Timeline for Pumping a few amperes for 100 µsec
Current License: CC BY-SA 3.0
31 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| May 13, 2012 at 10:44 | vote | accept | Ktc | ||
| Apr 27, 2012 at 11:51 | comment | added | Rocketmagnet | @Ktc - The only thing is, with a FET, there are many more ways to choose the wrong one. A BJT really only needs to be NPN. Of course, the even simpler way to do this is just with a current limiting resistor. But then it might be a bit too sensitive to variation in LED forward voltage drop. | |
| Apr 27, 2012 at 11:35 | comment | added | Ktc | @Rocketmagnet this is essentially a constant current circuit. I think there is no difference between BJT and Mosfet. Either case they act as a variable resistors. | |
| Apr 27, 2012 at 11:32 | comment | added | Rocketmagnet | @Ktc - Are you currently using a FET? The circuit recommends a BJT. | |
| Apr 27, 2012 at 11:31 | comment | added | Rocketmagnet | @Ktc - I say wiring because I know next to nothing about the application. There must be some copper from the LEDs to the rest of the circuit, even if it's a PCB track. You will need to take care that this doesn't cause EMI. | |
| Apr 27, 2012 at 11:18 | comment | added | Ktc | @Rocketmagnet also, if I short IN- to 5V, I see a super bright pulse. Somehow, either opamp selection or mosfet selection was not optimum is my guess. I will spend more time on monday to choose better components, perhaps try with BJT as well. | |
| Apr 27, 2012 at 11:16 | comment | added | Ktc | @Rocketmagnet I build this today on the breadboard. The circuit is extremely sensitive to the voltage on the IN+. If I touch with my hand to IN+, LED turns on. Truth is I couldn't manage to turn it off when uP is connected. Pulsing a high current turned out to be difficult. I used MCP6021 as the opamp. VN10LP from Diodes as the mosfet (I had that around, didn't use BJT). 1 ohm as RSense and the LED in question. Couldn't get it to pulse the way I wanted. (I see the pulse in IN+ but not from LED, I tried various pulse sizes, no luck) Need to look at it but it seems circuit needs more work. | |
| Apr 27, 2012 at 5:57 | history | bounty ended | Ktc | ||
| Apr 27, 2012 at 5:57 | comment | added | Ktc | @Rocketmagnet Thanks.. You asked for feedback. I tried to simulate the circuit above without good success. However I understand the circuit, the opamp and the constant current nature of the diagram. Could you please elaborate what you mean by wiring? These are SMT LEDs. | |
| Apr 26, 2012 at 1:28 | comment | added | Ktc | @Rocketmagnet I am not as precise you are right, 4.5A would mean lower light output, not preferred but a compromise that one could accept if the solution is robust. | |
| Apr 25, 2012 at 21:14 | history | edited | Rocketmagnet | CC BY-SA 3.0 |
Added another circuit suggestion
|
| Apr 25, 2012 at 14:16 | comment | added | Rocketmagnet | @Ktc - But you said you needed 5A. I am now confused about the spec. What is the actual tolerance on the current. If you can accept only 4.5A, why can't you accept a lower rise time? | |
| Apr 25, 2012 at 14:10 | comment | added | Ktc | @Rocketmagnet there is a version (1 channel I think) that can pump 4.5A. Assuming I can get 5.6V output, I can put two LED in series. The issue is starting and stopping it. This ic is designed in a way that you cannot stop the light. | |
| Apr 25, 2012 at 11:47 | comment | added | Rocketmagnet | That solution cannot provide enough power. It's only 1.45A per LED. You wanted 5A per LED. | |
| Apr 25, 2012 at 11:21 | comment | added | Ktc | @Rocketmagnet how about this Exar (exar.com/power/led-lighting/regulators/step-up-down-regulators/…) solution with super cap? Unfortunately, the solution cannot time the pulses but perhaps that can be done with the external mosfet and secondary control mechanism. All I am proposing is inserting mosfets before the LEDs and control those mosfets based on the pulse timing. | |
| Apr 23, 2012 at 17:43 | history | edited | Rocketmagnet | CC BY-SA 3.0 |
Mentioned using an inductor based step up regulator.
