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.
Revlac Guru Joined: 31/12/2016 Location: AustraliaPosts: 1024
Posted: 08:30am 29 Apr 2021
Copy link to clipboard
Print this post
It would also be worth checking the rectifier stage after the transformers, It looks like those are diodes, 4 of them and another 4 on the opposite corner. On the other board, if its an inverter charger then these would likely be IGBT's with a internal body diode, they could be mosfets, hard to see in the photo. These can also be damaged internally with out any showing any visible damage, if not checked it can blowup all the mosfets that are going to be replaced. Worth checking all of them, when your up to that stage.Cheers Aaron Off The Grid
wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156
Posted: 03:23pm 29 Apr 2021
Copy link to clipboard
Print this post
I think better than a small chance as they appear totally asynchronous.... Edited 2021-04-30 01:25 by wiseguyIf at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike
hary Regular Member Joined: 15/04/2019 Location: FrancePosts: 89
Posted: 04:47pm 29 Apr 2021
Copy link to clipboard
Print this post
I've checked the 4 diodes on each side of the boar and they seems to be ok.
I've also checked the 4 IGBT IXGX60N60B and 2 of them were blown off. They now disconnected before I'm sure every thing else is ok to put the replacements ones into action
I'm not sure to understand what you say. On the picture below, the spikes look synchronised on the SG3525 output and on its VCC.
I wonder how I could find the culprit. Powering directly that small control board gives the same result. Adding a big 2 1000uF capacitor on its power rail doesn't make things better . . .
I had answered you that there was no buck converter embedded on that control board as the inverter was DC12V in, but the control board still has 5V for the dsPIC30F2010 microcontroller : So YES, there might be a buck creating the 5V from the DC12V in ! Edited 2021-04-30 03:55 by hary
nickskethisniks Guru Joined: 17/10/2017 Location: BelgiumPosts: 458
Posted: 08:10pm 29 Apr 2021
Copy link to clipboard
Print this post
I think he believes they are not synced. If I review it again some pulses are indeed not synced. So it could still come from your power supply.
There are still other options to consider. Like disable other things in the circuit. You could for instance remove the 12V supply from the board to your pwm IC and try to power it with an external 12V. Maybe a little loading of your push-pull gate driver can't hurt. Edited 2021-04-30 06:12 by nickskethisniks
hary Regular Member Joined: 15/04/2019 Location: FrancePosts: 89
Posted: 05:38am 30 Apr 2021
Copy link to clipboard
Print this post
As I said, I powered a spare SG3525 with that same power supply and it gives perfect signal. Then that spare SG3525 powered from the control board and both SG3525 output the same synced spikes...
With these SMD components, it's quite difficult to disable some of them from the board
I've been misled : That board doesn't have a buck converter, instead, it's a boost converter that makes 17-19V. Maybe that's the reason for the KA7500 and one of these 2 small transformer on the control board. The spikes seem to be related with the pulse on that transformer, I pretty sure. I might try to disable it as it has a true hole transistor and should be easy to disable.
In any case, the spikes make the voltage to drop very close to 8V which is way too dangerous to power this board (I think). Could it be a design issue ?
As always, experimented fellowes' feedback would be greatly appreciated. Edited 2021-04-30 15:41 by hary
wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156
Posted: 08:20am 30 Apr 2021
Copy link to clipboard
Print this post
Hary, boost converters by their nature draw current pulses from the power supply line.
The glitches occur simultaneously as what happens on the supply line is simply transferred to the main drive pulses powered from the same supply line. In a sense they are synchronised with each other by default, but they are not synchronised to the main drive waveform pulses as they occur in different places relative to the main waveform. That is why Nicks and I surmised they came from elsewhere ie external power supply or on board buck or boost converter.
It is hard to say whether it is partly a measurement technique or whether the spikes are really that bad.
If we assume they are there perhaps the electrolytic capacitor or a HF bypass ( I would expect at least a 1u ) in parallel with the electrolytic close to the main inductor to be missing or faulty. It actually looks like there is no capacitor in the circuit at all. So either temporarily tack into place some capacitors in parallel with the existing ones (or just replace them) and hopefully then see some improvement.
Perhaps the main boost inductor has a shorted turn and saturates drawing large current spikes?
A "shoot through" of sorts could be causing this but without seeing the schematic that should not be taken too seriously, something appears to be either drawing large spikes of current causing the voltage to collapse or the circuit is basically ok but the supply voltage is just collapsing due to a fault or missing capacitor/s. Check the 12V supply line from the external supply if it is also collapsing with those glitches is it up to the job?
Keep plugging away - it will have a reason that is just waiting to be discovered. Edited 2021-04-30 18:42 by wiseguyIf at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike
hary Regular Member Joined: 15/04/2019 Location: FrancePosts: 89
Posted: 12:34pm 30 Apr 2021
Copy link to clipboard
Print this post
So, here is an oscillogram :
In yellow outputA of the SG3525,in blue the power supply line, and in purple the pulse on the boost converter. I think you can agree with me that the problem comes from the boost converter.
Here is another oscillogram after removing the transistor feeding the boost converter transformer :
Now, I'm thinking : The SG33525 drives the push-pull main mosfets driver. So, by the time this push pull is in stable state (already commutated ON or OFF), there shouldn't be any trouble that the push-pull's transistors' gates has some spikes. The only problem would be if the spikes occurs exactly when the push-pull driver is changing from ON to OFF or from OFF to ON.The main power mosfets of the first stage would take longer time to switch and it would make them working in the linear mode witch could lead to failure ? ? ?
So maybe these glishes are not so important.
I add a 2000uF capacitor on the control board power line, but still the spikes do occur.
Do you think this is acceptable as the state of the art of a proper electronic circuit ?
PS : That control board makes some 18V, -18V and -12V (and +5V for the microcontroller)
wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156
Posted: 10:58pm 30 Apr 2021
Copy link to clipboard
Print this post
Hary, the circuit may well be operating ok - I have seen similar glitches on a CRO many times that are caused by very fast switching edges & measurement technique. If you probe the supply line you can see the narrow glitches that appear to be superimposed on the 12V line.
But when you now shift the probe from the 12V line to a ground connection (yes CRO input effectively shorted out to each other) you can still see the glitches but maybe reduced in height slightly (play with the trigger level perhaps it might need readjusting to trigger).
Once you realise that you see almost the same glitch when shorted to ground they can safely be ignored. The long ground clip lead of the CRO probe can act as an antenna causing you to see a glitch but it is not really affecting the supply line if it is seen on both ground and supply, it is referred to as common mode noise and actually cancels out as far as the circuit being tested is concerned.If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike
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
