High voltage signal amplifier, with ±5kv DC power supply, how to amplify ±AC5v signal to AC5000V signal

colinhnu

High voltage signal amplifier, with ±5kv DC power supply, how to amplify ±AC5v signal to AC5000V signal [Copy link]

 

I would like to ask if anyone has made a similar high-voltage signal amplifier circuit. Can you give me a reference? I have also found some 1000V ones on the Internet. I am still a student, so I have not made many high-voltage things. I hope that those who have done it can give me a solution so that I can learn from it. Thank you!

maychang

What is the bandwidth requirement of your amplifier? Narrow-band amplification is easier, but high-bandwidth amplification is more difficult.

maychang

In addition, the original post did not say how much power or current the amplified 5kV signal needs to output, nor did it say the type of load after amplification (resistive, inductive, etc.). Depending on these conditions, the difficulty of this amplifier varies greatly.

chunyang

The original poster should change his mindset. There is no "amplification" problem here. If you want to get high-voltage AC, the easiest way is to use a transformer to boost the voltage. If you need a larger power, direct boosting cannot meet the requirement. If you already have a high-voltage source, you can consider the "modulation" method, that is, using a low-voltage signal to control the high-voltage switch to achieve it. If a sine wave is required, you can use a low-pass filter to modify it. However, since it is not easy to achieve high-voltage switching, the specific plan needs to consider frequency and power. However, for most applications, high-voltage AC signals are generally achieved by low-voltage oscillators boosted by transformers. A typical example is the two- and four-anode high voltage of CRT TVs/monitors. For large-size color TVs, this voltage is tens of thousands of volts, and this voltage is achieved by the line output transformer.

chunyang

The original poster's question is probably a "homework question" or a "creative question" rather than a practical application. If it is a practical application, it will not be limited by the conditions of the question. It can be powered directly by the 220V power grid, and the high-voltage AC source with the required power can be obtained through the switching power supply with oscillator and transformer. Of course, this is also the method of CRT color TV/monitor. As long as the frequency is not so high that the step-up transformer is too difficult to make, this method is the mainstream.

maychang

One of my replies was rejected for some reason. The original poster did not say the output power (or output current) of the amplifier, nor did he say the nature of the load (resistive, inductive, etc.). Depending on these conditions, the difficulty of making such a high-voltage amplifier will also vary greatly.

chunyang

maychang posted on 2018-7-6 13:07 One of my replies was rejected for some reason. The original poster did not mention the output power (or output current) of the amplifier, nor did he mention the load...

maychang's management authority should be able to pass the "review" by himself, try it on the management page.

colinhnu

maychang posted on 2018-7-5 17:43 What is the bandwidth requirement of the amplifier of the OP? Narrowband amplification is relatively easy, but high bandwidth requirements are more difficult.

Thank you for your replies. Bandwidth requirement: -3dB 1KHZ, current is about 5mA, not too large. This is a project for electro-jet printing, and I want to do it better. I also looked at the low-voltage circuit and can understand it. But this voltage is too high, and I am afraid that many parameters or devices are not suitable. After all, I have no engineering experience in this area. They are all theoretical things. If there is a high-voltage solution that can be close, I will refer to it and learn it myself.

colinhnu

maychang posted on 2018-7-5 17:43 What is the bandwidth requirement of the amplifier of the OP? Narrowband amplification is easier, but high bandwidth requirement is more difficult.

Thank you for your reply. Bandwidth requirement: -3dB 1KHZ, current is about 5mA, response: 100V/US (this parameter is very high, I feel). It doesn't need to be too large. This is a project for electronic printing, and I want to make it better. I also looked at the low-voltage circuit and I can understand it. But this voltage is too high, and I am afraid that many parameters or devices are not suitable. After all, I have no engineering experience in this area. They are all theoretical things. If there is a high-voltage solution that can be close, I will refer to it and learn it myself.

maychang

colinhnu posted on 2018-7-10 17:31 Thank you for your replies. Bandwidth requirement: -3dB 1KHZ, current is about 5mA, response: 100V/US (this parameter is very high, I feel). It doesn't need to be too large...

This requirement can be met by using silicon steel sheet core transformers. Ferrite core transformers can also be used, but the volume may be larger than that of silicon steel sheet core transformers. The voltage rise rate of 100V/us can be achieved as long as the transformer bandwidth reaches 5kHz. The key is transformer design. It cannot be simply like the industrial frequency power transformer, winding the primary pad with insulation and then winding the secondary. The primary and secondary sandwich winding method must be adopted. In addition, because the secondary voltage is very high, in addition to paying attention to insulation treatment, it is best to divide it into several sections to reduce the distributed capacitance of the winding. Such a high voltage is easy to generate corona, so it is necessary to be careful not to use polyester film insulation materials, and use paper (preferably capacitor paper) or yellow wax silk as the insulation layer. The distance between the winding and the core must be sufficient. The transformer must be impregnated, preferably vacuum impregnation followed by atmospheric impregnation.

maychang

colinhnu posted on 2018-7-10 17:31 Thank you for your replies. Bandwidth requirement: -3dB 1KHZ, current is about 5mA, response: 100V/US (this parameter is very high, I feel). It doesn't need to be too large...

