mic pre for laptop.

[Johan]

I found a little mic transformer in my drawer yesterday and thought Id build a micpre for going straight into the input on my laptop(for convenient home recording). I came up with this and it seems to work well... any pointer on improving would be welcome thou..
its just a small transformer (~1:6) a 6 position dual toggle rotary switch set up as a voltage divider a simple amp driven by the laptop input(3.2volt phantom through a 2k2 resistor) and a switch disabling the transistorstage if I plug into a regular line input...it all lives in a MXR sized box

at first I didnt think I'd need the transistor stage, but reading up on laptop input I found it's only 2k2 and not a 10k input as I expected..this seems to work well thou..

again..any ideas on improving it without loosing simplicity, would be welcome...so, PRR?.. R.G.? Merlin?...
J

[sault]

I'm used to seeing an active buffer, not a transformer. I would normally expect a higher input impedance... if you're using a microphone with a higher impedance, would that not skew the frequency response?


Sault

[Johan]

... if you're using a microphone with a higher impedance, would that not skew the frequency response?

Sault

probably yes. the transformer I used expects a 150-300 Ohm mic, like a SM57/58 and brings the signal up ~15db and impedance up to ~10k .I should have added that in the first post...
..I thought the laptop input would be a 10k one but it gets "phantom" power through a 2k2 resistor, so I added the transistor stage which uses the phantom resistor as a collector resistor and it all becomes a gain stage...but in trough, the transistorstage was an afterthought to solve the problem with the low input Z...
J

[PRR]

That will work, obviously does. If you are happy, I'm happy.

In theory.....

The mike should be high-Z loaded, typically 2K or more. The transformer, if good, will do that, _IF_ the secondary is not loaded. It's too dark on your desk to read the JPEG-ed text; 300:10K? Then the secondary "should" be loaded in 50K-100K to throw ~~2K at the mike.

The near-naked transistor, at say 1mA, has input impedance near 3K. Reflected back through the 300:10K transformer that's like 100 ohms input. That's acceptable, but some mikes sound different 100 ohms or 2K load. Some mikes sound "better" at low-Z load (a matter of opinion and use). Some mikes just do not care.

The low base impedance also skews your careful attenuator ratios, though since this is surely turn-to-taste not the calibration lab that's no big deal.

It is generally poor practice to attenuate mike-level signals, unless they are WAY hot. You reduce the mike signal, the transistor hiss is unchanged, signal/noise ratio is lower.

_IF_ the source feeding the transistor base were zero impedance (much-much-less-than 3K), the transistor gain is something near 50. The output must clip near 1V level. (The sound-chip inside the PC may clip sooner!) The maximum level at base is 30mV; at primary about 5mV. Taking SM58 or EV635A, input clipping is 102dB SPL which is not "loud" by modern studio standards. Using large condenser, input clipping is 90db SPL which is not at all loud musically.

Since the laptop "mic" input is a microphone input, and you have voltage gain in the transformer, and your musical efforts may be louder than generic laptop sound sources, it almost seems you want voltage LOSS, not gain.

When you want to try something different: put in some emitter resistance. Raises input impedance, lowers gain. A sequence like 33, 100, 220, 470, 1K, 2.2K ought to give useful settings. Cover your ears when you adjust gain: it will pop.

[Johan]

thank's PRR. I'll try your sugestions in the next few days. I also have a 150:600 transformer. it might be a better match then. I very rarely use anything other than a sm57, so 150 ohm input is just fine...yeah, I'm one of those guys..
J

[sault]

Hey, for the heck of it (and because I didn't have anything else on hand) I tried recording worship yesterday with just a pair of 57's in an XY pattern near the back of the room... I was very surprised to hear a moderately decent response! Even with signal levels below -20db most of the time (ick), I still got decent frequency response and a very low S-to-N ratio. I was very proud of my little Lexicon Alpha!

Anyways, working with a transformer is unfamiliar territory for me... time to hit the net again and see if I can educate myself...

The direction I would have instinctively gone would have been to simply sandwich a volume control between two buffers and called it good... something which can be easily done with a single TL072, for instance.


Question - if the selectable resistor control was put *in front* of a buffer, would that not allow us to change our input impedance? I'd think that a buffer following it would keep the impedance "stiff" (is that the right way to say it? stiff? loaded?).

In other words,

input -> selectable resistor (from 330Ω to 2k2Ω to 1MΩ I suppose) -> buffer -> volume control -> buffer -> output
(maybe putting a very high value cap between the resistor and buffer?)


Sault

[Johan]

just to finish this up...I tried the shanges suggested by PRR, and it's great...I liked it before, but this is better...also, it now has a 150:600 Ohm audio/mic transformer...
thank's
J

[tubelectron]

Hi and Good Night,

Interesting subject as I have a similar need - I'll be back !

A+!