T
treez
Guest
Hello,
The company I work for want to have on-board, mains powered, 3.8kw electric chargers in their electric cars. The spec is 230VAC input, and 300-410VDC output.
They only need 6 chargers for 6 cars. The number may rise to 100, but not likely any more than that. Electric cars, as you know, are not experiencing that much buyer interest globally.
Anyway, the company have bought in off-the-shelf 3kW chargers to do the job. (about £1000 each) These chargers are very small which is favourable, as the electric cars are quite small. However, these bought in chargers are unfortunately, occasionally tripping out for quite long periods, and then suddenly start working again, for unknown reasons. (Also, the bought in chargers are only 3kw instead of 3.8kw, but I guess our bosses were willing to put up with that because of the off-the-shelf convenience).
This “tripping out” makes the chargers unfavourable, and so we decided to open them up and see if we could see what was going wrong. We found that the off-the-shelf chargers are Half bridge LLC converters supplied by a dual PFC stage. The LLC converter, and both PFC stages are all controlled by a single microprocessor.
The design is very small, and this has been possible because they are very tightly controlling the output voltage of the PFC stage so as to provide an optimum input voltage for the LLC converter. That is, the microprocessor very tightly adjusts the output voltage of the PFC stage so as to allow the LLC converter to operate at its resonant frequency over all battery voltage levels. We saw the PFC output voltage varying right up to the 450V level of the PFC output capacitors, and right down to 330VDC, depending on the battery voltage level. (well, it actually only went up to 440VDC, not 450VDC).
We could not see what was causing the chargers to trip out, therefore, I was tasked to come up with a design so we could make the charger ourselves, -one that would not trip out repeatedly. I chose a dual PFC stage, followed by a Full Bridge LLC converter. I do not have great software capability, so I would not be able to quickly do a PFC control algorithm in software (at least not at all quickly). Therefore, I chose a FAN9672 dual PFC controller to control the PFC stages. The problem with this is that though the output voltage of the FAN9672 can be varied, it is not possible to vary it to the wide extent that our bought-in charger does. Therefore, the design that I did does not always operate the LLC converter at its resonant frequency….indeed, when near the extremes of battery voltage, I rely on frequency change of the LLC switching frequency in order to correctly set the output of the LLC controller. This is OK, but it does mean that our design is not going to be as efficient as the bought-in charger modules that we have. This in turn means that my design will not be as small as our bought-in charger modules.
…This in turn is making the bosses reluctant and hesitant about going ahead with the design. As a side effect, the general staff of this company, (ie people not directly associated with doing this charger project) are extremeley anxious that they will loose their jobs. They see the swift acquisition of a suitable battery charger to be crucial for the company survival, and thus the survival of their jobs. They are thus very keen to see me depart this company as soon as possible, “with or without a parachute”.
I find in many UK electronics companies, non-electronics staff are neurotic about the acquisition of what they think is a fully competent electronics engineer. They (quite understandably) see the lack of (what they think is) a fully competent engineer as a threat to their jobs. –You can sympathise. In one particular company, I openly and honestly stated that I was not a software engineer, and that it would take me about a month to write some control software for them. (the engineer who had just left them would have taken just a day to do it). I was also doing hardware work for them, in fact, that was actually my main job. The general staff of that company overheard that I could not write the software quickly and decided to set about getting rid of me themselves. They believed, like many non-engineers do, that a modern electronics engineer cannot possibly be a “real” electronics engineer if he/she isn’t a software wizard. Therefore, they repeatedly offered me cups of coffee. I gladly accepted all these coffees. After a few days, I began to notice a very mild burning sensation in the back of my throat whilst drinking these coffees. I passed it off as my imagination, and in fact wondered if I was developing some kind of allergy to coffee. (heaven forbid).
Then one day, the usual complementary cup of coffee was again delivered to my desk…I thanked the chap politely, and took a sip. It was only one sip, as the “coffee” immediately burned the back of my throat quite harshly….i could then feel the burning sensation travelling down my gullet, and then into the top of my stomach. This was on a Friday, and the throat pain lasted the day, and prevented me from sleeping for virtually the entire weekend.
Everybody at this company was always very pleasant, though I regularly over-heard often angry discussions that they wanted a “real” engineer who could write software quickly like their previous guy. The bosses would not listen to them, (or just couldn’t find a software engineer) and so I guess they thought their only option was to hammer me with “firey” cups of coffee.
