DC/DC Converter (Miniature Remote Control Solar Car)

[Nigel Goodwin]

Thanks for "saving" me. Thanks a lot. But I tried simulating the boost DC/DC converter without the zener diode and I couldnt obtain a constant or steady output. Is this simply because of my wrong selection of capacitor and inductor? Mind to give some hints on how should I determine the capacitor as well as the inductor value?

Did you simulate the circuit as shown above?, you don't have any control electronics - so there's no feedback. I'm also VERY dubious about the value of simulations, they rarely seem to work as the real world does.

 

[devonsc]

Rearrange

I'm afraid my posts are very messy. Sorry if it is, here I would like to rearrange a few of the latest post.

Did you simulate the circuit as shown above?, you don't have any control electronics - so there's no feedback. I'm also VERY dubious about the value of simulations, they rarely seem to work as the real world does.

Thanks, I will contruct the circuit in practical and do some measurements and observations on the waveforms through the oscilloscope as soon as possible.

:?: By the way, do you mind giving me a hint on how do I determine the proper capacitance value and inductance value to use? Thanks.

:?: By the way, do you mind briefly explain on the purpose of the existing diode in the MOSFET? Real sorry to trouble you. Thanks a lot

:?: Is the 741 Op-Amp suitable for me to use to implement in the Current-to-Voltage converter? I will post my circuit as soon as possible for advice as well. Thank you very much.

 

[Oznog]

1. You show a bipolar in the pic. A MOSFET is a much better idea for the high efficiency you require. Look for one listed as "logic level".
2. The Zener is unnecessary. It is not a good mechanism to prevent overcharging, and in any case these currents are far too low to damage the battery through overcharging. Frankly, it will be a miracle if you manage to charge the battery in the first place.
3. The capacitor is probably unnecessary. The battery will charge about the same if it's 30 mA pulsed at 50% or 15mA filtered into a constant DC current. I doubt it will do anything helpful, and if it's got a poor leakage resistance, will drain a small amount of current (under 1mA, but it's there) all the time.

The inductor size is important. If it's too low, the solar cell will be loaded with ripple current. This is not damaging, but the efficiency suffers since the solar cell is only optimally loaded at one voltage/current point and a ripple by definition means the current will vary over the period of the PWM.

 

[devonsc]

1. You show a bipolar in the pic. A MOSFET is a much better idea for the high efficiency you require. Look for one listed as "logic level".

Sorry, the Figure has not been updated. Real sorry. I'm now trying to use the BSH102 MOSFET, but I couldnt find this part here. Looking for a replacement, any suggestions?

2. The Zener is unnecessary. It is not a good mechanism to prevent overcharging, and in any case these currents are far too low to damage the battery through overcharging. Frankly, it will be a miracle if you manage to charge the battery in the first place.

Thanks for that advice, I will now remove the zener diode. But would you mind to suggest a method for me to prevent overcharging? The small current output from my panel might not overcharge the battery but I would like to get some info about methods to prevent overcharging. Mind to give some advice?

3. The capacitor is probably unnecessary. The battery will charge about the same if it's 30 mA pulsed at 50% or 15mA filtered into a constant DC current. I doubt it will do anything helpful, and if it's got a poor leakage resistance, will drain a small amount of current (under 1mA, but it's there) all the time.

Blur here. I thought the purpose of the capacitor is to perform some boost function during the charge-discharge process? :? Sorry, this seem to be a real nonsense from me. Mind to explain further?

The inductor size is important. If it's too low, the solar cell will be loaded with ripple current. This is not damaging, but the efficiency suffers since the solar cell is only optimally loaded at one voltage/current point and a ripple by definition means the current will vary over the period of the PWM.

Thanks a lot. I've tried simulating using 100mH and it seems that it works well, I don't get much ripple. But I will try again I practical.

 

[devonsc]

About the PIC that I will be using for my project, 16F819 recommended.

