On 11-Jan-12 21:21, Dwayne Reid wrote: > Good day to all. > > I've revisiting a design that I did a couple of years ago. Its a > small enclosure with 2- prismatic NiMh cells for power, 2- boost > converters to generate 5.75V @2A peak& 7.2V @ 50mA continuous. > > I used 2- National Semiconductor LM2623 boost converter chips to > generate those voltages. However, after the design was finished, > approved, accepted and the devices were built and delivered, the > customer complained that there was audible noise in the speakers when > the device was active. That was after using two identical prototypes > for a number of weeks before the final design was approved. > > You all know that goes. > > The noise is caused by the 7.2V converter running in 'burp' mode - > the voltage ramps up past the desired set point, the converter shuts > down, the voltage decays down past the hysteresis point, the > converter turns back on. The resulting ripple is in the audible > frequency range. > > Because these devices are completely sealed, we did a bandaid > solution by adding a very low ESR 330uF 10V electrolytic to that > voltage rail in one of the wiring harnesses that leaves the > unit. Its ugly but it works. > > We've now been asked to build more of these units. In addition to a > low-battery automatic cutoff, the customer wants us to move the > bypass cap to inside the enclosure. I have enough room to do this > (3- 100uF 10V ceramic caps in 1206 package) but I'm wondering if > there is a better DC-DC converter chip available. > > The specs needed are: 1.8 - 2.5 Vdc input, 7.2Vdc output, 50mA. The > real problem is the low input voltage. I'd like to be able to > purchase the parts from one of the North American distributors > (Arrow, Digikey, Future, Mouser, Newark, etc) if possible. Dwayne, It's funny I had to build a similar converter for a project and I just=20 finished yesterday. My design was for a 1.8-2.6V input 12V/350mA output. I used PIC12F1822 as a continuous mode controller, the PWM is 182kHz,=20 the peak inductor current is limited to 4A with internal comparator=20 which is connected to 25mOhm current sense resistor. The regulation is=20 by burp mode using fast A/D but I tested a hardware solution that worked=20 by linearly increasing the current sense resistor as the output voltage=20 rises and it worked pretty well. At the end, I have chosen burp mode=20 since there were less components and my 12V is for a noisy DC motor. For 7.2V/50mA I would just use a shunt regulator to dissipate excess=20 current if the load is fairly constant. Or, you can try using=20 discontinuous mode and skip pulses when the output voltage is too high. I am really happy how this little PIC works as a switcher. The=20 efficiency at full load is ~75% and I can use the remaining 16MHz PIC=20 resources to do some other useful stuff. Djula > > I'd gladly use Microchip's MCP1640 family, except that the output > voltage doesn't go high enough. In fact, I *really* wish that I > could use that family, because it includes an output-disconnect > feature that completely disables the output load when the chip is > disabled. I need that for the low-battery automatic cutoff that I'm > adding to the unit. As it is, I'll be using P-channel MOSFETs for > that function. > > Anyway, I'm hoping that someone can suggest a DC-DC converter chip > that meets those specs and runs in PWM mode rather than burp mode. > > Many thanks! > > dwayne > --=20 http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .