That1580/1510 mini mixer

Where we discuss new analog design ideas for Pro Audio and modern spins on vintage ones.
weroflu
Posts: 39
Joined: Wed Apr 19, 2017 9:19 am

Re: That1580/1510 mini mixer

Post by weroflu » Wed May 17, 2017 3:48 pm

LIttle update. 5 cards done and working, but not rigorously tested yet. Just from quick testing with a 57 the preamps sound great.

Waiting on the drv135 board.

I added a cypress usb uart to the rfduino board and was able to squeeze in a microusb footprint on the chassis, waiting on that board as well.

The last channel card was all soldered yesterday but these qfn's are killing me. I need to change methods. If there is a next round I'm definitely ordering metal stencils. There's one more method that might work out - running a thin line of paste across each row and blasting air.

I think it will be a good test when the chassis is filled, it might make some of these noise issues go away.

weroflu
Posts: 39
Joined: Wed Apr 19, 2017 9:19 am

Re: That1580/1510 mini mixer

Post by weroflu » Wed Jun 14, 2017 9:16 am

Catastrophic hard drive failure, had to rewrite arduino and android code. Thankfully I had all schematics backed up properly.

drv135 test tone board done
Android ble app now functional - transmitting bluetooth data to arduino. The dread in programming this was not imagined.
All 6 channels now finished, qfn soldering method with paste works.

I messed up the backplane or case measurements. Technically the whole assembled mess fits inside, but I can't actually get it inside the case.
https://1580mixer.shutterfly.com/pictures/30

Ordered/received some larger 2 oz. tobacco tins to house the current prototype.

It looks like rev 2, if it happens will entail a new p/s board with lower rail voltages, all of Wayne's panning suggestions, maybe even try to eliminate the line receiver and sum balanced (total of 2 chips eliminated), squeeze everything down a bit so it can properly fit in the original case - maybe have the input jacks mounted on the backplane, maybe change the switch types so that they are flush with the top edge of the channels.

If rev 2 happens then I can make the channel strips properly modular, ie they will be able to be plugged and unplugged, but not the power supply or the summing card, not sure about the bluetooth card.

User avatar
mediatechnology
Posts: 2922
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: That1580/1510 mini mixer

Post by mediatechnology » Wed Jun 14, 2017 10:01 am

Sorry about the hard drive failure.
While you were dealing with all that we had a dbf server crash and rolled back the forum a few days before we found a backup that worked.
After running it for a few days I upgraded the board.

Sounds like the project is moving along nicely.
Mack is naming names.
Big names in H-Wood / Gov’t (local/federal).

weroflu
Posts: 39
Joined: Wed Apr 19, 2017 9:19 am

Re: That1580/1510 mini mixer

Post by weroflu » Fri Jun 23, 2017 10:14 am

I spent some time testing the bluetooth to rfduino gain control. The following works:

All channels populating the chassis, rfduino connected to arduino serial monitor via uart cable. Usb powering the logic bus, all other rails unpowered.
gain settings printed on the serial monitor.

This also works: plugging in the power supply while usb is powering the logic bus, not sure which one has priority when both usb and on-board supply are providing 3.3v but it doesn't matter since it works. gain settings printed on serial monitor here too.

This one has me puzzled:

I left the uart cable plugged in so I can see all the gain settings on the serial monitor. Then plugged in a mic to every channel one by one and tested the gain settings. On channel 3 serial monitor output shows gain being incremented/decremented properly, confirmed since I'm reading from the 5171 gain register, yet actual volume is not increasing or decreasing. Signal is passing though, I think at a gain setting of 0 (5.6db)

Same thing happened on a different channel yesterday but it switched to channel 3.

User avatar
mediatechnology
Posts: 2922
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: That1580/1510 mini mixer

Post by mediatechnology » Fri Jun 23, 2017 10:57 am

Maybe one of the Rg connections on either the 5171 or 1570 are open.

BTW You helped inspire me to install Ubuntu Studio alongside an old XP installation.
I'm noticing myself gravitating away from the separate Win 7 machine I have and am doing more on the Ubuntu machine including this post.
I still have to keep XP running on the test bench but dual boot is really working out for me.

I have two Win 7 and two XP boxes and it looks like both XP will become dual boot.
I'm almost out of Win 7 licenses and Win 10's telemetry, forced updates and ads give me the creeps.
Linux seemed to make sense again.
Mack is naming names.
Big names in H-Wood / Gov’t (local/federal).

weroflu
Posts: 39
Joined: Wed Apr 19, 2017 9:19 am

Re: That1580/1510 mini mixer

Post by weroflu » Tue Jul 04, 2017 3:18 am

Probably you were correct about a floating Rg connection. These were all problems from bad qfn joints.
I retouched 3 more channel cards and it looks like they are all good.

