That1580 mini mixer

Where we discuss new analog design ideas for Pro Audio and modern spins on vintage ones.
weroflu
Posts: 56
Joined: Wed Apr 19, 2017 9:19 am

Re: That1580/1510 mini mixer

Post by weroflu »

Scoped the differential out on the 1580... only one side is showing signal.

Now to find out if the problem is on the input or output, though there's nothing going on at the ouput aside from two traces and two vias.

Also reflowed a new preamp card, same problem as the previous one.

Edit: found something.

I left out the .1u cap between Vcc and Vee on the 1580. Could that have been the culprit?

Edit: Ok got it working. There were a few issues. I could have sworn that I had the probes on 10x when I did the first measurement but one was on 1x which could have explained one differential leg not showing up, but it would not explain bad audio when not using the scope. Also I might have not tacked on one of the wires properly to an ic lead for the measurements.

The main thing which was most likely the cause is I reflowed the 1580 from the underside pad with an iron and retouched the perimeter with flux and iron.

Scope showed good differential signal and it passed the ear test too. Dynamic and phantom mics both working, and suprisingly very quiet, but I bypassed summing section for now.

I'll sit on this for a day, try to fix the other cards in the card graveyard. Got to think about the suggestions for virtual earth and panning. If the boards in hand work, and summing is not so noisy and crosstalk isn't horrible it might be best to keep this version as is.

No one answered the question a few posts ago about using ltc3265 charge pump for the rails vs inductor boost switcher.
weroflu
Posts: 56
Joined: Wed Apr 19, 2017 9:19 am

Re: That1580/1510 mini mixer

Post by weroflu »

Update: something innocuous like testing my noob summing card turned into a disaster. I had a bridge on one of the backplane connecters from 48v to left out. Lots of chips got real hot and died, no smoke though. Note to self, don't put 48v bus rail next to left/right outputs.

I must have also fried the tps7a47 regulator because it was only outputting 6v then -.7v, so I just pulled it, replaced it and + rail is good. I soldered up a new summing card and it worked. So now I have one working preamp card, summing section working.

I dead bugged a drv134 https://1580mixer.shutterfly.com/pictures/29
and will be running sweep tones through the preamp card tomorrow.

I think it was a good idea to move ahead on the current flawed design because it's forcing some of these errors to the light. I don't know if I'll finish it as it stands but probably make at least 1-3 more cards to facilitate more testing.

Also realized that lowering the rail voltage is completely self-contained on the p/s card so that can be done at any time. When that happens rails will be +-8v, with a 9VDC nominal input, dc converted to +-8.75 (any better suggestions here?), then regulated to +-8. I don't know how close the 1580 runs to the power rails. It will be a little bit of a challenge squeezing some of the extra passives around the regulators when this happens instead of using a fixed -15v reg and the anyouts on the tps7a47.

There is some low frequency clicking going on when the summing card is plugged in, it sits very close to the dc converter, so that needs to be looked at but post panning output from the preamp card sounds good and the master output sounds good too aside from that clicking. I thought originally it was from the rfduino card but that was wrong. Speaking of that I need to develop that android app more, I have a skeleton working but no bluetooth function. I was thinking of taking the easy way out with this card - if there is too much noise I'll just put in a hard poweroff switch for it, that can also act as a reset. Once levels are set there is no reason for that card to be powered.

LL1527 balancing transformer is on the way so I'll pick either that or the drv134 card as my signal generator balancer for testing.
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: That1580/1510 mini mixer

Post by mediatechnology »

Note to self, don't put 48v bus rail next to left/right outputs.
I think I said something similar when I noticed the +48V by the 3.3V on the PSU card.
Sorry you had to do all that extra repair work.
weroflu
Posts: 56
Joined: Wed Apr 19, 2017 9:19 am

Re: That1580/1510 mini mixer

Post by weroflu »

Picoscope signal generator 20hz-20khz sweep, into drv134, into preamp card, output post pan/pre summing into scope

Still getting frequency dependent gain unless i'm not measuring correctly. Gain is set to 30db, which is what I see here from about 600hz-20khz.
Attachments
sweep.png
sweep.png (111.39 KiB) Viewed 13236 times
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: That1580/1510 mini mixer

Post by mediatechnology »

Let's see if I understand correctly.