Nigel? The 16F819, I couldn't find this unit after trying to search for it in a few places in my country. Do you mind to recommend other PIC microcontroller? Please? Help needed.

 

[Nigel Goodwin]

About the PIC that I will be using for my project, 16F819 recommended.

Nigel? The 16F819, I couldn't find this unit after trying to search for it in a few places in my country. Do you mind to recommend other PIC microcontroller? Please? Help needed.

You don't have your location filled in!, so we don't have any idea where you might be?.

Search for PIC's you can get that have A2D built-in, the 16F876(A) is an 'older' one, but is 28 pin rather than 18 pin - but it should be available anywhere.

 

[Nigel Goodwin]

Just to complicate your life even further :lol:

If you look on the MicroChip website there is a programmer called the 'PICKIT1', this is a USB powered programmer for FLASH devices - using the 16C745 USB PIC.

The full circuit diagram and software source code is provided by MicroChip, as a USB application demonstration.

The reason I'm mentioning this is that they use the PWM module in the 16C745 to create a boost converter to generate the 13V required for switching PIC's to programming mode. The output of the converter is fed back to one of the A2D inputs which regulates the width of the PWM to keep the output stable. It does all this while at the same time receiving the USB data and programming the chip.

I'm not suggesting you use the 16C745 (as it's OTP) but the 16F876 has similar PWM modules and A2D inputs - so it opens up the possibility of doing the PSU and monitoring all in a single chip!.

Just a thought!.

 

[Oznog]

1. You show a bipolar in the pic. A MOSFET is a much better idea for the high efficiency you require. Look for one listed as "logic level".

Sorry, the Figure has not been updated. Real sorry. I'm now trying to use the BSH102 MOSFET, but I couldnt find this part here. Looking for a replacement, any suggestions?

2. The Zener is unnecessary. It is not a good mechanism to prevent overcharging, and in any case these currents are far too low to damage the battery through overcharging. Frankly, it will be a miracle if you manage to charge the battery in the first place.

Thanks for that advice, I will now remove the zener diode. But would you mind to suggest a method for me to prevent overcharging? The small current output from my panel might not overcharge the battery but I would like to get some info about methods to prevent overcharging. Mind to give some advice?

3. The capacitor is probably unnecessary. The battery will charge about the same if it's 30 mA pulsed at 50% or 15mA filtered into a constant DC current. I doubt it will do anything helpful, and if it's got a poor leakage resistance, will drain a small amount of current (under 1mA, but it's there) all the time.

Blur here. I thought the purpose of the capacitor is to perform some boost function during the charge-discharge process? :? Sorry, this seem to be a real nonsense from me. Mind to explain further?

The inductor size is important. If it's too low, the solar cell will be loaded with ripple current. This is not damaging, but the efficiency suffers since the solar cell is only optimally loaded at one voltage/current point and a ripple by definition means the current will vary over the period of the PWM.

Thanks a lot. I've tried simulating using 100mH and it seems that it works well, I don't get much ripple. But I will try again I practical.

1. Too many to mention. Just find an NMOS with a low threshold voltage.
2. Overcharge protection depends on battery type. For SLA, they use a constant "float" voltage. It is acceptable to charge at a higher voltage but once a full charge is detected- generally by the current going down with a fixed voltage, or if the voltage rises when you're using constant current.
3. The cap does nothing special. It charges up during the part of the cycle when the diode is conducting current to the load, and it discharges when it's not conduction. Otherwise the output is a string of current pulses. But a string of current pulses is probably fine for you.
4. 100mH may be physically large. Inductors are more complicated in the real world than capacitors and optimum selection is tough! Getting 100mH in a physically small pkg means a lot of turns of thin wire, making a large DC resistance. It could be 10-100 ohms. The voltage drop burns up power on both the charging and discharging phases. Also, an inductor will only carry so much current before it saturates and stops acting like an inductor and is just a DC resistance. But that's not likely with a few mA, like this in even the small ones.
Your simulation will give you a good idea how much inductance you need. The feature to design for is how smooth the current is off the solar cell. I'd look for under 5% ripple as a rough guess, preferrably half of that.

