Development of a low-power backup power supply with a sine output. Part 1. Statement of the problem

Uninterruptible power supplies (UPS) are widely used both in the home and in industry. They are designed to provide the necessary power to equipment from backup sources in the event of a “loss" of the main power supply. The backup sources in such UPSs are mainly batteries. Therefore, these UPSs provide power to the equipment for a limited time, from a few minutes to a couple of three hours. On sale there is a huge amount of such equipment, as they say, for every taste and color "pocket", with various characteristics and various functions.

Consider the scope of application in everyday life.

Each house has a refrigerator. The main models use a compressor driven by a bipolar single-phase asynchronous electric motor. The power of household refrigerators is 100-200 watts. The loss of the main power supply (a la 220 volts) for several hours can lead to defrosting of the refrigerator. This is not critical, but inconvenient. But a conventional computer UPS will not help here: the compressor engine will not be happy with the form of voltage provided by such a UPS. For this kind of consumers need a sine output UPS.



An example, frankly, the most real, but not the most necessary for the use of UPS.

Private house, heating system, circulation pump. The same problems. Modern gas boilers basically have it in their design. In the event of a power failure from the computer UPS, they also do not particularly want to work. True, you can survive a couple of hours without electricity and a working boiler, since the house does not cool down to freezing temperatures in a couple of hours.

We continue to look for the use of a UPS with a sine output at home.

The same private house, the circulation pump in the heating system, but the heating system itself is not on gas, but on wood. So, after working, you have melted the stove and warm the house, the pump pumps the liquid through the pipelines of the heating system. Bam! Turned off the electricity. The boiler began to overheat wildly, due to a lack of circulation, but this is not a gas valve to shut off, you will have to extinguish the stove, in the literal and figurative sense. And if there was a UPS, then calmly for an hour and a half it would be possible to melt the stove and continue to wait until the power grid restored the supply of natural gas to electricity. Already more real. Further.

The village, there is natural gas, there is an outdoor boiler of the type AOGV-11.6-3. Here is its description: it is intended for electrically independent autonomous heating systems for houses, cottages up to 110 sq.m.So for his work electricity is not needed. But for the circulation of the coolant, a circulation pump is used. The lack of electricity does not affect the operation of the boiler, but due to the lack of circulation, the water in the boiler begins to boil and is squeezed out through the expansion tank with all the ensuing consequences. So in such a situation you have to turn off the boiler. But if the owners are not at home, or does this happen at night?

Here for such a specific case, a UPS with a sine was needed. It makes no sense to develop a full-fledged UPS for such a task. If the transition time from the network to the backup source is 5-10 seconds, nothing critical will happen, both with the heating system and with the pump itself.

From the foregoing, the task follows: to develop a low-power redundant power supply 220 V, 50 Hz with a sine output.

Foreword


We will turn to the market and study the available offers. To do this, let us set some criteria for the chosen equipment. Requests will be modest.

  • Power - 200 W
  • "Pure" sine output,
  • Well ... that's enough.

Google at the request "UPS 200 W with a sine" did not please the results. Almost the very first link talks about the calculation and selection of UPSs for a gas heating boiler. The truth is, the price of a UPS is not pleasing to the eye, especially if you look at the batteries that are not included in the package of the UPS - it becomes generally sad. The rest of the links are the same story that does not please the ordinary mortal representative of the middle class.

Introduction


There are many different UPSs on the market, why develop something? There are several answers to this question:

  • The price of the proposed devices, albeit not sky-high, but still high,
  • The function of the “pure” sine is more expensive and is far from universal,
  • Understand the princes of building such systems,
  • Gain experience in the design, development of power electronics,
  • Experience in programming MK
  • Everyone can continue the list for self-motivation ...

For the time being, let’s keep silent about the price of the developed device, since so far there are no specific “digital cameras”. But after testing the device, we will calculate all the costs of consumables.

Lyrical digression
There is a series of articles on geektimes.ru devoted to the design of a powerful UPS with a pure sine output. Starting post can be found at this link . In this cycle, we will not develop a full reduced copy of this UPS, but will approach the specific task a little differently. Yes, and the device being developed does not turn up the name UPS. This will be a backup power supply with automatic input of a reserve (hereinafter RIP ABP), in which the transition time from the network to the RIP will be about a couple of seconds.

Initial data


Let's start with the necessary calculations. Let us estimate the power that the rip needs to give to the load, battery life, etc. Let's get started.

The load will be our circulation pump. Let's look at the common models on the market. Here's what Google gave out: a link to the characteristics of circulating pumps.

For houses of small area up to 100 sq.m, pumps with a capacity of up to 100 watts are used. Moreover, rarely does anyone use the third power mode of the pumps. So let's stop at 60 watts of power consumption. It is from this power that we will build on the calculations.

