A Modern Loftech TS-1 Audio Test Set

[pascal.verdet]

Maybe I will have to tell you exactly how to do this since I have only been thinking about this for more than 3 decades.

Why not ? There is more idea in two heads

[mediatechnology]

Maybe I will have to tell you exactly how to do this since I have only been thinking about this for more than 3 decades.

Why not ? There is more idea in two heads

Do it!

[JR.]

Maybe I will have to tell you exactly how to do this since I have only been thinking about this for more than 3 decades.

Why not ? There is more idea in two heads

Damn I just wrote a long detailed answer and the web gods ate it..... arghhhhhhhhhhhhhhh

I will try again later...

Hint: check out the DSPic that has a stereo 16b DAC built in... you can make nice sine waves with a lookup table (I know because I have done it).

More later

JR

[JR.]

OK before I forget what I wrote,, I will break this down into sections.

#1 sine wave... As I already suggested you can make a clean 16b sine wave with built in stereo DAC, 17b if you sum the two outputs together.

For higher performance you can pass the DAC output through a tracking narrow bandpass filter (opposite of notch filter) to reduce noise floor and low level distortion while this may reach diminishing returns quickly.

At some point design and layout may limit ultimate performance.

An obvious benefit from using DAC a look up table is you can make any waveform you can imagine... speaker testers like warble tones, you could also cover a function generate triangle waves, etc.

FWIW, I had a prototype next generation drum tuner using the DAC to simultaneously make 10+ different pitch sine waves (then used a FFT to read for drum resonances... it worked). When summing that many sine waves the amplitude must be scaled down to prevent saturating the DAC.

BTW a DAC sinewave has zero settling time, common to low distortion analog sine wave generation. An issue in the TS-1 where i could have easily delivered lower distortion but the longer setting time was unacceptable, IMO.

JR

[JR.]

#2 dB meter
Again I would embrace the digital technology for this.

The standard 12b A/D built into PICs in combination with front end gain ranging could easily capture steady state measurements over a wide dynamic range. The math to convert linear levels to dB is not hard for even a RISC micro. I would use DPOTs for front end gain ranging... Just like old school bench equipment used a range switch, this range switch could be automatically controlled by micro.

For better capture range a 16b codec can be had in the $1 cost range (last time I looked years ago).

I already mentioned the attractiveness in using a 2:1 compressor in front of an analog capture conversion. The THAT corp app note is a kind of rehash of a product that dbx sold decades ago,,, they used their rms and vca in front of a small mechanical meter to read and display a wide dB range. Years later Paul Buff did his own version (with his VCA no doubt).. both are now out of production which may tell us something about this market.

I would need to invest some bench time to see how good i can do with DPOT in front of 12B internal A/D, but external 16b codec is cheap if needed.

Once again design and layout of this front end will matter to the resolution floor. Inside bench quality test equipment the details matter.

JR

[JR.]

#3 control interface...

In my oldest scribbling I visualized a numeric keypad for frequency selection (similar to HP 8903). The frequency knob could be replaced by a rotary encoder with smart scale ranging... First turn CW is fast, turning back CCW is slower, and so on every time you reverse direction, to help zero in on target frequency with less overshoot.

Display could be semi standard numeric LCD.....

---------
Rethinking this decades later the display/control interface could be a graphic LCD with soft button touch screen overlay (outputs x-y matrix info). The touch screen could literally double as soft control buttons where the interface is only limited by your imagination to construct drill down menus. (I have one such display interface sitting untouched. interface design is reasonably straightforward with micro).

Also hip would be a smart phone interface, where some of the display/control cost could be off loaded to the phone. I do not have the chops to design a smart phone interface but it would be slick. If i was younger I would be cultivating that capability. I can imagine using smart phone interface for all kinds of products.

JR

[mediatechnology]

I'm starting to be reminded of the NTI products: http://www.nti-audio.com

Maybe its just me but I hate the thought of a Smartphone interface due to fat-finger.

It would be interesting to put the TS-1 beside an NTI and compare the amount of time it takes users to get "800 Hz at -23 dBu" or some other oddball combo into a load. I'm thinking the TS-1's "analog" user interface would get me there quicker. The single scroll wheel selecting level and function bothers me.

Not sure what Pascal's goal is but having a modern board in a TS-1 style box with big bright LEDs seems appealing to me. I've never had interest in the NTI stuff for analog.

If one did go whole hog DSP for tone generation and analysis I'd put a mic and low bit rate recorder in it to produce voice circuit-identifier messages. If you're in a broadcast TOC and you've got tone coming at you from 14 different sources its hard to identify where its coming from or if you have the right circuit. The generator could announce its circuit ID and location. A small monitor amp with miniature internal speaker tied to the input would also be useful at the far end of that tone.

I did a product I called "The Identifier" using ISD chips that alternated between message and tone. It ran for weeks on 4X AA batteries and was low-cost enough to be disposable. It read out the circuit ID, provided reference level and let other people know that circuit was in-use.

[pascal.verdet]

Hi,
In fact, I want just to have a modern test bench, to replace very old, gen, freq, and voltmeter.
I'm looking around the PIC, because with a handful of components, you can get the voltmeter,
the frequency counter the dB meter, and cherry on the cake, you can also directly generate a frequency ???
or drive an external frequency generator,
The AD9835 allows the user to make sine waves from 1Hz up to 25MHz with better than 1Hz resolution.
With which, we will certainly have a "cleaner" signal
Cheers,

[pascal.verdet]

Oops,
Continuing my research ... I realize that we can do all this with a good sound card, and the right software...

[mediatechnology]

I use both analog instruments and sound card-based tools.

Software: viewtopic.php?f=15&t=879
PCM4222 A/D for T&M Review: viewtopic.php?f=6&t=886

My most frequent goto equipment is a Kronhite-1600, Heath IG-18, Leader scope and AudioTester software.