Audacity Forum

Thanks for all the comments Koz. I haven't started thinking about it yet... I had some ideas a few monthes ago, which I didn't write down and that I have forgotten by now...
I will have to start searching again, read the datasheets, googling, etc...

My main goal is to build an amp for my large diaphragm t-bone mic. First something simple to connect to my soundcard analog input, and if it goes well later I'll think about building an usb interface...

I also have another project, which is to try to find a way to hack and improve the t-bone micplug I bought... I have no idea yet if this two projects will get together at some point or not...

I wanted to get a pair of 0.1% resistors too but it was out of stock on digi-key, so I couldn't add it to this order... I'll try the local stores and if I'm unlucky I'll add it to the next online order (maybe next order will be on farnell and they probably have it). I also need to check how much current I need to choose the proper transformer. But right now I'm quite busy with other things and I probably won't have time to look into that before xmas...

bgravato wrote:I wanted to get a pair of 0.1% resistors too but it was out of stock on digi-key, so I couldn't add it to this order...

If you get a pack or resistors, there's a good chance that there will be a matching pair. If you have a reasonable multimeter, testing is very easy. If not, there's the old Wheatstone Bridge method. http://sound.westhost.com/project96.htm#appendix

If you don't mind taking up a bit more board space, a pack of (E24) 5% 5.6k, a pack of 5% 1.2k resistors have a good chance of producing a very close match.
E48 series resistors are rated at 2% and reasonably inexpensive and are available as 6.81k - you would just need to find a pair that are within about 10 ohms of each other.
CPC Farnell do a 6.8k 1% resistor for under 1 Euro (£0.53 GBP) for a pack of 50.

steve wrote:If you get a pack or resistors, there's a good chance that there will be a matching pair. If you have a reasonable multimeter, testing is very easy. If not, there's the old Wheatstone Bridge method. http://sound.westhost.com/project96.htm#appendix

I thought of that too... Not sure about the reasonable multimeter... I have a cheap one made in china hehe That's why I had thought about the wheatstone bridge method too

If you don't mind taking up a bit more board space, a pack of (E24) 5% 5.6k, a pack of 5% 1.2k resistors have a good chance of producing a very close match.
E48 series resistors are rated at 2% and reasonably inexpensive and are available as 6.81k - you would just need to find a pair that are within about 10 ohms of each other.
CPC Farnell do a 6.8k 1% resistor for under 1 Euro (£0.53 GBP) for a pack of 50.

I think I saw those at digi-key too... but 1% and 2% are not 0.1% either hehe (though the chances of getting 2 matching ones are closer...)

The guys at the local store I usually go are very nice so I might ask them to let me check some at the store with one of their reasonable multimeters hehe

I might ask them to let me check some at the store with one of their reasonable multimeters

Not if you like them and want them to like you. That's serious drudge work. You don't know which two are going to come out, so you have to record each reading and then go back through the close ones to make sure you didn't make an error, eventually producing the hero pair. A good digital multi-meter or actual bridge.

And given where these are in the circuit, you need the lowest noise composition you can get in addition to matching.

Metal Film? Carbon Film? I forget which material has the least noise, but start there. Cheap carbon composition is laughably noisy and unstable with temperature.

Koz

I pulled down the data sheet. It looks like a very nice amplifier. Top quality noise specification, frequency response, etc.

However. They don't mention gain trim in the sheet. You can't run the thing at 60dB gain all the time. That's an invitation to input overloading and clipping. The chip I used (THAT 1512) attempted to deal with this by giving designs for gain trim by either a variable resistor or gain resistor switching -- none particularly graceful, but they reflect what's done in modern amplifiers.

http://www.thatcorp.com/1510-1512_Audio ... _ICs.shtml

They also talk about using Shottky Barrier diodes as protection...

we recommended using Schottky diodes ( 1N5819 types) at D 1 ~ D4 to protect the 1510/1512 inputs against overloads. ...

Oh, and I can tell you 60dB makes any imbalance at all shut the amplifier down, and the obvious solution of trimming the "Reference" input doesn't work beyond a very gentle trim to make up for slight product imbalances.

How do you tell if you're clipping? That's another design problem on the back burner. Aaaaany minute now I'm going to get to that.

Koz

I don't know yet how much gain I will need... maybe 40dB wiil be enough.

maybe 40dB wiil be enough.

That's almost certainly not true. Refer to my review of the Shure X2U...

http://forum.audacityteam.org/viewtopic ... =20#p81588

That device only has one control for volume in the sound channel similar to the other device I have, the Sound Device MixPre...

http://www.sounddevices.com/products/mx2master.htm

Both of them max out at 60dB, and both of them get blown away with the simple, cheap Peavey PV6...

http://www.sweetwater.com/store/detail/PV6

...which has three controls for each microphone: trimmer, channel fader, and master fader. It also has more gain and the ability to control it effectively, unlike the other two. Yes, I clearly understand the Peavey is not portable and the other two are.

My home-made unit had an interesting gain line-up. The transformer was good for 20, the amplifier was good for 20 and then I put a master amplifier and line driver good for 30. 70 total. That part worked, although I still expect to have overload problems.

I didn't think it was important to have serious control over the gain until the first time I had to record a Real Announcer/Presenter/Voice Talent and deal with his expressive presentation.

"Can we take that last again? You're fine. I need to make an adjustment..."

You are going to run into expression problems, too, given your microphone and theatrical playing.

Koz