Lyrical digression
If someone needs more power, then for this it is necessary in the calculations to proportionally increase the values ​​obtained, and according to the data obtained, recalculate the electronic components used in the circuit (diodes, transistors, etc.).

The voltage at the output of the rip 230 V. Power 60 watts. Therefore, the current will be I = P / U = 60/230 = 260 mA.

Now let’s set the efficiency of the converter 12 V DC -> 230 V AC in the region of 90%, then when the supply voltage from the battery is 12.4 Volts, the current consumption with the battery will be:



Here we will rely on these figures, both when choosing electronic components, and when PCB manufacturing.

Block diagram of the rip


For the converter 12-> 220, the scheme of increasing the IIP is basically used. That is, using a pulsed step-up transformer, 310 volts of constant voltage are obtained, and then a pure sine of 220 volts at the output is obtained by a bridge circuit controlled by a sine wave and LC filter. This approach uses many components from integrated circuits to high-speed diodes, etc. After all, impulse circuitry.

For this rip with its negligible low power, you can go a little different way.

From the so-called DC link, which in our case will be the battery bus, that is, 12 volts, through a bridge circuit controlled by a sinusoidal shim, apply to the primary winding of a conventional linear network step-up transformer. Remove the already required 220 volts of sinusoidal voltage from the secondary. Fortunately for such power and the dimensions of the transformer will not be large. The transformer itself will serve as a filter and smooth out the voltage shape almost to a recognizable sine. And if you put a low-impedance LC filter between the bridge and the primary winding of the transformer, you can get the voltage shape at the transformer output very close to the sine.

It turns out something like this scheme. Clickable image




In this example circuit, the components are taken to show the main idea of ​​the RIP circuitry, and their values ​​do not correspond to the calculations that we will perform below. The circuit itself will become more complicated as the device is designed.

During the time when the pump is operating from the RIP, the battery is discharged, and after switching from the RIP to the network, it makes sense to charge the battery to its rated capacity. When “electricity is present in the outlet”, the pump runs on mains, and the output circuit of the RIP can be used to charge the battery. That is, apply voltage to the secondary winding of the transformer (it is also high voltage), and remove the alternating voltage from the primary winding (it is also low voltage), straighten it on the diode bridge, smooth it with a capacitor and charge the battery with it. Consider the changes that need to be made to the circuit for this approach. Clickable image




That is, as long as there is voltage in the network, the relays are pulled up, and the mains voltage through the relay contacts enters the load, as well as the high-voltage winding of the transformer. Further, the voltage is removed from the low-voltage winding. The voltage is rectified by parasitic diodes of the transistors (for the sake of correctness, it is worth pointing out that we will not use them, we will install external high-speed diodes in parallel with the transistors for the required current), it is smoothed by the capacitor and, through the P-channel transistor controlled by the control circuit on the MC, pumps the necessary charging current in the battery through a smoothing inductor.

When the electricity "ends" in the network, the relay opens, and the circuit will work in the reverse order. From a 12 volt battery through a bridge of transistors, filters and a transformer, the voltage will be supplied to the load.

In order not to sculpt synchronization with the network, etc. for an almost instantaneous transition from the battery to the network and vice versa, if the network is lost, the load will be de-energized, the relay will open, the circuit will prepare everything and everyone for operation from the battery and start generating voltage to the load. When restoring the network, the circuit will stop the voltage generation, make sure that everything is fine, and close the relay to go to the network and charge the battery. The functions of the control circuit are outlined during the development process.

The structural scheme and the basic principle of the RIP were disassembled and, in this positive mood, we proceed to the calculations of the necessary components of the circuit and the choice of the hardware platform, both for the “brains” of the device and for power elements. The truth is in the next article.

And again the digression
Recently I published a series of articles on Habr called "Measurement of the weight of minerals in the mining industry." But apparently miscalculated with the target audience of Habr, so I publish the development of the RIP here at Geektimes . If anyone is interested in reading my previous publications, then I will leave links to this material here:

  1. Measurement of the weight of minerals in the mining industry. Theoretical basis
  2. Measurement of ore weight by stator current. Practice. Part 1. The signal processing algorithm in MK
  3. Measurement of ore weight by stator current. Practice. Part 2. Software implementation on MK



Conclusion


In the following parts, we will consider the calculations of the developed RIP, prepare the electrical circuit of the device, select the hardware platform, and develop the topology of the printed circuit board for the RIP. We will analyze the functions of the device, write programs for the MK, carry out a full cycle of setting up and testing the device on equipment, and also hand over all this to the real customer.

PS: True, the development will take some time, and further publications will be published as the project progresses. I focus on three more articles, with an interval of about 2-3 weeks.

Continued here:

Part number 2.
Part number 3.

Also popular now: