NE570-based NR compander (P-522) by JR

Where we discuss new analog design ideas for Pro Audio and modern spins on vintage ones.
Crusty
Posts: 69
Joined: Mon Aug 20, 2007 1:42 am

Re: NR compander (P-522)

Post by Crusty »

Ah, that's cool! I had to think it over for a while to 'see' what happens when the signal shifts from AC=0. The extra bit showed up then too. LOL Are you coding in Microchip assembly?

As far as Terrasonde, your pricing is about right - maybe a liitle high-ish. It's a good box, but like I said geared toward the acoustics crowd. The test bench upgrade is very weak and a bit kludgy IMO but is usable. I'd have rather given you the money for something that was better suited to what I do with a better feature set...

My wish list (everyone, feel free to add to it):

Full AC/DC metering including dB scaling, with peak and relative modes (Terrasonde doesn't have these modes). At least dual-channel I/O; more would be even better. Or the ability to slave units together and have one master controller.

Distortion analyzer - must be real-time. Not sure how many harmonics are practical. Reading/displaying two channels simultaneously would be very cool.

Spectrum analyzer - 1/3 octave, beyond 20Khz.

Phase and crosstalk, and amplitude sweep.

Logging and automation would be a nice option too.

Signal generator - 20Hz beyond 20kHz as practical. Sine/square/pink/white. +25dB output level, low Z. Balanced inputs and outputs.

Other features might be a programmable pulse generator for continuous trains or bursts, like we talked about for testing dynamic processors. Maybe even a Wow-and-Flutter meter would be a useful add-on. Easily upgradable via USB or similar - all of these functions could downloadable modules for extra $$.

Unless there is something on the market that does these things? I haven't come acrossed it yet...
User avatar
JR.
Posts: 3709
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: NR compander (P-522)

Post by JR. »

It seems you could cover a lot of the features you asked about with a decent computer sound card and some software. Probably some freeware out there that does some of that.

I wanted to KISS with my next Test set,, basically cover the old one with a few new bells and whistles but not remotely replace bench gear.. For now at least I have talked myself out of pursuing this..

Yes, I am working with PICs... my old tuner is working on 8b platform, new one is 16b and that is a night and day difference. For a hand held test set, I could use a DSPic to do FFT et al. (but not just yet).

JR
Cancel the "cancel culture", do not support mob hatred.
Crusty
Posts: 69
Joined: Mon Aug 20, 2007 1:42 am

Re: NR compander (P-522)

Post by Crusty »

Yep, probably not worth the effort unless you consider adding it to the tuner as an option or bonus. It would make it that much more valuable to add to the engineer's toolkit. I think the producer/engineer crowd would be a good additional market for a drum tuner, especially with the typical micro record budgets nowdays where a drum tech hanging out for the whole session is out of the question. Having a tuner would give the average engineer with no tuning experience a bit of an edge at freshening the drums up after a few takes. And engineers are are far more concerned with drum tuning than many drummers are. I don't mean that in a bad way either. Drummers are listening more to their performance a lot of times as opposed to tuning.
Reading the forums, it sounds like people are still recording drums on a regular basis. And the secret to recording good drums is simply a well-tuned kit with a guy who knows how (and when) to hit them properly. LOL
JR. wrote:It seems you could cover a lot of the features you asked about with a decent computer sound card and some software. Probably some freeware out there that does some of that.

JR
Yes, I think so - for a start, this looks really good for under a 100 bucks: http://www.trueaudio.com/index.htm Doesn't read distortion, but I guess I can figure it from the RTA trace...
A decent USB interface for about $350: http://www.m-audio.com/products/en_us/F ... Ultra.html
And a little computer for about $400: http://www.dell.com/content/products/pr ... l=en&s=dhs

Less than a grand, not bad(!). I could probably do some simple automation too, which would be a plus.
But my desoldering station is struggling with the lead-free solder, so that will probably be next on my list to upgrade...

Are you assembling the tuners yourself, or are you parting out that gig?
User avatar
JR.
Posts: 3709
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: NR compander (P-522)

Post by JR. »

I already have studio owners buying my tuners so they can get consistent results.. even a well tuned drum. tuned differently won't sound the same for over dubs or projects with time interruptions.

I am working to make my next generation tuner cheaper and even easier to use.. without getting ahead of myself, a future high end model might justify a LCD/graphics display and higher performance codec. so never say never...but not this revision
----
In the current model I used a contract manufacturer to fabricate and assemble the two PCBs inside, I had Alcoa make the aluminum extrusion, other companies make; the plastic (end) extrusion, steel bottom grille, and photo printed face plate. I personally do all the secondary operations, like cut aluminum extrusion to length, cut and drill holes in plastic end parts, solder wires to the two speakers, then program the microprocessor and assemble the sundry parts. I have made many changes to the software since I started building these, and several small changes to the processor board. Doing rework on small SMT PCBs is not my idea of fun, burt better than re-running them.

