SI22723A - Dedicated display with intelligent light points - Google Patents

Abstract

Subject of the invention is a dedicated display with integrated intelligent light points, where the positions of the light points are arbitrary and predefined, hence they are not necessarily positioned in a grid. The multitude of mutually logically connected light points with elements like (bulb, LED, LCD, oLED, mechanical display or any other element, which emits or reflects light) creates a light field, which represent a variable light-based traffic sign, information panel, ad panel etc. and is controlled via PCBUS (Power Communication Bus) of the supply communication bus, which links the light elements to the controller and requires only two conductors for this.

Description

PURPOSE DISPLAY WITH INTELLIGENT LIGHT

POINTS

The object of the invention is a dedicated display with intelligent light points, where the positions of the light points are arbitrary and predefined, so they are not necessarily positioned in the raster. A plurality of logically related light points with elements such as (bulb, LED, LCD, oLED, mechanical displacement, or any element that emits or reflects light) forms a light field that represents variable light traffic signs, information boards, advertising signs, etc., and is controlled via is PCBUS (Power Communication Bus) power communication bus, which connects the individual light points to the controller and requires only two conductors. The invention belongs to the class G09G 03/32 of the international patent classification.

A technical problem which the invention successfully solves is to enable the development of serial production of unique dedicated displays, which, despite a significantly more complex solution (compared to LED chains), would provide a lower price for unique displays, due to the possibility of simple and fast automated serial production, where the layout of the light points on each individual display may vary.

The solution to the technical problem posed is therefore the fundamental paradox of serial production of unique products. Previous solutions are based on connecting individual light points to wires. Another solution is to make a dedicated printed circuit board with the light points at the desired points at the same time as the controller for them.

Both solutions so far are not best suited for small-scale and prototype solutions. It requires a lot of manual work by the person who assembles the device or by the knowledge and development work of a specialist. This takes a lot of time and often costs too much. Manufacturing automation would be very complex, expensive and also unreliable.

In a dedicated intelligent light display device according to the invention, the light positions are arbitrary and predefined, so they are not necessarily positioned in the screen (such as in a matrix display) using a solution where the PCBUS also has the function of load-bearing structure, ie the assembly of individual iPix light points onto a carrier plate with two conductive layers.

The invention will be explained in more detail on the basis of an embodiment and accompanying figures showing possible applications and embodiments, with three basic types of solutions described below: a simplified system for displaying only two images, a simple system for displaying up to 2 ** N images and a complex system capable of controlling individual iPix light points:

FIG. 1 is a basic embodiment of a dedicated display with intelligent light elements in a complex system capable of controlling a single light point according to the invention; FIG.

Figure 2 is a circuit of the base unit of the light point for displaying only two images;

Fig. 3 is a code diagram of the control of the base unit of the light point when displaying four images;

figure 4 circuit of the base unit of the light point to display only 2 ** N images, hereafter N = 4;

Fig. 5 is a schematic illustration of a variant solution with wire connections; 6 is a schematic view of a variant solution on a support plate with two conductive layers;

7 is a circuit of a base unit of a light point;

Fig. 8 diagram of impulses in communication;

Fig. 9 shows the connection of the controller to the base unit of the light point; Figure 10 communication packets in a single communication cycle; Fig. 11 Fault detection circuit of the light unit base unit.

The basic implementation idea of a dedicated 20-point intelligent display for a complex system capable of controlling individual iPix points is shown in Figure 1. Each iPix point contains an LED element and an MPU microcontroller that communicates with the main controller via the PCBUS power bus. Power can be supplied via two conductive panels, which are isolated from each other. This kind of solution eliminates all the wires and the automation process is extremely simplified. The MPU contained in each iPix light point enables, in addition to communication, direct control of the iPix light point and controls its proper functioning. If the iPix point light is malfunctioning, it reports this to the master controller.

Figure 2 shows the iPix 'base unit circuit for displaying only two images. It consists of the parallel connection of two io sequential connections of diode D, resistor and LED 1 and diode D, resistor and LED 2. Such an embodiment permits the execution of a dedicated display adapted for a typical application intended for displaying only two images (eg, the cross-arrow traffic sign). In this case, the iPix 'base unit does not require any intelligence, since the power is from AC terminals A and B. In each case, only one light element (LED 1 or LED 2) lights up. D diodes are only required when VFmax <Bait LED. The physical implementation of the iPix 'light point is simple, the driver for the light point field can be easily implemented with the Hbridge FET power level, and the cost of individual iPix' light points is minimal.

