Aaron's Homepage Forum
Aaron's Homepage Forum
Home | Profile | Register | Active Topics | Members | Search | FAQ
Username:
Password:
Save Password
Forgot your Password?

 All Forums
 Electronics
 Power Supply
 12/120V inverter again
 New Topic  Reply to Topic
 Printer Friendly
Previous Page | Next Page
Author Previous Topic Topic Next Topic
Page: of 63

wasssup1990
Nobel Prize Winner

A Land Down Under
2261 Posts

Posted - May 04 2005 :  11:32:51 PM  Show Profile  Visit wasssup1990's Homepage  Reply with Quote
Hey John,

Use this site to host any file type no bigger than 2MB. You'll have 100MB of free space and your account won't expire. Totaly free and easy to join.

www.thefilebucket.com

J.C.

Edited by - wasssup1990 on May 04 2005 11:54:18 PM
Go to Top of Page

audioguru
Nobel Prize Winner

Canada
4214 Posts

Posted - May 05 2005 :  12:20:44 AM  Show Profile  Reply with Quote
Hi Johnny,
Thanks, but this site is only for chatting.On the other sites I attend we post projects, schematics and datasheets etc directly to the page all the time.

Go to Top of Page

hobby16
New Member

France
3 Posts

Posted - May 05 2005 :  03:30:30 AM  Show Profile  Reply with Quote
Hi all,
This circuit is fine http://www.web-ee.com/Schematics/110VAC%20Inverter%20for%20Automobile/inverter.htm
To simplify more, replace ICs 555+4020 with a CD4060.
Spread the word.
Go to Top of Page

hobby16
New Member

France
3 Posts

Posted - May 05 2005 :  04:20:51 AM  Show Profile  Reply with Quote
Hi Audioguru,
If you don't buy an inverter for yourself but have the occasion to give some advice, here are tips I found for commercial products I bought for my yatching activities:
- Square signals are the cheapeast and work not bad. Beware, it may blow your old tv. But with new electronic devices which have plenty of filter and surge protector, the overvoltage due to the square signal does no harm. Expect to use it at best 60% of the specified power. More and you'll have trouble (blow the fuse, blow your inverter, blow your 220V device, in order of preference!).
- In most case, you can increase the frequency to 100 hz (open the case and turn the right trimmer). The transformer will make less noise and heat and your 220V device will see no difference (forget the 50 hz digital clocks!).
- Don't be fooled by the pseudo-sinus marketing term. It's just like square signal, with lower yield!
- Sinus inverter are more sophisticated with all the bell & whistle protections. I have bought one because it includes a battery low voltage protection and it interfers less with my VHF. But most time, I think they are not necessary: they are considerably pricier which is NOT justified imho. Maybe, one day the Chinese (or the Indonesian) will master the microcontroller design and prices will drop.

Back to your 500W inverter, I simply don't believe it. Maybe it will consume 500W at the 12V side but as to finding 500W at 220V, humm. Do your people know the yield of their beast, the dissipation at the transformer, resistors and transistors. If they don't, then 500W (650W ?) must be in their dream.
And again, I think that saying there is no cheap mosfet in Indonesia is misleading. They have plenty of cellular phones and PC power modules. And those have plenty of spare mosfet. The people you are generously helping are beginner electronicians with bad ideas or solutions.

(Edit...Remove notification)

Edited by - Aaron Cake on May 31 2005 10:20:48 AM
Go to Top of Page

audioguru
Nobel Prize Winner

Canada
4214 Posts

Posted - May 05 2005 :  08:51:54 AM  Show Profile  Reply with Quote
Hi Hobby,
The Chinese inexpensive little inverters I can buy here have a processor and display all kinds of info (voltages, power output etc). They have a tiny fan for forced cooling. Some have a modified sine-wave output so that their stepped p-p and RMS output voltages equal a sine wave.

The 500W inverter that I helped design uses ten cheap 2N3055 transistors to produce 500W output with all transistors at their lowest guaranteed gain. With typical transistors its output power is considerably more.