My plan for the next version is to probably invest a little more in an IM case, and get a contract manufacturer to do 100% assembly. I can even get one to drop ship the products to customers so I don't have to do as much day to day work, or manage employees. I am frankly less enthusiastic in my marketing since sales success means more work for me. :roll:

I was pleasantly surprised by how well I could move 2D cad results between different PC based design programs to compare and integrate results. The last time I ran a one man shop back in the '80s it often took multiple prototype runs to get parts to fit together adequately.

JR
Cancel the "cancel culture", do not support mob hatred.
User avatar
mediatechnology
Posts: 5473
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: NR compander (P-522)

Post by mediatechnology »

JR. wrote:It seems you could cover a lot of the features you asked about with a decent computer sound card and some software. Probably some freeware out there that does some of that.
Crusty -

Take a look at Visual Analyser for basic tests. http://www.sillanumsoft.org/

And RMAA: http://audio.rightmark.org/products/rmaa.shtml

Both are free. I haven't tried the VA 2009 HR version (just downloaded it) but do like 10.5. There's a lot it can't do- but hey it's free - and what it does do it does very well. There are VA and RMAA results posted over in the M/S thread.

Wayne
Crusty
Posts: 69
Joined: Mon Aug 20, 2007 1:42 am

Re: NR compander (P-522)

Post by Crusty »

I can see using the tuner in recording school curriculum too.

Is the extrusion custom? How horrible are the die costs/minimums?

What sort of final test jig do you use to test assembled units? Is it some sort of electro-mechanical setup?

I agree about the CAD stuff, it's very satisfiying to have it all fit properly the first time. LOL I'd really like to learn Solidworks someday...

Wayne, those are some good options too. The Rightmark would be good for me if it was real-time. Keep your eye out for some good sound card options for me!
User avatar
mediatechnology
Posts: 5473
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: NR compander (P-522)

Post by mediatechnology »

I use an EMU-0404 internal.

Apparently Asus Xonar D2's are testing well at RMAA but I've never found the summary for it.

The product "specification" page is here: http://usa.asus.com/products.aspx?l1=25 ... odelmenu=2

This is the chipset information:

24-bit D-A Converter of Digital Sources:
Burr-Brown PCM1796 *4 (123dB SNR, Max. 192kHz/24bit)

24-bit A-D Converter for Analog Inputs:
Cirrus-Logic CS5381* 1 (120dB SNR, Max. 192kHz/24bit)

Visual Analyser is realtime. RMAA has a realtime FFT display during level set that is sort of useful. For realtime I use VA however. The nice thing about RMAA is the summary.

(John - Sorry to hi-jack this thread. It's going in about 3 directions. I'm still in awe at the simplicity of the noise reduction.)
Crusty
Posts: 69
Joined: Mon Aug 20, 2007 1:42 am

Re: NR compander (P-522)

Post by Crusty »

Well I sent the thread in more than three directions - at least four in my last post. And John did mention the TS1 and maybe doing another test set so...

Next time I'll start a thread called "Ask JR" as a catch-all for my follow up off-topic questions. LOL
User avatar
JR.
Posts: 3709
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: NR compander (P-522)

Post by JR. »

Crusty wrote:I can see using the tuner in recording school curriculum too.
If I was smart I'd give some away to schools... They are always asking me for free ones to give away as prizes, not to use themselves, so I ignore them... and everybody else bugging me for free units.
Is the extrusion custom? How horrible are the die costs/minimums?
The aluminum extrusion die wasn't that bad (<$1k) but minimum cost effective aluminum push was 500#. In hindsight I wanted my tuner to have some heft to feel substantial, but I wasn't counting the weight of the battery and two loudspeaker magnets, so I ended up weighing more than I wanted or need to. I could have designed a thinner profile for the aluminum extrusion but then I'd have even more length of it, since they'd push the same 500#.

The plastic extrusion tool was maybe 3x the aluminum, but I was able to run less of it per lot.

I expect tooling for my next version to be several times higher (injection molded) but part cost, assembly, and everything else will be cheaper, and at a lower price point I will sell more.

What sort of final test jig do you use to test assembled units? Is it some sort of electro-mechanical setup?
I pre test the speakers with a clean sine wave, and that the processor is functional when I program it. After I assemble the tuners I run an auto-calibration sequence to tweak the two microphones to read exactly the same, then confirm by running scans on three different sized drums.

But still more work than I'd like to do.. The PICs I use had extra A/D inputs so I designed some for self trouble shooting but I find the manufacturing of PCBs to be quite good... and never needed to use these.

JR
Cancel the "cancel culture", do not support mob hatred.
Crusty
Posts: 69
Joined: Mon Aug 20, 2007 1:42 am

Re: NR compander (P-522)

Post by Crusty »

Very cool. Selling by-the-pound makes sense. I imagine the upfront costs for the molded case will be pretty hefty, even with modern 3D CAD...
Have you talked to any contract manufacters yet? Someone told me that as rule-of-thumb, a turn-key contract manufacturer wants 50% of list price for building, testing, packaging, and shipping to a distributor. Does this sound about right?
Post Reply