help on DC-AC inverter for Public art Project. Thks

This is the place for any magazine-related discussions that don't fit in any of the column discussion boards below.
dyarker
Posts: 1742
Joined: Fri Aug 22, 2003 1:01 am
Location: Izmir, Turkiye; from Rochester, NY
Contact:

Post by dyarker » Tue Jul 22, 2008 5:57 am

MAX335 would have been fine with 10VAC RMS.

With TRIACs you need to find some with very low cut-off current. The typical 1 to 2A TRIAC may cut-off at 10mA, which is less than the pixels use.

Good luck with your project,
Dale Y

User avatar
Viking
Posts: 87
Joined: Mon Apr 12, 2004 1:01 am
Location: Aylesbury, England
Contact:

Post by Viking » Tue Jul 22, 2008 8:32 am

dyarker wrote:With TRIACs you need to find some with very low cut-off current. The typical 1 to 2A TRIAC may cut-off at 10mA, which is less than the pixels use.
Hello Dyarker,
Of course you are right, didn't think of that one!
Back to jolly old drawing board.
Federico, I will have to have another think about this one.
Regards
Rob
LIFE….the crappy bit between birth and death

federico muelas
Posts: 67
Joined: Thu Jul 17, 2008 7:54 am
Contact:

Post by federico muelas » Tue Jul 22, 2008 1:20 pm

Hi Guys,
Thanks for your great help,
actually I made a mistake, each cell uses more amperage than I thought.
with NPD-300
power=1W/m2
current=0,05A/m2
so each 5"x5" pixel uses
10V 60Hz (20V recommended but works at 10V too) @ 0,63W or 31,62 mA

with NPD-220
power=5W/m2
current=0,1A/m2
each pixel:
20v 60 Hz (50V recommended but works at 20V too) @ 3,16W or 63,29 mA

I think I did the math right this time. Do this fix the cut-off problem?

What about SSR instead of triacs.
thanks again
Fede

dyarker
Posts: 1742
Joined: Fri Aug 22, 2003 1:01 am
Location: Izmir, Turkiye; from Rochester, NY
Contact:

Post by dyarker » Tue Jul 22, 2008 6:33 pm

My try at the math:

5 inches / 39.37 inches per meter = 0.12700m

0.12700m * 0.12700m = 0.01612m^2

5W per m^2 * 0.01612m^2 = 0.080645W (at 20V I presume?)

0.080645W / 20V = 0.00403A or 4mA

At 10V less current.

Somebody please check my math too!

I think you need one pixel of film to experiment with and get real numbers. The specs on web aren't that clear.

C U L -
Dale Y

federico muelas
Posts: 67
Joined: Thu Jul 17, 2008 7:54 am
Contact:

Post by federico muelas » Tue Jul 22, 2008 7:15 pm

hi dyarker,
Thanks a lot for your math.
shouldn't be ?
if is 5W/m^2 then 5/39,37= 0.127W/inch^2
since pixel is 25 inch^2
0,127 x 25= 0.635 W

I guess you're right, my numbers are too high.
I have some samples of the film (not the final size though) but they will allow me to double check the numbers tomorrow.
thanks again
Fede

federico muelas
Posts: 67
Joined: Thu Jul 17, 2008 7:54 am
Contact:

Post by federico muelas » Tue Jul 22, 2008 7:43 pm

Hi Dyarker,
I Apologize, you're right, 5 * 0.01612, I got confused with the Sq stuff.
so I think 0.080645W is right, so i guess we still have the low cut-off current issue, right?
thnx
Fede

User avatar
Viking
Posts: 87
Joined: Mon Apr 12, 2004 1:01 am
Location: Aylesbury, England
Contact:

Post by Viking » Wed Jul 23, 2008 3:53 am

Hi Guys,

I need some time to think this through, and I'm up against it at work at the moment and can't think straight.

But regarding triacs (and SSR's as well), if you are switching DC, then you need to reduce the current below the hold-on level for the device to switch off, but we are switching AC, so the device will switch off on the next zero cross once the gate drive is removed?

Isn't this correct?

Regards
Rob
LIFE….the crappy bit between birth and death

dyarker
Posts: 1742
Joined: Fri Aug 22, 2003 1:01 am
Location: Izmir, Turkiye; from Rochester, NY
Contact:

Post by dyarker » Wed Jul 23, 2008 6:03 am

Rob,

Yes, but if the AC current is less than hold current the TRIAC will cut-off as soon as trigger pulse ends.

federico,

Power is given for square meter. So I converted 5 inches to meters. Then one side times other side (in meters) to get square meters for a 5 inch square pixel. No need to mess with square inches.

Low current isn't a problem, it's better. Current less than 30mA, and AC peak voltage less than 15, lets you use MAX335 like you wanted to in first post. With TRIACs you'll need more parts. The TRIACs, some logic to trigger TRIACs in sync with beginning of each half cycle, shift registers to control which TRIACs to trigger (which pixels), etc.

The MAX335 has the shift register, clocking, and serial to parallel in sync with CS. Even if testing with real film shows that you need TRIACS, the MAX335 might be a good choise to drive the TRIAC trigger inputs.

