June 6, 20242 yr Author 16 minutes ago, BritishRacingGreen said: My original theory that Voltronic has gone for a non-standard method to control SMPS negative feedback via the UC3845 (U10) COMP (compensation) input pin , holds no truth. In fact the datasheet describes it as a preferred method where secondary-side feedback is required (via the isolation barrier). In a primary side design (no isolation) the feedback pin VFB is used with a simple voltage divider from the output rail. An extract from the datasheet is shown below: Please take note that the 3843 device in the family is nearly identical to the 3845 , except for one important detail . The 3843 can span the entire output duty cycle 0-100 % . The 3845 is capped at 50% , regardless of the state of negative feedback on VFB and COMP pins. So don't use 3843 on the Voltronic boards unless where specified. See below : Edited June 6, 20242 yr by BritishRacingGreen
June 7, 20242 yr 14 hours ago, BritishRacingGreen said: Please take note that the 3843 device in the family is nearly identical to the 3845 , except for one important detail . The 3843 can span the entire output duty cycle 0-100 % . The 3845 is capped at 50% , regardless of the state of negative feedback on VFB and COMP pins. So don't use 3843 on the Voltronic boards unless where specified. See below : Thanks for the info. I am familiar with the UC38xx family having repaired other products that use them. Still need to design a flyback with them at some time tho. The parts have arrived, its this weekend's task to get this inverter up and running again. Will feed back my progress as and when.
June 14, 20242 yr nice...smps discussionBritishRacingGreen i was totally defeated with dead smps of a clone 6k solar inverter the main smps is fed by battery sps that produce 150vdc input ....the 150vdc comes in...but no12...no5vdc...no-12 dead replaced us3845 ...no use tested vcc.is 8.2vdc tested diodes ok no output short ..the ic is ok i fed 8.5vdc without even powering the inverter and got square signal at pin 6 and5vdc at refrence pin i tried to tweAk start up circuit by reducing some resistors to boost vcc...no use vcc still 8.2vdc i notice with a scope the ic produce some pulses at start up but no output[though pin 8 .never give 5vdc) ...actually there was around 2vdc at output at startup phase...then it all collapse i run out of ideas replaced coopler and tl431 no use i lastly suspected smps transformer...can i test it with variac...any other ideas..or shorted aux winding.. repaired many but this one exceeded my skills Edited June 14, 20242 yr by wael_fathe
June 14, 20242 yr Author 10 hours ago, wael_fathe said: tested vcc.is 8.2vdc Hi @wael_fathe I trust you are well. By 'vcc' are you reffering to the vcc pin 7 of the UC3845? 150vdc from the SPS is in my book is way to high, it should be about 60vdc max, if the SMPS is Voltronic standard one for Axperts. Does your smps relates to the standard one in my and @Coulomb 's schematics?
June 15, 20242 yr On 2024/06/14 at 1:05 PM, BritishRacingGreen said: Hi @wael_fathe I trust you are well. By 'vcc' are you reffering to the vcc pin 7 of the UC3845? 150vdc from the SPS is in my book is way to high, it should be about 60vdc max, if the SMPS is Voltronic standard one for Axperts. Does your smps relates to the standard one in my and @Coulomb 's schematics? this design is normally produce 150vdc input to main sps... its clone...but the outputs too low ... yes pin 7 is 8.2vdc uc3845b done all the doable time to change the transformer...any ideas to test it؟ Edited June 15, 20242 yr by wael_fathe
June 16, 20242 yr On 2023/04/27 at 5:41 PM, BritishRacingGreen said: I have a nasty solar related problem , but an interesting one also. I passed on my old MKS3 (only 2 years old though) to my son-in-law , he added a Pylon UP5000 and used it as a backup system for a couple of months. Last week we added 4x555W panels to the mix (one string supply (170-180V working voltage) . I encountered a nasty failure mode in that the PV will supply high loads or charge flat battery without any problems . When the PV demand is in the order of 1kw and higher , no problem. But when the load is only like 300watts and the battery is full, then the system becomes unstable . The battery voltage rise slightly , and of course affects a large charge current from the MPPT . when that happens , the PV drops again , and everything repeats at about 3 second interval. Add a large load , hairdryer at 1.5kw , everything is stable again. or add a flat battery , everything is stable again. weird. What worries me is that sometimes the AC