|
High Freq. Sinewave Inverter
Hi,
I have already tried transformerless( only small ferrite core of EE65X13) design upto 1000W but output was squarewave. Now I will use SPWM signal to convert DC to AC. In this type it easily possible to make 3Phase output without much circuit.
My main aim is to make sinewave inverter without using bulky iron core. I will upload some details very soon.
|
Great to know forum is alive!
Thanks a lot for information. For long time nobody posted and Saugata isOriginally Posted by dknagare
busy with his own projects. I also came across one commercial model which
didn't have any bulky transformer but only small ferrite ones like in smps.
I think I sent my block diagram to Saugata. I have changed OS in my PC and
it is difficult to locate it but I will get and put it here.
My idea was to have inverter with output switched by fast relay for crossover. Also if we don't use relay and instead use directly take out output
from inverter and meanwhile have heavyduty charger, we get cheapest
online UPS. For this there can't be output transformer and so output has to be HalfBridge type with centre tapped power supply. To get this advantage
we have to put premium on the Mosfet/IGBT voltage.(Require withstand voltage atleast 800 to 1000V) but that is better than some 10Kg transformer
for 1000W type and also price about 1000-1200/-.
I am also having one inverter with high frequency chopping but its output is
50Hz and so transfomer is very bulky.
Now I pray that sooner or later Saugata will tell us about new design !!!
|
High Freq. Sinewave inverter
Hi,
Can anybody provide me schematic diagram of output stage of HalfBridge as could not found application for dual voltage tapping +/- since the gate voltage should not exceed to say 20 V w.r.t. source of lower side Mosfet/IGBT. This can be achieved by using Bridge driver IC, but I could not find any practical application for it, however a separate bridge driver like TLP250 can be used but practically I have not yet used. Is anybody enlight/share on this with their experience.
|
Driving Mosfets in High Voltage.
Hello!
I already informed U in pvt mail that U have to use either Pulse txrs or
opto isolators for driving. Bridge driver ICs don't work more than 100volts.
All U have to do is put the txrs/optos in output of Ur circuit or use simple
bridgedriving ICs and txrs. In case of Opto's U have to steal driving current
with high resistors and zeners and use 1more transistor/gate to make fail-safe. Means when signal the FET remains off. I think that in this respect
the pulse transformers are ideal solution and cost no more than Rs.12-15 each. U can also borrow some technic and keep the gate reverse biased
during off period.
Many circuits are available on Fairchild's website as they have most the
driving ICs and POWER mosfets.
|
Sorry guys not being able to keep up.
Hi Dayama,
I am sorry I couldn't be in touch as I am under terrible pressure handling some railway projects. I am a one-man army as you know.
Back to the topic:
1. Any improvement or innovative idea is welcome.
2. My sinewave inverter is running successfully (24VDC-240AC~50Hz) but cannot share the schematic and code as I am commercially bound.
3. Should there be any PIC related problem I will surely try to help within my knowledge boundaries.
4. I have already done a PIC16F676 based squarewave UPS design (inspired by a forum member) and currently testing it.
5. I am working with PIC16F73 to port my design of sinewave inverter into a small offline UPS.
6. Double conversion is okay but generally more complicated and may be expensive in terms of ROI. What I have found is that boosting the voltage from 24VDC to 384VDC (to create a 240V AC) is less efficient. May be due to my limited knowledge on designing sm transformer. A 48V DC based boost converter gives better result.
7. I would be working on a different technology from June onwards that promises to deliver sinewave output using a single stage and ferrite transformer. I don't know whether I would be successful there as it is my own project and have limited funding. If I can spent time and get some fruitful results I will post them here.
8. Regarding MOSFET drive what I have realised is that a soft turn-on reduces harmonics and a fast turn-off is necessary to cope for switching losses. Also ringing should be well controlled otherwise bursts for no apparent reason. And never use the Microchip drivers. They are real crap. I followed every guideline and got terrible results. IR, Toshiba great. In fact I asked in the Master's conference that why Microchip is using IR gate drivers on their dSPIC motor control board when they have there own product line.
Regards
Sougata
|
How time flies!!!!
Dear Saugata,
thanks for writing. It has been long time to see U or some posts.
We all understand that U are helping the forum but U have Ur own army
to care and feed :-)
Now about Ur points:
1.& 2. We would never ask U to divulge Ur commercial secrets for FREE!
