I am currently having troubles with my electroluminescent panel circuit. The driving IC is an HV823 by Supertex. And the EL panel is by Beicoh. I am having many of these HV823 IC's blow in the field but cannot reproduce any of these failures at my site. I believe that I am at times providing too much current through the driver IC which is causing it to exceed its maximum power ratings, but I cannot prove this. I am still unsure if the problem exists in my circuit, or perhaps the panel itself might momemtarily be opening/shorting. Does anyone have any experience with these manufacturers? Perhaps one is having quality issues that I should know about. Thanks!
Here are a few tricks I picked up when dealing with the Supertex HV chips:
*) The big one: Read the appnotes! AN-H33 and AN-H34 are the two important ones. Read them - every page, every word, every note, every warning.
1) Get a current limited power supply and set it to about 30mA. If possible, make it shut down when the current consumption hits 30mA. That should stop the chip from blowing up during testing.
2) Try increasing the value of Rsw (the booster frequency control resistor). When picking a value for Rsw, start high, then reduce the value until the current consumption hits 20-25mA.
3) Most faults are caused by the inductor saturating. I honestly can't remember how to detect saturation though - maybe someone else can provide that bit of info.
4) If you want to make an 823 fail, flip the power on and off rapidly. Hook up a DSO and watch what the circuit does pre-failure. Get as much data as you can.
5) Put a series resistor between the HV823's output and the EL sheet. Around 2 to 5k should be fine. Also check that you're not subjecting the parts to ESD. Get a wrist strap, get an ESD-dissipative mat to work on and check the grounding on your soldering iron.
The big thing to watch out for is inductor saturation. Inductor saturates, HV drive transistor gets overstressed and goes bang. I used an EPCOS inductor on one of my circuits. I could dig out the part values I used for that if it would help.
Later.
--
Phil. | Acorn RiscPC600 Mk3, SA202, 64MB, 6GB,
philpem@despammed.com (valid address)| ViewFinder, 10BaseT Ethernet, 2-slice,
http://www.philpem.me.uk/ | 48xCD, ARCINv6c IDE, SCSI
No to DRM, software patents and the EUCD!
... This is 1 FM. - Quick Mr. Worf close the bloomin\' hailing frequencies.
Place a small resistor in series with the inductor or the switching device. Just so that it drops about a hundred mV to measure the current. If you can place it in a ground leg, fine. If not use another small transformer that offers a sufficiently low cutoff frequency and hook that up in parallel to the resistor, C-coupled if the DC would be too much.
Then look at the current ramp on a scope, preferably a good old analog one. A sharp rise at the end of the ramp (shark fin style) indicates core saturation. When core saturation sets in the inductance rapidly drops which sharply increases the dI/dt.
Many modern current mode PWM chips such as the LM3478 have a feature that sense such a condition and react accordingly.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.