output also fluctuates somewhat. The MKS3 MPPT was just working fine when I decommissioned it at my own site. So here is my theory . I think the MPPT regulation is faulty . I have seen it on the DC bus voltage curve on the HomeAssistant log. So when there is no load on the MPPT , the DC bus shoots up . But when there is load , the regulation improves for the wrong reason , dampening the rise in DC bus voltage. Could be a faulty MPPT current sensor , I am taking wild guesses. There are suggestions it may be a panel related issue , but I have my doubts . When a panel is weak , or we have dry joints in the PV cables , we would experience just about the opposite behavior as described above . Fortunately , courtesy of our dear friend @Steve87 , I have an MKS4 which I am going to test tomorrow at son-in-law site , in order to make sure in what subsystem of the installation the fault lies , panels or inverter . So I will keep posted on my findings . One thing I have learned , is that my PV commissioning procedure must be revisited in order to ensure that PV operates adequately at various load conditions. The interesting thing is that no error is offered , I think the poor regulation is still within the DSP's windows of forgiveness , otherwise it would have raised a DC voltage too high error or so. I hope its hardware related and not require firmware upgrade . I am sh#t scared of firmware upgrade , so far I had no need to touch that , it makes me wake up at 3 in the morning , eyes wide open. Then I realize we have a mentor in @Coulomb , nightmare is over and I fall asleep again. @BritishRacingGreen I was reading for 2 hours about your journey into repairing different inverters. This one particular defect I am experiencing also on a Easun SMX2 inverter. MPPT does not function as expected - more or less the same behaviour as you described. What was the conclusion you found ? The inverter I received had burnt IGBTs on DC-AC side, replaced them and seems to be working OK now. However, the owner said that this MPPT error was present before it finally died. Someone probably connected PV input reversed. I checked diode and IGBTs on MPPT controller - they are good... Edited June 16, 20242 yr by sethmad
June 16, 20242 yr 9 hours ago, wael_fathe said: done all the doable things to change the transformer...any ideas to test it؟ Not really, apart from checking that there are no shorts primary to secondary. You really need a shorted turns tester, and one that operates at high frequency. If you happen to have a RF signal generator, you could apply max output and 20 kHz frequency or the lowest it will go at the primary and check for something at the secondary. Ideally it should have voltage close to the ratio of the number of turns (same turns per volt at the input and output). I suppose an AF signal generator at the top end would also work. Don't test it at 50 or 60 Hz; it just won't work as it's ferrite.
June 16, 20242 yr 3 hours ago, sethmad said: MPPT does not function as expected I recently linked a messy partial schematic of a high PV voltage MPPT that I didn't realise someone had posted: https://forums.aeva.asn.au/viewtopic.php?p=101874#p101874 Edited June 16, 20242 yr by Coulomb
June 16, 20242 yr 4 hours ago, Coulomb said: I recently linked a messy partial schematic of a high PV voltage MPPT that I didn't realise someone had posted: Thanks a lot! @Coulomb Does anybody have a quick guess on what could go wrong if someone accidentally swaps the PV input wires(reverse polarity on PV input)? This seem to be the main cause of the defect. Unfortunately I have 2 boards with exactly the same behavior. This is why I firstly thought it's a software issue. (very big coincidence as these boards came from different persons) Board1 - Had the big 500V capacitors exploaded - I replaced them and now seems to be working fine - except the MPPT part. Board2- had defect IGBTs on DC-AC part, replaced them and seems to be working fine - except the MPPT part. Owner reported that this behavior appeared only after accidentally reversed polarity of PV inputs. All power IGBTs and power diode seems to be OK. Mcu reports no error - it just tries to lock the MPPT but voltage drops to 350V to 40V and then restarts. Will check more next week(I hope I can get a good board so I can do cross measurements also). Any hints would be greatly appreciated! I attached some photos. MPPT is included on the mainboard on this model. Edited June 16, 20242 yr by sethmad corrections