 

[devonsc]

Thanks a lot for that advice. I will read up the PIC 16F876 as soon as possible, I'm sure a slow learner like me, I will have lots of questions later on.

By the way, would like to ask for advice regarding powering the PIC microcontroller. Should I have an additional supply, say an additional battery being included in my Charging & Monitoring Unit to power up the PIC microcontroller or it is advisable to tap the voltage from my 12V battery and regulate it to the operating voltage level of the PIC?

Will "one" say that my Charging & Monitoring Unit is "no good" simply because of that additional battery to power up the PIC? :? Your advice or opinion is very much needed

Oh, by the way, I'm from Malaysia...sorry for not posting that earlier.

 

[devonsc]

:arrow: About the MOSFET, thanks. I will try searching them.

2. Overcharge protection depends on battery type. For SLA, they use a constant "float" voltage. It is acceptable to charge at a higher voltage but once a full charge is detected- generally by the current going down with a fixed voltage, or if the voltage rises when you're using constant current.

:arrow: Yup, I'm using a SLA battery (12V, 1.3AH). I still don't understand on how can I prevent an overcharge. Sorry.

3. The cap does nothing special. It charges up during the part of the cycle when the diode is conducting current to the load, and it discharges when it's not conduction. Otherwise the output is a string of current pulses. But a string of current pulses is probably fine for you.

:arrow: Got it! Thanks. Can we put it in this way : By having the capacitor, my charging process will be a little faster as it is constantly charging My phrases sounds a little funny.

:arrow: Regarding the inductor, does it mean that you suggest me to implement a 50mH unit?

Thank you very much.

 

[samcheetah]

devonsc this might be of some help to you

https://www.electro-tech-online.com/custompdfs/2004/11/slabatts.pdf

 

[devonsc]

devonsc this might be of some help to you

https://www.electro-tech-online.com/custompdfs/2004/11/slabatts-1.pdf

Thanks, does this mean that I can use a zener diode to maintain a float voltage for the battery? Is this the easiest way but power consumption method? Need advice...

 

[Nigel Goodwin]

Thanks, does this mean that I can use a zener diode to maintain a float voltage for the battery? Is this the easiest way but power consumption method? Need advice...

It all depends on the charging power you have available, if you have enough power to overcharge the battery you would need a zener diode large enough to absorb the extra power - a large zener, mounted on a large heatsink. This used to be done on Triumph motor bikes, with the zener and heatsink mounted between the forks in the airflow.

If you're still talking about solar panels it's not needed at all, you don't have enough power to cause a problem.

Try telling us EXACTLY what you are trying to do, with what voltages and currents, and why - it's been so long, and so many different threads, I'm losing count :lol:

 

[devonsc]

Try telling us EXACTLY what you are trying to do, with what voltages and currents, and why - it's been so long, and so many different threads, I'm losing count

Sorry, hope you're not mad. As shown is what I intend to do.

(Dotted lines indicate detachable while the box shows the Charging and Monitoring Unit)

All this while, we have been talking about Battery Charging Circuit 1, which is the Boost DC/DC Converter. Was thinking to develop this circuit in such away where one can use different types of solar panels. Meaning, as long as the panels are needed to be boost to a certain level, it can be used. Thus, I was thinking to develop a circuit that takes into the consideration of preventing overcharging as well. I thought there are certain panels that are capable in doing overcharging this 12V 1.3AH battery? Thanks.

And just to reconfirm about the circuit posted on page 7, entitled Boost Converter. Can I implement that, with the changes such as a MOSFET, a zener diode and the PWM from PIC?

