Please confirm if i am correct that IGBTs QB + QA are switching sinewave and IGBTs QD + QC are switching square wave pulses in the  inverter shown..

The IGBT referrences QB QA QD QC

P.S. I removed the large black cap in the manual photo in the top photo of unit I'm busy with..

 

P.S. I removed the large black cap in the manual photo in the top photo of unit I'm busy with..

@AndrewF yes that's correct, so one igbt in the halfbridge is on for the whole half cycle duration, while the other igbt switches the sine wave pwm. No need for both to switch pwm pulses. 

 

@AndrewF yes that's correct, so one igbt in the halfbridge is on for the whole half cycle duration, while the other igbt switches the sine wave pwm. No need for both to switch pwm pulses. 

Thanks that makes total logical sense now, As QB has sine wave But QA doesn't have , need to trace where that opposite cycle wave has gone missing  !!

Which GND should/must be used on the scope probe to measure signal input on the opto isolators driving the inverter IGBTs...

On the 5 kVA schematic, QA1 and QC1 switch the neutral with 50 Hz square waves. So a square wave on QA1 would be normal.

 

Which GND should/must be used on the scope probe to measure signal input on the opto isolators driving the inverter IGBTs...

It's very tricky measuring waveforms on upper transistors in a live inverter. Hence BritishRacingGreen's efforts to document testing the gate drivers with IGBTs not present, or my testing of gate drivers with no bus voltage present, only current limited power supplies on +12 and -12V and +5V, and shorting the inputs to ground with the control board disconnected.

You can attempt to use two traces and subtract the voltages, and use no ground reference at all other than the DSO's internal connection to mains earth. At those voltages, a few volts between neutral and earth can be cancelled out by the subtract function.

Hi would it be possible to isolate the DC to DC SMPS converter mosfet gate pulses so that there's no switching on DC to DC converter and therefore no BUS voltage and then do test method injecting RasbPI signals into IGBT inverter circuit.

So use the inverters SMPS to power all circuits instead of using external power supply to power up as described in  journey of repairs....., the only external power supply would be the 25v - 150v DC injected into the BUS to confirm that IGBTs are working OK..

Just a thought.....

 

would it be possible to isolate the DC to DC SMPS converter mosfet gate pulses so that there's no switching on DC to DC converter

Yes. Not plugging in a control board, and not shorting the output of the opto enabling the SG3525 chip, should disable all pulses to the MOSFETs. I'm assuming a model with the DSP on a separate, pluggable control board. Some of the lower power models have the processor directly soldered to the main board.

 

therefore no BUS voltage

Without the control board, there should be no enabling of the bus soft start circuit. As long as there is no AC input, the bus should not be driven.

 

So use the inverters SMPS to power all circuits instead of using external power supply

I thought that there was a reason that this would not work, but I can't think of it now. @BritishRacingGreen, can you think of one?

It would certainly be convenient not to have to hook up more than one bench power supply.

 

I thought that there was a reason that this would not work, but I can't think of it now. @BritishRacingGreen, can you think of one?

 

Hi coulomb , yes it is possible to use the onboard SMPS and feed it with a bench PSU in the order of 35VDC to 55VDC without feeding the battery bus (CHG+ or BAT+).

I connect the positive of the external PSU to SCC connector pin 2. D53 will block the feed going onto the battery bus. The negative terminal can be connected to BAT- terminal , but if  there is a MOSFET driven polarity  circuit fitted on older models , then this negative must be connected after it , i.e. CHG- .

You still need to close the AC terminals in order to start the SMPS.

I have used this feed  for the last couple of weeks on all main boards that I have been investigating or repaired.

Of course this is only possible if  only the bare main board is powered , controller card , MPPT al undocked from it.

 

Hi on the AC conversion circuit all the IGBT are measuring the required -5V at their gates on startup except for QA1 which has no voltage on the gate, any ideas what normally causes this  problem  ?

According to Maxos schematic... Opto's  U2, U4, U12, are getting 15V and -5V from TX7 and   ZD3 and ZD4 respectively. I'm not getting correct voltages here.

At switch on voltage on ZD4 5.6v is -2.5V and once DSP board activates sine wave this voltage increases to -170V,  these voltages should be at -5v on ZD4 and 15V  on ZD3...

its strange that it  increases to -170 only when sine wave signal activated by controller board.

  mentions in one of his submissions that he rewound TX7 , I don't remember the reason for the TX7 rewinding .....was it to solve a similar problem  ???

Hi so I have managed to get the AC circuit to start up, though when it reaches around 200V AC the DSP board shuts it off again.... I wonder what that could be...I remember at one stage the BUS voltage went to around 399V DC now it seems to go up to 425V DC , is there possibly protection circuit kicking in... 

Any ideas please...

BTW this model is 145 max solar input...5KV .....48V

11 hours ago, AndrewF said:

is there possibly protection circuit kicking in... 

There is, but it should result in a fault code, probably fault code 09 (bus soft start failed) or 53 (inverter soft start failed).

I assume that if there was a fault code, you would have mentioned it.

11 hours ago, AndrewF said:

I remember at one stage the BUS voltage went to around 399V DC now it seems to go up to 425V DC

I assume that's measured with a multimeter, not from data from the inverter commands. It would be good to know that the inverter thinks the bus voltage is, though as I type this I can't immediately see how this would produce your symptoms.

11 hours ago, AndrewF said:

BTW this model is 145 max solar input...5KVA .....48V

So that eliminates any possibility of the solar charger messing with the bus voltage.

I can't think of anything yet...

8 hours ago, Coulomb said:

There is, but it should result in a fault code, probably fault code 09 (bus soft start failed) or 53 (inverter soft start failed).

