Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me.
Tinker Guru Joined: 07/11/2007 Location: AustraliaPosts: 1904
Posted: 10:34pm 26 Oct 2017
Copy link to clipboard
Print this post
I know, oztules did a very good explanation a while back how this feed back works.
I'm making some spare inverter driver boards and prefer to test all the EG1810 functions before connecting the board to the Mosfet gates/ toroid/ etc. This requires to preset the over current trip, check the low battery voltage shut down, monitor temp when the fan switches on and when complete shutdown occurs by overheated thermistor.
In other words, I want to make sure the driver board works as it should.
I check the waveforms going into the IR2110's and then observe the soft start at COM & LO (pin 2&1) on the Hi side.
OK so far, but what I am not quite sure why it is so, the board soft starts without *anything* connected to the VFB terminal.
Pin 13 goes to a 5K pot wiper via a 100R resistor, there are 104 bypass caps at each side of the resistor. One side of the pot goes to ground, the other to VFB in via a 12K resistor. As mentioned, VFB terminal in is open circuit.
I thought there is a faulty EG1810 chip but I then plugged in a brand new one and it behaved the same. The first EG1810 ran my inverter fine BTW.
So my 1810's run without voltage feedback in the above testing, can anybody confirm theirs do the same?
Klaus
Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498
Posted: 10:40pm 26 Oct 2017
Copy link to clipboard
Print this post
I take you mean EG8010?
I accidentally connected up an inverter the other day without the AC connected to VFB and it shot up to full volts very quickly, I shut it down and called myself a few four letter words. Plugged the AC into the VFB and it soft started in about 4 times longer. Although the initial start raced up it still was a soft start but it just happened much quicker. There are only 10 types of people in the world: those who understand binary, and those who don't.
noneyabussiness Guru Joined: 31/07/2017 Location: AustraliaPosts: 513
Posted: 10:59pm 26 Oct 2017
Copy link to clipboard
Print this post
Every one of the EG8010 chips i have soft started to a preset point and then promptly shut down from undervoltage ... correct flashes etc. . The data sheet has an explanation on this, to test the waveform. ..I think it works !!
oztules Guru Joined: 26/07/2007 Location: AustraliaPosts: 1686
Posted: 02:35am 27 Oct 2017
Copy link to clipboard
Print this post
Mad, the fast turn on... but still technically soft start is due to the feedback loop having no input. As you know, it's voltage stability and recovery is very fast, so on start up it will look at the VFB and that rising voltage will determine the start and end points of the ramp.
Without the VFB, it will increase the pulse width at the maximum default rate for start, as I cannot see a corresponding voltage come up as the pulse width ramps up... so goes for broke within the parameters set in the chip... so fast soft start, then collapse 3 seconds later.
With VFB, the chip sees a constant feedback voltage rising, so controls the rise so that it equates to the mandatory 3 seconds to reach target voltage... which it can see... so a normal soft start, and continues to run after the 3 seconds, as there is VFB.
Not sure of Tinkers real question. If he is saying that the 8010 runs after 3 seconds without VFB, and the VFB pin has no voltage on it..... it is simply the wrong series of 8010 chip.
If one reads the original drivel I wrote on playing with the egs002, there is reference to that behavior there along with the chip designation that caused it in that case. Clockman also did a reference to this problem, along with the numbers that work, and those that didn't from memory..... in the same thread I think.
I used to use the ones I had for testing in the very early days.... I had no others for a time. They work perfectly well.
......oztulesVillage idiot...or... just another hack out of his depth
Clockmanfr Guru Joined: 23/10/2015 Location: FrancePosts: 429
Posted: 04:16am 27 Oct 2017
Copy link to clipboard
Print this post
The EG8010 LQFP-32 sinewave inverter chip, that we know that works, ... should have the code ........... HN1518HT1536 on it.
Its those last 3 numbers that can be different.
I ploughed through Aliexpress until I could actually see those correct numbers on the chip. Like oztules I bought plenty, 80 or so, from different suppliers in the end I got 50 good ones.
This supplier was good. ..... but sadly no longer stocking?
Trust this helps?Edited by Clockmanfr 2017-10-28Everything is possible, just give me time.
3 HughP's 3.7m Wind T's (14 years). 5kW PV on 3 Trackers, (10 yrs). 21kW PV AC coupled SH GTI's. OzInverter created Grid. 1300ah 48v.