June 16, 20242 yr Author 16 hours ago, wael_fathe said: this design is normally produce 150vdc input to main sps Really , why on earth would a clone 'design engineer' affect a fundamentally different SMPS design ?. The spec of TX9 , the analogue feedback loop , the current limit feedback to name a few, all needs to change. But most of all , the spec of 150V is just senseless. The SMPS has been originally designed to accommodate battery-dependent machines . In this case the raw battery is fed to the SMPS DC input rail. So this clone actually needs to have a completely separate 48V-150V DC-DC converter , that is not economical . In later battery-independent designs , a separate SPS module is required where both mains and PV (SCC)must supply the main SMPS input , in combination with the battery. Both mains and PV has dc-dc converters to generate 60VDC and this is connected to the SMPS input rail via steering diodes. In the standard Voltronic design the input range is not specified , but I can state that where an SPS is used , the voltage is 60VDC derived from mains , 60VDC derived from PV , and whatever the battery voltage is (48-56VDC). I have tested this SPS+ input from 42VDC - 70VDC . The schematic of the Voltronic SMPS refers as shown below: Note the DC input filter electrolytic capacitor C69 . It is rated at 100VDC , so it is not going to be 150VDC friendly . What is yours? The Voltronic part number of TX9 is 41-070194-02G . This is true for various machines I have worked on . Has your transformer a similar marking ? Edited June 16, 20242 yr by BritishRacingGreen
June 16, 20242 yr Author Hi @Coulomb , I trust that you are well . I have a question for you regarding the compatibility of different Axpert low voltage SCC modules within 5kW capability. I have an old MKS2 Mecer machine that has a large flat SCC module that is mounted on top of the main board. This module is beyond repair as it experienced pcb burns. I do have a compact low voltage 5kW MPPT that comes out of an Axpert King II machines. ALthough also faulty , it it still serviceable. Both are same specs , 5Kw , low voltage , battery SCC's. Here is my question : Is there a solidly high level of abstraction between SCC and its host in terms of the message protocol employed? In other words , is it possible to mechanically retrofit the bottom SCC into the MKS2 , and the SCC should 'happily' honor the host requirements ? Or is it not that simple.?
June 16, 20242 yr 5 hours ago, sethmad said: Does anybody have a quick guess on what could go wrong if someone accidentally swaps the PV input wires(reverse polarity on PV input)? First of all, there will be large electrolytic capacitors across the PV input. These do not like reverse polarity, and could fail spectacularly. But they may not get time for much in the way of fireworks. High PV voltage solar chargers are essentially boost converters. So: inductor in series, active switch in parallel, diode in series, capacitor in parallel. With reverse polarity, the free-wheeling diodes (back diodes, antiparallel diodes) in the switches (IGBTs usually; 500 V is too high for MOSFETs) will short the input via the inductor. So far, nothing too bad happens; the free-wheel diode is rated for the same current as the IGBT, and is fast as well. But then the inductor will saturate in a fraction of a second, making it very low impedance, and this will (I think) cause the PV diodes to fuse. But wait, the inductor has considerable stored energy (E = ½LI²), where does that energy go? It doesn't go towards the load, because the diode (the boost diode, not the IGBT free-wheel diode) is reverse biased now. The inductor is bursting with energy, and does not want to allow a rapid change of current as the fuse is demanding. So it generates an enormous (multi-kilovolt) back-emf to try and find a path for the current to continue to flow (in the "wrong" direction, opposite to normal operation). This high voltage will break down something, probably the boost diode, and that transfers the bad polarity to the DC bus. That does no favours to the bus capacitors. It now could be transferring energy to the buck stage's inductor, which is also hefty. From the carnage usually seen, I suspect that a pulse finds its way across the transformer to the MOSFETs and causes problems there. Sorry, I don't have great detail, but let's just say that it's highly desirable to avoid this situation if at all possible. Interrupting big inductors' current flow with a "mechanical" switch or fuse never leads to good things. There are movistors in there to absorb transients, but these things are the size of a few coins stacked atop each other, and can't compete with a ferrite toroid a hundred times its size and weight.