My Indonesian pen-pal recently discovered that he can buy compact-flourescent lightbulbs very cheaply. Inside are nice expensive Mosfets and other valuable stuff. He is beginning to design inverters with those Mosfets.
Apparently, inverters and car batteries are a big business over there. Most towns don't have electricity but everyone has a colour TV and flourescent lights. The millions of inverters are powered by car batteries and some companies do nothing but charge them.

Go to Top of Page

hobby16
New Member

France
3 Posts

Posted - May 06 2005 :  05:49:05 AM  Show Profile  Reply with Quote
Hi Audioguru,
For your friend's information, some additional notes:
- Inverters are the future also elsewhere due to renewable energy. Here in France, wind and solar use must be increased even if electricity is predominantly nuclear.
- Mosfets (and in a lesser extend IGBT) price follow the same price curve of PC computers because they are hugely used in cars, cellular phones, hifi and other modern electronics.
- Beware between high voltage (main) and low voltage (<40V) Mosfets: low voltage offers very low On resistance and so low loss. Also, with a 4 milli ohm mostfet, good wiring is CRUCIAL. In an inverter , the use of high voltage Mosfets (with high RsdON) or IGBT is inappropriate. But you can always try with any Mosfet, the control stage remaining the same.
- My method of choice is for implementing a circuit is sourcing THEN design and NOT the other way round. But well, when you have Farnell or Radiospares, it helps!

(Edit...Remove notification)

Edited by - Aaron Cake on May 31 2005 10:19:34 AM
Go to Top of Page

John Doe
Apprentece

Mexico
14 Posts

Posted - May 25 2005 :  04:49:04 AM  Show Profile  Reply with Quote
Hi! I am new in this theme, but have worked a little in other stuff. I am trying to build an inverter for my car, searched a lot and found this the simplest inverter. It is based on saturation and cut of the transistor basics. I tried two circuits first, with only one 2N3055 (I started first with a TIP41C but it overheats and won't work anymore) using only a 500Ohms resistance and a diode, it worked, but just delivered me like 10 watts; that circuit is based on a touch box (I dont know how to say it "Caja de toques" (Its a box where you grab two electrodes and feel high voltage until body resists)). Second I tried the one presented here, except that the caps were electrolytic and in the polarity shown in the diagram on this page. Both caps blowed in 20 seconds, but the inverter gived enough power to turn a 100watts bulb, but not as if it would be connected to the wall. I tried other cap values, no Tantalium ones. I concluded that if voltage of the cap is more than 50volts, the circuit works, but does not delivery more than 20 watts; the ideal value is 16 volts but caps explode, the cap value in uf must be in the range of 10-100uf, i found ideal value is 47uf; I tried one bipolar (non polarized) caps and appears it will work, so let me buy the second and see if it works, because bipolar doesnt overheat, its a 6.8uf 25volts cap. Other important thing, is that 10ohm resistences also overheat and goes off, so I buyed wire resistances that come in plaster blocks at 5 watts and they worked perfectly. Also, I got a transformer, those used for public lighting with sodium steam bulbs, put off the cap and igniter that comes with, and check that it could serve. Its input is 125 volts, and delivers 55 volts 4 amperes; so I will rebobinate the secondary to deliver 12 volts, its perfect for this work. Sorry for my long post but I use to write long. Greetings! Good Luck!
Francisco Javier Olvera
Go to Top of Page

John Doe
Apprentece

Mexico
14 Posts

Posted - May 25 2005 :  05:07:54 AM  Show Profile  Reply with Quote
Ups, I forgot something... 2N3055 is 15 amperes ratted, try 2N5886 or TIP35c instead, they are 25 amps maximum ratted; only if you want more than 300 watts and dont like that 2N3055 overheats. Thanks Aaron for putting at easy reach those valuable ideas and circuits!

Go to Top of Page

audioguru
Nobel Prize Winner

Canada
4214 Posts

Posted - May 25 2005 :  09:25:22 AM  Show Profile  Reply with Quote
Hi Francisco,
This project doesn't work:

1) The caps explode because they were backwards and because their ripple-current rating is exceeded. Each cap is charged to about 24V, one-at-a-time. Then they have the transistor's collector end driven to ground, but therefore the other end tries to drive the other transistor's base to negative 24V.
All silicon transistors have an absolute max reverse-biased voltage rating for their base-emitter junction of only about 5V to 7V. Therefore the transistor's base-emitter junction is avalanching and causing an extremely high current in the cap and a huge waste of power. Over and over again by the oscillation.
Lots of power in a small cap equals a nice explosion!

