Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Doubts regarding 100W inverter design

Status
Not open for further replies.

sriblue7

Member level 1
Joined
May 6, 2011
Messages
41
Helped
1
Reputation
2
Reaction score
1
Trophy points
1,288
Location
Chennai
Activity points
1,614
I have some doubts regarding the design of a 100w inverter
1)is it possible to use 12v 4.5Ah Lead acid battery to provide 5hrs supply continuously?As 100W requires 0.5A(100W/220V) at the output,can i use the above battery along with some MOS amplifiers such as P55NF06(to provide a amplification of ~12 times([0.5*20(turns ratio)]/0.9))?
2)If i connect two 12V 4.5Ah lead acid battery in parallel will the Ampere-hour rating be 9Ah?
3)What is the difference and use of MOS in H-bridge configuration and push-pull configuration?
3)Where should the PWM signal be filtered to get sine wave (input/output side)and how?

Thanks in advance..........
 

I have some doubts regarding the design of a 100w inverter
1)is it possible to use 12v 4.5Ah Lead acid battery to provide 5hrs supply continuously?As 100W requires 0.5A(100W/220V) at the output,can i use the above battery along with some MOS amplifiers such as P55NF06(to provide a amplification of ~12 times([0.5*20(turns ratio)]/0.9))?

My friend calc is wrong. Theoreticaly 100W / 12V = 8,33A then 4,5Ah / 8,33A = 0,54h. But this is theory Lead acid starter batteries are defined with /10 discharging time (for deep cycle bat /20), with this condition manufacturer capacity is guaranted. This battery will work shorten then this calc showing, this happen when you dont respect /10 time.


2)If i connect two 12V 4.5Ah lead acid battery in parallel will the Ampere-hour rating be 9Ah?

Yes. But thise batteries should be in pair, with almost the same internal resistance, age, usage life,...

3)What is the difference and use of MOS in H-bridge configuration and push-pull configuration?

3)Where should the PWM signal be filtered to get sine wave (input/output side)and how?

Search edaboard threads, there is lots of post about that.
 

I have some doubts regarding the design of a 100w inverter
1)is it possible to use 12v 4.5Ah Lead acid battery to provide 5hrs supply continuously?As 100W requires 0.5A(100W/220V) at the output,can i use the above battery along with some MOS amplifiers such as P55NF06(to provide a amplification of ~12 times([0.5*20(turns ratio)]/0.9))?
2)If i connect two 12V 4.5Ah lead acid battery in parallel will the Ampere-hour rating be 9Ah?
3)What is the difference and use of MOS in H-bridge configuration and push-pull configuration?
3)Where should the PWM signal be filtered to get sine wave (input/output side)and how?

Thanks in advance..........

1:

0.5 * 220 = 110. 110 * 5 = 550.

Battery -> 12 * 4.5 = 54Ah. Assuming you discharge only upto 60%, that is 32.4Ah. So, it can provide back up for only 32.4/550 *60 = 3.5 minutes. In practice it will be even lower due to inefficiencies and losses in the converter and the internal resistance. To provide 5 hours backup, use a battery of capacity minimum (550/12)/0.6 ≈ 70Ah, assuming you discharge the battery upto 60%.

2:

You will face problems due to factors mentioned by tpetar. If you're going to be using two batteries, it's better to use them in series, so that they function together as one 24V battery.

3:

It should be filtered at the output of the bridge converter. Of course, this means that you must have a sinusoidal PWM input. You should use LC filter due to the high power (high current) at which RC filter would have too high losses. First you need to tell us how you're driving the bridge. Are you using SPWM from a microcontroller? If you're using PWM for square wave, then you shouldn't try to get sine wave from the output by using filters.

You can filter the signal at the input and use a power amplifier (maybe using BJTs) to drive the load. But you'll find out that this will be very inefficient.

Hope this helps.
Tahmid.
 

