Texar Prism clone prototype Finished Version

[flyboy71]

One of my odd hobbies is to recreate or attempt to recreate vintage designs to see how much I can simplify them or take a different spin on an idea. I have a pair of Texar Prisms that I have been poking around in and trying to figure out what makes them tick. Since my units have aged out of spec and they aren't a matched pair they are more of a novelty at this point. The heart of the Prism, the M101 board, is really just a few comparators that toggle at preset audio levels, a linear ramp and hold generator, a log generator and a handful of digital gates and an LM3914 to drive the display. Based on that I was able to use a microcontroller, a display driver, an analog switch and an opamp and emulate this with a handful of parts. Since microcontrollers can output PWM easily and have an ADC built in, I took a PWM approach on this. The duty cycle vs attenuation is handled in a lookup table on the chip. Sure I could do this with DSP but thats boring plus this is fun just to sit and watch. I also used this as an excuse to get my surface mount setup up and running and needed a reason to try something. It a work in progress as I only have 3 of the 4 bands running and am working out details in the code but so far its about 90% there for something nobody really needs these days.

[mediatechnology]

That's very cool!

People will be very interested in this.

[flyboy71]

Its still in an early stage; I just populated these boards today and whipped it together so there are a few gremlins to work out like some distortion but I have that with my analog prisms also so I suspect my setup isn't the best yet. I need a good audio DAC and some decent source material or dig out some old vinyl and see how it works. I'm messing with the source code now to optimize the time constants in code. My approach was different than what Glen used on the prism but I don't expect any field repairs. Each board is selected for the band and the audio is on board so there's no board swapping here. This was done to cut down on the number and size of boards and cost. I also moved input and output pots to the boards to adjust each one individually so the level of compression can be better distributed. With all the ground plane I'm not hearing discernable digital noise which I had on the breadboard version. I have some video of it but cant post the extension plus its large but the still pic doesn't do it justice.

[mediatechnology]

Jeff sent me this video of the Texar Audio Prism Clone with three bands operating.

Looking forward to seeing the 4th band and hearing the result!

Video: https://proaudiodesignforum.com/content ... yboy71.mp4


A DSP-based Texar Audio Prism Clone from flyboy71

[flyboy71]

The 4 bands and both channels are up and running. Its comparable to the analog units in producing loudness and presence but one drawback is the ability to hard limit peaks. I will continue tweaking the design as time permits. Im actually surprised it performs as well as it does considering it is a minimal approach to emulating the Texar core processing board.

[flyboy71]

I had some time to mess with the design and a few code tweaks, fixed a bug and also changed the method I was using. It now peak limits quite well and the overall RMS level was greatly improved. The core functionality tracks alongside my analog prisms so Im fairly confident I can move this on to a final version now that is more DIY friendly.

[mediatechnology]

Awesome!
I like what I've heard so far.
The PWM gain control works really, really well.

[emrr]

very cool

[mediatechnology]

flyboy71 processed Sting's version of Little wing with the Texar Prism clone and sent it to me.

I chose this song and sent him the original because it has a very high dynamic range.
I'm really stoked about the results.

The original unprocessed version: https://proaudiodesignforum.com/content ... cessed.wav

The processed Texar Prism clone version: https://proaudiodesignforum.com/content ... _Clone.wav

The processed version has about 10 dB more power with essentially the same peak level.

Cool Edit Statistics



Larger version: https://proaudiodesignforum.com/images/ ... _Clone.JPG

MasVis Analysis:



The Texar Prism clone uses PWM gain control so there's no hard-to-find Vactrols.

[emrr]

Nice results