There were 2 sources of noise from the rfduino still audible and one from ambient noise sources/power supply. While in ble advertising mode the rfduino presents a clicking sound correlating to the advertisement interval, and when the ble stack is on there is a rapid clicking. I had hoped using new code to keep rfduino in ulp (ultra low power) mode would take care of this, but it didn't, it just attenuated both of these noise sources.

There is another native function to power off the rfduino, well actually it doesn't power off, it just goes into a different? ulp mode, and this has taken care of both of these noise sources. I'm working on the arduino and android code now to trigger it and hopefully wake it back up. I made some errors in the pcb design of the spi card, but the tx and rx pins are available as gpio's (only need one) so this should work. Worst case though is I can trigger this systemoff function easily and then do a hard poweroff/poweron sequence if I need to change more gain settings.

I added a volume reset button that zeroes out all channels, and an rfduino reset function. Had to experiment a little with timing issues and put some delays in the android code as the volume reset was not working all the time - not a big deal.

So I'm getting close to a first attempt at putting this in the enclosure, which will hopefully help out with the third noise source. Not optimistic though since hoping rarely works. The power supply gets hot! with all channels plugged in, but it doesn't blow and it seems somewhat stable from the rudimentary testing.

edit: re linux:

I used puppy linux for a while, it was very fast, then switched to arch which is almost as fast but a lot better. Ubuntu was too bloated and slow last time I tried. I assume wine is no good for you? I tried to run everything natively in linux but for Ltspice and a few others I needed wine.

User avatar
mediatechnology
Posts: 2922
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: That1580/1510 mini mixer

Post by mediatechnology » Thu Jul 06, 2017 7:03 am

Do you suppose that the RF activity is being heard due to rectification of the RF envelope?
That's common with wired USB and GSM phones.
You might need some internal screening.

On the subject of Linux I tried Ubuntu a long time ago on older hardware and it was bloated.
At the time I settled on Xandros which is long gone. It was the most Windows-like at the time.

Recently I tried a live CD of Linux Mint and a live USB of Ubuntu.
I ended up installing Ubuntu Studio and 64 bit Linux running on 8 year-old hardware is faster than Win 7 64 bit running on 6 year-old hardware.
Ubuntu Studio 64 bit is also much faster than Win XP 32 bit running on the same hardware.
Since Ubuntu worked so well I stuck with it.

It took awhile to get Samba running with writeable shares: viewtopic.php?f=15&t=895
Now that I've solved that I'm going to explore Wine.
Mack is naming names.
Big names in H-Wood / Gov’t (local/federal).

weroflu
Posts: 39
Joined: Wed Apr 19, 2017 9:19 am

Re: That1580/1510 mini mixer

Post by weroflu » Thu Jul 06, 2017 7:43 am

I'm not sure I explained things well.

The two obvious noise artifacts that were being generated from the rfduino are solved by putting it in a ulp mode (that can be woken up). So these are now a moot point. I never implemented the usb connection to the phone. Noise is now the same with the rfduino card in as it is with it unplugged.

The next level of noise troubleshooting is either ambient noise, internal from the power supply, improper grounding, layout, something along the summing bus path, or the summing amp itself. I had previously tested single preamp channels, direct out into the adc, and things were very clean and quiet so that may rule out the power supply.

User avatar
mediatechnology
Posts: 2922
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: That1580/1510 mini mixer

Post by mediatechnology » Thu Jul 06, 2017 8:00 am

So after it wakes up from the alternate ULP mode it is still quiet?
Mack is naming names.
Big names in H-Wood / Gov’t (local/federal).

weroflu
Posts: 39
Joined: Wed Apr 19, 2017 9:19 am

Re: That1580/1510 mini mixer

Post by weroflu » Thu Jul 06, 2017 8:34 am

No, it will likely still be noisy again as soon as the ble stack starts up. But that's not the intended use. Once I set the gains, that's it, on to recording. If I want to tweak gains after an initial setup, I don't care if noise is present again since I won't be recording. For a professional product I realize this would be not acceptable but such is the luxury of diy. If someone somewhere wanted to run with the ball here there would be some work cut out to deal with this noise. I had early in the design phase anticipated this rf noise to be a possible dealbreaker but also knew there were these ulp modes on the rfduino that might work out.

That reminds me I switched to using zero crossing gain setting. At low gains there was not much noise present in a gain update but at higher gains there was quite a large noise pulse. Easy enough to do with 2 lines of code. And along the same lines as above it doesn't practically matter for my use case since the red light will not be on while doing any adjusting.

Post Reply

Who is online

Users browsing this forum: No registered users and 14 guests