API-ish Mic Pre Project

[synthetic]

I just recorded some quick guitar tracks through the preamp. It sounds nice -- open and full. I did an A/B test with the preamps in my TASCAM FW-1884, which I like, and it sounded midrangey and soft in comparison.

If you're interested in hearing what it sounds like, I'm uploading a few sound files to my site. They are really bad recordings of really bad guitar playing, but what the heck. I used an MXL 960 tube mic at the bridge and an MXL 603 small condenser mic at the 12th fret. No compression, gain or EQ added.

MP3 format, 44.1k (457kB)
http://www.jefflaity.com/music/API/apitest_960_tube_441.mp3
http://www.jefflaity.com/music/API/apitest_603_small_441.mp3

AIFF format, 44.1k/24-bit (5.87MB)
http://www.jefflaity.com/music/API/apitest_960_tube_441.aif

WAV format, 96k/24-bit (12.7MB)
http://www.jefflaity.com/music/API/apitest_960_tube.wav

When I do some real recordings, I'll let y'all know. I can't wait to record my wife singing through this. I'll also post pics of the finished unit, but that won't be for a few weeks. I also ordered some new caps as chrissugar suggested, this should open the sound up even more. Thanks again, everyone. :guinness:

 

[Flatpicker]

[quote author="synthetic"]I just recorded some quick guitar tracks...[/quote]I'd love to hear them, but they won't download for some reason.

 

[synthetic]

Yeah, I think our server is down. Try again tomorrow, sorry.

 

[synthetic]

The site is back up.

 

[FrankSL]

Can you upload the traks recorded with the other preamps, just for comparison?
Thanks,
Frank

 

[synthetic]

OK, sure. Here are recordings made with my TASCAM FW-1884 for comparison. A different take, of course, but the same mics and placement.

44.1kHz, 24-bit, WAVE format
http://www.jefflaity.com/music/API/fwtest_603_small.wav
http://www.jefflaity.com/music/API/fwtest_960_tube.wav

44.1kHz, 16-bit, MP3 format
http://www.jefflaity.com/music/API/fwtest_603_small_16.mp3
http://www.jefflaity.com/music/API/fwtest_960_tube_16.mp3

 

[FrankSL]

Many thanks!
I had a fast listen with headphones, so not really serious, but the diy one looks really more detailed and clean, isn't it?

Frank

 

[synthetic]

Thanks. To me, the FW preamp sounds like it's going through a high and lowpass filter. The sound opens up when I switch to the API-ish (I have to come up with a name for this thing...) and I hear more detail.

I still think the FW-1884 sounds nice. (Full disclosure: I work for TASCAM Marketing.) It's just that this preamp sounds much better. And the recordings you hear are still going through the FW preamp, so future recordings will sound even better. I have to make a cable that will go out of the API preamp into the insert return of the FW, bypassing their mic/line preamp.

And next time, perhaps I'll have the patience to spend more than 30 seconds setting up mics.

 

[synthetic]

Well, it's been six months since I picked this project up, but I got it mostly working again last night. I have a few questions, though:

1. It stopped working when I added phantom power. I implemented it using the phantom power circuit in this diagram:

http://www.jensen-transformers.com/as/as016.pdf

When I disconnected this circuit, the preamp started working again. It didn't matter whether the 48V power was on or off. However, since there's effectively a short between the two legs of the input through 14k of resistance, is this not the best way to implement phantom power? Do I need some blocking caps or something?

2. I get noise whenever I move the gain pot. This shows up on both channels of my mixer, so if I move one gain control I get the noise on both outputs. Any ideas?

3. Most of the gain happens in the last 20 percent of the knob travel. I'm using a 10k linear pot in series with a 10k resistor in my gain stage.

Could I adjust the gain resitors so that more gain happens sooner?

Thanks for your help. [/img]

 

[Rob Flinn]

It didn't matter whether the 48V power was on or off. However, since there's effectively a short between the two legs of the input through 14k of resistance, is this not the best way to implement phantom power? Do I need some blocking caps or something?

If you are applying this to the primary of a mic tranformer it is not a problem to not have blocking caps, because the 48v is isolated from the rest of the circuit by the transformer.

The 14K in parallel with the transformer is also not a problem. This just adds in parallel with the input impedance, which is far lower. The result is that the input impedance is slightly lower than without the resistors, but this is generally insignificant. If it worries you then use a 2 pole switch & you can switch them out of circuit when you don`t require 48v. I suspect you may have not connected the phantom circuit correctly.


The gain pot should be a reverse log pot for this circuit. This will stop all the gain being down one ened of the pot. I think in my version it was 22k

 

[Flatpicker]

[quote author="Rob Flinn"]...I suspect you may have not connected the pahntom circuit correctly...[/quote]I agree ? one resistor should be tied to the ?+ Input? of your transformer primary (which should also be connected to pin 2 of the XLR) and the other resistor should be tied to the ?- Input? of your transformer primary (which should also be connected to pin 3 of the XLR). Tie the other ends of the resistors together and connect that junction to 48V. That?s really all there is to it. Check your wiring and see if this is done correctly. Also, the resistors must be 1% 6.82K.