In the variant solution, where up to a maximum of 2 ** N images need to be displayed on the dedicated display (the possibility of individually controlling N groups of iPix points), the implementation example is shown in Figures 3 and 4. In this solution, the iPix light point can appear in several images. The information for powering the light point is encoded as shown in the code diagram in Figure 3. On the Tslike image-changing period, which is usually a long time of the order of a second, modulation occurs at Tdata, which is substantially shorter than the time of image change, therefore, it does not interfere with the LED power supply and does not affect the luminance. During Tdata, N data pulses are sent, the sum of which determines the code that fires the corresponding iPix lightpoint base unit. A simplified circuit to decode and drive the iPix Light Point Base Unit is shown in Figure 4.

io The power reserve during data transmission is again made using diodes D and C2. Deletion of the counter (start of data detection) is performed with the differentiator C1 and R1. The code that lights up the LED is selected by connecting to the appropriate Qx clip. The receiving code logic can be implemented with a simple 4 bit series 74 counter, say 74HC163 (for example N = 4). is There are three main advantages to deciding whether to use the circuit provided, namely that the price for the 74 circuit by a factor of 3 5 is less than the cheapest CPU, no need to develop light point software, and that the driver and data pulse generator is very easy .

2o Figure 5 shows a schematic view of a dedicated solution of a dedicated display with intelligent iPix light points in O housing for a complex system with the ability to control a single iPix light point with wire connections, and Figure 6 shows a schematic representation of the same variant solution on a carrier plate in O housing two conductive layers P1, P2, and intermediate insulation I, while Figure 7 shows the base unit unit of the iPix light point.

The selected PIC controllers feature low power consumption below 200μΑ and 2 to 5V power supply. Both are very well suited to this project because in the two phases of the basic MPU cycle, only the capacitor C. will be energized. The capacitance C and the times of the individual phases of the cycle are therefore interconnected. The idea behind the whole operation is that it will be achievable.

io The current limitation of the MCU output is 20 mA of current through the LED, which will be sufficient in this case. The current can be further limited by selecting the Rled resistor. The brightness of the diode will be controlled by the pulse width modulation MPU in the READ phase. The impulse diagram for communication is shown in Figure 8. Reaching at least 16 luminosity levels is therefore not a problem. Due to the communication of is and synchronization, a 100% duty cycle is not possible, but in real terms it will approach 90%. Communication is always done cyclically and synchronously. The cycle period is constant. Since the power supply and communication are combined along the same line, the communication and power supply problem must be arranged in such a way that it ensures constant luminosity, regardless of communication. Therefore, we use the solution with communication pulses that are always present, regardless of the information transferred and the modulation of these pulses with useful information. One communication cycle is divided into three phases, namely (Figure 7) READ - Transmission from the controller to the iPix light point. The iPix light point controls and data go in this direction: status, luminance, operating modes VVRITE - transfer from iPix light point back to the controller. A minimal amount of information is transmitted in this direction, mainly in the status of "error LED"; SYNC - start synchronization.

These three cycles are made possible by the binding given in Figure 9, which shows the controller's connection to the iPix base unit.

In the READ phase, the supply voltage Un = UR is greater than Uf (LED voltage), so it can illuminate. The current through the diode defines the Rled resistor. Un in this case also provides power to the MPU via diode D. In this case, also the capacitor C is charged to the UR. At this stage, short negative communication pulses on the power supply are present, which are the carriers of information in READ communication (see detail). The length of these Tpulse pulses is significantly smaller than the period of occurrence of these Tbit pulses. Which allows them to not noticeably affect the brightness of the LED. The coding (modulation) of these pulses must be such that the number of communication pulses and widths do not change, ensuring uniform luminosity regardless of communication. Since all iPix light points can be lit at this stage, the power supply here is large.

In the VVRITE phase, the supply voltage Un is reduced to Uw- under L / fLED. Therefore, the current through the diode is automatically interrupted. The voltage of Un is also less

2o as the voltage of capacitor C which is now supplying the MCU. The current from the capacitor can only flow into the MCU and not back, which prevents diode D. The current drops practically to zero at this stage, allowing the power to be delivered at this stage via relatively high impedance. In fact, the power is through the Rfw resistor. All this allows the selected iPix light point to force GND on Un via an IN / OUT MPU pin, which allows communication back to the controller.

The SYNC phase is followed by the end of the cycle, where the iPix field controller sets the voltage Un to GND, which indicates to all the light points that a new cycle will begin. The length of the SYNC phase must be longer than the communication pulse, but not too long for the reserve in C not to be consumed.

Figure 10 shows the encoding - communication protocol when connecting the controller to the iPix lightpoint base unit.

