DN14 Question

powerfly -


1.) During CC operation or during a fault condition, the output voltage of this supply will be less than nominal. If Vout is used and the power supply enters the CC region of operation, Vout will be less than nominal and the current control circuitry will cease to function as intended. Vbs is independent of Vout due to the winding polarity and tracks the DC bus input voltage. This means that the CC control circuitry can continue to function even if Vout is all the way down to 0V.


2.) C3 provides the bulk of current for load transients. If you want your CC circuit to "see" the current from fast load transients, than your location is probably best. If you application will see lots of fast load transients and you don't want to prematurely enter CC mode, placing R6 behind C3 shields it from the faster load transients so that the supply only enters CC mode with load currents that are continuously above the CC mode threshold.

-The Traveler

How can I shield R6 from he load transients? I already have R6 between RTN and C3.

You'll most likely be fine. Just make sure that the RTN connection for your control circuitry is after R6 (closest to the PSU output).

-The Traveler

You mean, the control circuit ground should be close to other end of R6?

Would it make any difference if the R & C in the op amp feedback are interchanged?

I just finished designing a CC/CV ckt with TOP249YN. The application is a Battery charger. The circuit works fine when charging the battery.
Once it enters the trickle charge mode, the power supply makes a audible shhhhhkkkk noise and goes OFF. Comes up again and goes off.
I have no idea what's wrong.

See the attached PDF for my note on where you could place R6. You want to make sure that the ground for your control circuitry (U4A, U4B, etc) is referenced to the output return of the power supply. If you place the control circuitry ground on the other side of R6, the ground for your control circuitry will fluctuate with your load current due to the voltage drop across R6.


As far as C10/R13 and C9/R9 go, it won't cause any problems if the R and C change positions.


For your battery charger, can you attach a schematic, layout, transformer documentation, PIXLS/PIEXPERT files, etc? Is the audible noise coming from the transformer? Is it coming from your clamp circuitry? Have you looked at the major components' voltage and current waveforms to see what they're doing when this occurs?

-The Traveler

[quote]See the attached PDF for my note on where you could place R6. You want to make sure that the ground for your control circuitry (U4A, U4B, etc) is referenced to the output return of the power supply. If you place the control circuitry ground on the other side of R6, the ground for your control circuitry will fluctuate with your load current due to the voltage drop across R6 [/quote]

dang! In my circuit the ground is connected to the other end of R6 (not the RTN).
U4A,B ground has to be connected to RTN. But the reference and opto ground can be connected to other end of R6? Fluctuation will not be a problem for the reference. It might make a small difference for the opto. Correct?

I have attached the schematic. Ignore R15, R16 and L1 is 3.3uH, Cc is 2.2nF on the PC Board.

In DN14 Shouldn't R8 be connected to RTN? The reference voltage generated by R7/R8 would change with load current.

Any comments, Traveler?

powerfly-


You do want to make sure that you're using the output of the power supply (before the sense resister). The reason for this is that as load current increases, there is a voltage drop across your sense resister. If you're referencing the opamps and shunt regulator to a point after this voltage drop, you'll get a load-current dependent error in your output voltage.


There is a bigger problem that I see in your design however. The output after D5 is your secondary side bias supply. You have this winding wound with the same polarity as your main output winding. You don't want to do this! DN14 explains this in greater detail and you should read up on it.


The short version is that you want your secondary bias winding to have the same polarity as your PRIMARY winding. This makes your sec bias voltage independent of your main output voltage. With your design, as the output voltage drops, your secondary bias voltage will also drop. So as your PS enters CC operation, the bias supply will drop out and it's not going to function correctly.


-The Traveler

Traveler, the transformer DOT in the schematic is wrong. The transformer was designed using PI expert. I am pretty sure that's correct. Sorry for the mistake in the schematic.

Another question - why is the reference to the current control opamp connected to other side of R5.
The 24.9k and 2k voltage divider has an output of 0.18v. But the 2k resistor is connected to other side of sense resistor instead of ground?

powerfly -


Not sure what you're asking about R5...R5 is part of the output voltage sense circuitry.


The divider formed by R7 and R8 creates the reference voltage for the current control op-amp. In this configuration, without any DC negative feedback, the current control op-amp functions similarly to a comparator. When the voltage across R6 is higher than the reference created by R7, R8 and U3, the op-amp turns on, dumping current through the opto. If the inverting terminal of U4A was connected to GND, *any* return current through R6 would create a voltage higher than ground and the power supply would go into CC mode.

-The Traveler

Let me reframe my question.

The reference for the current control opamp is generated by a voltage divider formed by R8 and R7. Basically, this ref voltage is to be constant through out the operation regardless of the load current.

Why is R7 connected to the other end of R6. Shouldn't it be connected to RTN like R5 in the R4, R5 voltage divider.

If R7 is connected to other side of the sense resistor, then the ref voltage into the inverting pin of U4A increases as load current increases. The non-inverting voltage will never be higher than the inverting voltage. Correct?
U4A Pin 2 and Pin 3 connection are mixed up.

powerfly -


The output voltage resister divider R4/R5 needs to be at the output terminals of the power supply for the reasons you're saying. For the current control circuitry, the opamp is only interested in the differential voltage across the current sense resister.


"Ground" for all intensive purposes isn't a real thing. The idea of "ground" or "zero electric potential" was created to make circuit analysis easier.


If you choose the bottom of C2 to be "ground", then RTN will be higher than GND as load current increases. Having R4/R5 at the output of the supply compensates for this by increasing the positive output with respect to GND but results in a fixed output voltage at the PSU output terminals. For the current control circuitry, the reference voltage into the inverting pin is now fixed with respect to GND and as RTN increases with respect to GND a differential voltage is created across R6 increasing the voltage into the non-inverting terminal of U4A.


If you choose RTN to be "ground", the non-inverting terminal of U4A will now be fixed at "ground" potential while the bottom of C2 decreases with respect to RTN (GND) as load current increases. Since your current control reference is also connected at this node, it too will decrease with load current.


Either way, the result is the same in that U4A is sensing the differential voltage across R6. Once the voltage across R6 is sufficient to turn on U4A, you enter CC mode.

-The Traveler

Let me consider an example-

RTN is my ground reference.
Let's say load current is 1A, drop across R6 is 100mV.
The RTN side of R6 is 0V and C2 side is -100mV ?
Instead of voltage at U4A non-inverting terminal increasing, the voltage at inverting terminal just keeps on decreasing?
When it goes below zero, U4A goes into CC mode?

Also, if I reference R7 to RTN and connect U4A pin3 to left side of R6, would the ckt work.

powerfly -


If RTN is your ground reference and 1A is flowing through R6...then yes, the bottom of C2 will be -100mV...with respect to GND. It's just ohms law...current flows from higher potential to lower potential.


With respect to RTN (being the ground), yes.


Yes.


Yes but you would need to swap the inputs of the opamp.


The thing to keep in mind is that U4A is sensing a differential voltage...

-The Traveler

So my ckt in the schematic, with other side of R6 as reference should also work?
Output voltage changes a bit with load current. The current limit is 1.8A, drop across 0.1 ohm resistor will be 184mV, so output voltage will decrease by 184mV. This is not a big deal for my application.

Things are much easier, once I ignore the ground and just look at the current flow.

powerfly -


It's really up to you to decide what acceptable power supply performance is. With the way you're setup, you'll definitely have load regulation issues. But as you said...for your application this isn't a big concern...so there you go.

-The Traveler