Cheap Surround sound buffer/level shifter

Where we discuss new analog design ideas for Pro Audio and modern spins on vintage ones.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Cheap Surround sound buffer/level shifter

Post by JR. »

I decided to move this here as I soak this design in beer at night.

To present the design brief, and catch up everybody on current status,,, First I am a packrat so had enough old speakers sitting around to support a 6 channel system, front L&R, rear L&R, dialog center, and sub... I had to recone the woofers in one 3 way 8" pair of AMR studio monitors, and the woofers in my old 2 way 6" hone hifi speakers. The recone kits were relatively painless. The woofer surround remained intact.

For amps I started out with 3x cheap stereo amps. I miss the light show from the meters but not the mediocre sound. I have long been a fan of Bruno Putzey's class D amps (Hypex) so bit the bullet and purchased a 4 amplifier module and a 2 amplifier module.

I am cheap so went through a few, too cheap to work digital surround decoders. First upgrade was to one with a master volume control, recent upgrade even has a remote volume control.

I had some drama getting the surround channels working but a helpful customer service guy kept nudging me toward the solutions. Both my ROKU TV monitor and the Direct TV satellite had sound settings that affected operation. After that was sorted next problem was a low level hum/buzz that the previous digital decoder did not suffer.

I resolved the low level noise by powering the digital decoder from a 5V supply in the amp (a nice feature).

This leaves one last functional improvement (driving amps to full level), and one feature improvement (more blinky lights). I miss the meters from the cheap stereo amps.

The Hypex amp requires 2.5V input to realize full scale output. This is a little problematic as the cheap digital surround decoders running from 5V wall warts make 1V max output cleanly. Adding a simple op amp buffer with a second inverted differential output gives me a quick +6dB so almost there (2.5V) and reasonable from 5V rail.

I suspect I could grab enough rail voltage from the amplifier power supplies to generate a regulated +/- 15V but that is more effort than justified to make 2.5V. Alternately I could repurpose the now surplus 5V wall wart to provide a -5V rail... Old school op amps like i have laying around in my back lab are still challenged about swinging close to rails.

OTOH modern CMOS op amps easily git er dun with a single 5V rail.

The cons of using a single 5V rail is that I need to bias the audio up to 2.5V and cap couple in and out. The pros from +/-5V rails is I could DC couple inputs and outputs.

I am leaning toward KISS with just one single 5V rail. I am pretty confident I won't hear a few extra capacitors in series. I am leaning toward the opa 2156 TI cmos amp, not cheap but for a one off project why skimp?

Before I resolved the noise floor issues I was leaning toward a balanced or at least a differential stage between the decoder and amp grounds. Now that they are the same ground I can eliminate that stage. I am thinking of using pads on every decoder output to balance levels, while in over a year of use I have never needed to trim level balances yet.

==

Next design decision is about the blinky lights. I have already decided against full meters (like on the old hifi amps), further the only one 5V rail makes stacking LEDs in series ineffective. At a minimum I want one LED for signal resents, say down around -20FS... Then a second LED up around -3dB FS could indicate head room.

After too many beers I have thought about a third LED, a variant on my old FLS invention (The LEDs that light above the EQ slider). I could have a third, middle LED that illuminates for the loudest of the 6 channels. So Signal present and headroom for all channels and a loudest channel LED that moves around.

Any thoughts or suggestions?

JR
Cancel the "cancel culture", do not support mob hatred.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Cheap Surround sound buffer/level shifter

Post by JR. »

The plan is coming together... not ideal, but pretty damn close.

Running pairs of op amps (inverting) biased up to V/2 with a cross-connected feedback R from the hot input to the other gets me an easy +6dB with some CMRR that I don't seem to need. Adding the extra 2 dB of voltage gain to get me to amplifier FS only degrades the CMRR (that I don't need) a little.

Meter strategy is coming together.... 6x signal present LEDs at something like -20dB or more. 4x headroom LEDs at -3dB, one for center, one for sub, one shared for front L/R and another shared for rear L/R. 6x intermediate level LEDs somewhere between -20dB and -3dB... If I get really clever I may let the threshold for this intermediate level LED float following an average of all channels. I'm still soaking this idea in beer.

JR

PS; Making the system 6dB louder (4x the power) is probably more than I need... but the engineer in me says "give me more power scotty". :lol:
Cancel the "cancel culture", do not support mob hatred.
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Cheap Surround sound buffer/level shifter

Post by mediatechnology »

I am leaning toward the opa 2156 TI cmos amp, not cheap but for a one off project why skimp?
I'm glad that you now get the one-off concept of this and other DIY fora rather than falling into the mass-produced lowest BOM manufacturer mindset.
Combining that with analog blinky lights I see a real breakthrough.

IIRC TI had some low-voltage direct-coupled balanced line drivers that use an internal charge pump for the negative rail.
They are designed for home theater.
You might want to have a look in the TI online selector guide.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Cheap Surround sound buffer/level shifter

Post by JR. »

mediatechnology wrote: Thu Jul 30, 2020 7:14 am
I am leaning toward the opa 2156 TI cmos amp, not cheap but for a one off project why skimp?
I'm glad that you now get the one-off concept of this and other DIY fora rather than falling into the mass-produced lowest BOM manufacturer mindset.
I have made more one-off projects than mass production, this century. That said my philosophy hasn't changed about not using parts that are better than needed, and the 2156 is arguably better than I need with 30+ rail voltage, but the rail to rail 5V performance is very nice, marginally better than 6V process cmos op amps that don't suck.

