| Electronics > Projects, Designs, and Technical Stuff |
| Lab Power Supply - The Lost Current |
| << < (7/26) > >> |
| radoczi94:
--- Quote ---oscillations without a scope With no scope the best you can do is trying to measure AC with meter. On DC mode with meter, only last digit value should change. --- End quote --- Thanks for the idea, will try that. My meter has freq counter function, if the amplitude is high enough, it is able to estimate the osc. freq. Will try that too. If I have time. About the schematic. I think I understand the schematic as it is, but yeah, it could be a lot easier to read, definitely needs some optimizing. I put in those regulating circuits and the reference. http://www.electronics-lab.com/project/0-30-vdc-stabilized-power-supply-with-current-control-0-002-3-a/ Here is the original schematic. I want to redesign the PCB, it had some problems, the circuit needs some fusing and a handful of protection diodes. It went trough a lot of soldering and desoldering, now it looks like it's made by some blind one handed apprentice. |
| C:
Before you start changing thinks, would be good to understand the circuit. Most times each circuit has problem areas. There are normally many ways to work around a problem, some better then others. Some fixes help in one area while making other areas worse. In a lot of ways, the Electronics-lab link of last post is much better. Look at something simple. Last schematic Transistor labeled Q1 With Q1 acting like a switch, why short out the output of U2? Not good for U2 especially when there is another way to do same thing. Q1 could pull down on junction of R13 and Q2! --- Quote from: radoczi94 on February 05, 2018, 09:43:49 am ---About the schematic. I think I understand the schematic as it is, but yeah, it could be a lot easier to read, definitely needs some optimizing. I put in those regulating circuits and the reference. http://www.electronics-lab.com/project/0-30-vdc-stabilized-power-supply-with-current-control-0-002-3-a/ Here is the original schematic. --- End quote --- So if you really understand current schematic and compare it to this schematic some of the bad changes should really stand out. These days it is not uncommon to have a load that is doing high current PWM. Not uncommon to have many different PWM loads. During testing/design a Lab Power Supply is the power supply for this new circuit. Until you know exactly what value of series resistor to use for that PWM dimmed LED the law power supply CC mode is the protector of that high dollar LED, would be a good idea for it to function properly. With this being a simple cheap supply, you may not have the best but still want the best it can be. So think about a load that changes Initially with Max current of CC mode the output of this supply will be rapidly changing from VC mode at some voltage to some current value in CC mode. How does the lab power supply handle this? What changes in the lab power supply circuit. To make it easy the negative output of supply is connected to ground reference With respect to ground reference what is point labeled 7 on last schematic doing? What is cathode of D5 doing? What is anode of D7 doing? What is voltage across C5 doing? Is C5 a good idea or is it causing problems? What is base of Q2 doing? How does it change with low/high output voltage, low/high output current? Each little change can effect output of the supply. A change that makes one part of circuit better that at same time makes supply output worse is not a great change. Some changes that could effect output can be hidden by other parts of circuit, but there is a limit to how much. Try to understand last post schematic and then look at your first posted schematic. What does each change do? What is changes to circuit when load changes such that same CC mode current is two different output voltages? --- Quote from: radoczi94 on February 05, 2018, 09:43:49 am ---I want to redesign the PCB, it had some problems, the circuit needs some fusing and a handful of protection diodes. It went trough a lot of soldering and desoldering, now it looks like it's made by some blind one handed apprentice. --- End quote --- If you are going to redesign the PCB, I would strongly suggest a redesign of circuit first to fix the many problems both versions have. |
| radoczi94:
--- Quote ---Last schematic Transistor labeled Q1 With Q1 acting like a switch, why short out the output of U2? Not good for U2 especially when there is another way to do same thing. Q1 could pull down on junction of R13 and Q2! --- End quote --- You meant R15? Because if I pull down the R13-Q2 juntion, Q1 shorts C1 to ground, causing rather large currents and it would not be so good for the cap and the transistor too. The opamp should not care about that short, it has "infinite" output short circuit protection to each power rail and ground. But still, it's better to do that trough a resistor (if you meant to pull down the base of Q2). --- Quote ---So think about a load that changes Initially with Max current of CC mode the output of this supply will be rapidly changing from VC mode at some voltage to some current value in CC mode. How does the lab power supply handle this? What changes in the lab power supply circuit. To make it easy the negative output of supply is connected to ground reference With respect to ground reference what is point labeled 7 on last schematic doing? What is cathode of D5 doing? What is anode of D7 doing? What is voltage across C5 doing? Is C5 a good idea or is it causing problems? What is base of Q2 doing? How does it change with low/high output voltage, low/high output current? --- End quote --- On the last, linked schematic: All with the respect of the output negative -Point 7 It is the "high current" power rail. With a varying load it is also varying, depending on the main transformator and the capacity of the filtering cap. -Cathode of D5 The cathode of D5 is on the ground.If we measure that point in the respect of the negative output, there should be the voltage drop across the shunt resistors. A square wave if the varying load is on the output.(If the load is pwm) -Anode of D7 That's the negative voltage rail, there sould be the zener voltage+the shunt voltage drop.A square vawe with a dc offset. -C5 should be same as the point 7. I think it is a deoupling cap, but not shure. Therefore, I do not really know if it is necessary or it is bad. -Base of Q2 There should be the output voltage + Q4 and Q2 B-E voltage drop. At maximum voltage, there should be 31-32, at minimum voltage there should be the between 1-2 volts, depending on the transistors. At low output current there should be output voltage+ the B-E voltage drops, at max. current, there should be 0 volts. Or even negative voltages. I wanted to make a stabilised supply for the IC-s, because my transformer is 28Vac and capable of 5amps. The rectified voltage would be much higher, than the opamps would tolerate, that is why tose parts were put in. Changed the voltage reference also, because I did not liked that circuit at all, never understood, actually. Why are these changes so critical? --- Quote ---If you are going to redesign the PCB, I would strongly suggest a redesign of circuit first to fix the many problems both versions have. --- End quote --- I was intended to do so, but I can not really found the problems with this circuit on my own. If you guys could help me find those probelms, explaining, why are they a problem and how can it be solved, I will correct them. I'm just not at the level of do this on my own. The more I learn about electronics, the more dumb I feel myself. |