The noise may be because C6 (250uF) isn?t large enough. I would make it at least 10 times that (2500uF), but it doesn?t need a high voltage rating ? 6V is enough.

 

[synthetic]

Here is the latest version. I'm still hearing some periodic RF noise in the circuit, even with the new 3300uF cap. I can sometime make it happen by messing with the gain pot. Also, the circuit oscillates if I turn the gain all the way up. I can live with this, but it might help some of you to troubleshoot the circuit. Everything is pretty well star-grounded from input to output. I don't have a shield around my non-toroidal power transformer.

I added the resistor in series with the gain pot to come closer to a reverse log taper on the gain, but it doesn't really do much. If anyone has a source for reverse log 22k pots in the USA, please let me know. I'll probably have to break down and buy a bag of 10 of them for $25, which is Vimex's minimum order.

I'll be hooking up phantom power again soon, but the real problem is the RF noise. Any suggestions? Thanks, -jl

 

[CJ]

What type of output transformer are you using?
What frequency is the osc. at?

Those darn gain pots are a waste. I was doing some frequency sweeps on my Malcor mic pre and noticed that at some gain settings, the opamp just chews up the signal so bad that I have to turn the signal generator way up.

I think the best thing to do is find a gain setting your real happy with, and put in a fixed resistor. I do not think these things were designed to be used with a gain pot and a microphone, but first as a fixed gain line amp for a mixer or something. They do sound great as mic pre's, but you have to tailor the gain to the particular op amp you are using to get a good sound up and down the frequency spectrum. The gain pot messes with the frequency response too much, at least with mine.

 

[Flatpicker]

Ok, there is some weirdness with your schematic.

First, you have a 2k in parallel (not series) with the 10k. This only changes the min gain from ~2 to ~7. It?s not going to noticeably affect the taper. I?m with CJ ? toss the pot and use a switch! You can get 12 position Alphas from Mouser really cheap and they are decent switches. Calculate your resistor values to give you the gain steps you want.

Second, you have the transformer primary wired 1:10 and center-tapped. Wire it like the non-center-tapped version in diagram ?C? of the 1636 data sheet (the diagram on the left-hand side) since you are not using the center tap.

To stop oscillations, change C5 to 33pf and put a 100 ohm in series with it. If it is still oscillating, put a value between 33pF and 100pF between the ?-? input and ground.

 

[Flatpicker]

[quote author="CJ"]What type of output transformer are you using?[/quote]LL5402

 

[synthetic]

I'm not using an output transformer at all. The signal is unbalanced out of C7. The "optional" output transformer is for a future upgrade.

Sorry, I meant that the resistor is in series with the gain pot. You're right of course, I'm dancing around with three feedback resistors to get the taper I want. Once I get the system working the way I want it to I'll mess with the gain control. However, I'd prefer a linear control since I like to ride the mic gain when tracking vocals.

Is the oscillation related to the RF noise? Will rewiring the transformer or chaging the feedback cap get rid of this noise? Since the oscillation only happens when the gain is wide open, I'm not concerned with it unless it's affecting something else in the circuit.

 

[Flatpicker]

[quote author="synthetic"]...However, I'd prefer a linear control since I like to ride the mic gain when tracking vocals...[/quote]Change R2 to 5K and put a 5K pot in series with it in the fb loop. That will not only give you plenty adjustment to "ride", but more linear action.

[quote author="synthetic"]...Is the oscillation related to the RF noise? Will rewiring the transformer or changing the feedback cap get rid of this noise? Since the oscillation only happens when the gain is wide open, I'm not concerned with it unless it's affecting something else in the circuit.[/quote]If this thing is in a metal box, you shouldn?t be getting RF, especially since you are using an input transformer. You may experience slight 60Hz hum though, since your non-toroidal power transformer can more easily radiate into you input transformer.

As for the oscillations, your highest gain setting is only ~60dB. I can run my 992s at ~70dB with no oscillations and mine are configured as I described above. Anyway, stability issues with the 992 are more critical at low gains than high.

I don't know what effect it will have, but you really should change the LL1636 transformer to non ct. Just try it and see.

 

[chrissugar]

I have an API topology preamp with almost the same components, LL1636 input, LL5402 output and a Forssell jfet opamp based on Fred's white paper. It sounds great and I use it for more than two years.
R1=10ohm, R2=10K, like in your schematic, but C5 is too high for the Forssell opamp. C5 should be a value between 10-47pF.
Bypass C6 with a 100n polyprop.

chrissugar

 

[CJ]

I bet if you hung a chunk o iron on the end, it might behave differently.

 

[Flatpicker]

[quote author="chrissugar"]...Bypass C6 with a 100n polyprop...[/quote]Thanks - I forgot to mention that.