Constructing a "Grid" Circuit

Discussion in 'The Projects Forum' started by amcarr20, Apr 12, 2011.

  1. amcarr20

    Thread Starter New Member

    Apr 12, 2011
    4
    0
    I'm sorry if this post becomes long winded, but I wanted to include as much information as I can: I am part of a group that has had very limited experience with any sort of transistor's, AC voltage, Drivers, and very limited PIC programming.

    The goal of a project:
    The goal is to essentially create a "grid". This grid will operate at a safe 5VAC with a frequency of 60hz and hopefully a current around 0.5-1Ampere. Frequency and voltage should not vary within 2% tolerance. There will be a load that rectifies the power from the grid to power a couple of LEDs. At different times, alternative energy sources will generate DC voltage to go into the grid. This is assisted by a Connection which will both connect the alternative energy to the grid and shut off if unfavorable circumstances occur. These sources generate anywhere from 5-15 VDC along with 25mA of current (these are shaky numbers at best, the idea is to take dc and convert it to ac)

    The restrictions:
    We can not use ready built systems such as inverters. The grid is allowed to use one type of power electronic (ac-dc, ac-ac).

    Issues:
    We first tried to start with DC and create a sin wave generator (specifically a bubba oscillator) and amplify it using multiple op-amps. Our result was a 5vac with 25ma of current. We could not find a decent way of upping the current without hindering the wave form or putting it out of phase. We have very little (almost none) experience with PWM signals and how they can be used in a chip such as IR2110 with Mosfets in a Hbridge to generate the voltage we need (if anybody can explain it to me that would be righteous). We got frustrated with that and instead tried to go down the AC road.

    When attempting AC we could not find a decent way to take 13VAC (which is our only option) and turn it down to 5VAC while still keeping the same waveform and a decent current. I looked into options of a dimmer circuit (using Triacs and Diacs) but when I multisimed it, it did not look very promising. But I know there must be a way I can use a switch to get the voltage I need, I just don't quite know what it is. But when it comes to connections we still need to know how to construct AC from DC.

    The renewable energy systems are currently able to generate around 5VDC and 90mA... One renewable energy group can generate 90mA the other can generate 200mA... The connections team won't allow the renewable energies to hook into the grid until 5VDC is being generated... But we don't know how to invert it.

    We have been working on it for quite some time and may lack the technical experience to accomplish most of the tasks. Assigned to us, I have attached a schematic of an inverter we attempted to build and the actual guidelines for the project. If you need any clarification (that I may have forgotten) please ask.
     
  2. eblc1388

    Senior Member

    Nov 28, 2008
    1,542
    102
    The easiest way to transformer an ac voltage is to use a transformer, one that can easily allow user to change its output voltage. Try to get a toroid transformer with main voltage primary and 12V or 15V secondary. If your requirement is 5V AC at max. 1 amp, then any toroid transformer with rating higher than 6VA will work.

    Insulates the transformer high voltage windings connections with tape as you are not using them and works only with the low voltage 12V or 15V winding.

    Manually winds say 20 turns of PVC insulated wire on the toroid and connect the two ends of the pvc wire to an AC voltmeter. Apply 13V to the toroid secondary connection and note the reading on the voltmeter. You may get a voltage between 2V to 4V. This voltage is directly proportional to the number of turns of the PVC wire so you can get exactly 5V with the correct number of turns using ratio calculation.

    Remember to allow several turns extra so that you can have 5V at full load of 1A. This would mean 5.1V or more at zero load current.
     
  3. amcarr20

    Thread Starter New Member

    Apr 12, 2011
    4
    0
    So if I buy a torroid transformer with those specifications, I just need to wrap the wire's that come with it to get the voltage output I'm looking for? You'll have to excuse me, I've never manipulated a transformer before... I do have knowledge about their basic concept. I'll check to see if my lab carries a torroid transformer. Thank you for the quick response!
     
  4. eblc1388

    Senior Member

    Nov 28, 2008
    1,542
    102
    OK. Here is more information regarding the transformer.

    You have an AC13V source available to you and you want to obtain from it AC5V with about 1A current.

    There is virtually no existing transformer which suits that requirement available for off the shelf purchase. Therefore you will have to order one specifically, build one or modify an existing one.

    Normally the transformer everybody familiar with comes in the shape of a solid cube and it is very difficult to change the output voltage.

    Luckily some transformers do come in the form of a donut shape. They are known as toroid transformer. Because of the hole in the middle, user can add turns to the toroid to change its output voltage. Most often there are at least two set of windings on the toroid which are commonly known as primary and secondary winding. You connect the source to the primary and obtain the output from the secondary. However, nearly all toroid transformers sold would have the primary winding rated to be connected to the AC Mains as this is what most people would have wanted to use the toroid transformer.

    In your case, the voltage of the source is only 13V so it has presented some difficulties. Fortunately the toroid transformer, like all transformers, works also in the reverse configuration. That means if one put the source voltage to the secondary, voltage(high) would appear at the primary side. This high voltage you don't want so I have told you in previous posting to insulate the ends of primary connection wire with insulation tape.