|
| Apr 23, 2012 at 17:16 | comment | added | Rocketmagnet | @RussellMcMahon - This charge pump calculator is quite interesting. If I am using it correctly, the efficiency seems to depend a great deal on the load current. I assume that's what you meant? | |
| Apr 23, 2012 at 17:05 | comment | added | Rocketmagnet | @RussellMcMahon - Well, also if the flash rate is actually 10ms, then this puts the average current at 50mA, which is more than that MAX1682 can handle. | |
| Apr 23, 2012 at 16:43 | comment | added | Russell McMahon♦ | A capacitor charging a capacitor will be relatively inefficient if the pump capacitor voltage drops by a substantial percentage during discharge or charge. (Work out 1/2CV^2 before and after discharge and be surprised). Best efficiency will be obtained with a switching regulator using an inductor. | |
| Apr 23, 2012 at 12:44 | history | edited | Rocketmagnet | CC BY-SA 3.0 |
Suggested a comparator.
|
| Apr 23, 2012 at 11:40 | comment | added | Rocketmagnet | @Telaclavo - The switching frequency is only 66kHz, which is very slow for a buck regulator. It's certainly possible because I have had this working myself. | |
| Apr 23, 2012 at 11:29 | comment | added | Telaclavo | I thought you wanted to keep the buck always on (because, usually, those are not blocks that can be turned on and off very quickly). Ok, so you indeed want to provide short pulses, by powering on/off the whole buck. It might work, but the feedback needs to be very fast and accurate because, given the low value for L1 and (therefore) the high dI/dt you have at it, you may blow the LED, the inductor, or the PMOS if your feedback reacts a bit late. Also, R1 should be lower, to discharge Cgs in much less than 3 us (which is your off time). | |
| Apr 23, 2012 at 11:06 | comment | added | Rocketmagnet | @Telaclavo - 12.5W on the resistor for only 0.001% of the time = 0.125mW. | |
| Apr 23, 2012 at 11:05 | comment | added | Rocketmagnet | @Telaclavo - Whoops, yeah. I forgot to add a signal from the MCU in the second one. | |
| Apr 23, 2012 at 10:41 | comment | added | Telaclavo | 1) The LED cannot stand 5 A for more than 100 us. With these topologies, it is impossible to provide just a short pulse (5 A, 100 us) to the load. 2) Your 2nd schematic (the one with the comparator) does not provide a PWM signal to the PMOS. The PMOS might end up not switching, but partially on, making it dissipate too much (given the 5 A rate). 3) 5 A through RSENSE would mean 12.5 W on it. | |
| Apr 23, 2012 at 9:12 | comment | added | Rocketmagnet | @Telaclavo - I have updated the schematics. I hope they are correct now. Please let me know what you think. | |
| Apr 23, 2012 at 9:12 | history | edited | Rocketmagnet | CC BY-SA 3.0 |
Fixed schematics (I hope)
|
| Apr 23, 2012 at 0:07 | comment | added | Telaclavo | Be careful, the Q1 you drew has its source on the left (connected to C2). Real-world silicon MOSFETs have a parasitic diode that goes from S to D, so Q1 will always conduct. Even if you flip it horizontally, it will be difficult to turn it on, because none of S and D are ground. There are other issues, too. Too late here. I'll comment tomorrow. | |
| Apr 22, 2012 at 22:43 | history | edited | Rocketmagnet | CC BY-SA 3.0 |
Added simulation results.
|
| Apr 22, 2012 at 22:14 | history | edited | Rocketmagnet | CC BY-SA 3.0 |
added 107 characters in body
|
| Apr 22, 2012 at 22:05 | history | answered | Rocketmagnet | CC BY-SA 3.0 |