[SOLVED] Help with Hybrid car Converter.

Hi there.

I've been assigned a task of designing a DC converter to charge the battery of an hybrid vehicle. The problem is, the motor through regenerative braking will be supplying Voltage/current between 20-24V, 40-100Amps to the converter. So please is there converter i could use to produce 28V and 10Amps output?

I was going to use synchronous buck converter using LTC3789 but the chip couldn't handle the input current, so i'm confuse how to go about it.

Thanks for any suggestion in advance.

I believe you need a boost converter.

There are IC's that will control a mosfet in one or both switch positions. However in the early development phase, it will be easier to use a diode going to the output stage.

The screenshot is a simulation running a simple schematic.



All components will need to be robust since they must endure several amperes peak.

The capacitor may or may not be needed. It is there to show that the volt level can reach high enough to charge a 24V battery.

BradtheRad said:

I believe you need a boost converter.

There are IC's that will control a mosfet in one or both switch positions. However in the early development phase, it will be easier to use a diode going to the output stage.

The screenshot is a simulation running a simple schematic.



All components will need to be robust since they must endure several amperes peak.

The capacitor may or may not be needed. It is there to show that the volt level can reach high enough to charge a 24V battery.

Click to expand...

Hi, thanks for your reply, please is that circuit capable of achieving up to 90% efficiency or more? Because it is among the requirement for the design of the DC converter.

Acccording to your numbers, the converter is fed with 800 to 2400 W input for 280 W output. So if you want 90% efficieny, where's the breaking resistor?

meartisch said:

Hi, thanks for your reply, please is that circuit capable of achieving up to 90% efficiency or more? Because it is among the requirement for the design of the DC converter.

Click to expand...

It is only during diode On-time, that any charge goes to the battery.

The waveform through the diode appears to be 60 percent of the area of the waveform through the coil. This implies that only 60 percent of the current consumed goes toward doing useful work.

Seeing this, I was going to say that 40 percent of the power is wasted.

However the current is recirculating back to the motor. The 40 percent is not necessarily being dissipated as heat.

I put low-ohm resistors in the circuit, to represent the fact that there is some amount of resistance. However that is not sufficient to act as a load.

Anyway if I were to estimate the efficiency of this converter, it might be as much as 90 percent.

Braking time being short / very short phase and to recovery as much possible energy I think you need to extract braking energy in this short time; battery is not able to absorb energy is so short time; one solution is to use ultracapacitors and other (cheaper but not so good) to increase charging current at maxim allowed by battery design/model. Both method need custom converter to extract braking energy and "send" it in storage device.

BradtheRad said:

It is only during diode On-time, that any charge goes to the battery.

The waveform through the diode appears to be 60 percent of the area of the waveform through the coil. This implies that only 60 percent of the current consumed goes toward doing useful work.

Seeing this, I was going to say that 40 percent of the power is wasted.

However the current is recirculating back to the motor. The 40 percent is not necessarily being dissipated as heat.

I put low-ohm resistors in the circuit, to represent the fact that there is some amount of resistance. However that is not sufficient to act as a load.

Anyway if I were to estimate the efficiency of this converter, it might be as much as 90 percent.

Click to expand...

Hi BradtheRad, A very silly question, please can you confirm my attached schematic is similar to your proposed design?

In addition, I've been having problems designing the Resistor, Inductor and the Capacitor to produce the required output current(Even in simulation). Can you please suggest ways to achieve this? Most text books seems to have different formulas for the boost converter.

Thanks.




Big thanks to others for their suggestions

- - - Updated - - -

iop95 said:

Braking time being short / very short phase and to recovery as much possible energy I think you need to extract braking energy in this short time; battery is not able to absorb energy is so short time; one solution is to use ultracapacitors and other (cheaper but not so good) to increase charging current at maxim allowed by battery design/model. Both method need custom converter to extract braking energy and "send" it in storage device.

Click to expand...

Thanks for the suggestion, i think using ultracapactors will make it more complicated, especially for me just learning about power electronics? The whole idea of this project is to show the regenerative braking works.

Regarding schematic, I need to tell that generated voltage is not constant, it's function of motor (generator) speed. For DC motor with permanent magnets or separate coil field it's a liniar function of speed (e = k*n); so, if voltage it's not constant may need some mechanism (to extract maxim braking energy) to increase/mentain voltage or current injected in battery or ultracaps.

iop95 said:

Regarding schematic, I need to tell that generated voltage is not constant, it's function of motor (generator) speed. For DC motor with permanent magnets or separate coil field it's a liniar function of speed (e = k*n); so, if voltage it's not constant may need some mechanism (to extract maxim braking energy) to increase/mentain voltage or current injected in battery or ultracaps.

Click to expand...

True that's another concern i'm having. Please do you have a suggestion as to which mechanism i could use or design for this constant voltage and current.

meartisch said:

Hi BradtheRad, A very silly question, please can you confirm my attached schematic is similar to your proposed design?

Click to expand...