Thank u Tahmid and tpeter!
But i still have some doubts.
1)In the following link,does increasing the no. of MOS and wattage of transformer increase the output power for the same rating of battery?How does connecting MOS in Darlington improve the design? How can i modify this circuit to produce a sine wave output?
https://www.edaboard.com/threads/205764/
2)How to get the sine wave after comparing a reference sine wave(50Hz) with a high frequency triangular wave in SPWM technique?is SG3524 an alternative to the comparator ?
3)What is switched mode inverter and how to design it ?
 

1) Increasing number of MOSFETs and wattage of transformers increases output power capacity for the same battery. But since battery is the same, backup time will reduce for a greater power output. But the circuit can be improved. R2 and R3 should be about 10 to 22 ohms. Resistance of R3, R5 should be reduced as well.
2) You need to use SPWM. I think it would be difficult to do that without a microcontroller. Circuit will become complex.
3) You should read here:
https://www.industrial-electronics.com/switching-pwr-supply_4.html
Here is the basic concept:

Hope this helps.
Tahmid.
 

1)In the following link,does increasing the no. of MOS and wattage of transformer increase the output power for the same rating of battery?
If you're asking whether having higher power mosfets or transformers allows you to break the laws of thermodynamics, then the answer is no.

Your fundamental issue is that your stated specification (100W for 5 hours) can't be met with one 12V 4.5Ah battery, no matter what kind of circuitry you use. You will need at least ten times the energy capacity to meet your spec. So either get a bigger battery, or combine many small batteries.
 

Thanks for all!!!
Now i have got an outline of the design.
I am going to use 12V 135Ah battery and design a 1000W inverter(i don't mind even if it comes for less than 1 hour)
The inverter will be a switched mode type.
*I will convert first 12V dc to 320V dc(DC-DC converter) using SG3525 ,a High frequency transformer(such as ferrite core) and rectifiers(fast switching diodes).
*I will use SPWM technique to closely approximate a sine by comparing a high frequency triangular wave with a sine reference and use a full H-bridge to pull the PWM waveform to 320VAC(peak value for 220Vrms) and use a LC filter to get the final approximated sine output.
Is my design ok?will it be efficient ?
Can someone please tell me the frequency that i have to use in DC-DC converter,rating of the ferrite core transformer(if possible the calculations) and design of LC filters at the output?
I am not going to design a ferrite core transformer so it will be better if u can suggest me a practically available transformer in the market along with its rating?
Thanks in advance!!!!!
 

The basic idea is fine. Efficiency will depend on your implementation of the idea and the parts used.

You can start with 50kHz. That should be okay. You can change the frequency if you think necessary. You can go down to 30kHz or up to 100kHz or more.

LC filter will depend on your implementation of SPWM. You can use trial and error to find the necessary capacitance and inductance.

For 1000W, you can use an ETD54 core. You can increase the frequency to use the transformer at higher power. But at 50kHz, you should be able to use ETD54 core. Maybe even ETD49 core, depending on the core. But you say that you won't design the transformer. But maybe when purchasing a core, you can take a look at the core used.

Hope this helps.
Tahmid.
 
Thanks Mr.Tahmid...
is there any other specs regarding the ferrite transformer such as primary/secondary current,voltage,turns ratio,frequency of operation,saturation etc?
to provide a stable output voltage independent of load changes,should i sample the output (across C) ,step down and rectify it
and feed it to SG3525 of SPWM section or the one at DC-DC converter section?
 
Last edited:

Well, you need to select a transformer that can operate at the required current and voltage.

If you're using push-pull topology, you'll need to choose a transformer with centre-tapped primary.

The transformer must be able to operate at >100kHz. You should check the maximum flux density before saturation, but in general, it shouldn't be a problem.

For stable output, measure the output across C, and feed it to SG3525 section. You can keep the SPWM section fixed to simplify the design. The SG3525 can make adjustments to keep the output fixed by varying the high voltage DC bus voltage, provided you choose a transformer with a high enough voltage margin to allow the adjustment.

Hope this helps.
Tahmid.
 