2) Adding a protection diode in series with the base of each transistor will prevent the above problem. Then each transistor will need a resistor to ground at its base to turn it off.

3) You are correct that to produce 300W, the current from the 12V battery and conducted by each transistor is 25A, so TIP35C 25A transistors can be used.
But the inverter will have losses which reduce its output to about 210W.

The circuit's 180 ohm base resistors can provide a base current of about 61mA.
The datasheet for the TIP35C shows it conducting poorly with a max 4V loss with 5A of base current! The 2N3055 is much worse.
You can't simply use more-powerful transistors without increasing their input power.

4) You can use extremely low value base resistors and enormous caps to keep the oscillator's frequency down, to fix the above problem. But the resistors will cause a huge power loss. Therefore you need driver transistors or Mosfets.

So you see, it ain't easy to fix this poor design.



Go to Top of Page

John Doe
Apprentece

Mexico
14 Posts

Posted - May 27 2005 :  04:33:57 AM  Show Profile  Reply with Quote
Hi AudioGuru!

Thanks for your answer!
You are right in all points, I performed some changes in design.

1- First, I removed the two 10ohms resistor with their respective diodes, those who go from ends of transformer through center tap, they only avoid recoil of transformer and to stabilize oscillation frequency.

2- I added two diodes, (like you suggested me) from Base of both transistor to ground. Remember ground goes to emitters. It worked better.

3- I added one more 2N3055 in parallel with one of the transistor. (I was just random trying things! lol). It improved a little bit more output from the inverter. I really don't know if its correct, both transistors get less current from the circuit because current divides in two, and causes both transistor not to overheat.

3- I changed caps. The one, in side of the alone transistor its a 140uf 330 volts (those used for flash units); in the other hand, cap that is placed in the side of the two transistors, its a 100uf 250volts cap. That were best values I could get more output current, and like that, caps DO NOT overheat or explode! I used a 125/12 volts 3 amps transformer, I have not rebobinated the big one yet. But I have connected this transformer directly, and it delivers much more current than it would, for sample, it lighted a 100watts bulb almost perfect, it means capacity of it, is greater that label says. Thats why I am using it temporary.

One important thing, my current source for the inverter is a 125/12 volts 10 amps, so I really can be sure that its working at full capacity. I have not tried to power it from a greater source as a car battery. Other thing, I dont have a tool for measuring current frequency, to check that is in the range from 50-70 max. You are correct in the aspect that greater current transistors will need bigger input current, so I will not mess with them.
When I will connect the inverter from a bigger source, then I will conclude if it worked or not.

I really am trying to do my best to put this thing working better. Like you said back in the forum, if I want, I could buy me a new inverter, but this is like a challenge to me.

Go to Top of Page

John Doe
Apprentece

Mexico
14 Posts

Posted - May 27 2005 :  05:12:58 AM  Show Profile  Reply with Quote
Look at this
members.misty.com/don/flvdc.html
At the last sheet, it shows a medium power inverter, 200 watts rated; it was taken from a SolaBasic No-Break, that in fact is an inverter with a voltage regulator. It seems to be simple and must work. The problem is that rebobinate a transformer is a hard job without special tools. And you MUST do it in the correct winding phase, that is, if must be wired clockwise or counterclockwise, that are basic problems.
Greetings!

Go to Top of Page

audioguru
Nobel Prize Winner

Canada
4214 Posts

Posted - May 27 2005 :  05:32:24 AM  Show Profile  Reply with Quote
Hi Francisco,
I can't read the snakes and ladders of your TEXT schematic. This inverter should work fine and produce about 20W if its capacitors have the correct polarity and the supply is a 6V battery. Then the transistors won't avalanche breakdown.

