# High voltage power supply design with off the shelf components?

#### icecats

Joined Feb 20, 2023
27
Hi,
I have been reading a lot about flyback driver circuits for generating high voltages. Tutorials on Youtube are a dime a dozen describing minimalist circuits to drive automotive ignition coils and similar. However these tutorials rarely describe the underlying design theory and considerations when creating the circuit.
I want to approach the design of a high voltage supply a bit more rigorously and understand the design of the circuit. Additionally, I would like to design the circuit around an off the shelf transformer with published specifications (such as this one or this one). I have seen various design notes for flyback driver circuits, but few of these are for high voltages. Here are my design specs:
• Input Voltage: 5 VDC
• Input Current: ?
• Output Voltage: >= 2000 VDC
• Output Current: ~10 mA

Edit #1: Based on some of the responses so far, I have added some additional context: The intent of the project is to send signals along an existing wire used for livestock fencing. I would like the project to be a battery operated and handheld device. My first thought for input voltage was 5v so as to be close to the same level as the micro I will use; however based on the feedback so far perhaps 5v is not the best suited for the application. The load device is an electrical fence wire. High voltage is necessary so the signal can overcome small discontinuities (for example a splice in the wire) and poor ground connections. My output current specification was an estimate since I read that electric fence energizers typically output around 100 mA of current. Perhaps lower current in my application would be suitable.

Last edited:

#### Dodgydave

Joined Jun 22, 2012
11,255
5V supply to 2KV is a big jump, what do you need it for?

#### MisterBill2

Joined Jan 23, 2018
17,885
That is a bit of power. 10 mA at 2000 volts is 20 watts out from the supply. The supply will have at least three sections, an oscillator section, a power driver to the step up transformer, the step up transformer itself, and then the rectifier and filter sections.
Usually for a high voltage supply a input voltage of more than 5 volts is used to keep the current levels reasonable and the wire sizes thinner.
So there you have a list of the separate segments in the power supply.
And I suggest avoiding what you see on the cartoon channel because so much of it is fake, and much of the none-faked stuff is not very accurate.

#### LowQCab

Joined Nov 6, 2012
3,957
Hi,
I have been reading a lot about flyback driver circuits for generating high voltages. Tutorials on Youtube are a dime a dozen describing minimalist circuits to drive automotive ignition coils and similar. However these tutorials rarely describe the underlying design theory and considerations when creating the circuit.
I want to approach the design of a high voltage supply a bit more rigorously and understand the design of the circuit. Additionally, I would like to design the circuit around an off the shelf transformer with published specifications (such as this one or this one). I have seen various design notes for flyback driver circuits, but few of these are for high voltages. Here are my design specs:
• Input Voltage: 5 VDC
• Input Current: ?
• Output Voltage: >= 2000 VDC
• Output Current: ~10 mA
.
Do You realize that You will need ~6+Amps at 5-Volts ?

Can You access a higher-Voltage Power-Supply ?, like maybe ~12-Volts or ~24-Volts ?
.
.
.

#### icecats

Joined Feb 20, 2023
27
.
Do You realize that You will need ~6+Amps at 5-Volts ?

Can You access a higher-Voltage Power-Supply ?, like maybe ~12-Volts or ~24-Volts ?
.
.
.
Hi and thanks for the response! The intent of the project is to send signals along an existing wire used for livestock fencing. I would like the project to be a battery operated and handheld device. My first thought for input voltage was 5v so as to be close to the same level as the micro I will use; however based on the feedback so far perhaps 5v is not the best suited for the application. The load device is an electrical fence wire. High voltage is necessary so the signal can overcome small discontinuities (for example a splice in the wire) and poor ground connections. My output current specification was an estimate since I read that electric fence energizers typically output around 100 mA of current. Perhaps lower current in my application would be suitable.

#### LowQCab

Joined Nov 6, 2012
3,957
"Handheld" !!!!
I don't think that You realize how much Power You are Asking for.
Unless, maybe You consider a Car-Battery "Hand-Held", it might last ~3 or ~4 hours.
Normally, Electric-Fences are AC-Powered, for a good reason, and they usually run 24/7.
There is no reason I can think of to use a Micro-Controller.
.
.
.

#### ronsimpson

Joined Oct 7, 2019
2,904
The intent of the project is to send signals along an existing wire used for livestock fencing.
2000 volts is hard to work with. It can only be "modulated" slowly so the data rate will be slow. Why 2000V and not 200V for 20V?

How far do you want to send data? Can you see from one end to the other?

#### LesJones

Joined Jan 8, 2017
4,174
If you are looking at the data on the two transformers in your links the 4000 and 5000 voltage ratings are the ratings of the INSULATION between the primary and secondary windings. The turns ratio of these transformers is far too low for your application.

Les.