 

[Nigel Goodwin]

All this while, we have been talking about Battery Charging Circuit 1, which is the Boost DC/DC Converter. Was thinking to develop this circuit in such away where one can use different types of solar panels. Meaning, as long as the panels are needed to be boost to a certain level, it can be used. Thus, I was thinking to develop a circuit that takes into the consideration of preventing overcharging as well. I thought there are certain panels that are capable in doing overcharging this 12V 1.3AH battery? Thanks.

I would suggest that's not a good idea, trying to design your solar panel charging to accomodate any panel size is going to require comprimises - the worst point of which is that you are going to waste much more power than you can afford with the small panels.

And just to reconfirm about the circuit posted on page 7, entitled Boost Converter. Can I implement that, with the changes such as a MOSFET, a zener diode and the PWM from PIC?

MicroChip do in their PICKIT1 programmer, have a look at it - it even generates 13V which is pretty close to what you want!.

You could prevent over charging with the PIC as well, monitor the battery voltage and if it gets too high then reduce the charging rate via the PWM.

 

[Oznog]

Regarding the inductor, does it mean that you suggest me to implement a 50mH unit?

See what it takes to get a low input ripple current in your simulation. It would be good it have it under 5%, since any power you don't use isn't "saved" for later. And start looking for what inductors you can actually get ahold of and physically fit into your project.

The zener is not accurate enough to stop an overcharge. With a current of only mA, the voltage difference between charging and overcharging is small, less than the tolerance of the zener. Worry about overcharge if your current is going to be more than the 20-hr charge rate for weeks on end. Problem is, you're going to need a voltage divider (inaccurate and consumes current off the battery) and a more accurate voltage ref.

Would like to ask regarding MOSFET. Mind to briefly explain the purpose of the diode found in the MOSFET?

It's a incidental feature of the way they're manufactured. It can usually carry a lot of current and some designs forward bias it on purpose.

By having the capacitor, my charging process will be a little faster as it is constantly charging

But the current is basically the same over time. Pulses of 30mA of current 1/3th the time (when the transistor is off) or a smoothed out 10mA 100% of the time does the same job. Actually, there are obscure, minor efficiency issues here, some for the cap, some against. It won't hurt to leave it in but it probably won't improve things.

By the way, would like to ask for advice regarding powering the PIC microcontroller. Should I have an additional supply, say an additional battery being included in my Charging & Monitoring Unit to power up the PIC microcontroller or it is advisable to tap the voltage from my 12V battery and regulate it to the operating voltage level of the PIC?

Excellent question. The PIC needs at least 3v for the lowest of freq, and may not drive the MOSFET well. 5v is standard, 5.5v will cause it to latch up. Check your spec sheet for what your part needs. A 5.1v zener can keep the supply where it should be. Also diodes can be used in series to drop the voltage. And of course a reg may work, but it consumes quiescent current and may have minimum dropout issues.

Also, a solar cell is going to drop down to a voltage which won't cause it to turn off but it won't run properly either. When the sun comes back and voltage is restored, it may not have a valid state and won't run. There are several mechanisms to enable that will compensate for this, look up "brown out detection".

Did you ever decide what your panel voltage and current were going to be? I missed that.

 

[devonsc]

I would suggest that's not a good idea, trying to design your solar panel charging to accomodate any panel size is going to require comprimises - the worst point of which is that you are going to waste much more power than you can afford with the small panels.

Sorry for my late reply. Thanks for that advice. I will now stick to one panel, at this moment, most probably the one with output voltage 7.5V.

MicroChip do in their PICKIT1 programmer, have a look at it - it even generates 13V which is pretty close to what you want!.

Thanks a lot! I will read up the PICKIT1 programmer.

You could prevent over charging with the PIC as well, monitor the battery voltage and if it gets too high then reduce the charging rate via the PWM

Hope you wont mind, I'll learn up the PIC for the PWM as well as the ADC first at this moment? Slow learner here. Need help...

Thanks for everything.

