Fiber optic audio transmission for SSTCs / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Fiber optic audio transmission for SSTCs

Move Thread

LAN_403

uzzors2k

Fri Oct 17 2008, 06:45PM

Registered Member #95 Joined: Thu Feb 09 2006, 04:57PM
Location: Norway
Posts: 1308

After the wave of instrument modulated SSTCs that have shown up lately I thought I would make my own. Since I plan on using both a keyboard and an ipod with it, I thought fibre optic would be safest. I've spotted out a pair units I think will work, the TOTX173A(F) transmitter and TORX173 receiver. The transmitter being a (F) won't cause problems will it? The SSTC only has interrupter control, ei duty cycle. So modulation will have to be class D.

Now I thought I could make a small class D converter, just something simple to create PWM from the audio. The PWM could then be fed directly into the optical link, which controls some enable pin in the SSTC driver. Since the receiver is just connected to the enable of the driver, this would also allow me to use a standard interrupter if desired. Sound good so far? Now my question, this transmitter/receiver pair is specified for 6Mb/s maximum in the datasheet. The audio PWM I feed into the link will vary between near 100% and near 0% duty. What is the highest frequency I can use for the PWM with reasonable distortion on the audio? I'm no audiophile, and the SSTC will distort plenty I assume.

aonomus

Fri Oct 17 2008, 10:19PM

Registered Member #1497 Joined: Thu May 22 2008, 05:24AM
Location: Toronto, Ontario, Canada
Posts: 801

Another thing to keep an eye out for is if the Tx/Rx goes down to DC (ie: if you want CW, or the Tx percieves high duty cycle as DC, or if you want to shut it down). Apparently some transmitters will spit out random noise with no input.

Also, from

the optoisolator used was HFBR1412T by Avago.

Another thing that you may want to consider is how you plan to shield the connector on the reciever end in particular, since the receivers appear vulnerable to EMI, many people have either used panel-mount ST connectors with metal shielding, or just wrapped the connector/cable assembly in tin foil and connected to ground.

Granted this doesn't answer your question.... d'oh.

Steve Conner

Sat Oct 18 2008, 09:53AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Hi uzzors,

I always wanted to try this

You see people building simple self-oscillating Class-D amps all the time, and I think you could use one of these as your Class-D converter. They're technically known as "Sigma-delta modulators" AFAIK. You could just connect the fibre optic TX in place of the power stage, unless the output filter was part of the feedback loop in the design you picked.

Your question about frequency and duty cycle is more involved than it seems. The easy answer is about 1MHz

You'll struggle to find a commercial PWM chip that can work this fast, hence why I suggested the sigma-delta approach. Someone like HFSSTC-Freak may be able to give more info.

If you want a more comprehensive answer, you can take the receiver's spec for pulse width distortion, use that with the frequency to estimate % distortion of the audio signal, and choose a maximum frequency based on some maximum level of distortion that you decided you can put up with.

I used the HFBR series things for my Tesla Guitar setup, with a 10 meter ST-ST patch cable from a computer store (they're used in fiber optic networking) I chose them over Toslink because they have metal housings, and the glass fiber lets more signal through than the plastic stuff used in Toslink. Also the 10 meter cable was cheaper than a 10m Toslink.

uzzors2k

Sat Oct 18 2008, 01:42PM

Registered Member #95 Joined: Thu Feb 09 2006, 04:57PM
Location: Norway
Posts: 1308

Hi Steve and Aonomus,

I've used RF modules before and had problems with noise during CW or no input, and it's areal pain. I just saw now that the TOSLINKs aren't any good for DC, so I think I'll use a HFBR-1528 and 2528 pair which are. They're good for 10Mbit/s. I can't tell what kind of cable to use, is ST the right kind? I can't see how it connects to the modules though...

For the class D modulation I was just thinking of plain and simple open loop PWM, with a triangle wave and comparator. I forgot to mention my SSTC runs at 600kHz or so, so the PWM frequency should be kept well below that, otherwise it might start interfering with the resonance. So really that moves the limiting factor to the SSTC instead... The more I think about it the more of a problem it seems to become. Even with a PWM frequency of 100kHz there will only be six periods in one PWM period. What do you guys think?

Maybe some kind of analog power modulation will have to be used afterall. What kind of modulation did you use in your Tesla Guitar amp, Steve? Frequency modulation like in your PLL circuits?

Steve Conner

Sat Oct 18 2008, 02:44PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Hi all,

The Tesla guitar was a DRSSTC, so the possibilities for modulation were a bit limited. It can't be FM'd without losing zero-current switching, and it can't run at duty cycles more than about 1% without self-combusting.

I used a Schmitt trigger to fire one burst on each positive zero crossing of the audio signal, and an envelope detector hooked up to a 555, to make the burst length proportional to the volume of the audio signal.

The HFBR series come in various connector options, make sure you order one that suits whatever patchcords you can get. In my case, it turned out to be ST.

If the PWM frequency is different to the SSTC frequency, you will hear audible beat tones caused by aliasing between the harmonics of the two frequencies.

Maybe for analog modulation, you could use a Class-D power amp with its modulator split from its power stage by a fibre optic? The output filter would suppress the beat tones.

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob