Page 9 of 11

Re: That1580 mini mixer

Posted: Fri Oct 23, 2020 3:10 am
by AnalogJoe
weroflu wrote: Mon Oct 19, 2020 8:00 am c12,14.16 gone.

Metering will be done post unbalancing right before the 5k mixing resistors. Going with attiny84 somethings on each channel card, 10 leds, will try to emulate an lm3916 with the mcu. low current leds and peak metering should not be too much of a current draw. (led color render is messed up, it's supposed to show as 8 greens, 1 yellow, 1 red)

I'm hoping that those mini jacks (which are soldered to the motherboard) will squeak into the tins.

That's about it for the design I think. The rest is making it work and writing code.
Hopefuly OLED technollogy will die soon, OLED displays have their days numbered, my U1273A meter has an OLED display, I regret buying it every day, at first it lwas beautiful to look at, now its impossible to read because the display is so dark, it makes sense since it has organic materials which decay over time, I had to buy a new display from a Chinese dealer, it wasn't cheap, it cost me around $100 USD, I just wish I had the LCD version instead.... My point: avoid OLED https://en.wikipedia.org/wiki/OLED#Lifespan

Re: That1580 mini mixer

Posted: Sat Oct 24, 2020 10:17 am
by JR.
AnalogJoe wrote: Fri Oct 23, 2020 3:10 am
weroflu wrote: Mon Oct 19, 2020 8:00 am c12,14.16 gone.

Metering will be done post unbalancing right before the 5k mixing resistors. Going with attiny84 somethings on each channel card, 10 leds, will try to emulate an lm3916 with the mcu. low current leds and peak metering should not be too much of a current draw. (led color render is messed up, it's supposed to show as 8 greens, 1 yellow, 1 red)

I'm hoping that those mini jacks (which are soldered to the motherboard) will squeak into the tins.

That's about it for the design I think. The rest is making it work and writing code.
Hopefuly OLED technollogy will die soon, OLED displays have their days numbered, my U1273A meter has an OLED display, I regret buying it every day, at first it lwas beautiful to look at, now its impossible to read because the display is so dark, it makes sense since it has organic materials which decay over time, I had to buy a new display from a Chinese dealer, it wasn't cheap, it cost me around $100 USD, I just wish I had the LCD version instead.... My point: avoid OLED https://en.wikipedia.org/wiki/OLED#Lifespan
not to feed the veer but IIRC OLED displays were supposed to be the next great display technology. That short life seems like a swing and a miss...

JR

Re: That1580 mini mixer

Posted: Sat Nov 21, 2020 4:37 am
by weroflu
Oled: Show me how I can use lcd in a smaller footprint than what I'm using now.

I managed to squeeze the 6th channel back in by offloading the p/s onto a mini card.

Getting close to a board run and then I saw the That 6261, hmmm.

It wouldn't be hard to add a master channel with gain/attenuation now. a pair of 1646/1606's after the summing and one 6261.
edit: i didn't look closely enough, the adc driver is not bypassable. Not sure if this is a good idea.

Re: That1580 mini mixer

Posted: Mon Dec 07, 2020 9:47 am
by weroflu
Boards ordered. First next step is just a physical dimension test on the jacks and edge connectors.

I caught one error, which might turn into a positive...

I rearranged the phantom blocking caps, which were previously 22uf - as I understand it this about the low end of what you want.
I can only fit 10uf now, well no I can Kludge in 22uf but not prettily and probably higher if I mash them.

Also realized that I never had .1u film caps in parallel in any of my past designs. This will be a challenge to squeeze in but I think it's possible on the next board run. Is it worth it? Wima, Kemet, any recommended brand?

Regarding these current 10uf blocking caps, I noticed the that 5171/1580 sample schematic does not include a T-bias 22k resistor as in the 1510/12 datasheet. Is this a simpler way to bring the corner rc frequency down lower without using a larger phantom blocking cap?

Re: That1580 mini mixer

Posted: Tue Dec 08, 2020 4:33 am
by mediatechnology
weroflu wrote: Mon Dec 07, 2020 9:47 am Boards ordered. First next step is just a physical dimension test on the jacks and edge connectors.

I caught one error, which might turn into a positive...

I rearranged the phantom blocking caps, which were previously 22uf - as I understand it this about the low end of what you want.
I can only fit 10uf now, well no I can Kludge in 22uf but not prettily and probably higher if I mash them.

Also realized that I never had .1u film caps in parallel in any of my past designs. This will be a challenge to squeeze in but I think it's possible on the next board run. Is it worth it? Wima, Kemet, any recommended brand?

Regarding these current 10uf blocking caps, I noticed the that 5171/1580 sample schematic does not include a T-bias 22k resistor as in the 1510/12 datasheet. Is this a simpler way to bring the corner rc frequency down lower without using a larger phantom blocking cap?
I would leave off the 0.1 uF films since you are space-constrained.

T-bias will raise the common mode impedance to reduce the effect of cap mismatch so it will improve LF common mode rejection. It won't lower the differential Fc. The best way to lower the differential cutoff frequency under your circumstance would be to raise the bias resistor values somewhat.

The downside to raising the bias resistor values is increased 1/f noise when the input is open circuit. Since open inputs should be muted I don't see this as a big penalty.

What are the values of your bias resistors?

Re: That1580 mini mixer

Posted: Wed Dec 09, 2020 11:13 am
by weroflu
1.2k bias straight from the datasheet.

I don't have mute on this version.

Re: That1580 mini mixer

Posted: Wed Dec 09, 2020 11:38 am
by mediatechnology
I calculate a differential mode 13 Hz cutoff with 5uF (2X 10uF in series) and 2400Ω (2X 1200Ω).

I would raise Rbias and continue to use the 10 uF.

Re: That1580 mini mixer

Posted: Mon Feb 01, 2021 3:49 pm
by weroflu
Lots of progress.

328p mcu working, oled display working, mini joystick working, power supply working, and mcu section of the channel cards is working.

Wayne you will like this new power supply, it is half the size of the last one. Double sided board about the size of two thumbnails

New board order coming up as i used the wrong footprint for the ixys lcc120 relays. Had to make a breakout board just for a single dip8 since space is very tight.

This week or so I will try to get some signal into a channel card.

Re: That1580 mini mixer

Posted: Sun Jun 06, 2021 4:47 am
by weroflu
one preamp card is built but have not passed audio yet. i have some beta code for a linear audio meter loaded, have not gotten around to making it log yet.

i paused building due to:

spending some months trying to turn this from an analog mixer into a mixer/stereo adc/stereo recorder. right now the design is 5 channel inputs feeding a stereo mix that feeds cs5381 adc card i2s out to the samd. there will be a very minimal interface for stopping and stopping recording at fixed rate 24/96k. the new motherboard will have a microsd slot, samd21 to hopefully handle the i2s stream. schematics attached. it's not really in the league of professional design, but i think i'd be happy if it worked and sounded halfway decent. the schematic is close, but i haven't looked into how to write the micro sd logging/recording code.

i'm keeping the preamp cards the same as the current just-mixer version. signal is singled ended after the polarity switch and then there is another single ended >differential buffer stage at the adc. it is probably better to keep things differential all the way to the adc but that means a new iteration of the channel cards and squeezing more traces in to the summing section. if i get feedback here that it definitely should be done this way then i'll make the necessary revisions. the channel card schematic is not changed but i included it here for ease of viewing.

any suggestions welcome.

Re: That1580 mini mixer

Posted: Mon Jun 07, 2021 6:16 am
by mediatechnology
Nice work!

Could you perform polarity reversal in software and save a bunch of switches/relays?