| C:
Yes R15 Q1 controls base of Q2 Think this through a step at a time. What happens if I short Q2 base to the positive output? U2 supplies current to other side of R15. Assume for a moment that U2 output is shorted to point labeled circle 7, In practice U2 will be limited to less then this. What is max current possible. After you think you have some answers look at end of this post. --- Quote ----Base of Q2 --- End quote --- If you think about it, Base of Q2 is output control. It will change based on voltage in VC mode and try to change by current in CC mode. So all of the circuit is changing based on current drop across R7 except for what is directly connected to circle 4 ( negative output. C5 C5 is trying to prevent voltage changes between circle 7 & circle 4. You stated that circle 7 changes based on current demand of output. So C5 is also trying to prevent fast CC mode response to a current change. My initial thoughts is that C5 is bad, makes supply worse in both CC & VC modes. --- Quote ---I wanted to make a stabilised supply for the IC-s, because my transformer is 28Vac and capable of 5amps. The rectified voltage would be much higher, than the opamps would tolerate, that is why tose parts were put in. Changed the voltage reference also, because I did not liked that circuit at all, never understood, actually. Why are these changes so critical? --- End quote --- look at last schematic at U1 Think of how an op amp works, Output is like a variable resistor to +V and a variable resistor to -V. Effects of Changes to +V are reduced by power supply rejection ratio of op amp Your first posted schematic has R13 changing by load current & filtered some by C10 You have power supply rejection ratio of U1 vs rejection of TL431. But by creating the stabilised supply for IC's you now have all IC's having to fight against Load current changes on their supply lines vs IC output. You have a bunch of capacitors connected to what is noise for what you want on output. Examples are C3, C4 C5,C7,C9,C6. So you are starting from last schematic that has a poor CC Mode and then increasing source voltage creating more problems. Might be a good idea to fix CC Mode in such a way that more source voltage becomes less of a problem, instead of hack the hack the hack. There are many ways to do this. With no scope and limited testing capability you will get a "I think it is good lab power supply". With changes you have made from last schematic, your limited testing is saying "I know it is broke/bad" A simple circuit for a power supply that has a CV mode and a CC mode is as follows. A pull up resistor that turns on the output to MAX. Your Voltage reg & Current reg then limit the max. Your current regulator's diode is setup to do this. Needs to control output directly not via voltage regulator as this adds a second control loop that is talked about. Output of Voltage regulator needs to connect to diode with anode of two diodes common and controlling output. So you have one diode for CC Mode & one diode for VC mode & you could add a third diode or more diodes for shutdown or other limit conditions. With low side current sense you have a problem that any control circuit current changes connected to negative output also effects your current sense. If this is a 0V to 30V supply, You could think of it as a 5V to 35V supply. The negative output regulates the -5V supply you have. With -5v Controlled by output negative, you could then use this to control +V to op amps. You have just removed supply bounce by current. This leads to control circuit power around negative output. You could think of floating control circuit power with positive output as the reference. This can make it easy to have higher voltages and make high side current sense easer. The down side is that today most think of negative ground. What I am trying to show is how you think about a circuit can change what you see. A small change can make a huge difference. So step one is understanding both circuits better and comparing what the differences do. Any change needs you to understand the effects of the change. Answer Q1 controls base of Q2 question. You missed that when Q1 turns off Q2, collector of Q2 can rise and will not short out C1 D10 keeps Q2 base from going below circle 3 (positive output ) by diode drop. At a Diode drop below Q2 is off, This in tern means that Q4 is off. The only thing left trying to hold base of Q2 high is R15. Max current is R15 with voltage of circle 7 - circle 4 Now Think of what U2 will do. Say output was set at 1 V. Pulling Q2 base low will cause output voltage drop which will cause U2 to turn on harder trying to make U2 output more positive. So at start very little current through R15, but will increase to U2 max output voltage current with Q1 having to conduct this current. |
| radoczi94:
Huh, that is a lot of information in one go. Now I see, why this circuit is a total failure. Maybe with extra, stabilised supply voltages for the opamps,with a decent voltage reference and with the OR diode gate, this thing will be able to work like as it should. :horse: Or maybe I should just look for another schematic. Designing for another 4-5 days, taking away from my precious time, building it, for 2 days, experimenting for weeks, because why would it work for the first time. Or just getting an old, well proven circuit from somewhere, understanding how it works and building it as it is. I need to think this trough. I really wanted to build an indestructible, foolproof power supply, that is why I used that massive transformer and heatsink. With the case I built, it's weight is around 20 kgs, and not even used the full 5 amps of the transformer. Anyway, thanks for all help, it's really appreciated! |
| Navigation |
| Message Index |
| Next page |
| Previous page |