Your ideas:
1) Without the 10 ohms resistors and diodes, the 2N3055 transistors will blow-up if a load is turned off with the inverter still running.
2) The reversed diodes you connected from each base to ground still cause a huge current in the timing capacitors when they are discharged into them. Also, the capacitors cannot swing to negative 24V then discharge slowly into the 180 ohm resistors for the timing period.
3) When paralleling transistors, each should have an emitter resistor to help equalise the different base-emitter voltage of them. They provide more output even though their drive is halved because their current gain nearly doubles with the lower collector current. Adding a paralleled transistor to the other side will increase the output even more. All transistors should each have an emitter resistor of about 0.1 ohms.
4) You won't need huge 330V capacitors if their discharging current is reduced by adding diodes in series with the base of each transistor. Then you can easily calculate the inverter's frequency since the discharge is only an RC network.


Go to Top of Page

John Doe
Apprentece

Mexico
14 Posts

Posted - May 27 2005 :  11:56:46 PM  Show Profile  Reply with Quote
Hi!
Again, thanks for your patience and your answer. Unfortunately when I send a post, no answer is given, so you dont know if it was posted or not, and always post same message twice or triple. I send a schematic in text format, but because the line width of this textbox, it appeared distorted, but next, I send the web link of it and delete the other one post.

In your point one, I dont agree with you. Why? because first, if I put those resistors with diodes, total output seems to be reduced, and second, I believe that nothing that could happen in the secondary winding could affect the primary, because they are insulated; that is why there are many transformers with same number of turns in both windings, to insulate two circuits; altough I see that when you connect a charge to the inverter, oscillation of the circuit slows down (you can hear it), so I have doubt. But I have disconnected the charge with the inverter on and nothing happens. Maybe because of the low amperage I am handling.

In the second point, caps cannot be charged to 24 volts, because it is not my input voltage, and my transformer is 125/12.

In third point, I put 1ohm resistors to each emitter of the transistors (that was the minor value I found for wire resistors), going from the emitter of one transistor, to the emitter of the other one (those one that are parallel connected). Total output power goes down significantly. But you are right, that there must be different voltages in the base-emitter junction of each one to make them do the switching. Electronics are sometimes difficult to understand. Also, if I put another transistor in parallel with the other alone transistor; output is reduced.. Is strange, but it happens. I think because that transistor is a TIP3055, the other side paralleled are 2N3055, theorically are the same, but one says 80mw and the other 115mw, frequencies are the same.That could be the reason, but I dont have another 2N3055 to put in that place. I will try smaller values for the resistors, I hope I found them, because coal resistors are maxium 1 watt. But minium tolerance values, and those one wire resistor, are 10% tolerance, that can cause differences in frequencies I think.

In point number four, I dont understand where do the diodes goes to? They go from Bases of each transistor to where, to negative side of each cap, or where? And how can I measure discharge time of the caps to know the inverter frequency..

Other thing is, that on monday I will get an UPS (No-Break), so I will disassembly it, and do a reverse engineering procedure; to see how it works. If you want, once I got it, I can send you the schematic.
I really appreciate your help. Thank you.
Go to Top of Page

audioguru
Nobel Prize Winner

Canada
4214 Posts

Posted - May 28 2005 :  02:46:13 AM  Show Profile  Reply with Quote
Hi Francisco,
1) All inductors produce a very high voltage when their current is stopped, like a car's ignition spark coil. Although the primary winding is electrically insulated from its secondary, the high voltage spike is inductively coupled from the secondary to the transistors in its primary. The resistors and diodes cause an enormous loss so perhaps zener diodes would be better like most power Mosfets have.
Your low current is reducing the voltage spike.

2) The center-tapped primary winding is a transformer all by itself. Since the center-tap is connected to +12V, when one end is driven to ground then the other end swings to +24V. So the caps charge to +24V.

3) When you parallel transistors, each one needs its own emitter resistor to ground. Because each transistor will have a slightly different base-emitter turn-on voltage and the one with the lowest voltage will take most of the current, which causes it to heat and reduce its turn-on voltage even lower while the higher voltage turn-on transistor does nearly nothing. Some manufacturers match transistor so they are about the same. You don't have a big box full of transistors to pick and match them so you should use 0.1 ohm emitter resistors to raise the turn-on voltage enough so that your transistors perform nearly the same.

