HartleySmith

's profile
image
Electronics Engineer & Technical Author

Biography has not been added


HartleySmith

's contributions
  • 01.20.2014
  • Latching power switch uses momentary pushbutton
  • Hi Barkuti, Glad you like the circuit - it's really great to get such positive comments. You are right about the values of C1 and R2. The values are not critical and you should be able to experiment with different values to get the best performance for your application. I hope you find the circuit useful. Best regards, Anthony H. Smith.
  • 01.20.2014
  • Latching power switch uses momentary pushbutton
  • Hi Mcbain1, There is no easy way to eliminate, completely, the current flowing through R1 when the circuit is latched on. However, you can minimise the current by increasing the value of R1. The circuit shows a value of 10k, but you could increase this by a factor of ten, or even a hundred, or possibly more. However, there are two factors to consider if you choose a very large value for R1. The first is collector leakage current through Q1 when this device is off. A very large value of R1 combined with a significant leakage current could drop enough voltage across R1 to bias Q2 on when it should be off. However, most good quality NPN bipolars have leakage currents of just a few nanoamperes, so this is unlikely to be a problem. The second thing to consider is the gate-source capacitance of Q2. In order to turn off the MOSFET, this capacitance must be discharged via R1 in series with R3. So increasing the value of R1 will increase the time taken for Q2 to switch off fully. Whether or not this would be a problem would depend on the type of device used for Q2 and on the turn-off speed you require for your application. I suggest you experiment with a few large values of R1 to see how your circuit behaves. Regards, Anthony H Smith.
  • 01.20.2014
  • Latching power switch uses momentary pushbutton
  • Hi Davidk, The circuit as it stands was designed to be used with a simple, momentary pushswitch having volt-free contacts, so I think you may have problems interfacing it to the SL353LT Hall sensor. I don't have a sample of the SL353LT, but the datasheet shows that the sensor's output is essentially a logic-type output which swings between the sensor's supply rails. For this reason, I think the circuits in my idea would need a fair bit of redesign to accommodate the SL353LT. However, don't despair! It's fair to assume that your application needs to be activated by a magnet of some kind, so why not consider a reed switch? You would need a SPST-NO reed switch connected in place of the pushswitch. You'd probably be able to source a suitable reed switch that is cheaper than the SL353LT, and it would meet your requirement of having zero current draw in the off state. Hope this helps, Anthony.
  • 01.20.2014
  • Latching power switch uses momentary pushbutton
  • Hi GeorgeG404, Sorry for the delay in getting back to you. If you wanted to control the circuit of Fig.1 with a microcontroller port, the simplest way would be as follows: (1) Dispense with the push switch, C1 and R2. (2) Combine R1 and R3 into a single resistor. Value not critical - say 470k or thereabouts. (3) Disconnect R4 from Q2, R5 and the load. Reduce its value to, say, 10k and connect it to your microcontroller port. With the circuit modified as described above, Q1 and Q2 will turn on and energise he load whenever the micro port goes high. You will need to latch the port high to keep the load energised. This modified version of the circuit would allow for a relatively low level signal from the micro (say 3V or 5V) to control a load operating at a much higher voltage. Make sure that the transistors are properly rated to deal with the DC supply voltage. The circuit of Fig.2 could also be adapted to allow control from a micro port, but I'll not go into that right now. Regards, Anthony H Smith.
  • 03.16.2015
  • Schmitt trigger adapts its own thresholds
  • Hello Tyler, For signals swamped with noise (like the example in Figure 2) the thresholds VTU and VTL should be set as close as possible to the peak levels, VU and VL. I chose 5% because it suited the example shown in the figure, but it could just as easily have been 2% or possibly even 1%. For signals with less noise content, a value of 10% or more could be adequate. However, when dealing with relatively small signals (like the example in Figure 3) it's important to take account of any input offset voltages in the buffers (IC3) and output comparator (IC4). In that case, if the thresholds were set too close to the peak levels, the combined offset voltages could prevent reliable triggering. The offsets in the input comparators responsible for setting VU and VL also play a part and they, too, could conspire to cause problems when dealing with very small input signals. There are no hard and fast rules when setting VTU and VTL and it would probably be best to determine them empirically based on the particular application you have in mind. Kind regards, Anthony.
  • 03.16.2015
  • Schmitt trigger adapts its own thresholds
  • Hello VladYer, Many thanks for the clarification. Unfortunately, my Russian ain't too hot (er, non-existent actually) but it looks like you've gone through several interesting iterations to get to the final circuit. Have you breadboarded it? It would be great to see some test results from an actual circuit. Would I be right in thinking that the minimum input signal amplitude that would produce triggering is roughly equal to twice the Schottky drops? Cheers, Anthony.
  • 03.16.2015
  • Schmitt trigger adapts its own thresholds
  • Hi VladYer, Thanks for the link to your schematic - looks like a neat little circuit. Could you explain the significance of points X and Y ? Anthony Smith
  • 01.20.2014
  • Latching power switch uses momentary pushbutton
  • Hi Tfboy, Thanks for your comments. I'm glad you like the circuits. If you need a CMOS logic output, it may be possible to use the Fig.2 circuit and connect the output at Q2's drain to a suitable logic gate. A Schmitt trigger device such as a 74HC14 (or similar) would be best. There's no reason why the circuits couldn't be used to generate logic signals. However, it's important to recognise that they were intended to be used for switching power rails that might be incompatible with logic circuits. In fact, with the correct choice of components, the circuits could be adapted to switch rails of tens, or even hundreds, of volts. Therefore, if you need just a latching logic signal, it may be better to use a circuit based only on CMOS logic devices. For example, there are plenty of textbook circuits that use an SR latch or JK flip-flop where the output can be toggled high and low using a single push switch. If you go down this route, remember that you'll need to debounce the switch to avoid false triggering. Hope this helps. Good luck. Anthony Smith.
  • 01.20.2014
  • Latching power switch uses momentary pushbutton
  • Hi Simon, As author of this Design Idea, I'd like to thank you for your kind comments. As to my name on the PDF download, my user name is HartleySmith, so I suspect that the web author has mixed up my real name and my user name. For many applications, it should be possible to build the circuit for pennies. Your comments about Q2 suggest that you're interested in the circuit shown in Figure 2 intended for high-side loads. If your small LED is a low- to medium-power type (that is, one that doesn't require hundreds of milliamps of forward current), then you should have no difficulty finding a suitable N-channel MOSFET. I recently designed the Figure 2 circuit into the power supply section of a digital multimeter. The requirements were to switch a supply voltage of 12V max., and the supply current was no greater than 100mA. I used a 2N7002 (the surface-mount version of the ubiquitous 2N7000) for the MOSFET. The 2N7002 is dirt cheap and can switch currents well above 100mA at voltages up to 60V. You may like to consider it for your LED circuit, although I'm sure you may be able to find plenty of other N-channel devices that are equally suitable. The circuits could be adapted to use BJTs instead of MOSFETs mainly by lowering the resistance values and increasing the capacitance value accordingly. It would also be necessary to include current limiting base resistors for the replacement bipolars. However, for most cases you should be able to source suitable MOSFETs without any difficulty. Kind regards, Anthony H. Smith.