Well chaps, thanks so much for the help. I'd have been really stuck otherwise!

I'll keep the thread updated with my progress (still waiting for some transistors in the post).

 

I have some bad news. I forwarded the circuit a few days ago (so it was the one with the second np cap added before the diodes) to Rod Elliott (the guy who designed the circuits) to double check that everything was okay, and he says the following:

Matthew,
Attempting to use the same battery for phantom and to power the preamp is a really bad idea, for a multiplicity of reasons.

First and foremost, the input XLR, mic body, etc. will be at a negative voltage referred to earth/ground. Should the mic or connector contact the chassis of the recorder you'll have a partial short on the battery. Exactly what will happen is unknown, but it won't be good.

Your circuit compromises the effectiveness of the ground connection - it is meant (and expected) that ground really does mean ground!

For the sake of an extra battery you'll end up with a circuit that will cause you grief at some stage.

Cheers, Rod

Any thoughts?

 

Make sure you send him the final design, bottom from Post #38
http://img191.imageshack.us/img191/7903/phantompowerpreamplowno.png

It sounds like he saw one of the previous designs
The Gnd (and the mic body) are connected to the battery negative in this design.
He could miss seeing the connection just as I did earlier on. (#36, #37)
The grounding has been one of the main points of our discussion all along.

 

Yeah, he saw this design. He mentions that the mic ground will be at a 'negative voltage', which is of course -ve for the preamp, so that does indicate that he is aware that the mic is connected to 0v (which is -11v for the pre). However, because the output ground is also connected to 0v it would be impossible for there to be a short, so it does indicate that he might have missed something.

I don't want to be wasting his time, though, and he's already been a massive help, so I'm trying not to bother him with this too much (in the contact section of his website he cautions against lengthy back and forth emails). As a result, I want to make sure that it's not the circuit that's at fault before contacting him again, because he surely knows his stuff.

I've been looking at the pinout of the IC, and the circuit we've come up with does seem to be correct to me (which isn't saying much because I know very little about all this stuff).

The output of the IC is, according to the datasheet, pin 6, so that's okay. Pin 4 is Vcc-, which is the ground for the IC, right? Because that would mean the the output ground should be (and is, in our diagram) 0v (-ve). If the output ground was meant to be the 'virtual ground' of the preamp, the output ground would effectively be pin 3 of the IC, which according to the data sheet is IN+, and that doesn't make sense to me.

I also wonder what he means by the effectiveness of the ground being compromised. The only change really is that the mic's ground isn't the same as the pre's ground, but it shouldn't matter because they're not connected (unless that's exactly what he means about it being compromised). The only other thing would be that the output ground is -ve, but as I mentioned previously, that seems correct for the IC's output. Quite confusing!

 

He's saying that if you touch the metal body of the microphone to the metal body of the preamp (if the preamp has a metal body which is connected to its ground (1/2 supply)) you would create the short that you had previously circled with a red oval.

Edit: this would be the same as connecting a wire directly from the midpoint to the lower point of your battery pack.

 

The output of the IC is, according to the datasheet, pin 6, so that's okay. Pin 4 is Vcc-, which is the ground for the IC, right? Because that would mean the the output ground should be (and is, in our diagram) 0v (-ve). If the output ground was meant to be the 'virtual ground' of the preamp, the output ground would effectively be pin 3 of the IC, which according to the data sheet is IN+, and that doesn't make sense to me.

I also wonder what he means by the effectiveness of the ground being compromised. The only change really is that the mic's ground isn't the same as the pre's ground, but it shouldn't matter because they're not connected (unless that's exactly what he means about it being compromised). The only other thing would be that the output ground is -ve, but as I mentioned previously, that seems correct for the IC's output. Quite confusing!

For the IC to function in this particular circuit pin 4 (Vcc-) must be at the battery ground, pin 8 must be at the battery V+, (Vcc+) and the input must be biased 1/2 way between those points.

The confusion here is the phantom voltage connection. It must be the same as the op amp Vcc- voltage which must be earth or "outside world" Gnd.

Hope this helps, gotta go, life is calling. .

 

The output of the IC is, according to the datasheet, pin 6, so that's okay. Pin 4 is Vcc-, which is the ground for the IC, right? Because that would mean the the output ground should be (and is, in our diagram) 0v (-ve). If the output ground was meant to be the 'virtual ground' of the preamp, the output ground would effectively be pin 3 of the IC, which according to the data sheet is IN+, and that doesn't make sense to me.

Vcc+ goes to 22V and Vcc- goes to 0V exactly as in the most recent schematic.

