias voltage becomes active at this point. Then you hit play and it actually starts recording. On a camera, both the bias voltage and the recording must start at the same time. Maybe there are dc-bias blocking capacitors and they are just popping as the power is applied.

farss wrote on 9/30/2011, 7:50 AM

If the DC voltage on the input is being switched high when the camera goes into record then it's very likely that any mic plugged into the camera is also going to cause a "pop".
There's another problem here. Even with the external capacitor to block the DC voltage you will still get the "pop". As the DC voltage is switched on that capacitor has to charge up i.e. for a short period of time the potential on both sides of the capacitor will be the same. As the capacitor charges up the potential difference decreases. What you get is a transient pulse which will be heard as a "pop".

None of this makes any sense to me for another reason. If there's no DC to power the mic except when the camera is in record then you would not be able to monitor your audio until you went into record. Not exactly a very useful "feature".

Bob.

paul_w wrote on 9/30/2011, 7:52 AM

"On a camera, both the bias voltage and the recording must start at the same time. Maybe there are dc-bias blocking capacitors and they are just popping as the power is applied."

Could be. And the camera recording is simply delayed a little after the supply voltage goes high. That would stop the thump on the recording. In effect, allowing the voltage to settle before recording to file. That sounds correct.
But its doesnt help your preamp.
A voltage controlled switch would do it, cam goes high (record), close the switch (solid state switch like a 4066) after a small delay, then audio in connected. You would then hear no thump at all in your headphones working the same way as the cam does. This may be over complicated!! Would try the cap first.

[edit] and this feature, assuming this is the actual fault, should be already built into the preamp!

Paul.