Yes.

It's only a start, of course.

If the window to recover energy is short, then you need to absorb as much energy as possible (per posts #4 and 6).

You might consider using twin or triple interleaved boost converters.

In addition, I've been having problems designing the Resistor, Inductor and the Capacitor to produce the required output current(Even in simulation). Can you please suggest ways to achieve this? Most text books seems to have different formulas for the boost converter.

Thanks.

View attachment 84313

Click to expand...

I forgot to state the controlling frequency.
1 kHz.
40% duty cycle

1 kHz is slow. It is common to use a much higher frequency. Also a lower Henry value, and a smaller coil.

Hi, After testing the suggested boost converter with the Dc motor, i was unable to get a constant output and in some case i got current going up to 99 Amps. So please what can i do to get a constant output?

In addition, is there any book you can recommend on Dc converters, as i can't seem to find textbooks on Boost or Double interleaved converters so it's just a game of playing around with the values to get result on simulations which is not working.

Any help by others will be greatly appreciated

Wow, there seems to be several stages and several interleaved converters in schematic #2.

I was picturing this sort of interleaved converter:



It shows the battery getting 54 A peak. To reduce this amount, reduce the duty cycle on all the clocks.

To get all the pulse trains right will require some effort.

The capacitor may or may not be necessary. It is there to show that the converter can deliver as high as 29 V, in order to charge the battery.

Of course this is only a theoretical simulation, to illustrate the concept.

BradtheRad said:

.

Click to expand...

Hi BradtheRad, (Final requests).
I have resorted to using the simulation(falstrad) u are using. However, as shown by the attached Picture, i couldn't get the same result with the same specification you presented. I was just wondering if you can send me the export link of the design, or otherwise specify possible ways to acheive the same result.

And finally do you have any textbook you can recommend with similar design, because i need to understand it better for my technical report.

Thanks for all your help.

Edit power supply is +20 Volt same as ur design.

Good work. You just about had it all there.

Click this link to open falstad.com/circuit, load my schematic, and run it on your computer. (Click Allow to load the Java applet.)

http://tinyurl.com/amy6dqt

The clocks are staggered at 90 degree intervals. Each has to be adjusted individually in its Edit window.
Also the frequency and duty cycles must be adjusted individually. In a hardware version a microcontroller would be able to provide the proper pulse sequence, and vary duty cycle as necessary.

The design might be improved by taking out the diodes and substituting mosfets/transistors (known as synchronous topology). This would reduce losses due to diode drop.

I guess you know how to copy a schematic to the clipboard by selecting Export.

And finally do you have any textbook you can recommend with similar design, because i need to understand it better for my technical report.

Click to expand...

This is a very demanding application, namely to take 1 or 2 kilowatts of energy and pack that into a battery in real time. There may very well be better designs that can do the job, although they may be a 'trade secret' or somesuch.

In order to avoid wasting any braking power, the most efficient solution may involve supercapacitors or ultracapacitors (per post #6).

Here's a couple of websites that discuss boost converters.

http://www.simonbramble.co.uk/dc_dc_converter_design/boost_converter/boost_converter_design.htm

Hi meartisch,

I am also struggling with regenerative braking in my project. But, in my case, i need a bidirectional DC-DC Converter in order to charge and discharge the storage battery.
(you could take a look in the following link if you wish)
https://www.edaboard.com/threads/279800/

I am using Matlab/Simulink for my simulations and after a lot of searching i have found a simulink model of a HEV, which is attached below. I don't know if you would like to use Simulink (it is a good program, but it is really demanding), but the attached model is fully functional and the DC-DC converter used for charging the battery in this model is what you need for your project (in my opinion). I would be happy if the attached file could help you.

Best regards,
strkar

strkar said:

Hi meartisch,

strkar

Click to expand...

Thanks a lot for that file sharing, i was going to model my circuit in simulink after i finish this part of the project. That will surely go a long in having an idea about the modelling design. I'll try and help you ask my friends if they can find similar model to help you in your project.

- - - Updated - - -

Hi, Thanks!! The Circuit has been finally done except for the clock, please how do can i design a square wave that will give 90 degree interval respectively? My guess was to use an or Gate at the output of the square wave generator to achieve 90 while also using two 'OR gate' in series to achieve 180 and soon?

Final question!!
For educational purpose. Can u please briefly explain how the circuit works if u have time, because while i understand how conventional Dc converter operates, i don't understand how this circuit manages to have better efficiency and low ripple current compared to the conventional ones. Does the same principle of twin converter apply, apart from the extra phase shift, which i guess is used to reduce to the voltage stress on the switches, hence improving efficiency?

Thanks a lot for your help and apologies for the trivial questions.

You are very welcome my friend! It's my pleasure to know that i was able to help someone.
I would also be very grateful if you could send me the simulink model of a bi-directional DC-DC converter. It will really help me a lot, so please ask your friends if you wish. Thanks in advance!