For a higher volume design I would probably build up a few prototypes to vet them, but for one-off not worth the effort to save a few dollars. There are lots of op amps with that same foot print so i could, but nah... :lol:
Combining that with analog blinky lights I see a real breakthrough.
I have been making blinky lights, including with 339 quad comparators for almost 50 years.
IIRC TI had some low-voltage direct-coupled balanced line drivers that use an internal charge pump for the negative rail.
They are designed for home theater.
You might want to have a look in the TI online selector guide.
Thanks, I checked them out. Interesting to add a charge pump cap to replace a coupling cap. :o

I have seen the onboard charge pumps built into digital interface chips to get some extra voltage swing. If inclined to go that way a single charge pump chip could generate a -5V rail, but I already dismissed the even simpler hanging a 5V wall wart to make a second rail... KISS.

I need to scratch out a schematic to share my current thinking. I suspect operating the op amps inverting biased up at 2.5V will reduce input stage issues. Differential output means I don't need full rail to rail output swing. But 4.5V swing from 5V rail is not typical for jelly bean GP op amps.

JR
Cancel the "cancel culture", do not support mob hatred.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Cheap Surround sound buffer/level shifter

Post by JR. »

5vbuffer.jpg
This is rough first pass at my 5V buffer...

at top DECFR+out is decoder front right + output feeding top of attenuator. DECFR-out is decoder RCA jack ground. This RCA jack ground is bonded to PS ground through power supply.

U4A is inverting with gain of 1.33x. It's output is back fed into U4B creating a unity gain inverted version of that 1.33x for 2.66x total or just over 2.5x needed to drive amp to FS.

R39 might deliver better CMRR if lowered to 7.5k, but alternately could be eliminated and circuit would still deliver FS voltage gain.

The FR-in and FR+in are feeds to the hypex amp input.

JR
Cancel the "cancel culture", do not support mob hatred.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Cheap Surround sound buffer/level shifter

Post by JR. »

After soaking it in beer last night I think I have come up with a simpler plan for my blinky lights that does not even involve LM339s.

A compound 6 NPN transistor long tail pair (long tail sextet?) with bases connected to each audio channel op amp output. The highest voltage channel steers current to light one LED at a time connected to each one of the six collectors. This is a variation on my FLS invention I did back at Peavey that lit the LEDs above the loudest GEQ channel.

In addition a 2 PNP LTP with one base connected to this common emitter (that represents highest channel output), and the other base to an overload DC voltage reference steers current between a green normal "ON" LED and a red Overload LED...say 2 or 3 dB before amp clipping.

This is very easy on current draw and simple... Diodes and hold caps on each base make a more stable display.

I can add a 7th NPN to the long tail sextet biased just on, to stop hunting around during quiet passages... I can design this in and then populate or not depending on how it looks. Bulk transistors are not very well matched for Vbe, so may not look pretty with no/low input.

JR
Cancel the "cancel culture", do not support mob hatred.
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Cheap Surround sound buffer/level shifter

Post by mediatechnology »

Sounds/looks similar to the Moog Ladder Filter concept: https://proaudiodesignforum.com/forum/p ... 148&p=1214
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Cheap Surround sound buffer/level shifter

Post by JR. »

mediatechnology wrote: Fri Jul 31, 2020 11:47 am Sounds/looks similar to the Moog Ladder Filter concept: https://proaudiodesignforum.com/forum/p ... 148&p=1214
Similarity perhaps in number of transistor involved. Think 6 or 7 transistors with all the emitters connected together to same one node. Highest base steers current to it's collector and LED.

I like discrete design especially when you can do more with less...

JR

PS I'll add to schematic later
Cancel the "cancel culture", do not support mob hatred.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Cheap Surround sound buffer/level shifter

Post by JR. »

Damn the dog ate my homework....

late last week while out cutting my rain ditch I heard the loud report of the fuse blowing on the power pole across the street... It sounds like a gun shot, but I didn't need electricity to cut grass..

Unfortunately my PCB/schematic design session evaporated when the power went away. Apparently the auto-save function only works after the first session has been saved.

I didn't lose much, but start over yet again.

JR
Cancel the "cancel culture", do not support mob hatred.
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Cheap Surround sound buffer/level shifter

Post by mediatechnology »

Sounds like the work of a squirrel.
In my old neighborhood squirrels were constantly coming into contact with primary and blowing fuses.
Why that doesn't happen here I don't know - we have just as many trees.

BTW most utilities now are moving to MOV-based lightning arrestors which have way more leakage current.
I'd be very cautious touching the copper drain wire running down the pole.
Whereas they used to cut the drain wire on poles without equipment I think they're now bonding all of them on new construction.
In Dallas many of our air switches are now under remote RF control.
I'm seeing the pop up everywhere.

How long did it take for them to replace the fuse?
Post Reply