4) Add a diode in series with the base of each transistor with its cathode to the base and a new 1k resistor to ground to turn-off the transistor. The anode of the diode connects to the cap and 180 ohm resistor. Then when the other transistor drives the junction of the cap and the 180 ohm resistor to about -24V, the diode will be reverse-biased and protect the base-emitter of the transistor from avalanche breakdown and allow the cap to discharge into the 180 ohm resistor for its timing period.
The discharge time of a capacitor into a resistor in parallel is RC where R is in ohms, C is in Farads and the time is in seconds. R times C is called one time constant and the voltage at a discharge time of RC equals 37% of the cap's fully charged voltage. The discharge is exponential so the amount of time that the cap discharges to the voltage to turn-on the transistor must be calculated. Making it more complicated is the fact that the cap isn't discharging into a parallel resistor. The resistor is connected to +24V. It is too late now for me to figure it out. I would just try it and measure the frequency.
Use electrolytic caps, not tantalum because the cap will be charged to about 1V in reverse for a moment.

When you have the inverter producing more power this way, you can try reducing the value of the 180 ohm base resistors (and increasing the value of the caps to keep the same frequency) for more base drive to the transistors. You will reach a point where the increase in power simply heats those resistors.



Edited by - audioguru on May 28 2005 02:56:20 AM
Go to Top of Page

John Doe
Apprentece

Mexico
14 Posts

Posted - Jun 02 2005 :  12:46:56 AM  Show Profile  Reply with Quote
Hi again AudioGuru!

Didn't answer before, because I wanted to make more tests. First, I got the UPS, its a TrippLite 280 watts, it doesnt work from wall outlet side, from the battery inverter it works great, results: clean and stable output. Obviously I didnt buy it new!.I disassembled it, and discover that, the transformer used, is much more bigger than mine, but I was believing that it was going to be like the one shown on the power inverter for powering flourescent lights web (the last one I told you), to have the primary tapped winding (drive), and a feedback winding, but not... It only is a center tapped 12 volts transformer, and in the 125 volts side, it has 2 wires for main output, and other three wires for voltage control regulation I supose. Second, I got surprised that transistor it uses, are similar sized as a little TIP 31c (for 6 amps) BUT, it has 4 and with great big heatsinks. I could not see the model of them. And schematic seems to be complex, because its not only an inverter but a surge protection and voltage regulator; it also has 2 PICS for control and auditive alarms, and to give connectivity to USB interface, and a serial memory.

It has an electronic component I couldnt recognize, it says "voltage 125 VAC, 12VCD", but I think its a Thyristor. Circuit is complex and may be I wouldnt be able to draw it out. But it works great for my purposes, but I keep working on my homemade inverter.

In other stuff, I powered my inverter with the battery of the UPS that provides like 15 amps I think. Changes to the inverter are:

1) 4 2N3055, connected in parallel. I didnt use the emitter-ground resistor you told me in point 3, because they are already wired to ground in aaron's schematic.

2) Caps where the same, one 100uf 250v, the other 140uf, 330 volts. They never heat.

3) Added diodes (for the two sides only), between the cap and the 180 ohms resistor that goes to the base of the transistor.

4) Diodes from sides of the transformer to the center tap with the 10ohm resistor remained.

I powered the inverter with the UPS battery and got like 75 watts. But better than if it was powered by my 12 volts source.

You are right what you say that caps charge to 24 volts, because of the tap, its exactly the middle of the winding. And you also are right that if you disconnect the charge from the secondary, it induces high voltage in the other side. Only when you relase the charge, thats why platinums (for the cars) got burned in a time, condenser avoids a little that coil effect.

In the other side, those 10ohm resistor, waste too much power, I disconnected them for short and power increased. In point number three, I cannot connect those resistors, because emitters are already connected to ground. But you are correct that not all transistor have same turn on voltages, because of little differences on materials being made.

I will test reducing the 180ohm resistor, and putting bigger caps, to see what happens. I will try to get an oscilloscope to watch output frequency, because, altough voltage and current can be fine, frequency can burn out delicate electronic devices. Greetings!
Francisco Javier Olvera
Go to Top of Page
Page: of 63 Previous Topic Topic Next Topic  
Previous Page | Next Page
 New Topic  Reply to Topic
 Printer Friendly
Jump To:
Aaron's Homepage Forum © 1995-2020 AARONCAKE.NET Go To Top Of Page
This page was generated in 0.19 seconds. Snitz Forums 2000