My friend, there are no trivial questions... there are only questions which need answers, so now:

For your first question:
If you are referring to your attached circuits(photos), i'm sorry but i can't help you because i'm not familiar with this simulation program. But, if you are referring to the simulink model which i attached, what you need is already implemented. You can make your configurations in the control blocks and the 'powergui'.

For your second question:
I'm a bit confused as i can't really understand to which exact circuits you are referring. But, generally speaking, i will tell you that (i'm sure you already know most of these):
There are a lot of different types of DC-DC converters. There are isolated(with transformer) and non-isolated ones. There are one way(for the power flow) and bidirectional ones. There are conventional and some with novel circuit topology ones. Some have half bridge, while others have full bridge. What i'm trying to say is that every circuit topology for a converter has its own pros and cons. But, apart from the topology, the control of the gates is of paramount importance in order to improve efficiency. Soft switching techniques (such as zero voltage switching-ZVS) improve efficiency for sure. But efficiency depends on a lot of different parameters.

Hope these help you a bit...

meartisch said:

Hi, Thanks!! The Circuit has been finally done except for the clock, please how do can i design a square wave that will give 90 degree interval respectively? My guess was to use an or Gate at the output of the square wave generator to achieve 90 while also using two 'OR gate' in series to achieve 180 and soon?

Click to expand...

I have used a 4017 IC to accept a pulse train, and rotate the pulses among four channels.

That way you would need just one pulse generator, and one adjustment of frequency and duty cycle. It would be easier than creating four PWM clocks each needing individual adjustment.

Final question!!
For educational purpose. Can u please briefly explain how the circuit works if u have time, because while i understand how conventional Dc converter operates, i don't understand how this circuit manages to have better efficiency and low ripple current compared to the conventional ones.

Click to expand...

By adding a second coil, you have a larger conduit to absorb the energy coming down the line. With a single converter you operate at a certain duty cycle (say it's 50 percent). Heavier wires are needed if you are going to carry all that current in one converter.

By adding a second converter you absorb a greater percentage of the energy in the limited time window that the energy is available. You can also get by with thinner wires, and use smaller coils.

With multiple converters you are also presenting less resistance to the incoming current. Reducing I_squared_R losses.

The battery benefits as well since it is fed a continual, less drastic current waveform, instead of being forced to absorb heavier bursts of current on a 30 or 40 percent duty cycle. (The Peukert effect comes into play here.)

I am not certain how large is the jump in efficiency when going from one converter to two, or from two to three, etc. The quadruple interleaved layout is an exploration, showing (theoretically) how several coils could divide the load. Your project may work just as well with a triple interleaved... or, might find you need a quintuple.

Does the same principle of twin converter apply, apart from the extra phase shift, which i guess is used to reduce to the voltage stress on the switches, hence improving efficiency?

Thanks a lot for your help and apologies for the trivial questions. View attachment 86992View attachment 86992

Click to expand...

It occurs to me that a good comparison is a half-wave power supply to a full-wave power supply.

It's hard for me to be sure how well my theoretical design would work in a real application. So I can only speak theory. I would not be surprised if there is a more efficient way to accomplish what you want to do.

Hi Bradtherad, i did manage to successfully complete the project, so a big thanks to you. Here are my few observations, under breaking at low speed, i did get low input power which translate also to lower output power to charge the battery. After a few research, i stumbled on feedback current sensing normally used to stabilize the output voltage. The problem with all the current sensing i found, was that they are mostly used on low power input. Please can u, suggest ways or links to a design to stabilize the output voltage regardless of the input power.

Thanks.

meartisch said:

Hi Bradtherad, i did manage to successfully complete the project, so a big thanks to you.

Click to expand...

Fantastic! Congratulations. This sounds like it could be a breakthrough energy-saving system that might be installed in the next generation of fuel-efficient vehicles.

Here are my few observations, under breaking at low speed, i did get low input power which translate also to lower output power to charge the battery. After a few research, i stumbled on feedback current sensing normally used to stabilize the output voltage. The problem with all the current sensing i found, was that they are mostly used on low power input. Please can u, suggest ways or links to a design to stabilize the output voltage regardless of the input power.

Thanks.

Click to expand...

There is voltage-based feedback and there is current-based feedback.
The typical method is for the control IC to take feedback across a sense resistor. Or, a hall sensor, or an extra wire in the coil/transformer, etc.

As for stepping up low power so it will charge a 24V battery...
If you get the duty cycle adjusted to 95% or more, you can feed the boost converter a supply as low as 1V, at just a few A, and provide 28V at a few milli-amps. (Theoretically.)

Screenshot:



I have no experience with control IC's, and I don't know what range of operation you will be able to achieve, as far as duty cycle. For a boost converter it can get complicated. You need to monitor the load, and then you also need to monitor current through the coil so you can switch it off and on at the correct moments.

In order to do battery charging, I am not sure which control method you need to use (since it's not likely you can use both)... whether you want the converter to step up the supply (a) to a particular volt level, or (b) so as to maximize current going into the battery.