TI E2E Community
Non-Isolated DC/DC Forum
LM3488 possible latch-up mode
I have the LM3488 designed in many of our standard products configured as a simple boost converter. One in particular, is our series of lower power (<10watts) LED drivers. I have the LED return current flowing thru a sense resistor that develops the feedback voltage on pin 3. I like this current-mode controller due to its size and cost. It has worked flawlessly with over 10,000 of our products fielded. The other day a customer called complaining of an issue that hinged around the fact that he was powering our driver with a limited current source power supply. What appears to happen is at Time=0, his power supply smoothly ramps up to +12V as seen on pin 8 of the controller as well a one side of the boost inductor, you can see the drive output, pin 6 become active (Fswitch = 150khz). In about 5ms the output reaches it nominal output current of 200mA thru a load that drops 40V (8watts). No issues and input current to booster = .75amps @ 12V or 9watts input, yielding an efficiency of 88.9%, all is well.
I mentioned the customer’s power supply has current limiting and its set to 1.06A. What happens every so often on power up is as the input voltage comes up, it triggers the over current function. Upon close inspection it happens at 8.5V or lower, which is what you might expect as the input current to supply 9 watts in is over 1.06A. Now the not so good part, when the current limit kicks in on the user power supply, the output voltage folds back to around 5V. The whole system hangs drawing 1 amp. Closer inspection reveals the switching action to the boost mosfet stops. Looking at pin 6, the drive pin for the gate, it sits at around 4 volts forever!...no switching action, nothing but leaving the boost fet on and hence current flowing continuously thru the boost inductor. I would expect the circuit to act erratic because we are "starving it" during power-up, but shouldn't the drive pin still switch, regardless if the feedback pin (3) or the coil current sense pin (1) is not above their thresholds? - I'm thinking this is a state that is not appropriate for any switching control chip, because now the coil never see's any off time. Going around all the pins of the chip, I do not see any levels that are in appropriate to cause this type of behavior. As you know the part is designed to operate down to +3V and in this case it is still running at around +4 to +5V.
Any thoughts or insights would be greatly appreciated. Thanks in advance
as you mentionend before this is due to input power supply current limit.
Could you please provide me the schematic and the specs (Vin range, Vout and Iout) for this application?
Thanks for the quick response. Since the complete schematic that includes other aspects of our driver product is proprietary, I will "sanitize" and provide just the boost circuit aspect of the design. I'll include all key electrical details per your request and send it over to you tomorrow.
Included is the schematic of the boost circuit. I've annotated it with the key parameters during operation.
please modify the schematic as attached, please let me know if that solves your problem.
Thanks for your quick response. Sorry, I do not see an attachment.
the attachment was actually an image, anyway i am going to send it to you as a zip file.
Thank you Giuseppe.
I see how this fix will work. I'll verify in the lab.
So what I was experiencing with the LM3488 would be expected, given my situation?
The LM3488 is low side N-Channel MOSFET controller which is able to reach a 100% duty cycle. Unfortunately, the LM3478/88 gate driver could gets stuck generating a high driver signal that keeps the MOSFET turned on all the time when the IC hits 100% duty cycle if the input power supply is current limited.
The problem occurs because the LM3488 doesn’t have an Under Voltage Lock Out circuit which turns off the device when the input voltage goes below a certain value. Without the UVLO the device will keep switching as long as the input voltage is higher than 2.97V (minimum input voltage). If the customer has set the maximum input current for an input voltage higher than 2.97V what happens is that the input current value will be higher than the maximum allowed at the start up causing the input current limit. When the input power supply goes in current limit, the output of the converter decreases then the duty cycle will increase because the device is trying to get more energy from the current limited input supply. Consequently, the low output voltage will cause the error amplifier to generate a high signal which doesn’t allow the intersection with the sense signal and then turning off the MOSFET. Furthermore the sense signal is no longer a ramp due to the current limit of the input power supply (the input power supply is current controlled instead of voltage controlled). This will cause the device to reach 100% duty which is basically a short circuit at the input.
Hope that will help to understand the problem.
All content and materials on this site are provided "as is". TI and its respective suppliers and providers of content make no representations about the suitability of these materials for any purpose and disclaim all warranties and conditions with regard to these materials, including but not limited to all implied warranties and conditions of merchantability, fitness for a particular purpose, title and non-infringement of any third party intellectual property right. TI and its respective suppliers and providers of content make no representations about the suitability of these materials for any purpose and disclaim all warranties and conditions with respect to these materials. No license, either express or implied, by estoppel or otherwise, is granted by TI. Use of the information on this site may require a license from a third party, or a license from TI.
TI is a global semiconductor design and manufacturing company. Innovate with 100,000+ analog ICs andembedded processors, along with software, tools and the industry’s largest sales/support staff.