I assume that if there was a fault code, you would have mentioned it.

The inverter started up normally then at around 200V output switches off ..

On 2023/11/02 at 8:09 PM, AndrewF said:

then at around 200V output switches off ..

Oh, that's interesting; per the "diagram", the DC-AC converter is running, but there is no load. Maybe that's always the case during inverter (not bus) soft start.

Does the whole inverter and display turn off at that point, or does it stay like the above?

On 2023/11/03 at 12:59 PM, Coulomb said:

Oh, that's interesting; per the "diagram", the DC-AC converter is running, but there is no load. Maybe that's always the case during inverter (not bus) soft start.

Does the whole inverter and display turn off at that point, or does it stay like the above?

The whole inverter shuts down with alarm  as the AC voltage reaches between 150v - 200v AC , but no error codes shown... , I'm busy reading through  "A journey Started"  hoping to see if he experienced a similar problem.. I removed the BUS caps measured them and they fine.. 

There might be a feedback voltage issue where the processor can't monitor output levels although the display shows the output voltage. I tested the few series resistors that appear to be a feedback system but need to study the schematic to confirm this, I was just hoping that the problem was common and fairly "easy fix" 

On the schematic for AC conversion U3 & U1 T350 IGBT driver opto's have 820nf ceramic caps across pins 5 & 8 , I damaged /lost these caps when replacing opto's , If I use 1mfd ceramic caps replacement would it be ok ??  I can't source 820nf...

10 hours ago, AndrewF said:

If I use 1mfd ceramic caps replacement would it be ok ??

Yes, they probably used 820 nF because they could not source 1 μF at the time of manufacture.

Edit: Or they always were 1 μF; caps don't tend to get values marked on them (grrr). 820 nF is possibly the closest E12 value to what Maxo measured on his test board way back when.

Edited November 7, 20232 yr by Coulomb

Is there a schematic of the DSP Module board for these  Axpert 5Kw inverters, I need to find why there's no low signal via CN11-3 to switch U17 on via pin2 to pull U16 pin 2 ( FB )low to enable the BUS soft start SPS to operate and pre charge the BUS caps.... U17 has 12 V on pin 1...And the BUS soft start SPS has about 13.5V on U16 pin 7.

 

Any thoughts why the DSP module not switching a low via CN11-3 to U17 pin 2.. ? 

Not sure if other inputs are required on the DSP board to allow the low signal from CN11-3 to activate Opto U17 

5 hours ago, AndrewF said:

Any thoughts why the DSP module not switching a low via CN11-3 to U17 pin 2.. ?

My schematic shows CN-11 pin 14; that's third pin from the pin 16 end, not third from the pin 1 end. Could this be confusing things?

As per the other post, check the power Darlington chip ULN2003 with diode and resistance measurements, and maybe try turning it on with say a 1 kΩ resistor to +3.3V. You might have to hold the processor in the reset state to stop the DSP from pulling its output low.

Quote

Is there a schematic of the DSP Module board for these  Axpert 5Kw inverters?

I'm not aware of any. I started a few times, but became frustrated with tracks running under large parts, and not being able to figure out where they go.

Edited November 19, 20232 yr by Coulomb

.

On 2023/11/19 at 3:36 AM, Coulomb said:

You might have to hold the processor in the reset state to stop the DSP from pulling its output low.

1. Hi Coulomb , how do I hold the processor in reset state ... ?

 

2. Are there any conditions on the main board that can stop soft start circuit from testing for shorts by pre charging BUS caps ?

22 minutes ago, AndrewF said:

how do I hold the processor in reset state ... ?

By pulling the reset low (/XRS, pin 78 of a '2809). Pin 77 of a '2809 is Vss, and it is recommended that XRS be driven by an open drain device, so it should be safe to short pins 77 and 78 (near one corner).

Looks like for the '2806x processors, /XRS is pin 11, with no adjacent pin connected to Vss (zero volts, digital ground).

You may find it easier to work with the triple voltage supervisor chip, U7. It has a master reset input on pin 7, internally pulled up with a 14 kΩ resistor. Pull it low (e.g. short to pin 4). For a '2809 processor, this part will have marking code 30718; the 18 at the end signifies a 1.8 V threshold. The '2806x based control boards won't have this power supply and may not have this chip at all. I don't have such a control board handy to examine.

 

24 minutes ago, AndrewF said:

Are there any conditions on the main board that can stop soft start circuit from testing for shorts by pre charging BUS caps ?

I would say if the main switch is off, e.g. the power supply is started because the solar charger said to (145 V max PV models only). Then the firmware would probably stay in standby mode, and would not progress to pre-charging the DC bus. Certain fault conditions can also keep the firmware in standby mode.

On 2023/11/19 at 3:36 AM, Coulomb said:

As per the other post, check the power Darlington chip ULN2003 with diode and resistance measurements, and maybe try turning it on with say a 1 kΩ resistor to +3.3V. You might have to hold the processor in the reset state to stop the DSP from pulling its output low

ON the DSP board I traced CN11-14 to the ULN2003 chip output then to its corresponding input, and soldered a test wire at the ULN input so that I could measure a low or GND on power up once DSP plugged into main board ....

 

On power up I get immediate GND connection on the test wire on DSP relative to U17 pin 1 measuring expected 12 V.  I then check CN11-14 again , power down , power up then GND works fine at CN11-14, then I get ready to test Soft Start BUS circuit and.... once again no GND on CN11-14 at power up... did some research, but all explanations on ULN2003 show +ve input switching on ULN not GND input switching, with 5V + at pin 9 and GND at pin 8, I cant trace any connection at pin 9(measuring power on pin 9 with board plugged in is impossible IMO) .. Any thoughts...