C U L -
Dale Y

federico muelas
Posts: 67
Joined: Thu Jul 17, 2008 7:54 am
Contact:

Post by federico muelas » Wed Jul 23, 2008 7:04 am

Thanks guys,
Then the MAX335 sounds like a winner to me, whatever system we end using.
I should be getting the 335 samples today or tomorrow and I'll give it the first try.
thanks
Fede

User avatar
Viking
Posts: 87
Joined: Mon Apr 12, 2004 1:01 am
Location: Aylesbury, England
Contact:

Post by Viking » Wed Jul 23, 2008 8:03 am

Hello dyarker, but we would not be using a trigger pulse. If a pixel is set to on, then the shift register output associated with that pixel will be switched in the on condition permanently. So the triac will retrigger on each cycle.

The absolute maximum switch current for the MAX335 is 30mA. The actual maximum safe current is probably 10-15mA, which should still be ok for this application. But I'm concerned by the voltage. If the peak input to the switch is 14.5 (clamped) and the switch has an on resistance of 150R, then at 7.3mA, you will drop just over a volt, which means the pixel will see about 13.5V p-p or 9.5V RMS.
I think Federico will have to try this on a number of real pixel elements, using the actual MAX335 device, and see if they do really switch completely.

But dyarker is correct, if the MAX335 driving the pixels directly is a non-starter, then we can always use it to drive some external pixel drive circuitry.

Regards
Rob
LIFE….the crappy bit between birth and death

User avatar
Viking
Posts: 87
Joined: Mon Apr 12, 2004 1:01 am
Location: Aylesbury, England
Contact:

Post by Viking » Thu Jul 24, 2008 10:03 am

Hello Federico,
The electrical equivalent of the pixel is probably capacitate in nature, may similar to that shown:-

Image

The 6818 32-bit shift register feature sourcing output (active pull up at 40mA maximum to 60V maximum) and would be ideal drives for triac's, so worth considering as an alternative. And you would only need about 115 of them compared to 462 MAX335’s or 7 6818’s per display module.

Regards
Rob
LIFE….the crappy bit between birth and death

federico muelas
Posts: 67
Joined: Thu Jul 17, 2008 7:54 am
Contact:

Post by federico muelas » Thu Jul 24, 2008 11:14 am

Hi Rob,
thanks a lot,
so you think that if we use the 6818, the cut-off voltage is not a problem anymore because the shift register will retrigger the triac on every cycle, right? do you think this can short the lifespan of the triac or get a flickering effect on the pixel?
I just received some samples of the MAX335 so I'll try it with the NPD-300 and let see how it goes. I'll let you know.
I'm ordering some 6818 to but they are harder to get.
the diagram you sent me is it to drain the capacitive leakage in each pixel? so 2R is the pixel?
thanks again
Fede

User avatar
Viking
Posts: 87
Joined: Mon Apr 12, 2004 1:01 am
Location: Aylesbury, England
Contact:

Post by Viking » Thu Jul 24, 2008 12:09 pm

Hello Federico,

No, it is all part of the pixel element. The 100k resistor is the pixel leakage, the capacitor represents the capacitance (main component of the pixel) and the 2R resistor the series resistance of the capacitor. There will also be some inductance, but not worth bothering about.
But this is just guess work on my part.

Triac life should not be a problem. My only concern with the triac circuit, based on a simulation, is that the resulting AC across the pixel is not very symmetrical, in other word there is a small DC component. I’m not sure how tolerant the pixel is of DC? With LCD’s their life is shortened if not driven with pure AC. This may be a flaw with the simulation and may not actually be a problem.

It would not be difficult to test this by getting a low current triac (one with 5mA or so gate trigger current) and testing this. This would also show if the 6818 idea would work. By the way, if you can find a SN75518, then this is a near equivalent and would do for testing purposes, and is also available in DIP package form, unlike the 6818, which seems to be available only in surface mount. This would make prototyping easier.

Regards
Rob
LIFE….the crappy bit between birth and death

federico muelas
Posts: 67
Joined: Thu Jul 17, 2008 7:54 am
Contact:

Post by federico muelas » Thu Jul 24, 2008 12:49 pm

Thanks Rob!
I already order the 6818 and the PLCC sockets, it seems that the SN75518 is obsolete and out of samples, but i can try the 5818 that is also a DIP IC.
The problem with the film is that a DC square signal would create too much heat and destroy the chemical composition somehow, even with a square wave, because you don't have the cool down stage that an AC can give you.
I'll ask the manufacturer tough.
thanks again
f

User avatar
Viking
Posts: 87
Joined: Mon Apr 12, 2004 1:01 am
Location: Aylesbury, England
Contact:

Post by Viking » Fri Jul 25, 2008 3:40 am

Hello Federico
If you are going to test the MAX335 idea, remember that a lightly loaded transformer gives a higher output voltage than that stated. I would try a 9V transformer and use a voltage clamp to limit the input voltage to below +/-15V peak-peak.
I have modelled this and you actually get quite a distorted drive voltage fed to the pixel element due to the clipping of the waveform. The MAX335 is being run to close to its maximum for this to be a practical solution. I would definitely investigate the triac option and us a 20-25VAC as the source supply.

Image

Image

Regards
Rob
LIFE….the crappy bit between birth and death

Post Reply

Who is online

Users browsing this forum: No registered users and 9 guests