SMPS circuit for 230V AC to 221V DC 20mA output, 1% ripple, DCM, N-channel MOSFET

243 13

#1 21660807 21 Jan 2012 01:44

Charith Mendis Charith Mendis

Anonymous

Post #1
21660807 21 Jan 2012 01:44

I have been given to design a SMPS for the following specifications.

Vin = 230V ac 50 Hz
Vout = 221V DC
Iload = 20mA DC
output ripple = 1% of the output voltage

N-channel High Side MOSFET, and should work in DCM. I think this specification is little awkward and I am unable to design a specific circuit.

Can anyone help me on this?
ADVERTISEMENT
#2 21660808 21 Jan 2012 01:45

Charith Mendis Charith Mendis

Anonymous

Post #2
21660808 21 Jan 2012 01:45

I forgot, they wanted this done by a buck converter. It's become even harder.
#3 21660809 21 Jan 2012 12:37

Todd Hayden Todd Hayden

Anonymous

Post #3
21660809 21 Jan 2012 12:37

At what point are you stuck? Wouldn't you normally use a buck converter to regulate to a lower voltage, why is that harder?
ADVERTISEMENT
#4 21660810 21 Jan 2012 15:35

Cody Miller Cody Miller

Anonymous

Post #4
21660810 21 Jan 2012 15:35

The design then should be broken up into two parts.

* The 230 VAC to an intermediate DC voltage.
* The intermediate DC voltage to Vout.

What kind of budget are you looking at for this project. Can you buy modules like these "AC to DC modules":http://www.vicorpower.com/cms/home/product-selector?productCategory_filter=AC-DC%20Power%20Systems
ADVERTISEMENT
#5 21660811 21 Jan 2012 19:24

Steve Lawson Steve Lawson

Anonymous

Post #5
21660811 21 Jan 2012 19:24

Silly question, but, for that small amount of current and slight voltage drop, isn't SMPS overkill? How about a simple linear regulator? We're talking a mere 5th of a watt, here!
#6 21660812 21 Jan 2012 22:13

Charith Mendis Charith Mendis

Anonymous

Post #6
21660812 21 Jan 2012 22:13

Yes, but the absurdly high voltage at the output makes it hard to make 5-15V needed for the controller Vcc. I can't use an external power. With an isolated topology I could have achieved this.
#7 21660813 21 Jan 2012 22:14

Charith Mendis Charith Mendis

Anonymous

Post #7
21660813 21 Jan 2012 22:14

I am stuck at finding a good MOSFET driver and a controller. I have done the design for the buck and it works in PSPICE.
#8 21660814 21 Jan 2012 22:17

Charith Mendis Charith Mendis

Anonymous

Post #8
21660814 21 Jan 2012 22:17

It can't be a isolated supply. The budget is
#9 21660815 21 Jan 2012 22:21

Charith Mendis Charith Mendis

Anonymous

Post #9
21660815 21 Jan 2012 22:21

Yes, you are right. But, I need to design according to their specification.

Can you specify good MOSFET drivers and a controller IC for this? Also I need their PSpice models. I need to do a simulation myself to get sure. Also, method to drive the Vcc of ICs is a problem. I can't use external power.
#10 21660816 21 Jan 2012 22:24

Charith Mendis Charith Mendis

Anonymous

Post #10
21660816 21 Jan 2012 22:24

Hello, by intermediate voltage do you mean 325V(the peak) got from a smoothed bridge rectifier?
I am using it currently.
ADVERTISEMENT
#11 21660817 21 Jan 2012 22:31

Steve Lawson Steve Lawson

Anonymous

Post #11
21660817 21 Jan 2012 22:31

How much regulation is needed?
#12 21660818 22 Jan 2012 02:35

Charith Mendis Charith Mendis

Anonymous

Post #12
21660818 22 Jan 2012 02:35

1% of the Vout as the output ripple.
#13 21660819 22 Jan 2012 11:45

Todd Hayden Todd Hayden

Anonymous

Post #13
21660819 22 Jan 2012 11:45

Why are they mandating an N channel high side switch (why do they care)? Is this homework, paid work, or hobby?
#14 21660820 22 Jan 2012 13:06

Charith Mendis Charith Mendis

Anonymous

Post #14
21660820 22 Jan 2012 13:06

Actually this is a favour asked from a member in our faculty, eventhough it seems absurd.
Create an account, log in here. You will receive points by participating in discussions.
Join this discussion.

Install Elektroda application

Didn't find an answer? Ask Artificial Intelligence

*I agree to send the question to OpenAI, Anthropic PBC, Perplexity AI, Inc., Kagi Inc., Google LLC - owners of language models in order to prepare the best response. The companies may monitor and log information entered into the form.

*I agree to publicly display my question and answer. The question and answer will be publicly available to everyone. The process may take a few minutes. Upon completion, you will be redirected to the page with the answer.

Wait...(2min)

Topic summary

✨ The discussion centers on designing a switched-mode power supply (SMPS) converting 230V AC at 50 Hz to 221V DC output at 20mA with a 1% output voltage ripple, using an N-channel high-side MOSFET operating in discontinuous conduction mode (DCM). The design challenge is compounded by the requirement to use a buck converter topology, which is unusual given the output voltage is close to the input peak voltage (~325V DC after rectification). The designer is constrained by a non-isolated topology and cannot use external power for the controller Vcc, complicating the choice of MOSFET drivers and controller ICs. Simulation in PSPICE has been done for the buck converter stage, but selecting suitable driver ICs with available PSPICE models remains problematic. The project is a favor rather than commercial work, with a limited budget and strict adherence to specifications despite the low load current (20mA) and minimal voltage drop, which some suggest might make SMPS overkill compared to linear regulation. The output ripple specification is confirmed at 1% of 221V. Suggestions include splitting the design into two stages: AC to intermediate DC voltage (around 325V) and then to the regulated output. The rationale for mandating an N-channel high-side switch is unclear.
Generated by the language model.

