6x2 mini* mixer

The only knobs I found that  will work are apem 4200272B1/4.  9.4mm diameter.

http://www.digikey.com/products/en/hardware-fasteners-accessories/knobs/568?k=&pkeyword=&v=679&pv73=37&pv76=467&FV=ffe00238&mnonly=0&newproducts=0&ColumnSort=0&page=1&quantity=0&ptm=0&fid=0&pageSize=25

If I even find them anywhere, maybe from a chinese supplier, the only way to adjust them will be holding from top and bottom since there will hardly be any clearance between them. It is acceptable to have no knobs since the 1/4" alpha is big enough, but not preferable. I could engrave a  groove on the shaft as position marker.

I won't be doing any gain riding, so if it's reasonably possible to set gains before a session that works for me, even if it's slightly more difficult than with a proper/larger format.

Isn't it missing the lytic cap on that Master switched gain control?

?

I don't have a cap inside the gain loop on the master output card. Can't remember why, maybe I thought that since dc offset should be low already, and since it's not much gain on the master it could be left out.

The whole summing section was wrong anway. There was a/c path from channel to channel so the panning didn't work. I hopefully fixed it. The path now is output from 1510 into this:

http://www.analog.com/en/analog-dialogue/articles/ultralow-distortion-panpot-amplifier.html

but I replaced the pot with fixed resistors.

Ordering new boards very soon. Also working on a that1580 version with digital controlled gain. Finished power supply this week and working on layout for the preamp cards.

Is that a slip in fit for something like an Altoids tin?

You're a madman, I love it.

First post updated. Switched to That1580/5171 - no pots.

Pic of simple Android gain control app. Have not yet started with bluetooth connection coding due to dread.

backplane render

Just revisiting this thread. I notice in the first post you are using a  ltm8049. I looked at the data sheet for this and it appears to be a ball grid array (BGA). How do you attach that to the PCB?

Cheers

Ian

ruffrecords said:

Just revisiting this thread. I notice in the first post you are using a  ltm8049. I looked at the data sheet for this and it appears to be a ball grid array (BGA). How do you attach that to the PCB?

Click to expand...

He has to use a stencil for the solder paste.  Put the board over a pre-heater, place the part in the footprint, and turn on the heat. The solder paste will get more fluid. Heat with a hot-air tool at low velocity, and the solder will melt and part will settle into the pads. If you watch the part from the side, you can sit it bob up and then settle down. It's pretty neat to watch! Remove hot air, turn off the board pre-heater, let it sit for a moment to cool.

The alternatives include using a proper assembly oven, or a toaster oven with some sort of control.

I have dealt with pads on the bottom of ICs but haven't been brave enough to solder BGA on my own bench...

One minor tidbit with the hot air station be careful about getting too close with a small IC the air pressure could move it around. With light pressure and heat the solder should be self-aligning.

JR

Finished first round of boards.  All rails on power supply working. It's running perhaps a little hot though for no or light loads at about 63c. Rfduino breakout board works and  able to load sample sketches.  Summing card done, but untested.  Preamp board not working, no signal passing. Found incorrect pin assignment on the 5171 on the rst pin, so spi address initialization could not happen.  I could not  even read the chip id. Hoping the fix will take care of the no signal but that might be too optimistic. New boards ordered today to correct the rst pin.

RST problem fixed with new boards. Reflowed one today, signal is passing on the left stereo buss. Right channel sounds like low frequency buzzing and hash. SPI bus is working though, able to read/write gain settings and chip id.

Back to the power supply...

The first p/s  had all rails working and stable, but it was running too hot at about 65c or over, unloaded. Also there was something like 70mA of current draw with all rails unloaded. I tried to track down the problem but failed.

So I thought perhaps since this was my first bga that I had damaged it during reflow. I tried a second one using a hotplate, and it looked like a perfect reflow. Populated a second board with just the passives surrounding the ltm8049 and still had the exact same problem.

I found one problem: forgot to populate the bulk input electrolytic capacitor, and was hotplugging from a dc brick. This seemed to solve some of the heat problem, but then again maybe not, on the second time measuring heat it was just slightly lower. It's possible I damaged both chips in the same way by hotplugging, but both are outputting good rails, so who knows.

The same excessive current draw is exhibited on both boards. I'm inclined right now to just bail on the ltm8049 and go for one of the dual output dc converters from LT or AD, unless a smps guru can help me figure out what's eating that 70mA of current.

It's alive.

Spi bus is working for reading and writing gain settings. I had some intermittent problems with startup/reset on one card, soldered up another card that worked. and then the first one started working out of the blue. So two cards on the spi bus were working well - just as far as the serial communication goes.

Audio passed through one channel, the other is dead.

Using sm57 for testing, signal is  brittle as if it's not being loaded properly, or there are other problems.

Power supply fed from dc brick is horrendously noisy.

9V battery gets rid of all the low frequency noise but there is some buzzing in very high frequencies.