In the READ (Master Communication - Controller -> iPix) phase, two problems need to be solved in the preparation of the io communication protocol: the suitability of encoding (since the signal and power are together, only encoding is appropriate so that the level is high most of the time) and the constant number of communication pulses so that the luminosity will not change with respect to communication.

is Figure 10 shows the first type of information that we want to encode. For coding, we use the known and proven Manchester code (second line in the figure), whose main feature is that it carries CLK information in addition to the information. Another feature is that the number of passages is always the same regardless of the information. The problem with the Manchester code in our case is that the duty cycle is 50%, which is not permissible because of the power supply. Therefore, we modify the Manchester codes so that each passage generates a short low-level pulse, as illustrated by the 3 lines in Figure 10. Decoding is done by a simple algorithm measuring the time between two pulses. If the time is short there is no change in the information signal, but if we measure a long time, the level of the output signal is inverted. Since we do not know what the initial bit of information will be, it is necessary to send some Ί ”at the beginning to synchronize the decoder and then one“ 0 ”to indicate the start of the packet.

Any negative impulse as a disturbance can cause the communication pack to be desynchronized. Problems can be significantly avoided if we do not react to the edge of the communication pulse but detect the pulse by scanning the level. The second trick is that because of well-defined times, we know when to expect the next pulse and scan it only then.

io Many disturbances can be filtered out in this way.

Since the MPU does not have the ability to process IRQ anyway, the whole program will run in one loop, we can use the following program principle:

• waiting for a communication impulse;

• At least 2x during the Tpulz time;

is · if the pulse is not present, assume twice the length;

• decoding the output from the measured time;

• now we have Tbit time to process everything that is needed;

• loop repetition.

This principle is good because we do not need IRQ operation, we spend a little time scanning the input and have a relatively long time to process other functions. Of course, the Tpulz and Tbit times must be appropriately selected to optimize MPU load.

In the VVRITE phase (direction of light point -> controller), standard serial communication encoding can be used. The enk and zeros ratio doesn't matter because the LEDs are off. Only a few bytes are transferred, so this phase is short enough for the power capacitor C to last until the end.

Any parameterization (setting the mode of operation, color luminance, ...) due to the modest MPU resources, cannot be performed in some permanent memory, but only in the MPU RAM. This means during the initialization phase after switching on the light point field or during operation. For most settings, this is not a problem. However, there is a problem addressing the iPix light spot.

Obviously, in the system in terms of communication, all iPix light points are on the same bus. So all units receive the same communication. Because we want individual control of each iPix light spot, it is clear that there must be some way of addressing it. The question is how to make each iPix light different from any other. In short, a unique code or address must be inserted into each unit, and this information must be permanently remembered. Because we want the lowest possible price for MPU variants with EEROM are eliminated. The code can only be written as a program constant to be entered with the firmware. Fortunately, we have in

The PIC to the MPU family has two options.

• The firmware is loaded into the flash memory of the MCU before mounting in the light element. A couple of flash locations can be assigned to the device address.

• In mass production where current programming is not desirable, MICROCHIP offers the possibility of entering unique code and fixed firmware at the chip production stage.

In both cases, therefore, we have an ID that is different for each light element in the field and can be used for a device address, which must have the following properties: because we are extremely limited by resources, only two bytes should be used for this purpose, 65536 different IDs, which is enough because we are targeting a system that will have an order of magnitude of light elements of about 1000. So significantly less than the possible codes. Furthermore, addresses must be continuously monitored one after the other throughout the field of light elements because in some portable communication packets, addressing with the bit position in the string is used, which will be explained later. This requirement allows for the simultaneous activation of all light elements on the display or, simply put, instantaneous image change by downloading a single communication package. In the case of pre-programmed addresses in range 0 - FFFF provided by the MPU manufacturer, there is a problem with the activation packet, where the light point unit needs to know the offset of the first unit in order to adjust the address according to the bit position in the activation packet. This can be avoided by fixing the DATA length in the activation packet to 2 ** N bits. In our case, 1024 is optimal. Now the offset is no longer needed. This also implies the value of Tcyc.

As stated earlier, a fixed number of communication bits are sent in one communication target, but not all carry the useful information. However, it is also possible to download several communication packages in one cycle. One package is composed as follows:

No. bits

8 16 16 0-1024 16 SYNC HEADER ADDRESS DATA CHK

The maximum length is therefore 1080 bits. And from here the link:

io Tr = 1080 * Tbit

SYNC

7 6 5 4 3 2 1 0 1 1 1 1 1 1 1 0

It is a constant set of 8 bits, where 7 units initially provide synchronization with the decoder. The last 0 indicates the start of information that must be processed

HEADER

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 No bits: ADD RESS + DATA + C HK 0 Ved no 0 0 0 Package with command 0 1 Activation package 1 0 Reserve 1 1 Reserve

ADDRESS

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 0 0 0 Goal 0 Title -1FFF (8191) 0 0 1 AND Unit group mask the rest Undefined - reserve

CHK

15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 The sum of all units in the package from the HEADE segments R, ADDR ESS, DATA

DATA

The data are compiled in two ways;

PACKAGE WITH THE COMMAND

N - bits M bits TEAM PARAMETERS

ACTIVATION PACKAGE

1024 bits_ "1" at appropriate position (index) triggers action on iPix unit named "index"

The size of the commands and parameters, as well as the commands and parameters, are still undefined at this stage and will be determined upon implementation of the final solution, io It is important to treat some of the DATA information as a command executed on a light element using parameters located in the PARAMETERS field. Command packs are suitable for controlling an individual light point or a group of iPix light points.