At a different company, (fuel pumps place), I was asked to get a 600W buckboost current source PCB that they had designed running. They had forgotten to add open-load overvoltage protection to it, so whenever it was powered up on no load, or the load removed, it would just blow itself up. Also, despite a 40Amps input current , the PCB trace for this was a long winding 1mm wide track. Also, they had not “Kelvin connected” the source current sense resistor, and in fact the connecting PCB trace was connecting to a thin copper track some 2.5 inches away from the actual current sense resistor pad. Also, they had strangely placed a common mode choke between the source sense resistor and the buckboost controller’s current sense input. Also, they had general analog monitoring circuitry liberally intermingled within the power switching loops of the buckboost converter, and no attempt whatsoever to manage the high di/dt power switch and rectifier current loops. There were also multiple other circuit errors. I informed them that a new PCB would be needed, but it took the bosses 3 weeks to sanction this. During this three weeks, the general staffing of this company decided that I was just making it up that a new PCB was needed, in excuse for not being able to do my job, and so decided to drain the brake fluid out of my back brakes whilst my car was in the company car park. The following day, I could hear a cackle of laughter, and jokes about brakes not working. The strange thing was, I eventually re-layed the PCB myself, got it working, and the chap who almost certainly had arranged for my brakes to be “done” came up and said he hoped I would be staying there now.
Anyway, back to the Electric car company of the present….
Another point about the 3.8kw charger that I have designed for them is that it obviously will use components sourced from Farnell or Digikey. As you will appreciate, for making just six chargers it simply isn’t possible to involve a big component distributor and get volume price reductions. The bosses actually have stated that the total materials and production cost of the units that I am tasked to make must be no more than the bought-in chargers. The bought-in chargers , however, are mass produced items, obviously attracting large volume cost reductions. The demand for my six chargers to be able to compete on price with the bought-in modules is obviously totally absurd.
Another point about the charger that I have designed for them is that I have used off-the-shelf Epcos TDK ferrite cores and off-the-shelf formers to go with them. However, the bought-in charger saves space by not using formers for the LLC power ferrites……they either wind the coils directly to the ferrite cores, or they have some custom made ‘coil-holder’. Either way, it would cost me far more to get the ferrites wound without using formers. The bosses have prohibited any extra ferrite production costs, and yet are insisting that the chargers that I am designing must be no bigger than the bought-in charger modules. This just isn’t possible.
Are there any magicians out there who fancy a job?
The company I work for want to have on-board, mains powered, 3.8kw electric chargers in their electric cars. The spec is 230VAC input, and 300-410VDC output.
They only need 6 chargers for 6 cars. The number may rise to 100, but not likely any more than that. Electric cars, as you know, are not experiencing that much buyer interest globally.
Anyway, the company have bought in off-the-shelf 3kW chargers to do the job. (about £1000 each) These chargers are very small which is favourable, as the electric cars are quite small. However, these bought in chargers are unfortunately, occasionally tripping out for quite long periods, and then suddenly start working again, for unknown reasons. (Also, the bought in chargers are only 3kw instead of 3.8kw, but I guess our bosses were willing to put up with that because of the off-the-shelf convenience).
This “tripping out” makes the chargers unfavourable, and so we decided to open them up and see if we could see what was going wrong. We found that the off-the-shelf chargers are Half bridge LLC converters supplied by a dual PFC stage. The LLC converter, and both PFC stages are all controlled by a single microprocessor.
The design is very small, and this has been possible because they are very tightly controlling the output voltage of the PFC stage so as to provide an optimum input voltage for the LLC converter. That is, the microprocessor very tightly adjusts the output voltage of the PFC stage so as to allow the LLC converter to operate at its resonant frequency over all battery voltage levels. We saw the PFC output voltage varying right up to the 450V level of the PFC output capacitors, and right down to 330VDC, depending on the battery voltage level. (well, it actually only went up to 440VDC, not 450VDC).
We could not see what was causing the chargers to trip out, therefore, I was tasked to come up with a design so we could make the charger ourselves, -one that would not trip out repeatedly. I chose a dual PFC stage, followed by a Full Bridge LLC converter. I do not have great software capability, so I would not be able to quickly do a PFC control algorithm in software (at least not at all quickly). Therefore, I chose a FAN9672 dual PFC controller to control the PFC stages. The problem with this is that though the output voltage of the FAN9672 can be varied, it is not possible to vary it to the wide extent that our bought-in charger does. Therefore, the design that I did does not always operate the LLC converter at its resonant frequency….indeed, when near the extremes of battery voltage, I rely on frequency change of the LLC switching frequency in order to correctly set the output of the LLC controller. This is OK, but it does mean that our design is not going to be as efficient as the bought-in charger modules that we have. This in turn means that my design will not be as small as our bought-in charger modules.