 

Yeah, I follow that. I was just wondering whether the 3.5mm jack ground should be connected to Vcc- (as it is in our diagram), or whether it should be connected to the pre's virtual ground, as Rod's original circuit diagram seems to indicate.

Edit: This is the reason the 22uf np cap was added, isn't it? To eliminate the voltage offset of 'output', and allow us to use 0v as our ground instead? Man, I wish I wasn't so dumb at this. Chaps, I'm sorry that you're having to put up with me.

 

I think the original reason it was suggested was because you were going to use a resistive rail splitter so 0V would provide the best path back to the battery. If you're using a direct (not virtual) ground midpoint (your ground wire directly touches the batteries) you may use either. I don't think it matters as long as an output cap is present. The output circuit won't present the issue that Rod Elliot mentioned because that circuit will not be sharing a power supply, only a single ground reference connection.

 

I think the original reason it was suggested was because you were going to use a resistive rail splitter so 0V would provide the best path back to the battery. If you're using a direct (not virtual) ground midpoint (your ground wire directly touches the batteries) you may use either. I don't think it matters as long as an output cap is present. The output circuit won't present the issue that Rod Elliot mentioned because that circuit will not be sharing a power supply, only a single ground reference connection.

It doesn´t reall matter whether the virtual ground is buffered or not. The main point is that the mic ground + shield needs to be connected to the -V supply, so you´d then want to have the box and the output ground connected to the same potential.

 

The "box" being what? There are three components to this circuit network, the mic, the preamp and the output device/power amp/computer/what-have-you. Each has a ground which may or may not be connected to its external casing. The mic is at 0V, the preamp is at 11V, the output ground is _currently_ at 0V. To me, the most correct solution seems to be to house the preamp in plastic so there's no exposed ground and leave the output at 0V so, in the case that the output equipment exposes its ground to its casing, sticking the mic on top of _that_ will not cause the 11V short circuit.

 

The box being the metal case. Both input and output are AC coupled, so it doesn´t matter what is the voltage on the inner virtual ground, because the negative pole of the battery is the common ground to the input and output.

 

MattP - To achieve lowest noise, including from outside sources, it's best practice to house the preamp in a metal enclosure which is grounded. This provides a shielded enclosure around the circuit - the same way as the shield on the mic cable.

And, in my humble opinion there is no better way to learn than to work on a project like this. You don't learn anything if you don't ask questions.

 

Roger that. I'll use some aluminium to shield/encase it, I think, but I'll make sure I then encase the aluminium in something non-conductive (a few layers of electrical tape should do, I reckon), to help prevent any possible shorts.

It's going to be fixed inside a case, so "

" (shorts, tehe) should be unlikely anyway.

My spare time will be a little short p) until the other side of the weekend (work stuff), but I'm going to go ahead and build it as soon as I'm free! I'm looking forward to it. Going to stick with the latest design, with 0v as ground for both inputs and outputs.

And you're right about learning, tubeguy. I've already learnt so much from you chaps, and I haven't even gotten my soldering iron out yet.

 

It's going to be fixed inside a case, so "

" (shorts, tehe) should be unlikely anyway.

Yep, I think you fit right in here.

A suggestion.
Build the preamp and test it before hooking it to your very nice costly mic. You can test it with a standard non-condensor mic or it can be tested with other signals (as from an ipod). An ipod signal is stronger than a mic signal and should be padded with resistors, so just ask when you reach that point.

Good Luck !

 

I have a dynamic mic hanging around, so I'll test it with that. Thanks for the suggestion!

 

I've finished my circuit. I just need to go through it one last time before I apply power, just to make absolutely sure that everything is okay!

Still got to build the phantom power bit (will do that on a separate board), but for now I'm just going to test it with a dynamic mic. It's only got two contacts, so I need to look up how I hook it up.

 

I've tried the board with 2x 9v batteries (one is new, the other is somewhat used). The result is that I get hiss, but nothing else. Even when I short the inputs with the ground (virtual ground), nothing changes. Mic or no mic, it's still the same.


The only things I can hear, are a tiny bit of 'knob noise' as I turn the VR, and a tiny click (only sometimes) when I touch the mic-in contacts. I should have guessed that it wouldn't work first time. Sheesh.

I cannot think what is wrong, as I've gone over the circuit several times and everything seems ok to me.

 

Thanks for posting the photo's.
Back of board - it appears there may be a short on the board between row 5-6, and near the middle row 12-13.
Please post the voltages from the virtual ground to V+, V-, the bases of Q1, Q3, emitters of Q2, Q4, and the voltages on the IC pins.

 

This is how in theory the DC voltages in the circuit should be, plus minus some component variation. Check if your values are the same.
edit: attachment