June 16, 20242 yr On 2024/06/16 at 8:32 PM, BritishRacingGreen said: But most of all , the spec of 150V is just senseless. That surprised the heck out of me too when I first heard about it. I think it has something to do with batteryless operation. I think that they just have to use a higher voltage MOSFET or IGBT, and of course the bypassing capacitor. The "transformer" turns ratio possibly needs to be tweaked as well, but that's not a huge issue. These flyback converters can apparently accommodate a fairly wide range of input voltages. On 2024/06/16 at 8:32 PM, BritishRacingGreen said: So this clone actually needs to have a completely separate 48V-150V DC-DC converter , that is not economical . I'm not convinced that this EASun is a clone. And I'm 98% sure that the 150 V bus is a Voltronic design, not a clone design. EASun seem to have a slightly different relationship with Voltronic than the usual resellers; they appear to be manufacturers in their own right. Quite possibly "reformed" clone manufacturers (or threatened with legal action, whatever). Inverex seems to be another manufacturing reseller, that licences the basic design and firmware from Voltronic, but makes the product themselves (I'm guessing). But requiring another DC-DC converter does seem crazy. 60 V seems to be a much better choice, but I guess it makes the battery a power source of last resort. Often you want that, perhaps there are times when you don't. Edited June 23, 20242 yr by Coulomb
June 16, 20242 yr 3 hours ago, BritishRacingGreen said: I do have a compact low voltage 5kW MPPT that comes out of an Axpert King II machines. No, the King IIs are the high PV voltage models. The "standard" King (they actually use the word "standard" or abbreviation "STD" in firmware for what I call King 1s, to distinguish them from King IIs) is the one with the 145 V max PV solar charger. 3 hours ago, BritishRacingGreen said: Here is my question : Is there a solidly high level of abstraction between SCC and its host in terms of the message protocol employed? In other words , is it possible to mechanically retrofit the bottom SCC into the MKS2 , and the SCC should 'happily' honour the host requirements ? That's a great question. I can tell you that King 1 firmware uses ASCII commands that look very familiar to me. I would guess that the commands are 100% compatible, though there may be devil in the detail. For example, the MKS models came in 60 A and 80 A variants; the MKS firmware goes to a lot of trouble to remember this limit and acts accordingly. The King 1s came out when 80 A was established; there were never any 60 A Kings. So there is no need for the "60 A agility". My wild guess is that there would be about a 95% chance that you could just bolt an MKS charge controller into a King 1 or vice versa, and it would just work. I reverse engineered and documented the Solar Charge Controller protocol many years ago; see the AEVA PIP-5048MS etc index if you are curious. [ Edit: It's a little hard to find: see this post. There are fewer details than I remembered. ] No surprise, it's 2400 bps serial commands with the same CRCs and general format as other commands, so the same parser and command handler system is used for both. Serial (SCI) port 1 of the DSP is for the solar charger, and port 2 is for either the PC in the case of machines with no removable display, or to the display for models (like the Kings) with removable displays present. One of the commands is VERFW, which I often confuse with the QVFW and QFVW2 (sometimes even QVFW3) commands. The QGS and GS commands are the main commands once the protocol has started. The early charge controllers probably evolved from Voltronics' stand-alone MPPTs like this one: I can't quickly find the model number, but something like GX-60 comes to mind. People have noted the remarkable likeness of the PCB inside these things. It may even be the same (perhaps apart from connectors), but with different firmware. The stand alone version drives a small LC Display; the one for the Axperts does the serial comms. Edited June 16, 20242 yr by Coulomb As noted.
June 16, 20242 yr Author 16 hours ago, sethmad said: MPPT does not function as expected - more or less the same behaviour as you described. What was the conclusion you found ? Hi @sethmad welcome to our discussion. We actually never got to the bottom of this . We introduced an MKS4 machine for a second opinion , the result were slightly better , but still did not resolve our issues. We subsequently added another panel that increased the open voltage pv input from 190V to 230V , and that solved a lot of issues. It appears that below 230V the mppt does not seam to be in a happy space. There is still pv power jitter when the battery is full and load is at 100W , but we decided to live with that . My own MAX7.2 at home has exactly the same pv voltage , but much more stable at low (100W) loads. So whether there is an issue with my son-in-law string I don't know , because the panels delivers very well when the power demand increases.