…This in turn is making the bosses reluctant and hesitant about going ahead with the design. As a side effect, the general staff of this company, (ie people not directly associated with doing this charger project) are extremeley anxious that they will loose their jobs. They see the swift acquisition of a suitable battery charger to be crucial for the company survival, and thus the survival of their jobs. They are thus very keen to see me depart this company as soon as possible, “with or without a parachute”.
I find in many UK electronics companies, non-electronics staff are neurotic about the acquisition of what they think is a fully competent electronics engineer. They (quite understandably) see the lack of (what they think is) a fully competent engineer as a threat to their jobs. –You can sympathise. In one particular company, I openly and honestly stated that I was not a software engineer, and that it would take me about a month to write some control software for them. (the engineer who had just left them would have taken just a day to do it). I was also doing hardware work for them, in fact, that was actually my main job. The general staff of that company overheard that I could not write the software quickly and decided to set about getting rid of me themselves. They believed, like many non-engineers do, that a modern electronics engineer cannot possibly be a “real” electronics engineer if he/she isn’t a software wizard. Therefore, they repeatedly offered me cups of coffee. I gladly accepted all these coffees. After a few days, I began to notice a very mild burning sensation in the back of my throat whilst drinking these coffees. I passed it off as my imagination, and in fact wondered if I was developing some kind of allergy to coffee. (heaven forbid).
Then one day, the usual complementary cup of coffee was again delivered to my desk…I thanked the chap politely, and took a sip. It was only one sip, as the “coffee” immediately burned the back of my throat quite harshly….i could then feel the burning sensation travelling down my gullet, and then into the top of my stomach. This was on a Friday, and the throat pain lasted the day, and prevented me from sleeping for virtually the entire weekend.
Everybody at this company was always very pleasant, though I regularly over-heard often angry discussions that they wanted a “real” engineer who could write software quickly like their previous guy. The bosses would not listen to them, (or just couldn’t find a software engineer) and so I guess they thought their only option was to hammer me with “firey” cups of coffee.
At a different company, (fuel pumps place), I was asked to get a 600W buckboost current source PCB that they had designed running. They had forgotten to add open-load overvoltage protection to it, so whenever it was powered up on no load, or the load removed, it would just blow itself up. Also, despite a 40Amps input current , the PCB trace for this was a long winding 1mm wide track. Also, they had not “Kelvin connected” the source current sense resistor, and in fact the connecting PCB trace was connecting to a thin copper track some 2.5 inches away from the actual current sense resistor pad. Also, they had strangely placed a common mode choke between the source sense resistor and the buckboost controller’s current sense input. Also, they had general analog monitoring circuitry liberally intermingled within the power switching loops of the buckboost converter, and no attempt whatsoever to manage the high di/dt power switch and rectifier current loops. There were also multiple other circuit errors. I informed them that a new PCB would be needed, but it took the bosses 3 weeks to sanction this. During this three weeks, the general staffing of this company decided that I was just making it up that a new PCB was needed, in excuse for not being able to do my job, and so decided to drain the brake fluid out of my back brakes whilst my car was in the company car park. The following day, I could hear a cackle of laughter, and jokes about brakes not working. The strange thing was, I eventually re-layed the PCB myself, got it working, and the chap who almost certainly had arranged for my brakes to be “done” came up and said he hoped I would be staying there now.
Anyway, back to the Electric car company of the present….
Another point about the 3.8kw charger that I have designed for them is that it obviously will use components sourced from Farnell or Digikey. As you will appreciate, for making just six chargers it simply isn’t possible to involve a big component distributor and get volume price reductions. The bosses actually have stated that the total materials and production cost of the units that I am tasked to make must be no more than the bought-in chargers. The bought-in chargers , however, are mass produced items, obviously attracting large volume cost reductions. The demand for my six chargers to be able to compete on price with the bought-in modules is obviously totally absurd.
Another point about the charger that I have designed for them is that I have used off-the-shelf Epcos TDK ferrite cores and off-the-shelf formers to go with them. However, the bought-in charger saves space by not using formers for the LLC power ferrites……they either wind the coils directly to the ferrite cores, or they have some custom made ‘coil-holder’. Either way, it would cost me far more to get the ferrites wound without using formers. The bosses have prohibited any extra ferrite production costs, and yet are insisting that the chargers that I am designing must be no bigger than the bought-in charger modules. This just isn’t possible.
Are there any magicians out there who fancy a job?