yes but it also reduces the signal into the camera and hence its own internal gain would need boosting - more noise as a result. Can the preamp's output be boosted to compensate rather than increasing the gain in the camera?
This is not really the answer guys, but at the end of the day it may at least help and i guess thats what matters. I still stand the simple cap idea, but its a question of rise times.
Was just reading an article about some hi fi guys with similar DC problem from an old CD player to an amplifier. Popping whenever the track changed! he a fitted a blocking cap and the job was done. Clean. Thats my point. Even though its a transient and people think a cap wont work, it can work. Depends on the rise time of the record voltage.

[edit] maybe I need to build a test rig to simulate this. Will post any results. Too much guess work :)

Paul.

Well according to the Beachtek tech, the Nikon is too sensitive and with the attenuation there will be a better match. Actually, the levels seem fine right now, but I will certainly try it.

I cut a 1/8" to 1/8 cable exposing the ends, got some capacitors and some alligator clip jumpers from Radio Shack and tried all manner of wiring in some blocking capacitors and no configuration seemed to have any effect on the pop. I am assuming that that is because while a capacitor would stop and remove the continuous dc-bias, it won't stop the initial pop as the power is first applied and the capacitor gets it's first charge. I expect that the Beachtek already has blocking capacitors and that what I was adding was redundant.

Yes Laurence glad you have tried this, saves me the bother of testing it.
So it does sound like the rise time from the camera record voltage is fast. Thats unfortunate because indeed, it means the blocking cap in your case will not work. I assume the value you tried was 10uF? or thereabouts. Any less than that and you run the risk of losing low end response. The beachtek may or may not have internal blocking at its outputs. Its designed to go into a high impedance mic socket (camera) which will always have dc decoupling anyway. The design is of course up to them and without actually opening the box or finding a diagram for its hard to say.
Nikon too sensitive? but as you say, the levels seem right as they are now. This will have to be increased by 25dB to compensate. Hopefully, this does not introduce excessive hiss.

Paul.

I used two 10uF capacitors, one for each channel.

The attenuating cable should limit the momentary direct current to "roughly" 2mA per channel if it uses a traditional L-pads.

[EDIT]
Product Highlights

My guess is that it would be a T-pad so that the cable would be symmetrical and not behave differently if plugged in the other way round.

I still think that a unity gain amp that is designed for microphone signal levels would be the best way to go. It would also need to have suitable input and output impedances as well but they would not be very critical.

Peter, do you know what current is available from the nikon? I dont have a Nikon with a mic socket so I just measuered the current from my laptop socket here, meter measuring current (dead short current) from mic pin + to ground and got 500 uA (micro). So i think the nikon would be similar. .5 to 10mA perhaps?. Its a tiny current, just enough to drive a mic.

Paul

1.2 mA

Ok so here is a suggested buffer amp that should do the trick.
Two things happening here: the op amp has very low output impedance typically about 10 ohms. Note there is no dc blocking cap between the op amp and the camera, this is intentional to 'shunt' the 2.5v to ground. This makes the 2.5v in effect, go away.
Second, its a buffer amp with a hi impedance input, by its very nature this will isolate the camera from the preamp unit.
Sorry the diagram is hand drawn and scribbled but i hope you get the idea. Thoughts and comments welcome! :)

[edit: link removed]

Edit: Actually I just realised that the output of the op amp will be at 2.5v always due to the way its biased. So rather than shunting to ground, its holding it at 2.5. A bit like having the camera in record mode always.

drawbacks to this design,
1 ) yet another box to put on the camera rig.
2) it needs a 9v battery

Paul

Right gentlemen, i am getting wrong for spending too much time thinking about this problem today, lots of other jobs I am supposed to be doing around the house!! Inlcuding editing a project with MBL! i got side tracked so i am now signing off. She who must be obayed has spoken - and I know my place lol. :)

Paul.

It must be at least 15 years since I was involved in this sort of thing and so I am unfamiliar with what components are available and what might be best these days. The main comments I would make on Paul's circuit relate to the power supply. The rest looks good.

The data sheet uses plus and minus 15 volt supply for the TL072 operational amplifier. That being so, why do we need a 5 volt supply? If we use a 9v battery, we could create plus and minus 4.5 volts with the voltage divider resistors as shown (2 x 4.7 kilohm, shown as 4k7). Will the op amp still work OK at this reduced voltage? My guess is yes, given that mic signal voltages are very small, and with full negative feedback around the op amp, distortion would be low.

If Laurence has not tinkered with electronics before he would probably feel more comfortable seeking the help of someone more experienced. If he doesn't know anyone, then maybe someone he knows does, or maybe someone at the local electronic component store could help.

After thought.
Following my suggestion would make the static output voltage of the op amp be 4.5 volts (or less as the battery drains). I think therefore that it might be prudent to put a DC blocking capacitor between the op amp output and camera input and a drain resistor to ground so that when unplugged, the voltage would be zero..

My choice of op amp was not correct. The TL072 is indeed a dual supply amp and does not work correctly with a single supply - thats what happens when you are tired and in a rush to draw a diagram! A better choice could be a NE5532. Very low noise and works fine with a single 9v supply.
I chose the 5v regulated bias supply for these reasons (note its not a power supply to the op amp, just bias). The potential divider voltage is 5volts / 2 = 2.5. Which is very close to the supply voltage at the Nikon mic socket. This was perfect for holding the voltage at 2.5. And because its regulated, its noise free. A smoothing capacitor should be added from the +5 to ground, i missed that.
Really the diagram should be re-draw, its a mess! appologies for the layout.

