Dual polarity SMPS currents and voltage control loop design

[HighTechPower]

Hi guys. I'm designing a +80V & -80V dual polarity SMPS with current sense resistors on high sides. The simulation file is doing perfect and figure is attached herewith. I followed application note (slup114) that I used to design current control loops as shown on it's page 22/23. However when checked on vero board, -ve rail current control loop output from U2 output is showing some negative voltages instead of +ve which I was expecting as done in simulation. Why is it so? What if I want it's output also some +ve value to trigger some circuit? Also to mention that reference voltage 1.24V is not showing on non-inverting pins of U1 & U2 when opamp IC MC34074 (LT1014 is used only for simulations and I actually used MC34074 on vero baord) is removed from socket and voltages checked. Maybe due to large 1Meg resistors R17 & R6 it's not showing 1.24V? What if I want 1.24V on non-inverting pins of U1 & U2 as well?
Also is there any better way to design dual polarity current s and voltage control loops? I tried to implement this way because I needed constant voltage and constant current triggering circuirty in action using single optocoupler by using ORing diodes at outputs of U1, U2 & U4.

 

[KlausST]

Hi,

some thoughts:
* you say "SMPS", it means SwitchModePowerSupply. I see no "switch mode", nor a "power supply".
* in your circuit I see no inputs and no outputs. Where are the loops, what should we look for?
* if the simulation differes (outside tolerance) from the vero-board, then maybe you don´t have a suitable GND concept (like star point) on your Vero-design.
* my saying is: "A circuit without capacitors is no reliably working circuit". In my eyes an analog circuit needs power supply decoupling capacitors, filters, loop stabilisation...
* you say "is showing some negative voltages". Withou giving values and timing, or scope picture. Do you expect us to guess?
* what does "+ve" mean?
* what does "to trigger some circuit" mean?
* "reference voltage 1.24V is not showing on non-inverting pins". The 1.24V are not directly connected to the pins. There are resistors. And for resistors it´s quite usual that there is voltage across them (else they would be useless). Anyways: I don´t understand why you are worried. Without saying what you expect and what you see instead ... a discussion is impossible. Maybe you expect it to be 1.24000V but it is 1.24001V .. so quite within it´s tolerance.
* What if I want 1.24V on non-inverting pins of U1 & U2 as well? --> If you want the 1.24V at the pins, then simply connect them directly to the pins.

Klaus

 

[BradtheRad]

due to large 1Meg resistors

My meters (analog & digital & oscilloscope) have internal 1 Megohm input resistance. This means it loads any volt level I read through it. If that volt level comes through in the circuit through a 1 Meg resistance then I only get half the reading I expect.

Does your meter have 1 Meg internal input resistance? Then you may find readings are thrown off. Lesser ohm values in your circuit should eliminate this tendency. Say 5 percent of meter input resistance you're using.

 

[HighTechPower]

Hi,

some thoughts:
* you say "SMPS", it means SwitchModePowerSupply. I see no "switch mode", nor a "power supply".
* in your circuit I see no inputs and no outputs. Where are the loops, what should we look for?
* if the simulation differes (outside tolerance) from the vero-board, then maybe you don´t have a suitable GND concept (like star point) on your Vero-design.
* my saying is: "A circuit without capacitors is no reliably working circuit". In my eyes an analog circuit needs power supply decoupling capacitors, filters, loop stabilisation...
* you say "is showing some negative voltages". Withou giving values and timing, or scope picture. Do you expect us to guess?
* what does "+ve" mean?
* what does "to trigger some circuit" mean?
* "reference voltage 1.24V is not showing on non-inverting pins". The 1.24V are not directly connected to the pins. There are resistors. And for resistors it´s quite usual that there is voltage across them (else they would be useless). Anyways: I don´t understand why you are worried. Without saying what you expect and what you see instead ... a discussion is impossible. Maybe you expect it to be 1.24000V but it is 1.24001V .. so quite within it´s tolerance.
* What if I want 1.24V on non-inverting pins of U1 & U2 as well? --> If you want the 1.24V at the pins, then simply connect them directly to the pins.

Klaus

Why I'm supposed to attach whole schematic here? Is this forum meant to teach novice people? and are you novice that you need full schematic to understand? and I already have labeled the simulation properly. Similarly I find people sometimes behaving as they are more superior people, maybe I know better than you in this field. If this is the case then this is my last post on edaboard. Unfortunately I can't find an option to deactivate my account.

 

[D.A.(Tony)Stewart]

What did you simulate for Vio? You have a very large attenuator. Just saying without checking specs or your calc’s.

When you put in realistic tolerances and real caps , zip your .asc file.

 

[KlausST]

Why I'm supposed to attach whole schematic here? Is this forum meant to teach novice people? and are you novice that you need full schematic to understand? and I already have labeled the simulation properly. Similarly I find people sometimes behaving as they are more superior people, maybe I know better than you in this field. If this is the case then this is my last post on edaboard. Unfortunately I can't find an option to deactivate my account.

It has nothing to do with superiour or novice. I just tried to solve circuit problems. I´m focussing on technical problems.

It´s not meant personally. But it´s a matter of fact, if there is a loop with an SMPS then the SMPS´gain and phase shift determines the whole loop operation and stability.

I will not bother you again.

Klaus

 

[D.A.(Tony)Stewart]

Better info leads to a better answer