When a dedicated display with intelligent light elements is in operation, the processor may break the circuit and the iPix light point due to interference. The problem is solved with the built-in WDOG function. Although the fixed iPix light spot is not good, it is not yet critical, in terms of another problem. The MPU can become pinched or broken by forcing a pin to communicate to the GND. This is not a big problem in the READ phase because only max l / O current of about 20 mA will flow through the pin, which is not a problem. The problem, however, is that in the WRITE phase, it will prevent all other iPix points from returning feedback.

This problem can be solved with the built-in MPU WDOG function which in this case will reset the MPU and the operation will normalize. In the case of a hardware failure, when one broken pin would prevent communication to all other units, there is no other help than detecting such a condition on the controller side and triggering a major system error. An error will be detected when no return communication is possible. Of course, the display will continue to display the information correctly, only fault detection will no longer provide information about each iPix light point. Figure 12 shows the fault detection circuit of the base unit of the iPix point light. Although experience shows that most LED faults are intermittent, the solution below also detects a short circuit. Because the PIC family has MPU variants with a built-in voltage comparator, it can be used for LED fault detection. Figure 12 shows a circuit utilizing a built-in comparator for fault detection on LEDs. is The detection is made when the LED is on (LED output MPU OUT) gives a low level, so the current flows through the resistor Rled and LED.

In terms of LED status, we have 3 voltage options at the comparator input

C +:

C + LED status Near Uf (drop to LED) OK Near GND Short circuit Near V + interrupted

In order to determine in which of the three areas the measured voltage is located and we have only one comparator, we must be able to set the reference on the comparator C- to two values RefHi and RefLow. Two experiments can thus determine the area where the measured voltage falls. For the lower reference, we take the internal reference that is already present in the MPU. Unfortunately, this reference is about 0.6 V because the Uf LED is much larger, you would need a slightly higher reference. The problem is solved by reducing the voltage with the splitter R2 and R4. For the second measurement, however, when we need a reference higher than Uf, we use input C- and divider R1 and R3.

The LED status is determined by the table:

The mind is at the common point R2 and R4.

Uref (C -) condition CMPout LED status RefHi Um> RefHi 1 Interrupted RefHi Um <RefHi 0 OK RefLow - Internal Um> RefLow 1 OK RefLo - internal Um < RefLow 0 Short circuit

is If the answer is OK in both attempts, the LED is ok, otherwise we have an error type.

A dedicated intelligent point-of-light display according to the invention that successfully solves a technical problem can be implemented with the latest technology, using processors that meet two essential elements: low power consumption (which is indispensable for multiplex power and communications on the same bus), and extremely low MPU price.

Both are a prerequisite for the proposed solution to make economic sense. Technology that allows the price of an MPU to be even less than the cost of a light element (LED) means a significant savings in the production of the device.

Claims (3)

PATENT APPLICATIONS 1. Dedicated display with intelligent light points, 5 that the iPix positions are arbitrary and predefined, each iPix containing a microcontroller (MPU) that communicates with the main controller via a power communication bus (PCBUS) and has a power communication bus (PCBUS) also the function of the load-bearing structure so as to allow the mounting of io individual light points (iPix) on a support plate with two conductive layers, and allows direct control of the light point (iPix) while controlling the correctness of its operation. Dedicated intelligent light point display device according to claim 1, characterized in that the base light point unit (iPix ') for displaying only two images consists of a parallel connection of two consecutive diode (D) circuits, a resistor and a diode (LED) 1) and diodes (D), resistors and diodes (LED 2) connected 20 to allow for the execution of a dedicated display adapted for a typical application intended for displaying only two images (eg the cross-arrow traffic sign). A dedicated intelligent point display according to claim 1, characterized in that in a variant solution where up to a maximum of 2 ** N images need to be displayed on the dedicated display 5, the point light (iPix) may appear in several images and the light-point information (iPix ”) is encoded in such a way that, during a Tslike image-shifting period, which is usually of the order of magnitude of a second, modulation occurs at Tdata, which is substantially shorter than the image-changing time, while io v at time Tdata sends N data pulses whose sum determines the code that fires the corresponding light point base unit (iPix).