Tinker Guru Joined: 07/11/2007 Location: AustraliaPosts: 1904
Posted: 10:20pm 27 Oct 2017
Copy link to clipboard
Print this post
Thanks oztules & clockman, your info is spot on. It appears I have the wrong type of EG8010 chips. I must have forgotten about the different chips, I'm blaming senior moments...
Anyway, did some more tests with the chips I have today. They are: HN1702HT1718
I eventually worked out a misunderstanding about how the 8010 works, should have read the application notes more thoroughly .
When I pre test my control board I was using a VFB voltage that was *not* in sync with the VFB voltage feedback wave shown on my CRO. It could not, since it was generated from a different source, doh! I powered the little 12v tranny from a variac for this test. So the chip was looking for feedback peaks that virtually never were in sync. with the chip output. My theory is, it just kept running and did not shut down from low voltage feedback.
I tried a different approach then and fed straight, very well filtered DC voltage in at the VFB socket. No peaks on that one at all. Interesting result. The EG8010 started up (no low voltage shut down) and I slowly increased the DC voltage towards 3V. Switching off and on at various stages, the chip always soft started, but did so rather fast.
Approaching close to 3VDC the soft start got very nice and slow. At 2.98V (on my meter) it stopped soft starting, only the square wave on the CRO now. Ditto anything over 3V. The chip never shut completely down with this test and no status light flashing sequence BTW.
So it appears some of my inverter problems were to blame on the wrong EG80110 chip. But it also did run under high load with the same chip??
Anyway, I'll get on Aliexpress and see if I can buy the right part number chips, they are not expensive.Klaus
oztules Guru Joined: 26/07/2007 Location: AustraliaPosts: 1686
Posted: 12:23am 28 Oct 2017
Copy link to clipboard
Print this post
The "wrong chip" works perfectly...except for the lv and hv shut down...so still very usable
.........oztulesVillage idiot...or... just another hack out of his depth
poida Guru Joined: 02/02/2017 Location: AustraliaPosts: 1420
Posted: 01:41am 29 Oct 2017
Copy link to clipboard
Print this post
I did some work looking at Vfb behaviour on a EGS002. The tests used a function generator putting a small extra voltage pulse onto a steady 3V which was connected to Vfb. I could vary when the short pulse occurred relative to the 50Hz output waveform, as well as vary the pulse width and height.
The aim was to see - if Vfb is controlling output at all times throughout the 50Hz cycle. result: yes. of course. - when is Vfb sampled by EG8010 result: only during top of one of the 1/2 cycles of the 50Hz output - what is the time window when Vfb is effective? result: approx 1ms. For the rest of the 50Hz cycle, Vfb is not sampled.
I had the EGS002 board running stable with 3V Vfb DC and I added small narrow width pulses of DC (both positive and negative), seeing how small a pulse could be and still effect output PWM modulation.
I showed the results of this work either on this forum or another one which I can't recall.
I want to make it clear that in my tests, I saw Vfb to be not effective at any point in time of a 50 Hz output cycle except the 1ms (from memory, it could be different) window coinciding with the top of one of the 1/2 cycles of the output. This means the EG8010 is not controlling output using a voltage feedback throughout the 50Hz cycle. It only checks Vfb momentarily (1ms each 50 Hz) and then adjusts modulation up/down as needed see: https://youtu.be/SdkPxG5mNOY Purple trace is Vfb Yellow and Blue 50Hz and PWM output You can see the PWM modulation depth change only when pulse is at a particular time in 50Hz cycle.Edited by poida 2017-10-30wronger than a phone book full of wrong phone numbers
Tinker Guru Joined: 07/11/2007 Location: AustraliaPosts: 1904
Posted: 12:24am 30 Oct 2017
Copy link to clipboard
Print this post
Thanks Poida and yes, you did post these results a while ago here and I had read them.
The EG8010 data sheets also mention that VFB sampling is done only at the peak of the wave form.
What I want to do is test all the functions of the EG8010 (as far as possible) before connecting any Mosfet gates to the IR2110 outputs. Thus making sure my control board works correctly.
Some of us have had enough of blowing EGS002 boards (they smoke far too easily) and, even if they are cheap, built our own version of them. You may have spotted pictures of oztules control board and others of mine, which have the SMD EG8010 chip on a little sub board to make a more manageable 32 pin plug in version.
What this post was about, I had forgotten that not all EG8010 chips are equal, some apparently do not shut down at over or under VFB feedback. But they maintain the output AC voltage correctly non the less. Klaus
poida Guru Joined: 02/02/2017 Location: AustraliaPosts: 1420
Posted: 03:11pm 30 Oct 2017
Copy link to clipboard
Print this post
No worries. I seem to recall that during over or under voltage, the 50Hz and 20kHz outputs were still enabled and visible on the DSO screen. What changed was the timing of the outputs. The change (during over of under voltage) was such that the H bridge gates would receive signals times to both switch on or off as the same time. This drives both ends of the primary winding to the same voltage and so providing nearly zero voltage difference at the H bridge output.