I guess it would be possible to go the other way, have a dc blocking cap at the output to the camera and use a pull down resistor to reduce the camera voltage to near zero. The problem i see there is the pull down resistor would have to be low in value - so the op amp then need to feed this low impedance. You would need a value like 100 ohms or less really (against a 2k pull up in the camera). So holding it high i think reduces battery current. Another way is to get the op amp biased in such a way where the output is zero volts and with no DC cap at output. (this is actually what i was originally aiming for!). But the 2.5 hi output would work just as well.

I want to appologise to Laurence for almost hi-jacking this thread and probably blowing the issue out of proportion. Your attenuator cable will help to a certain extent and we are really just tryting to help. But it may be getting a bit OT.

I would like to build one of these units, just for kicks. So if anyone would like to continue this as a project or test one built from me you're welcome. But this forum thread may not be the best place to continue it. Suggestions?

thanks guys - what a wonderful forum this is. And a great set of people here.

Paul.

I did have a couple of simple fixes for Laurence's problem.

1) Back to back diodes accross the feed to the headphones to limit the plop.
2) An inline volume control for the headphones.

Neither are elegant.

Bob.

Bob, yeah, a simple mute switch inline with the headphones! its so simple it made me laugh.

Paul.

In case anyone is still interested, here is an updated diagram with value changes.

http://cuenet.co.uk/sony/nikonantithump.pngSuggested Anti Thump diagram (not tested)[/link]

Paul.

I would be just a teeny bit worried about the effect of reverse current flow into the Nikon mic port, if the op amp output voltage were even slightly higher than the camera when active. It will certainly be higher when inactive.

I would therefore prefer blocking capacitors in this connection. The output static voltage of the op amps then becomes unimportant, so you could remove the Zener diode 5 volt regulator.

No its fine. As long as there is a 2k resistor bewteen opamp and camera socket, the max current (worse case) is 1.2mA. Thats the same current as the camera feeds. Not really an issue. Any developed voltage would simply be absorbed by the 2k external and internal resistors. However on the diagram i make a suggestion of 100 ohm for driving it high under low impedance (if the cam works that way) - i am retracting this idea just for safety until we have data. Its probably fine but because we dont know exactly how the camera is switching the voltage, its best to keep on the safe side. So forget point 2 of the Notes.
The whole point of this circuit is to counter act the DC bias transient produced by the camera going into record. Its not just a buffer. Thats why it is DC coupled (2k res). Treating this as a normal buffer amp with dc blocking caps at its output simply allows the camera to produce the transient again, which could feed back through the op amp. Back to square one at that point. So my intention was a two prong attack, yes its a buffer to help isolation but its also a DC match for the camera stopping the transient.

My main problem is this: I do not have a 5100 to do my own tests. So assumptions are being made. That make me feel uncomfortable about the design.

Paul.

"Treating this as a normal buffer amp with dc blocking caps at its output simply allows the camera to produce the transient again, which could feed back through the op amp. Back to square one at that point."

The whole point of the buffer amp is that the transient would NOT feed back to the Beachtek output. It would stop harmlessly at the output of the buffer amp. I don't see how any significant transient would get transferred to the input of the buffer amp and hence the output of the Beachtek device.

It looks to me as if the Beacktek output amp has significant output impedance, and the monitoring circuit is connected to it. That's why a plop can be heard at switch-on.

So far as your idea of stopping the transient is concerned, you are assuming that the equivalent circuit of the mic port is that of a 2.5 volt grounded battery, a simple on/off switch and a 2k ohm resistor, all in series. This may well be true but I thought it best to play safe just in case it wasn't. I measured approx. 2.5 volts on my camera, but Laurence's might be 2.4. If reverse current were to flow who knows what might happen. An apparent open circuit inside the IC might in reality be the equivalent of a diode biased to its nonconducting state. Once you disturb the DC conditions inside the IC, the high impedance might suddenly become low impedance, transistors might lock up in a funny state, etc. In view of that uncertainty I thought it best to play safe.

There would certainly be no harm in trying your blocking capacitor output and just relying on it as a simple buffer amp. No harm at all. And even better if it works.
I agree, i dont really enjoy the idea if DC at the mic socket, but like i explained, its really harmless given its 2k in series. Theres really no chance of lock up. 2.4v, 2.6v, doesnt matter, the opamp will hold at 'its' voltage ie.2.5. You end up with 0.1 volt drop maybe 0.2 at best. You are talking in terms of micro amps at that point.
Anyway, thanks Peter for your input and ideas, love Bob's simple headphone volume or mute idea - simple is always best!. And maybe if the dc blocking outputs works then its job done.
This is only a suggested diagram, and i was trying to defeat the DC and buffer in one go. I am sure others will have their own take on my suggestion here. Thats totally fine :) Not really expecting anyone to actually build one of these. More of an exercise for my old brain cells really! Kinda enjoyed the challenge. But i think i will jump out now and leave poor Laurence's thread alone.
Think i have spent enough time at this, made my point and now i'm done :)

cheers
Paul.

Well I ordered the 25 db attenuation cable. It has ends marked for the camera and source ends so I am sure it is a simple L or H pad. I expect that this will bring down but not eliminate the pop when the camera goes into record and if it matches the levels better, cool.

If the cable is not symmetrical in use (has specified which end goes where) then it is probably an L pad (voltage divider). H pads are only used in balanced circuits, which this isn't.

That makes sense. So, assuming it is an L pad, what would be the value of those two resistors in order to give 25db of attenuation?