 

[devonsc]

It would be good it have it under 5%, since any power you don't use isn't "saved" for later.

Thousand apologies as I don't quite understand this portion. Do you mind explaining? Please?

some designs forward bias it on purpose.

Does it mean that it is buit in such a way to assure that the MOSFET does not damage when one applies a high current at the Source? Where at this point, it performs as a normal diode? But if yes, what is the purpose of having such construction? Sorry, just asking...ignore this question if my question is meaningless. Real sorry.

But the current is basically the same over time. Pulses of 30mA of current 1/3th the time (when the transistor is off) or a smoothed out 10mA 100% of the time does the same job. Actually, there are obscure, minor efficiency issues here, some for the cap, some against. It won't hurt to leave it in but it probably won't improve things.

Thanks for the detail explanation.

Excellent question. The PIC needs at least 3v for the lowest of freq, and may not drive the MOSFET well. 5v is standard, 5.5v will cause it to latch up. Check your spec sheet for what your part needs. A 5.1v zener can keep the supply where it should be. Also diodes can be used in series to drop the voltage. And of course a reg may work, but it consumes quiescent current and may have minimum dropout issues.

Does it mean that you advice me to tap the voltage supply from the 12V battery and regulate it to a 5V level instead of using additional batteries?

Also, a solar cell is going to drop down to a voltage which won't cause it to turn off but it won't run properly either. When the sun comes back and voltage is restored, it may not have a valid state and won't run. There are several mechanisms to enable that will compensate for this, look up "brown out detection".

Thanks! I was about to ask the meaning of this term, "Brownout Detection" as I found it in the datasheet. But I've yet to understand. Will read up further, thanks again.

Did you ever decide what your panel voltage and current were going to be? I missed that.

At this point, it would be an output voltage of 7.5V with the load current of 40mA. Thanks.

 

[Oznog]

Say you want to use the average, optimum 40mA from the panel. If you have a 20mA peak-to-peak ripple due to a small inductance, the current will vary from 50 mA to 30mA. Look at the i/v curve for the cell. The voltage will drop to next to nothing if you try to draw 50mA, the power output is low. At 30mA you're not using the capacity of the cell. Now an input capacitor will help this to some degree, but it's not a magic bullet since the ripple voltage is much less than the current ripple due to the i/v curve of a solar cell. Caps are not as effective in this scenario.

But if yes, what is the purpose of having such construction? Sorry, just asking...ignore this question if my question is meaningless. Real sorry.

As far as I know, it's unavoidable the way the silicon is layered.

Does it mean that you advice me to tap the voltage supply from the 12V battery and regulate it to a 5V level instead of using additional batteries?

This is inefficient. If the PIC needs 10mA, it would take 10mA of the input current if source off of the 7V solar cell, but due to the boost ratio this would consume 20 mA of the solar cell's current if it pulled it off the 13.6v battery output. No, source the power off the 7v solar cell. Don't use additional batteries, this is indeed "cheating".

By the way, don't use a normal silicon diode for the diode, that takes 0.7v. Use a Schottkey, it's around 0.3v forward drop. More efficient.

 

[devonsc]

Say you want to use the average, optimum 40mA from the panel. If you have a 20mA peak-to-peak ripple due to a small inductance, the current will vary from 50 mA to 30mA. Look at the i/v curve for the cell. The voltage will drop to next to nothing if you try to draw 50mA, the power output is low. At 30mA you're not using the capacity of the cell. Now an input capacitor will help this to some degree, but it's not a magic bullet since the ripple voltage is much less than the current ripple due to the i/v curve of a solar cell. Caps are not as effective in this scenario.

Erm...I tried reading it again and again but nothing goes in my brain. Sorry. I dare not ask for further explanation I will try to search for more info to understand.

By the way, don't use a normal silicon diode for the diode, that takes 0.7v. Use a Schottkey, it's around 0.3v forward drop. More efficient.

Thanks for such direct guides.