I saw this as a sort of "idle mode", ready for adding some SPWM as required when intending to create an output of the bridge.
It's hard to verbally explain. Maybe this is what you are seeing and worrying about?wronger than a phone book full of wrong phone numbers
Tinker Guru Joined: 07/11/2007 Location: AustraliaPosts: 1904
Posted: 11:58pm 30 Oct 2017
Copy link to clipboard
Print this post
Thanks again Poida. Its nice to have somebody like yourself on this forum,, who has the equipment and knowledge to explain how things work.
I think you are spot on with the 'idle' mode -as far as my primitive testing can tell - and it did confuse me as with the other shut downs (over temperature, over current) there is no output from the 8010 at all once triggered.Klaus
Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406
Posted: 09:44am 31 Oct 2017
Copy link to clipboard
Print this post
[quote]I want to make it clear that in my tests, I saw Vfb to be not effective at any point in time of a 50 Hz output cycle except the 1ms (from memory, it could be different) window coinciding with the top of one of the 1/2 cycles of the output.[/quote] I can confirm that. This short sampling window only seems to occur right at the positive peak of each positive half cycle. If there is any noise or crud at the peak, Vfb becomes erratic, it needs to see a nice smooth sine hump completely noise free.
One other effect I noticed, any low pass filtering of Vfb should not introduce excessive phase delay into the Vfb waveform, or the voltage sampling point may not occur right at the peak of the hump where its supposed to be.
Initial testing with +3.00v dc on Vfb works. It just samples that, and seems happy.Cheers, Tony.
Tinker Guru Joined: 07/11/2007 Location: AustraliaPosts: 1904
Posted: 12:45am 01 Nov 2017
Copy link to clipboard
Print this post
One of the reasons I started looking into that VFB signal is that oztules had rectified the 12V output of the little sampling transformer on his control board.
Why??
On the application notes of the EG002 and the EG8010 circuits the VFB sampling signal is AC. None rectify the VFB signal in any way.
I tried 3V peak AC and its full bridge rectified version and both appear to work, there are just twice the number of positive sine wave halves on the rectified AC.
I shall conduct more experiments and choose the VFB wave form sine hump with the least noise present. I will also try to use shielded cable as in my inverter the little transformer is located about 300mm from the control board.
Feeding 2.98V clean DC into VFB certainly shows a very nice soft start. I think warpspeed made a good point about the noise on this signalKlaus
noneyabussiness Guru Joined: 31/07/2017 Location: AustraliaPosts: 513
Posted: 09:06am 01 Nov 2017
Copy link to clipboard
Print this post
Actually as u can see they do rectify the feedback signal when using the " low power frequency transformer " which we are effectively using... also in a earlier post of oz's there was an issue about grounding, hence the small transformer for isolation. . I think it works !!
noneyabussiness Guru Joined: 31/07/2017 Location: AustraliaPosts: 513
And the post about the grounding issue. . Bout halfway down...I think it works !!
Madness Guru Joined: 08/10/2011 Location: AustraliaPosts: 2498
Posted: 11:00am 01 Nov 2017
Copy link to clipboard
Print this post
The voltage on my Inverter is like Oztules describes it that the numbers are painted on the meter it does not change ever. The circuit I am using is 100% Oztules design which is DC to the chip. I think the only reason it samples at the peak as Poida discovered is that it takes one tiny sample in it's program at the top of the wave and then completes the next cycle based on that data. It would take a huge amount of processing power to try to regulate the voltage at 20KHZ and is completely unnecessary, once a cycle is more than enough to have a stable output. Edited by Madness 2017-11-02There are only 10 types of people in the world: those who understand binary, and those who don't.
poida Guru Joined: 02/02/2017 Location: AustraliaPosts: 1420
Posted: 11:22am 01 Nov 2017
Copy link to clipboard
Print this post
Tinker, I do not like the idea of 3V AC applied to Vfb. EG8010 datasheet on page 11 tells us the maximum rating for any I/O pin is -0.3V to 5.5V with respect to ground.
3VAC is going to be -3V to ground and it's low impedance too. It would be unwise to expect the IC to survive this for any length of time.wronger than a phone book full of wrong phone numbers
Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me.
Print this post
.