When designing the core, you can't calculate it according to the 1kHz frequency, but at the lowest possible frequency. For example, if 50Hz~1kHz is required, it must be calculated according to 50Hz. In addition, due to the high voltage, the insulation layer occupies a large proportion of the window area, and sufficient margin needs to be left when designing the core.

maychang

colinhnu posted on 2018-7-10 17:31 Thank you for your replies. Bandwidth requirement: -3dB 1KHZ, current is about 5mA, response: 100V/US (this parameter is very high, I feel). It doesn't need to be too large...

As for the power amplifier that drives the transformer, its supply voltage can be any value that is convenient for you, such as 48V, 72V... In addition, you said on the 8th floor that it is used for electro-jet printing. The current of electro-jet printing does not seem to be as large as 5mA. Of course, if the current consumption is less than 5mA, there is no problem if you design according to 5mA. There is no such thin enameled wire for the current of 5mA. You can only design according to the thinnest enameled wire, which will take up more window area. Therefore, sufficient margin must be left in the transformer core, and after the design is completed, it is necessary to check whether the window can be wound around.

colinhnu

maychang posted on 2018-7-10 19:18 As for the power amplifier that drives the transformer, its supply voltage can be any value that is convenient for you, such as 48V, 72V... In addition, you are on the 8th floor...

Thank you for your reply. Because I have a DC power supply of positive and negative 5KV, I also looked at some information about foreign equipment. Their solution is to use a common source and common gate amplifier circuit (I think that's what it's called). I feel that the circuit principle is relatively simple, but it feels so difficult for me to design it myself!

maychang

colinhnu posted on 2018-7-11 11:16 Thank you for your reply. Because I have a DC power supply of positive and negative 5KV, I also looked at some information about foreign equipment. Their solution is to use common source common...

Two tubes divide the 5000V DC voltage, and each tube bears at least 2500V. Have you seen MOS tubes with a withstand voltage of 2500V on the market? The tubes that are easy to buy on the market now generally have a maximum withstand voltage of 1000V (there are tubes that exceed 1000V, but the price is much more expensive and it is difficult to buy). Then at least 5 tubes are needed to divide the voltage to be able to bear the 5000V DC power supply voltage (it is best to use eight or even more tubes to divide the voltage). It is not impossible to design such a circuit, but with so many tubes requiring independent linear drive, one can imagine how complex it is.

colinhnu

maychang posted on 2018-7-11 11:39 Two tubes divide the 5000V DC voltage, and each tube bears at least 2500V. Have you seen MOS tubes with a withstand voltage of 2500V on the market? It is easy to find them on the market now...

Yes, the voltage is very high, and the requirements for the devices are very high, or the complexity increases a lot. It is too difficult for me. I can only learn and explore! Thank you for your advice!

colinhnu

maychang posted on 2018-7-11 11:39 Two tubes divide the 5000V DC voltage, and each tube bears at least 2500V. Have you seen MOS tubes with a withstand voltage of 2500V on the market? It is easy to find them on the market now...

Yes, the voltage is very high, and the requirements for the devices are very high, or the complexity increases a lot. It is too difficult for me. I can only learn and explore! Thank you for your advice!

maychang

colinhnu posted on 2018-7-11 12:00 Yes, the voltage is very high, and the requirements for devices are very high, or the complexity increases a lot. It is too difficult for me. I can only learn and explore! Thank you for your...

Learning is necessary. Explore is not necessary. The predecessors have already explored a lot of experience and conclusions. As application engineers, we just use them. I am afraid that I don’t know where to find these knowledge that the predecessors have already concluded.

maychang

colinhnu posted on 2018-7-11 12:00 Yes, the voltage is very high, so the requirements for devices are very high, or the complexity increases a lot. It's too difficult for me. I can only learn and explore! Thank you for your...

Maybe you have heard of it? China has now achieved positive and negative 750kV DC transmission. 750kV DC transmission has to be converted into 50Hz sinusoidal AC at the power consumption end to power various power consumption equipment. Obviously this is an inverter process, and the power supply voltage is 750kV. The insulator string on the 750kV transmission line is longer than two people stacked together. 750kV and your 5kV are 150 times different. You can imagine how complicated it is.

maychang

colinhnu posted on 2018-7-11 12:00 Yes, the voltage is very high, and the requirements for the devices are very high, or the complexity increases a lot. It's too difficult for me. I can only learn and explore! Thank you for your...

I still recommend that you use a transformer to step up the voltage to achieve an AC 5kV peak output. Of course, this transformer cannot be customized by the manufacturer of the transformer. You must design the parameters yourself. If you can wind the transformer yourself, then wind it yourself; if you do not have the conditions to wind it yourself, design the various parameters (including the structure) and then hand it over to the transformer manufacturer for winding. It is very likely that the transformer design and winding are not good at the first time, so start over. This is inevitable during the trial production process.

maychang

colinhnu posted on 2018-7-11 12:00 Yes, the voltage is very high, and the requirements for the device are very high, or the complexity increases a lot. It's too difficult for me. I can only learn and explore! Thank you for your...

In addition, you can consider piezoelectric transformers, which may be a possible solution to your problem. Piezoelectric transformers use the piezoelectric effect to convert electrical energy into vibration (mechanical energy), and then convert mechanical energy into electrical energy. Its output voltage can be very high. But I don't have any information on this, you can search it.