3. I have yet to commence the exploration of PIC. As I told U I am
doing all prokects in other world but ultimately it is a 8bit and uC thing.
So just nomeclature changes but basic things are same. I find that ATMEL
has also put up ARM type which are just as good as PIC and I don't have to change the assembly language. But sure will ask about PIC in few days after
finishing old projects.
4. I am also thinking about same PIC16F73 but I think other 676 has more PWM outputs which can of great help.
6.Ferrite transformer design is no big deal if U have designed iron core one.
There are many websites which have actual examples.Also if U can convert
48V to 300+ volts then 12-24 V to same is also not differrent. Just have some
more conductors and that helps for high frequency. For example I had make
primaries of 11SWG but instead I used 2 conductors in parallel of 14SWG.Nice thing about ferrites is that U need only few turns as turns/volt are as much as
40times less.
7. am hoping to know what 'new technology" U are talking about. Surely it will be very interesting for the thread of "sine wave inverter"
8. U have correctly stated the position of fast and slow turn on. Everybody
has to make decision about switching losses and surge supression and cost off higher voltage device. As far as Bridge drivers i hate those and I rely solely
on pulse transformers or opto couplers.
regards,
psdayama
|
IR2184 High/Low Side Driver
IR2184 is a high side AND low side driver. It generates the extra 10-15V needed above the bridge supply voltage to turn on the top FET, and gets you dead time protection and logic-level inputs, too. It's good for up to 600V, unlike every other FET bridge driver I saw. Two are needed for full bridge operation. If you need programmable deadtime, use the 21844.
Half-bridge schematic is included in the datasheet.
http://www.irf.com/product-info/data...ata/ir2184.pdf
Good luck,
Bryan A. Thompson
[email protected]
Last edited by batee; - 1st May 2006 at 21:04.
|
Half Bridge driver
Hi,
thanks but this driver may not be useful as the half bridge ckt we are discussing. I went thru the datasheet and driver ICs Vcc is referenced to
lower mosfet. However in our case the lower mosfet source is - 350 volts so there may be problem. D.K. Nagare has used pulse transformers which are
proper and no headache for driver configuration. He had problem due to voltage rating of Mosfet (600V type used in 700volt ckt). I hope he has found
the circuit working with say atleast 800V rated mosfets. I have suggested a
high value resistor network so that at output of bridge each mosfet has nominally 350V to withstand.
Hope to hear from U again,
psdayama
|
A capacitor pump of +12V referenced to the center of the bridge allows the top FET to turn on. I tried it at +170V top, 0V bottom and it works great, but agree the 700V would exceed its ratings.
Why generate a split source to power the inverter? I've not seen that done before. It seems like the cost of the copper and the extra DC link capacitor outweighs the cost of a couple of extra transistors.
If you're custom-winding a transformer to achieve this, can't you just add a tap at +/- 350V and another at +/- 360V to allow a simple level shifter transistor to turn on top FET?
Last edited by batee; - 2nd May 2006 at 08:51.
|
DDS Sine Waves...
Has anyone looked into using a DDS IC like the AD9833 for an inverter application? What were the results? How did you achieve PWM from the analog output?
Bryan A. Thompson
[email protected]
|
Split source
Problem is U haven't caught the thread! We are discussing eliminating
bulky 50Hz transformer for Inverter/UPS. Now there are upteen circuits with psudo sinewave output which are not upto mark in many situations like
running fans. What I have been proposing that instead of using PWM for
sinewave generation from battery there should be high frequency DC to DC
converter and the PWM drive the Half Bridge ckt. Now the load can be connected to output of inverter or normal mains supply just by switching
with relay and referenced to Neutral. Also if U don't want to switch then
U just arrange for proper battery charger and keep inverter always on making
online UPS.
I have found that the transformer costs whopping Rs.1200/- and weighs
more than 7kgs for 500VA output. On the other hand whole inverter costs as
much that amount as we use Ferrite txr of small size operating at 40KHz.
Instead of costly driver ICs one should use Pulse transformer as they cost
just Rs.12 and have 2 windings also. I just checked cost of 800V mosfet which is 6-7 times costly than 600V type 6N60. So what I think that in this
case 2 Mosfets in serries in one leg of HalfBridge would work well with pulse
transformer having 2 secondary windings.
Members who have read this thread : 2You do not have permission to view the list of names.
Posting Permissions
Bookmarks