What is the use of push-pull?
Does the following links use push-pull config?
**broken link removed** (posted by tahmid)
https://www.edaboard.com/threads/227764/ (posted by tpeter)
https://obrazki.elektroda.pl/45_1334092015.jpg (posted by goldsmith)

Should i use a center-tapped HF transformer if i use a 12v-> sg3525 -> H-bridge ->high DC voltage(320v)?
Are ferrite transformers available with center tapped for the specifications i mentioned?
 

And what MOS should i use for the 2 H-bridge in my design?
is IRF3205 sufficient for 1000W power?
 

What is the use of push-pull?
Does the following links use push-pull config?
**broken link removed** (posted by tahmid)
https://www.edaboard.com/threads/227764/ (posted by tpeter)
https://obrazki.elektroda.pl/45_1334092015.jpg (posted by goldsmith)

Should i use a center-tapped HF transformer if i use a 12v-> sg3525 -> H-bridge ->high DC voltage(320v)?
Are ferrite transformers available with center tapped for the specifications i mentioned?


I can answer your push-pull question. I think that ,in all the links you have mentioned, they have used push-pull converters. See wikipedia for push-pull converters. They are like two forward converters. In push-pull converters the direction of the voltage in the primary coil changes in both directions (which is done using those FET switches that you can clearly see in the diagrams ). This helps in giving an a.c. like voltage to the transformer. Also this has lesser ripple when compared to the flyback converter mainly because of the fact that we have both directions for voltage in the primary unlike the flyback converter where one part of the primary has no current flowing through it. I am also a beginner in Power Electronics so better wait till you get answers from the experts like Goldsmith, Tahmid.

Hope this Helps.

SherlockBenedict.
 

What is the use of push-pull?
Does the following links use push-pull config?
**broken link removed** (posted by tahmid)
https://www.edaboard.com/threads/227764/ (posted by tpeter)
https://obrazki.elektroda.pl/45_1334092015.jpg (posted by goldsmith)

Should i use a center-tapped HF transformer if i use a 12v-> sg3525 -> H-bridge ->high DC voltage(320v)?
Are ferrite transformers available with center tapped for the specifications i mentioned?

The third link does not use the 2-switch push-pull configuration mentioned.
If you're using the 2-switch push-pull topology, you will need a center tapped transformer. If you use H-bridge, you don't need center tapped transformer.

If you're using H-bridge in the 12V to 320V DC-DC converter stage, then you can use IRF3205's. Connect a few in parallel. Decide on the number based on efficiency requirements, losses and heatsinking/cooling.

In the 320VDC to 220VAC DC-AC converter, you need to use MOSFETs rated for >320V VDS. You can safely use 500V MOSFETs.

Hope this helps.
Tahmid.
 

Thank u Tahmid...
1)i have ordered for EE55 core center-tapped(50khz).So can i use IRF3205 with Vds=55V,Id=110A,rds=8 mohms(2 in the upper leg and 2 at the bottom) in push-pull configuration in Dc-Dc section and similarly IRF740 with Vds=400V,Id=10A,rds<0.55 ohm for 320VDC-220VAC section?
If i go for larger Vds,i am getting lower Id.Is it possible for IRF740 to withstand 1000W power or should i use any other MOS with Vds>320V and Id>100A ?

2)To get 50khz square wave in DC-DC section and 50hz SPWM wave at DC-AC section can i use microcontroller like atmega16(because SG3525 has many external components) or i should use only SG3525 ics ?Does this have any special feature which can't be provided by Atmega16?

3)How can i control the output voltage changes due to varying load ,using a microcontroller?

Thanks in advance.......
 
Last edited:

Thank u Tahmid...
1)i have ordered for EE55 core center-tapped(50khz).So can i use IRF3205 with Vds=55V,Id=110A,rds=8 mohms(2 in the upper leg and 2 at the bottom) in push-pull configuration in Dc-Dc section and similarly IRF740 with Vds=400V,Id=10A,rds<0.55 ohm for 320VDC-220VAC section?
If i go for larger Vds,i am getting lower Id.Is it possible for IRF740 to withstand 1000W power or should i use any other MOS with Vds>320V and Id>100A ?

