NEW DRSSTC9 / Projects / Forums

Forums

4hv.org :: Forums :: Projects

« Previous topic | Next topic »

NEW DRSSTC9

previous 1 2 3 4 next last

Move Thread

LAN_403

Dago

Thu Jul 10 2008, 03:37PM

Registered Member #538 Joined: Sun Feb 18 2007, 08:33PM
Location: Finland
Posts: 181

Reaching wrote ...

mhh. why should a toslink receiver pick up noise ??
for sure, a gdt with a long connection cable can pick up noise too and no one talks about that. just put the gdt close to the igbts and connect them with a cable to the driver.

do you want to say that a 1x1cm small plastic package can pick up as much noise as a 30cm long gdt connection cable???

Because the receiver has an amplifier inside with very small currents (microamps iirc).

Reaching

Thu Jul 10 2008, 04:30PM

Registered Member #76 Joined: Thu Feb 09 2006, 10:04AM
Location: Hemer, Germany
Posts: 458

mh, weÂ´ll see if it works or not and how much noise it will pick up.

GeordieBoy

Thu Jul 10 2008, 04:44PM

Registered Member #1232 Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881

GDT's are relatively immune to external fields provided you used twisted pairs for the ends of each winding leaving the core. The actual transformer is no more prone to picking up interference than a single loop of wire if you chose the right core material because the core has such a high permeability compared to that of air.

Marko

Thu Jul 10 2008, 05:08PM

Registered Member #89 Joined: Thu Feb 09 2006, 02:40PM
Location: Zadar, Croatia
Posts: 3145

Yeah - GDT is very immune to external noise since it's shunted with very high gate capacitance, and large amount of charge needs to be moved in order to affect the voltage.

Long wires will only add some leakage inductance, thus more ringing, but it will in almost all cases be tolerable even if just a twisted pair of wires is used.

People are concerned about toslink because it's internal circuitry is highly sensitive, as it needs to be to amplify weak photo diode signal.

Mathias

Thu Jul 10 2008, 05:54PM

Registered Member #1381 Joined: Fri Mar 07 2008, 05:24PM
Location: Hungary
Posts: 74

The receiver is bound to produce/pick up much noise if it's anyway near the "intermittent" TC and not in some cast iron block (need to shield it magnetically also)

Steve Ward

Fri Jul 11 2008, 07:30AM

Registered Member #146 Joined: Sun Feb 12 2006, 04:21AM
Location: Austin Tx
Posts: 1055

I had problems with my standard (lower speed even) OPF2412 optical receivers picking up noise from the tesla coils. I was never 100% sure, but i think it was mostly from the sparks and not from the H-bridge. Putting them in a metal enclosure seemed to fix the issue. The problem i was getting was of the worst kind... output of the Rx latching ON!

Ive also run across documents suggesting the use of the metal-bodied receivers (instead of the typical plastic stuff) whenever they are used in high EMI environments.

Reaching

Sat Jul 12 2008, 07:49AM

Registered Member #76 Joined: Thu Feb 09 2006, 10:04AM
Location: Hemer, Germany
Posts: 458

Tested the whole driver with high side powersupply and driver and everything together and it works wonderful. from frequencies down to 15khz, up to around 400khz, there is no problem. test with load was really surprising. with 27nF load on each high side driver i get risetimes of around 50nS! thats really fast

cause some of you seem to be worried about this emi thing, how can i test it without the risk for the bridge to get destroyed?

Can i build some sort of transmitter to simulate the high em field conditions ? A simple circuit like an oscillator, mosfet and some sort of series resonant circuit, with the resonant frequency i except for the drsstc to work with??

and if the receiver is getting affected by the fieldstrenght, what will be the best solution to shield it? should i just build a metal "cap" or is it better to shield the whole high side driver with all its components?

Steve Conner

Sat Jul 12 2008, 09:48AM

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

To test for EMI, I just run one of my other Tesla coils near the circuit, sparks, ground arcs and all, while monitoring it to see if it does anything funny.

One evil test I like to do to isolated gate drivers is hook the floating side straight to the tank circuit of my mini OLTC with the other side grounded. That puts a good 700V of high-frequency spikey crud and ringing across the isolation barrier.

Reaching

Mon Jul 14 2008, 07:51AM

Registered Member #76 Joined: Thu Feb 09 2006, 10:04AM
Location: Hemer, Germany
Posts: 458

Ok,
Cause the main intention for this project is to produce music with a drsstc, i started thinking.
i dont like this idea to control a 1k$ bunch of electronics with a simple 555 or something else.

the idea is really easy. why not go digital now? the main disadvantage in my analogue modulators was the lack of possibilities for adjustment and control. with a midi interface you dont have this disadvantages, but i have no clue how midi works and i dont wanted to spend weeks on it.

so i started working out my own technique.
cause im working with analog squarewave signals generated on a computer and some simple sequencer program, i though, why not use the signal voltage, known as volume for generating different on time lenghts. with this way i can seperate each signal with a resolution of 1Âµ to generate a binary "staircase" on each incoming signal. with the volume from 0 to 1 volt, the modulator can generate the on time pulses in 512 steps.
its completely digital and opens new possibilities. now its possible to control the on time or "volume" for each note to be played and that with a full analog signal consisting of squarewaves a computer produced.

the pic shows my theoretical idea. a staircase, generated by a binary counter with 8 bits, and a bitlenght of 1ÂµS generated by a crystal oscillator is fed into a voltage comparator. the signal from the computer is fed into the other input of the comparator, pruducing a signal like the "resulting signal" in the pic. with a complex logic, the staircase is only 512Âµ long seperated in 1 bit "steps" and then off for the next cycle of the incoming music signals. so the on time cannot be longer than 512Âµ with full voltage applied.. that is controlled by some simple gate logic and flipflops.

in reality it looks like the scopeshot above. the lower signal is the generated staircase from the counter and the r2r network, the upper signal is the resulting pulse, generated by the comparator and the other logic.

i should have thhis thing finished in the next few days, and then i can give you the results. i think it will be nice to control everything with just one simple signal.

of course, the modulator will have a switch to modulate manually via potentiometers etc, but with the advantage of digital signals, limit by 512Âµ max etc.
maybe iÂ´ll add a 7 segment display to count the on time pulses in ÂµS like i did in some previous projects. weÂ´ll see...

GeordieBoy

Mon Jul 14 2008, 02:32PM

Registered Member #1232 Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881

You could do the whole thing digitally using binary counters and binary comparators, instead of generating an analogue staircase waveform and then trying to use an analogue comparator.

Using the analogue method you may experience problems with glitches and noise on the ramp signal. Particularly if the analogue ramp is not monotonic. I can already see some variation in the step sizes around the centre of the staircase presumably due to bad trimming of the MSB.

previous 1 2 3 4 next last

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