UC3854B: UC3854

Lack of response means either ignorance or incompetence or both

Hi Russell,

A number of the TI e2e support team have tried to help you with your design but you have not shared any information in terms of schematic or waveforms which makes it very difficult to provide support. Your design may be based on the an exiting UC3854 design but we need to see the component values that you have used to understand how you have modified the design for you application specification.

In one of you earlier post you stated that the output voltage of the PFC stage was less than in the input voltage, even when the PFC controller is not active, no PWM drive, there will be an output voltage which will be equal to the peak of the ac line voltage minus a couple of diode drops due to the bridge diodes and the boost diode.

Regards

Peter

Hi Peter,

I can send the schematics that I used but before we get there I need to know that we addressed other possible issues.

For example ,the input voltage range that I am using is 5-100 VAC,is this chip able to adapt to that, I asked this question before multiple times and was not getting any clear answer.

The inductor that I am using is a 1mH inductor,Co is 180 mF

Russell

Hi Russell,

The UC3854B should be able to work over 5V to 100VAC, note that this is a 20:1 input voltage range which is a wide input voltage range for any boost converter. Whats your output voltage and power level?

It should be possible to select component values to get the UC2854B to work over this input voltage range, although the Vcc required for turn on has a max value of 11.2V so this begs the questions of what DC bias rails do you have available to you?

I suggest you try and get the IC to run open loop first, disconnect the gate for the mosfet from the output of the IC and tie the mosfet gate to ground with a 10k resistor. This prevents the mosfet gate from floating up and turning the mosfet on where it will be destroyed if there is an input voltage to the PFC stage.

I would the connect the AC supply through an isolation transformer or from an AC source and increase the voltage to say 30V and then check voltages on the pins of the IC. Vref should be up and in regulation, the voltage error output Vao should be high as the output voltage should be lees than your desired output. Vao should drive the Current amp, CAO to demand maximum duty cycle, check the CT pin for the ramp waveform.

What switching frequency are you running at ? At 5V input the peak voltage is about 7V, if we allow for two diode drops through the bridge then the voltage on inside or DC side of the bridge is about 5.6V. If you use a 1mH choke then the current will ramp through this choke at a rate of 5.6mA/us which is pretty small and it will be difficult to store significant energy in the boost choke unless the on-time of the boost mosfet is long time. Stored energy in the boost choke is 1/2*L*I^2.

I suspect once we get the IC running then may be limitations as to what performance we can achieve given such a wide input range for you boost stage and the boost choke you have fitted.

Maybe you can tell me what you are trying to achieve from this power stage in terms of input and output volts and amps and if you actually need power factor correction or just a boost function?

Regards

Peter

Hi Peter,

Thanks for your response.

I am using an isolated AC power supply and started low ,at about 10 v,the switching frequency value that I used for the calculations is 100k,

I agree with you regarding the inductor value being too large for my input voltage range values.

My main purpose here is a PFC method other than using a bulky and heavy PFC capacitor.

Attached is the schematic that I and another engineer have used.

Russell

Regarding the power level ,its in the 50-100 w range.

Regarding the DC bias , I do have a calibrated DC voltage supply .

Russell

Hi Russell,

There are some fundamental limits that you are facing here, at 5V input if you still need to delivery between 50W to 100W, that's an input current of 10A to 20A in rough figures.

The output divider is 510k and 15k, the Vsense level of the UC3854B is 3V, so the PFC BUS should regulate at 105V dc . If your max input voltage is 100Vac then the peak value of the ac line is 141V which means it will conduct straight through the bridge rectifier and boost diode and increase the PFC voltage to 140V.

TI has a power stage designer tool which will allow you to run through various designs for a boost stage its DC to DC but it gives very quick calculations of the voltage and currents in the key components. For instance at 5V input the RMS current in the boost PFC choke will be in the order of 20Arms for a 100V, 1A output.

www.ti.com/.../POWERSTAGE-DESIGNER STAGE DESIGNER TOOL&tisearch=Search-EN-everything

I would encourgae you to use this tool to get an idea of the power stage behaviour before trying it on the bench. Also please tray and get the IC running open loop first and then connect it to the mosfet to start processing power. There is a lot of design work required to get meet the specification you have outlines. Both the input bridge, boost mosfet and diode will have to be scaled considerable to cope with the current stresses at low input voltages.

Hope this helps.