June 16, 20242 yr Author On 2024/06/14 at 2:29 AM, wael_fathe said: at startup phase...then it all collapse Yes that is normal , because the initial bootstrap is only there for short while. In that period the SMPS should produce , and amongst other , will generate the +15VDC supply over C75 , which will power U10 via D49 even after the bootstrap via D50 falls away. But because your SMPS does not deliver , when the bootstrap via D49 falls away , everything collapses. The following test is provided to statically test the SMPS negative feedback loop. While no attempt is made to verify the linearity of the transfer function , it is adequate to test the extreme low point of feedback (resulting in highest output duty cycle) , as well as the high point of feedback (resulting in lowest possible output duty cycle) . To do so , we will manually power up the U10 and support circuitry with a power supply of 15V , and manually inject a variable 12V supply on the +12V output rail in order to fool the U10 that it is 'regulating'. 1. Remove all ancillary connections external from the main board , including controller board , fans , mppt , battery , everything. 2. You will require 2 x 12V variable bench supplies and an oscilloscope . Both of them must be current limited and both must be able to vary up to +15VDC. 3. Connect psu 1 positive to D57 cathode and negative to BAT- and set to +15VDC . Current limit to 100mA or so. This will provide a steady VCC on U10 pin 7 and power the chip . 4. Check there is steady +5V on U10 pin 8. 5. Check with oscilloscope that there is frequency and 50% (max) duty cycle on the gate of mosfet Q36 . 6. Connect psu 2 set to 11.7 V between D54 cathode and GND . Set current limit to 800mA. This will false feed the +12V bus. At 11.7V the U10 output duty cycle should still max out at 50%. Check that the voltage drop over U8 diode is well below 1.1V . PLEASE NOTE : by injecting this 12V supply remember that the 12V loads on the main board is still connected . So please make sure that you dont push this voltage higher than say 12.8V. 7. Increase the psu to 12V and measure the TL431 bias voltage over R209 . It should be very close to 2.5V . Increase the psu to 12.5V and make sure the bias voltage is above 2.5V. at this stage TL431 turns hard on and current flows thru the U8 diode. The voltage across this diode should saturate at approximately 1.1V. This will in turn turn the U8 transistor hard on on the SMPS primary side . In turn this will pull the COMP pin 1 of U10 well down to 0V . As a result of this , U10 must correspond by outputting its lowest duty cycle on its output to the mosfet gate. 8. If all of this passes , the feedback loop circuit is pretty much in order. ... I will continue with this in another post whereby you can bring the SMPS up with starting with a small 10V voltage at its dc input and working our way up to 40V if all goes well ... Edited June 16, 20242 yr by BritishRacingGreen
June 18, 20242 yr On 2024/06/16 at 5:52 AM, Coulomb said: Not really, apart from checking that there are no shorts primary to secondary. You really need a shorted turns tester, and one that operates at high frequency. i have bob paker ring tester..i used it it gave me 4 rings it is a marginal result could be good could be bad the device have 8 led indicator it seems 8 led lit means fine 4 led marginal and probably fine above 4 is fine less than or equal 3 is bad ... i closed the unit down ...in despair..my last try in thought about injecting a forced 13vdc to pin 7 vcc of the uc3845b using or-diode and observe the output....alas the output gave me 4vdc in what should be 12vdc output some thing is bad ,,,and it is not ... a feedback is i knocked a resistor in feedback cirucuit out and stilll get the same result open loop failure it have to be a transformer about your suggestion of feeding in 20khz i got a nice idea how about stealing some ac high frequency/from another inverter feed that in ... 200vac...f= 20khz and see what sort of junk output i am going to get?! Edited June 18, 20242 yr by wael_fathe
June 18, 20242 yr On 2024/06/16 at 7:49 PM, BritishRacingGreen said: Yes that is normal , because the initial bootstrap is only there for short while. In that period the SMPS should produce , and amongst other , will generate the +15VDC supply over C75 , which will power U10 via D49 even after the bootstrap via D50 falls away. But because your SMPS does not deliver , when the bootstrap via D49 falls away , everything collapses. The following test is provided to statically test the SMPS negative feedback loop. While no attempt is made to verify the linearity of the transfer function , it is adequate to test the extreme low point of feedback (resulting in highest output duty cycle) , as well as the high point of feedback (resulting in lowest possible output duty cycle) . To do so , we will manually power up the U10 and support circuitry with a power supply of 15V , and manually inject a variable 12V supply on the +12V output rail in order to fool the U10 that it is 'regulating'. 1. Remove all ancillary connections external from the main board , including controller board , fans , mppt , battery , everything. 2. You will require 2 x 12V variable bench supplies and an oscilloscope . Both of them must be current limited and both must be able to vary up to +15VDC. 3. Connect psu 1 positive to D57 cathode and negative to BAT- and set to +15VDC . Current limit to 100mA or so. This will provide a steady VCC on U10 pin 7 and power the chip . 4. Check there is steady +5V on U10 pin 8. 5. Check with oscilloscope that there is frequency and 50% (max) duty cycle on the gate of mosfet Q36 . 