FAQ

TL;DR: Design uses 230 VAC → buck to 221 VDC with 1% ripple; “output ripple = 1% of the output voltage.” [Elektroda, Charith Mendis, post #21660807]

Why it matters: Engineers searching “how do I build a tiny high‑voltage buck from mains” get constraints, pitfalls, and thread‑verified context fast.

Quick Facts

Specs: Vin 230 VAC, Vout 221 VDC, Iout 20 mA, non-isolated brief stated by OP. [Elektroda, Charith Mendis, post #21660807]
Ripple target: 1% of Vout (≈2.21 Vpp at 221 V). [Elektroda, Charith Mendis, post #21660818]
Topology request: Buck converter; designer noted increased difficulty. [Elektroda, Charith Mendis, post #21660808]
Mandates: DCM operation and N‑channel high‑side MOSFET switch. [Elektroda, Charith Mendis, post #21660807]
Intermediate bus used by OP: ≈325 VDC after rectifier and smoothing. [Elektroda, Charith Mendis, post #21660816]

What’s the simplest high-level architecture for 230 VAC to 221 VDC at 20 mA?

Use a rectifier and bulk capacitor to create a high-voltage DC bus, then a non‑isolated buck stage in DCM to 221 V. This mirrors the thread’s requested buck approach and maintains low output current with manageable magnetics. Keep clearances consistent with high voltage practices. [Elektroda, Charith Mendis, post #21660808]

Why not just use a linear regulator here?

A participant suggested SMPS might be overkill, but constraints include high output voltage and controller power challenges. Linear drop from a rectified bus would waste heat and still not solve auxiliary Vcc. “We’re talking a mere 5th of a watt, here!” reflects skepticism, not a final design choice. [Elektroda, Steve Lawson, post #21660811]

How do I meet the 1% ripple requirement at 221 V?

Target about 2.21 Vpp ripple. In DCM buck, size the output capacitor and switching frequency to keep the inductor current fully discontinuous while meeting ripple. Validate with simulation and measure at load. “Output ripple = 1% of the output voltage.” [Elektroda, Charith Mendis, post #21660818]

Is isolation allowed for this supply?

No. The requester explicitly disallowed an isolated topology. That restriction drives the choice toward a non‑isolated rectified bus plus buck stage and complicates safe referencing and measurements during development. [Elektroda, Charith Mendis, post #21660814]

What intermediate DC voltage should I expect after rectifying 230 VAC?

Expect about 325 VDC from a smoothed bridge (230 VAC RMS × √2). The OP confirmed using this intermediate node before the buck stage. Ensure components are rated for this bus. [Elektroda, Charith Mendis, post #21660816]

Can I power the controller Vcc from the high output if no external supply is allowed?

The OP notes Vcc generation is difficult due to the high output voltage and the no‑external‑supply rule. Plan startup and housekeeping power early, since this constraint impacts driver selection and reliability. [Elektroda, Charith Mendis, post #21660812]

Why mandate an N‑channel high‑side MOSFET in DCM—what’s the impact?

Mandating an N‑channel high‑side device in DCM narrows driver choices and startup strategies. The OP flagged the requirement, which increases design effort for a small‑power supply. Choose parts and topology with this constraint in mind. [Elektroda, Charith Mendis, post #21660807]

What MOSFET driver and controller ICs were recommended in the thread?

The designer asked for specific MOSFET drivers, controller ICs, and PSpice models but none were named in‑thread. The open request highlights parts selection and simulation models as a blocker. [Elektroda, Charith Mendis, post #21660815]

Is using off‑the‑shelf AC‑DC modules acceptable here?

One reply proposed splitting the job into two parts and considering AC‑to‑DC modules. That can reduce risk and speed development if specifications and budget allow. [Elektroda, Cody Miller, post #21660810]

What do other engineers think about using a buck here?

One respondent asked why a buck would be harder, since bucks normally step down. This underscores that perceived difficulty stems from the high voltage, Vcc generation, and high‑side drive—not the step‑down action itself. [Elektroda, Todd Hayden, post #21660809]

How tightly regulated must the output be?

The thread clarifies the requirement as 1% ripple at the output voltage level. Treat this as a hard target during simulations and component selection. [Elektroda, Charith Mendis, post #21660818]

What special constraints did the requester impose beyond specs?

The designer couldn’t use an external power source for controller Vcc, and isolation was disallowed. These constraints directly affect driver choice and startup circuitry. [Elektroda, Charith Mendis, post #21660812]

Who asked for the N‑channel high‑side switch, and why?

A participant questioned the mandate’s purpose. The designer said it was a favor for a faculty member, acknowledging the requirement may be arbitrary. [Elektroda, Todd Hayden, post #21660819]

Did the OP simulate the buck stage successfully?

Yes. The OP reports a working buck design in PSpice, though hardware driver/controller selection remained unresolved. Use simulation to verify ripple and DCM boundaries before prototyping. [Elektroda, Charith Mendis, post #21660813]

Quick 3‑step plan to de‑risk this design?

Rectify 230 VAC to a smoothed HV DC bus and verify ≈325 V.
Implement DCM buck to 221 V with mandated N‑channel high‑side switch.
Validate 1% ripple and address Vcc without external supply. [Elektroda, Charith Mendis, post #21660816]

What’s an edge case to watch out for during bring‑up?

With no external Vcc and a high‑side N‑MOSFET, startup and driver biasing can stall progress. The OP explicitly struggled with driver and controller selection; plan this path early. [Elektroda, Charith Mendis, post #21660813]

Generated by the language model.