    As the AC13V source is now connected to the secondary winding, that winding becomes the primary winding, in electrical terms. To get your AC 5V output, you have to construct another secondary winding and place it on the toroid. You can do that by placing several turns of PVC wire or enamel wire or as they usually call them "magnet wire". The exact number of turns is different with each transformer but a trial of say 20 turns and the actual voltage measurement would give you the relationship between number of turns required to get the 5V output.

    An example of the toroid transformer and the modification required are shown below. The blue wires shown in the second image is how you would add the PVC wires to the toroid. I showed only three turns in the image. The wire is extra to the toroid and you have to obtain it somewhere else.

    Note that there is no standard coloring scheme for wires forming the primary or the secondary winding so you have to read the toroid label to find out or ask someone who actually knows which is which.

    [​IMG]

    [​IMG]
     
    Last edited: Apr 13, 2011
  5. t_n_k

    AAC Fanatic!

    Mar 6, 2009
    5,448
    782
    It's not clear as to why 13V AC is your only option - is it derived indirectly from the normal reticulated AC mains supply to your building/lab?

    One of the problems of running the 5V "grid" from the 60Hz AC mains (albeit through another interposed transformer) may be your voltage spec requirement of a maximum of 2% variation. Mains frequency is usually maintained well within this spec by the generating companies, but the voltage may not always be so tightly maintained by your local supply authority. So a 5% change in the primary mains supply will show up as a 5% change in your nominal 5V "grid" voltage.

    How well regulated is the 13V source?
     
  6. t_n_k

    AAC Fanatic!

    Mar 6, 2009
    5,448
    782
    If voltage regulation is an issue you may be able to make use of special transformer known as a ferro-resonant type. Another forum member 'someonesdad' made a similar recommendation on another (unrelated) project thread recently.
     
  7. t06afre

    AAC Fanatic!

    May 11, 2009
    5,939
    1,222
    Do not forgett that the AC voltage is measured in RMS. The peak voltage will be about 1.41 times higher
     
  8. wayneh

    Expert

    Sep 9, 2010
    12,123
    3,048
    I've read your project doc and I don't think any fixed transformer is going to meet your needs, since the input voltage is variable. Right?

    I believe the way grid tie-ins are done now is by inverting a DC source to a true sine wave and then shifting the phase of that wave as needed to match up to the grid. Devices to accomplish those task are expensive at real scale. Maybe the low powers of your model will allow keeping costs in line.

    You might want to consider "mixing" DC power sources before inverting them, if this is allowed. I mean, put PV panels in series or parallel, for instance. That's very simple and cheap compared to mixing AC sine waves. I can even a imagine a DC "grid" of all the RE sources supplying a single grid-tie-in inverter. But maybe that's cheating.
     
  9. amcarr20

    Thread Starter New Member

    Apr 12, 2011
    4
    0
    Maybe a little cheating... but would have been a lot easier I'm sure. We have tried creating a Sin wave and matching it... with poor results. Primarily because we have no working knowledge of doing such a project. Our current idea (as asinine as it sounds) Is to put high wattage resistors and simply divide the voltage to get 5VAC. There will be an LC filter in there as well, but I think it might blow up in our face (quite literally) This was against my idea of trying to find a switch that could regulate the voltage. I have suggested using a transformer to step down the AC but they suggested there was a problem with connecting other ac sources.

    The problem is with control, we need a simple way to lower the voltage or current coming from the grid (5VAC) by using a pic. We have a way to monitor (which we haven't implemented yet, but I hope will work) but no way to control an ac voltage. Their idea is to use a potentiometer controlled by a motor. To me, this sounds slow and cumbersome. But I have no better ideas... Thank you all for all your replies, I assure you it's an on going struggle.
     
  10. GetDeviceInfo

    Senior Member

    Jun 7, 2009
    1,571
    230
    Let's explore this further;

    You have a battery for storage. It only makes sense that your PV devices charge the battery. You then invert your battery back to grid, consuming the accumulated charge, but no more, as the bulk of the battery is for emergency power, and in real life systems the battery charge volume would be comparibly small.

    You've not stated the characteristics of your AC grid in terms of cleaniness. A chopped DC will drive your loads without any noticeable problems, and can be phased over an AC source. You wouldn't do it real world, only because the utility would kick you off. You have no indication of such so where's your measure? If you like a classic approach to sine waves, you will be PWMing your inverter.

    Which comes back to the grid voltage. If your intent was to generate the grid power, then fixed frequency inverter is what you probably should employ. If your thought was a quick step down from the mains as a source, then you should probably employ the real world tools of renegotiating your source spec's, so you can move on with the body of the project.
     
  11. wayneh

    Expert

    Sep 9, 2010
    12,123
    3,048
    That has some merit and historical precedence, if you replace potentiometer with Variac. A Variac is a transformer with a variable ratio, allowing you to efficiently get continuous voltage control. The negative is simply the response time, which is in seconds. That could be OK for RE sources as long as they are operating continuously, but I would think connect/disconnect surges would shoot on through. Maybe if you can solve that, you could indeed use a Variac.
     
Loading...