6. Connect psu 2 set to 11.7 V between D54 cathode and GND . Set current limit to 800mA. This will false feed the +12V bus. At 11.7V the U10 output duty cycle should still max out at 50%. Check that the voltage drop over U8 diode is well below 1.1V . PLEASE NOTE : by injecting this 12V supply remember that the 12V loads on the main board is still connected . So please make sure that you dont push this voltage higher than say 12.8V. 7. Increase the psu to 12V and measure the TL431 bias voltage over R209 . It should be very close to 2.5V . Increase the psu to 12.5V and make sure the bias voltage is above 2.5V. at this stage TL431 turns hard on and current flows thru the U8 diode. The voltage across this diode should saturate at approximately 1.1V. This will in turn turn the U8 transistor hard on on the SMPS primary side . In turn this will pull the COMP pin 1 of U10 well down to 0V . As a result of this , U10 must correspond by outputting its lowest duty cycle on its output to the mosfet gate. 8. If all of this passes , the feedback loop circuit is pretty much in order. ... I will continue with this in another post whereby you can bring the SMPS up with starting with a small 10V voltage at its dc input and working our way up to 40V if all goes well ... nice trick thanks never before know any dynamic test for feedback such as yours thanks again i do have 2 power supply will try to do it ...much easier is to fed the transofrmer with 120vac at high frequency..and see what the transofrmer gives at the outputs ..i higly suspect the transofrmer
June 18, 20242 yr On 2024/06/16 at 12:32 PM, BritishRacingGreen said: Really , why on earth would a clone 'design engineer' affect a fundamentally different SMPS design ?. The spec of TX9 , the analogue feedback loop , the current limit feedback to name a few, all needs to change. But most of all , the spec of 150V is just senseless. The SMPS has been originally designed to accommodate battery-dependent machines . In this case the raw battery is fed to the SMPS DC input rail. So this clone actually needs to have a completely separate 48V-150V DC-DC converter , that is not economical . In later battery-independent designs , a separate SPS module is required where both mains and PV (SCC)must supply the main SMPS input , in combination with the battery. Both mains and PV has dc-dc converters to generate 60VDC and this is connected to the SMPS input rail via steering diodes. In the standard Voltronic design the input range is not specified , but I can state that where an SPS is used , the voltage is 60VDC derived from mains , 60VDC derived from PV , and whatever the battery voltage is (48-56VDC). I have tested this SPS+ input from 42VDC - 70VDC . The schematic of the Voltronic SMPS refers as shown below: Note the DC input filter electrolytic capacitor C69 . It is rated at 100VDC , so it is not going to be 150VDC friendly . What is yours? The Voltronic part number of TX9 is 41-070194-02G . This is true for various machines I have worked on . Has your transformer a similar marking ? c69 is rated 200vdc green one ...i actually tested one good working clone inverter and it gave me 120vdc input so it seems 150vdc is not that far and i assume the system normally give 120 in some clone and 150 in others (left smps in picture) source for 150 to the (right smps in picture) whcih xt for main smps which should work and produce all good vdc voltage 5 12 -12 but that never happen there is one more strange thing about those clone inverters even if both smps are good they work for about 3 sconds produce 120vdc ...-12 +12 5 then they shutdown only when the control board inserted , at that only case the work fine ,and continue witout a shutdown by the way clone that give me 120 output is 48vdc the one that give me 150 is 24vdc Edited June 18, 20242 yr by wael_fathe
June 18, 20242 yr On 2024/06/07 at 8:46 AM, Mach4 said: Thanks for the info. I am familiar with the UC38xx family having repaired other products that use them. Still need to design a flyback with them at some time tho. The parts have arrived, its this weekend's task to get this inverter up and running again. Will feed back my progress as and when. The saga continues... after replacing identified dead components D40 U16 Q6 and the main PSU caps I power up on the bench and all primary supplies and HFPW+ are present and accounted for. BUS+ also rises to dizzy heights when U17 is enabled via CN11P14, so all good so far. The display tho says HS and error 81 but I dont have PAR pcb connected so I'm not worried besides the current limited 50V supply is not hitting 500mA limit, it goes up to 200mA approx when the power supplies are turned on via AC Start connector. So time to put it back in the box and test. All connected incl. PAR pcb and SCC etc, and try power up via AC Start button the bench PSU hits 500mA limit, I check everything its the same as on the bench now what? Disconnect PAR board but no change. I'm monitoring Vbus but its not starting up. So I up PSU current lim to 1000mA (thats 48W) and try power up via AC Start multiple times each time current limit is reached, I switch off and check everything again. Then on next try suddenly display comes on, no current limit but after few seconds display shows HS and ERROR 09! Then shuts down. What the hell that means bus undervoltage. I strip everything fearing the worst sure enough QB and QD IGBTs are short circuit E-C. This implies enough energy existed to blow these monster IGBTs short circuit! Possibly there was a shoot-thru across the bridge how could that happen under control of the DSP? The 4066 IGBTs have a rugged spec including guaranteed short circuit SOA for at least 10uS. How could this happen? And with a current limited bench PSU? In truth I dont know. I am now at the point of recommending to the client to scrap it and start again with a Sunsynk. Hopefully he agrees because the engineer in me wants this unit to further analysis.
