0
\$\begingroup\$

I am currently in the theoretical (until I am very sure that it is safe) situation that I have an alternator powered by a gasoline engine that provides 220V 1500W AC power with RPM dependent frequency. My goal is to get this into 50/60Hz AC (also 220V). I do have a 24V -> 220V 4kw inverter that would be capable of converting back 24V DC to 220V 50ish HZ AC. I also have 8AWG cables that should be able to handle the roughly 60A that 1500W would be at 24V DC. My problem now is, I am really confused by the rectifiers I find online. I am not really sure how to select the right one. It should be able to rectify high frequency 220V AC to 24V DC (is this even possible? Or will it be 220V DC?) and obviously in theory up to 1500W.

So essentially my target setup would be: 220V AC generator (not 50Hz) -> rectifier 24V DC -> inverter 24V DC to 220V 50Hz AC -> utility

Is this possible and how do I find an appropriate rectifier that can handle this? Simply searching for rectifiers I found for example this: https://www.amazon.com/Aodesy-Bridge-Rectifier-Module-MDQ100A/dp/B07N3Y6LVN which apparently handles up to 100A at 1600V (160kW?) but it seems so small for that task...

edit: If there are easier/better ways to get the generators variable frequency output to fixed 50Hz output I am open to that!

\$\endgroup\$
7
  • 3
    \$\begingroup\$ 220VAC will not produce 24V DC with just a rectifier. It'll produce more like 320V. \$\endgroup\$ Aug 26, 2020 at 18:53
  • \$\begingroup\$ Yea, I wanted to write that one other option would be to have 220V AC -> 220V DC via rectifier and then 220V DC to 24V DC via step down converter, but I did not want to overload the question. \$\endgroup\$ Aug 26, 2020 at 18:55
  • \$\begingroup\$ Actually, the rectifier will increase the voltage? I instinctively read 230V in your comment, but its 320V? \$\endgroup\$ Aug 26, 2020 at 19:03
  • 2
    \$\begingroup\$ You appear to have missed the point that the DC obtained by rectifying will be \$\sqrt \$2 times the AC voltage. \$\endgroup\$
    – Transistor
    Aug 26, 2020 at 19:12
  • 1
    \$\begingroup\$ Given that level of knowledge, I would have to say this is not a safe project for you. Reducing RPM will also not do what you want. Consider buying an "inverter generator" which already has safe, working electronics built in. \$\endgroup\$ Aug 26, 2020 at 20:14

2 Answers 2

1
\$\begingroup\$

If this is your 1st inverter design , You need more experience, but these parts work with low F PWM >20kHz but you will need a 100W heatsink minimum.

Consider a multiphase inverter. Go re-search on google.

\$\endgroup\$
2
  • \$\begingroup\$ It is, I will consider this then I guess, or go a different path of regulating RPM to get a 45-70Hz output from the generator :) \$\endgroup\$ Aug 26, 2020 at 19:33
  • \$\begingroup\$ Learn how to use a 4046 PLL to regulate the RPM using a Xtal and 4040 or sim. To get 50.00 Hz \$\endgroup\$ Aug 26, 2020 at 19:37
1
\$\begingroup\$

(is this even possible? Or will it be 220V DC?)

From this line, I think you need to have a bit more knowledge with regards to AC/DC electricity. When describing AC signals/power lines, since unlike DC it has a varying value w.r.t time, we use a value called "Root Mean Squared" to describe the characteristics. This is because the average of the signal would be zero, so it would not be of interest. take a look at this: enter image description here (Image from: https://www.electronics-tutorials.ws/waveforms/waveforms.html)

If you just find the average of the waveform, it will be zero. intuitively, the area under the graph for the positive half cycle is the same in amount as the negative half cycle, however they are different in sign, and thus, the average is zero.

If you want to find the average of the root of the signal squared (mathematically), it will look something like this:

enter image description here

The top one being the original signal and the bottom one being the squared values. long story short, the maths work out like this:

enter image description here

with T being the period of your wave form function and f(t) being the function. so to answer your question, no! rectifying a 220V voltage (so like I said, 220V is the RMS value), depending on how much ripple you have, yields a voltage around the peak of your waveform (Vp). so you need to reduce your voltage.

Look into transformers, step-down to be specific. This way, you reduce the voltage, to say 20V r.m.s. That means 28V peak. But that is not the end! you need to make a stable 24V, and for that you need to find a suitable DC-DC converter. should not be hard to find one, try one of the many electronic component distributors in your country and use the filters to find products matching your specs. the circuit will be something like this:

enter image description here

However, for your application, voltage regulators would mean massive power loss and a lot of over heating, so try to look into as efficient as possible ways (switching converters).

Is this possible and how do I find an appropriate rectifier that can handle this?

Well, there are tons out there! the most important aspect is average forward current. again, look into electronic distributors. but if, say your load draws 100A average current and the diodes have 1V forward drop voltage, the heat generated is just too much and you defo need a heat sink.

Very Important: I just wrote this answer to give you a better understanding of AC voltages. the mains voltage is also lethal and VERY, VERY dangerous if you don't have prior experience, and based on your question I don't think that you do, the best option for you would be to up your budget a notch and buy the pre-designed electronics as it is not a safe process what you wish to do. Hope this helps you.

\$\endgroup\$
1
  • \$\begingroup\$ This was a very insightful answer, thanks a lot! Actually buying a 2kW generator from a local shop would be reducing my budget but theres not much to experiment there, however I agree that it is not within safe limits to do this project on my own now :) \$\endgroup\$ Aug 27, 2020 at 19:47

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.