You have an unbalanced PicoScope output feeding a DRV134 which then feeds the balanced mic input with gain.

If so, where is the pad/attenuation?
If the attenuation is internal to the PicoScope how is that performed?
Does it have an output attenuator or is the DAC input value being scaled?
You could be amplifying and measuring DAC noise.

What you want is Picoscope > DRV134 > Pad > DUT.
Drive the PicoScope output near FS to overcome its noise floor.
Pad the DRV134 output by 20, 40 , 60 dB depending on your DUT gain.
weroflu
Posts: 56
Joined: Wed Apr 19, 2017 9:19 am

Re: That1580/1510 mini mixer

Post by weroflu »

<<You have an unbalanced PicoScope output feeding a DRV134 which then feeds the balanced mic input with gain.>>

Correct. I think attenuation is not necessary because the sig gen output range is 1mV - 2V.

The output of the preamp is connected to 10x probes which attenuate the signal by a factor of 10. ie with 1V output fed into 10x probes, the pico sees .1V.

I can post more plots, but I also tried feeding single tones starting at about 400hz. The results were the same, higher frequency tones were amplified more.
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: That1580/1510 mini mixer

Post by mediatechnology »

I think attenuation is not necessary because the sig gen output range is 1mV - 2V.
But how does the PicoScope produce an attenuated 1 mV output?
Does it scale the DAC code or does it attenuate a DAC FS output?

I can set a sound card to generate a 1 mV output but I'm listening to it with a constant noise floor.
At FS I may have a 100 dB DR but at -60 dB output I have a 40 dB DR.

In your original plot it looked like I was seeing rising HF noise floor.
weroflu
Posts: 56
Joined: Wed Apr 19, 2017 9:19 am

Re: That1580/1510 mini mixer

Post by weroflu »

I bypassed the preamp and just fed the picoscope sig gen sweep back to 10x probes. First file is Y axis set to average. Second file is Y axis set to peak hold.

I don't know the answer to how does p.s. produce attenuated 1mV signal.
Attachments
sweep.average.png
sweep.average.png (117.54 KiB) Viewed 13234 times
sweep.peak.hold.png
sweep.peak.hold.png (117.57 KiB) Viewed 13234 times
User avatar
mediatechnology
Posts: 5442
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: That1580/1510 mini mixer

Post by mediatechnology »

It looks like to me that you need a pad at the output of the DRV134 and run the PicoScope out near FS.

Look at how the noise floor rises 20 dB from 1-20 kHz in loopback.
The ARB/Function generator is only 12 bit.
You're in the noise floor.

I checked Pico's docs and they make no mention of an output attenuator but they do say the Zout is 600R which is a little high to drive a DRV134 well.
If you don't want to add a buffer so the DRV134 sees 0Ω you could open the sense lines of the DRV134 to turn off it's cross-coupled output.
If you do however the DRV134 gain may change from 2X to 4X so that may work against you further.
weroflu
Posts: 56
Joined: Wed Apr 19, 2017 9:19 am

Re: That1580/1510 mini mixer

Post by weroflu »

It looks like to me that you need a pad at the output of the DRV134 and run the PicoScope out near FS.
Run picoscope sign gen full scale (2v) to increase s/n ratio ?

Already too hot for the mic input, but add in gain of 2 from the drv134 and it needs a pad.

600ohm output of the sig gen into 10k input on the drv134... not sure if this is much of a problem for basic testing?

What would you suggest for the easiest way to run tones into the preamp cards from the picoscope sig gen? I'm not married to using the pico signal generator.

I want to make a proper test board for drv134 so I could put a pad on there easily. Or maybe this is a better solution, but I'll still need a pad if using a mic level signal from the sig gen is a no go.

http://www.lundahl.se/the-magic-1527/

I think it's 150ohm output wired 2:1.
Post Reply