June 18, 20242 yr Author 53 minutes ago, Mach4 said: but after few seconds display shows HS and ERROR 09! The dreaded ERROR 09 , it always spells nasty failure of the components in the power chain. 09 is Bus Soft Start fail , it means the controller has detected a short circuit on the High Voltage DC bus. SO expect a number of IGBTS blown , together possibly with driver circuits as well. 53 minutes ago, Mach4 said: This implies enough energy existed to blow these monster IGBTs short circuit! Yes , even if you are using a current limited bench supply , you must remember it initially charges the low and high DC bus capacitor banks to their full capacity. This storage has enough energy then adequate to destroy the IGBT's and MOSFETS spectacularly . I feel your pain , I have had similar experiences in the pass , so I have a couple of walkthroughs in this thread to bring up the dc-dc converter in a very graceful and soft manner. Also the DC-AC converter can be verified in this manner . Edited June 18, 20242 yr by BritishRacingGreen
June 18, 20242 yr Author On 2024/06/16 at 4:42 PM, Coulomb said: No, the King IIs are the high PV voltage models. The "standard" King (they actually use the word "standard" or abbreviation "STD" in firmware for what I call King 1s, to distinguish them from King IIs) is the one with the 145 V max PV solar charger. The SCC from the King I have is definitely low voltage , so the King must have been MKI . On 2024/06/16 at 4:42 PM, Coulomb said: My wild guess is that there would be about a 95% chance that you could just bolt an MKS charge controller into a King 1 or vice versa, and it would just work. I am going to try that out .Thank you for the all great info on the low voltage SCC. Apart from the links you gave regarding the protocol , I also noticed in your repair index that you have created partial schematics for this SCC . That is manna from heaven , so once again thanks. Inspired by your available resources on the LV SCC and also in lieu of the fact that he LV SCC's lifecycle in terms terms of support and spares is at its tail end , I have decided to start learning more about them , and start repairing them .. So I have decided to repair the module I have received as faulty . This is the large flat module with the burn marks on the PCB. When I disassembled the board from from the main board , I noticed that bad burn / heat marks on PCB on the BAT- terminal . This terminal screws , along with BAT + and other mounting posts was loose , and the previous repair guy forgot to torque them. But I soon realized that there an additional burn mark . That was a blown PCB track . The schematic below refers. The piece of track blown is from the D7 cathode to the C75 capacitor. It also had underlying burns of the board itself. Further to this I tested the semiconductors of the RED interleave driver circuits. All semiconductors blown short or open circuit , except the 3 mosfets . The IR2011 driver blown and all six of the buffer amplifier transistors blown short circuit! What is not shown above is the CPU drive interface for the PWM_1_H input. From the CPU IO pin there is a 100R resistor in series to feed one input of a 2 input AND gate IC. The output of this AND gate IC feeds the PWM_1_H input of the IR2011 via a 100R resistor . The AND gate output has also blown , it has a very low resistance to 0V. The AND gate inputs impedance matches that of the BLACK circuit , so hopefully the CPU drive IO pin has been spared damage. This is a risk I will only be able to verify once the SCC is integrated back into the host machine. i have powered up the SCC on the bench with 48V DC on the battery terminals as well as false feeding the SMPS. In the circuit below I have connected a BAT+ supply to the upper connections of R26/R21 . The CPU switches the supply off after a couple of minutes , but then it restarts again . That's ok by me. I have tested the 8VDC , and the 16VDC for the BLACK circuit . The 8VDC feeds the 78L05 which produces good 5.0 VDC. I then repaired the RED circuit power supply by replacing the diode and repaired the broken track. Also removed all the driver transistors as well as the IR2011, and also the 2 input AND gate IC . Switched on and the RED PSU delivers a good 16V , slightly higher than the RED one. I replaced the IR2011 with a good one and false fed the PWM_1_H input of the IR2011 with 5V supply via 4k7 resistor. I tested the output of the IR2011 via switching the input logic hi and