2)To get 50khz square wave in DC-DC section and 50hz SPWM wave at DC-AC section can i use microcontroller like atmega16(because SG3525 has many external components) or i should use only SG3525 ics ?Does this have any special feature which can't be provided by Atmega16?

3)How can i control the output voltage changes due to varying load ,using a microcontroller?

Thanks in advance.......

1) With 1000W load at 10.5V (minimum battery voltage) and assuming 80% efficiency, the current flowing will be approximately 120A. So, you would need to parallel at least 2 IRF3205's for each of the H-bridge legs. With 2 IRF3205's in each leg, the maximum power dissipated is about 58W. That is too large a loss. So, you should parallel more MOSFETs to obtain greater efficiency with lower losses. Using 5 IRF3205's in parallel gives maximum loss of about 23W. That is better. So, you could parallel 5 or more IRF3205's in each leg. So, you would need a total of 20 IRF3205's. Or maybe even 4, depending on the allowable losses.

However, as you see here, the current is huge. Maybe it's better if you use 24V or 48V input. Then, you'll have lower currents and thus will need less MOSFETs and have lower losses with the same number of MOSFETs.

2) In the DC-DC section, it's better to use SG3525 as it is quite simple to use and you can have greater control than by using microcontroller for 50kHz output. For the 50Hz SPWM, use the microcontroller.

3) Use the SG3525 for the DC-DC converter section. Take feedback sampled from the sine wave output so that SG3525 can compensate for output voltage variation by PWM (changing the duty cycle) and by adjusting the high-voltage DC bus voltage. So, the SPWM section can be kept fixed, with SG3525 doing the error correction.

Hope this helps.
Tahmid.
 

Thank u Tahmid...
so i should use 10 IRF3205 for the DC-DC section.
1)How about the DC-AC section can i use IRF3205(10 nos) again?its Vds is only 55v ,but the dc supply is 320V.can it withstand?
2)Is it possible to reduce the number of MOS by replacing IRF3205 with any other MOS type?
3)What diode type can i use in rectification in the DC-DC section?
Thanks in advance.....
 

Thank u Tahmid...
so i should use 10 IRF3205 for the DC-DC section.
1)How about the DC-AC section can i use IRF3205(10 nos) again?its Vds is only 55v ,but the dc supply is 320V.can it withstand?
2)Is it possible to reduce the number of MOS by replacing IRF3205 with any other MOS type?
3)What diode type can i use in rectification in the DC-DC section?
Thanks in advance.....

1) If you're using H-bridge, you need 20 IRF3205s. But if you're using push-pull, you need 10 IRF3205s. You can increase the number to reduce conduction loss and improve efficiency.

IRF3205 can not withstand 320V VDS. You can use IRFP450 or IRFP460. You can also use IRF840 with higher conduction loss.

2) IRF1404 has about half the RDS(on) as IRF3205. So, you can use 3 IRF1404s in parallel instead of 5 IRF3205s. If you're happy with higher losses and greater heat, 2 IRF1404s in parallel may also suffice.

I've said this before and ask again, why not use 24V instead of 12V?

3) You have to use high-voltage ultrafast diodes. You can use MUR860. Use 4 to make a bridge rectifier.

Hope this helps.
Tahmid.
 

1)
3) Use the SG3525 for the DC-DC converter section. Take feedback sampled from the sine wave output so that SG3525 can compensate for output voltage variation by PWM (changing the duty cycle) and by adjusting the high-voltage DC bus voltage. So, the SPWM section can be kept fixed, with SG3525 doing the error correction.

Hope this helps.
Tahmid.

What will happen if feedback is taken from the 320V DC (i.e. output of DC-DC stage) instead of taking feedback from the ac output ?
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top