Regards

Peter

Hi Peter,

I can design around an input voltage range of 20-80 volts if that makes more sense in terms of the component values and the ability of the chip to function .

I think that the inductor value should be in the 100 micro henry range which is 10 times less than what we used .I am also changing the MOSFET to one with a higher current rating,(also because the one on there has blew up already).

I still didn't get to the diode ,what do you think I should go for .

Also thanks for mentioning the relevant TI design tool.

Russell

Hi Peter,

While the voltage levels that I am working with are in the 5-100 VAC range,the range that the circuit needs to optimize the PF can be in the 20-70 VAC range.

We initially selected the design components' values around the full voltage range but I don't think we have to do that.

Also,the power levels can be around 70 watts.

Also,do you think that the values of CCP,CCZ can cause the distortion that I got in my output Current.

Russell

Hi Russell,

Since your output voltage is less than 200V you can use a 200V diode, I dont know what device package you are using but I assume its TO-220?

The part needs to be an ultra-fast device to limit the reverse recovery charge as the boost will operate in CCM.

You could try this part from OnSemi, FFPF30UP20STU, its in stock in digikey.

Just curious why you need the power factor correction feature as this is generally only required for equipment connected to the AC grid and your supply does not appear to be grid tied.

Without see the input current waveform I cannot comment on the likely root cause of the distortion but give the fact that you are in the early stages of development there may be a number of factors affecting the ac line current waveshape.

What was the operating condition that you observed the ac line current in the prototype?

Regards

Peter

Hi Peter,

Thanks for the info so far.

I need to correct the PF for a generator output.

Unfortunately, I was not able to use the design tool that you mentioned because it needs IT approval.

The input line current was not distorted ,only the out one was and that indicates a low PF .We connected the circuit to the generator and measured the output at the scope.

I am not sure if that answered your question about the operating condition.

I am currently wrapping up the component value re-calculation and selection because some of the values that my college have selected where not optimal .

I will be retesting the circuit on Monday so any advice from your end would be appreciated.

Russell

Hi Peter,

One quick question I had is regarding the value for the hold time used in the calculations,My college has used a value of 6ms but the data sheet used 34ms.

Should I stick to the 34ms value,or is there a way to actually calculate that value?

Thanks

Russell

Hi Peter,

I so far have not been able to download the design tool on my work computer due to delay from IT,would you be able to run it and let me the recommended inductor value for that circuit?

Russell

Hi Russell

Is the hold up time important to your application? Hold up time is the normally important in application where the load needs to stay alive, like a PC, during an short AC line outage for 1/2 a cycle or maybe 1 full cycle.

What conditions do you want me to use as inputs to the design tool, i.e. min input voltage, output voltage, load power, freq etc

Regards

Peter

Hi Peter,

I don't need a long hold up time,if any.

The conditions that I have are as follows:

Vin min:20 VAC.

Vout :90 VAC

Load power:70 W

Freq:47-65 Hz.

Switching Frequency:100 KHz

Russell

Hi Russell,

I made a couple of changes as the tool works in DC,

Vin min 28Vdc (20Vac * 1.414)

Vout 130Vdc (90Vac * 1.414) note output will always be DC

If we set Vout at 130Vdc then we should leave some head room between the output and the input voltage so I suggest Vin max of 80Vac or 113Vdc.

See slide set attached for power stage designer component values and waveforms

Regards

Petere2e Boost Circuit.pptx

Hi Peter,

I finally got the tool installed,but when I entered the values I  got a much smaller recommended value for the Inductor(50uH),the only difference from your design was the 1.4 factor and the output current value where I used 1.5 which is too high,.5-.8 makes more sense.

I also am wondering if the tool is capable of giving suggested values for the capacitors in the circuit.

Hi Peter,

Manually ,I got 160 uH for the inductor using the formula L=(Vin,minixD)/(FsxIripple)

Vinmin =20VAC

D=.8

This inductor value is still much less that the value you got from the tool.

Any feedback would be appreciated

Russell

Hi Russell

I recommend you match the values in my design and then change one item at time to migrate the design to one that you believe matches your requirement better.

If you still have difficulties please send a snapshot of your design from the tool.

Thanks

Peter

Hi Russell,

I hope you are making progress on your project. Since there has been no activity on this post for a number of days now I am going to close this post. Please open a new post if you have more questions.

Regards

Peter