low. Next I start repairing the buffer amplifier but only inserted the top NPN transistor. This allowed me to switch the gate on and off and measure drive voltages. A very quick and dirty way to check conductance of source-drain of the MOSFET is to measure pure resistance between source and drain. So my MOSFET seems to be working ok for a start, and so my driver. It is reasonably difficult to test the turn-off circuit of the 10R resistor and the PNP transistor . But eventually I soldered a 220uF capacitor across gate and source of the MOSFET . Without the PNP transistor , the cap discharges slowly , but with the PNP transistor it discharges virtually instant , which verifies the 10R via PNP transistor switch-on. So still to do the other 2 MOSFET circuits. And reinstate the 2 input AND gate. So far it looks promising . I think I have covered all bases . Only risk is the CPU PWM drive , whether its good or not . I cannot thank you enough for all the resources you provided us. Thanks. It quite a exciting for me as this is the first time I have been exposed to MPPT hardware. Always had been a black box to me. As to why the circuit failed I do not know . The RED circuit could have failed due to the very bad dry joint on BAT- terminal when PV yield was high. Conversely , the RED circuit could have failed initially , and the resultant short circuit imposed a massive load on the terminal dry joint. It appears that the SMPS has no current limiting sense circuit (I might be wrong. Edited June 18, 20242 yr by BritishRacingGreen
June 19, 20242 yr Author There are a couple of observations that I have made on the SCC modules and it raises some questions for the untrained eye. 1. The power stage is relatively small for a 3/4/5kW capable SCC in relation to that of say the LV DC side of the main board. The mosfets are less, the heatsinks are smaller. Is the PWM frequency of SCC less than that on main board? If so, while I am aware that switching losses is dependant on frequency, is this also true for conduction losses? (source-drain). 2. I see the use of 5 pin 2 input AND gates on the SCC pwm drives. Is this some kind of failsafe gate to negate drive pulses during times when the CPU outputs are indeterministic, eg at bootup etc.? If memory serves they are also employed on the DSP board of the host machine. 3. I take it the totempole style buffer amps after the IR2011S gate driver is required because of the fanout of 3 mosfets as opposed to a single one. 4.I notice debates regarding LV SCCs versus HV SCCs. I know @Coulomb prefers the former because, amongst other, of its exposure to AC live. But I reckon that because the HV SCC is integrated to host HV dc bus, that its possible failure mode is more forgiving to the host power chain than in the case of the LV SCC that is directly connected to battery bus. The buck nature of LV SCC operation can cause a failure of output far in access of the max battery voltage. This to me is particularly bad for lithium batteries. The predominant boost nature of HV SCC has less chance to affect an overvoltage on the HV dc bus. Voltronic Axpert pv limit is 450 volt and this is a sweet spot for safe operation. I have received 2 x Pylontech bms's last year that had been connect to Axperts with LV SCC. Both bms's are of newer models where Pylon introduced a new layer of protection, that of voltage controlled fuses (8 of them, one for every mosfet lane). The fuses had blown because of short periods of high voltage on the input terminals. I also received the Axpert King inverter in one of the cases, and the SCC had blown mosfets to short circuit. I replaced the SCC with new one. Edited June 19, 20242 yr by BritishRacingGreen
June 20, 20242 yr On 2024/06/18 at 9:22 PM, wael_fathe said: about your suggestion of feeding in 20khz i got a nice idea how about stealing some ac high frequency/from another inverter feed that in ... 200vac...f= 20khz and see what sort of junk output i am going to get? Your basic idea is ok, but it should be 12V, nothing like 200V. In most inverters, it is labelled as HFPW+ or HFPWR. It's used to power the comms and paralleling boards too. You should see something at the output that if half wave rectified (positive side only, as it's a flyback power supply) should give you about 20V DC. Don't be surprised if the HFPWR at the input has ugly spikes, especially negative ones. That's just the way those flyback converters work.
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.