#### icecats

Joined Feb 20, 2023
27
"Handheld" !!!!
I don't think that You realize how much Power You are Asking for.
Unless, maybe You consider a Car-Battery "Hand-Held", it might last ~3 or ~4 hours.
Normally, Electric-Fences are AC-Powered, for a good reason, and they usually run 24/7.
There is no reason I can think of to use a Micro-Controller.
3-4 hours would be acceptable run time for this device.

2000 volts is hard to work with. It can only be "modulated" slowly so the data rate will be slow. Why 2000V and not 200V for 20V?

How far do you want to send data? Can you see from one end to the other?
Hi! I understand that the modulation rate is slow. I have done some testing with lower voltages and found that they do not work well on a realistic electric fence with splices and poor contact in places. I selected 2000 volts because this should be able to jump gaps in the fence of ~0.5 mm. The fence will be up to a couple of miles long, so I will not be able to see from one end to the other.

Nah, that's not me. However, this is an interesting post.

If you are looking at the data on the two transformers in your links the 4000 and 5000 voltage ratings are the ratings of the INSULATION between the primary and secondary windings. The turns ratio of these transformers is far too low for your application.

Les.
Why do these transformers have such high insulation rating if they are not intended to be used at these voltages?

#### LesJones

Joined Jan 8, 2017
4,174
Pulse transformers are very often used to drive the gate on thyristors. The drive circuit will be close to ground level but the cathode and gate can be many hundreds (Or possibly thousands) of volts above ground level.

Les.

#### MisterBill2

Joined Jan 23, 2018
17,885
Transformer isolation voltage ratings are important because transformers are often used for the isolation that they can provide in circuits that they are not powering, but providing coupling. The pulse transformer is one example. Going back to vacuum tubes, the rectifier tube's heater voltage supply transformer in a DC power supply was very often at a quite high DC potential relative to all the associated circuits.
In mains powered control applications the power supply for the control circuits must be isolated from the mains power by a transformer, which might be part of a switch-mode power circuit.

#### MisterBill2

Joined Jan 23, 2018
17,885
For sending digital data over a high voltage wire the easy option is to use a radio frequency carrier coupled to the wire through an adequate isolation scheme.
For additional discussion of this, read about the "broadband over power lines" experiments done several years ago. THAT stupid project was dropped because it would have interfered with a vast amount of wireless communication.
For your application interference would be much less likely. BUT the coupling schemes could be similar. You could utilize a locally unused CB channel for instance, and modulation schemes similar to what HAM radio operators have used for digital communications. And since the system would only transmit when sending data it would be quiet most of the time.

#### icecats

Joined Feb 20, 2023
27
For sending digital data over a high voltage wire the easy option is to use a radio frequency carrier coupled to the wire through an adequate isolation scheme.
For additional discussion of this, read about the "broadband over power lines" experiments done several years ago. THAT stupid project was dropped because it would have interfered with a vast amount of wireless communication.
For your application interference would be much less likely. BUT the coupling schemes could be similar. You could utilize a locally unused CB channel for instance, and modulation schemes similar to what HAM radio operators have used for digital communications. And since the system would only transmit when sending data it would be quiet most of the time.
That is an interesting concept! Would the high frequency modulation somehow bypass the need for a higher voltage? As in, would a low voltage but high frequency signal be able to couple between breaks in an electric wire?

#### MisterBill2

Joined Jan 23, 2018
17,885
Consider that if you used a 27 MHZ radio, ( an old 23 channel small mobile, for instance, designed to talk for a few miles, jumping a small gap and delivering a signal should not be an issue. The coupling at the receiver end would be a small value high voltage capacitor, possibly even a few inches of coaxial cable, Then connected to a low impedance coil to allow only the RF to enter the receiver radio. A similar arrangement on each end would allow two way communication.

#### LowQCab

Joined Nov 6, 2012
3,957
Nothing will bypass the need for very High-Voltage.

Nothing will bypass the need for purposefully designed
Glass or Porcelain-Insulators along the full length of the Electric-Fence.
Poor, or dirty, or wet, Insulators will increase the Current requirements of the system.

2-miles of Fence coverage may not be practical from a required Current standpoint.
1/4-mile would be my estimation of the longest practical length per Power-Supply.
.
.
.

#### LesJones

Joined Jan 8, 2017
4,174
Reading though the thread again I am not sure if you want to send data on the electric fence wire that is still in use for it's original purpose. (I.E connected to several thousand volts.) and if so is this hand held device to also provide this power to the fence or just power the communications link. Is the communications voice or is it digital data ?

Les.

#### icecats

Joined Feb 20, 2023
27
Reading though the thread again I am not sure if you want to send data on the electric fence wire that is still in use for it's original purpose. (I.E connected to several thousand volts.) and if so is this hand held device to also provide this power to the fence or just power the communications link. Is the communications voice or is it digital data ?

Les.
The purpose is to send data on a wire that is still in use for its original purpose. The hand held device does not need to the power the fence and is only needed form communications. The intended communications would be digital data.

#### LesJones

Joined Jan 8, 2017
4,174
What sort of bandwidth does your data require ?