Author Topic: Concerned about the reliability of Eltam EHID electronic ballasts for HID lamps  (Read 1803 times)
dor123
Member
*****
Offline

Gender: Male
View Posts
View Gallery

Other loves are computers, office equipment, A/Cs


WWW
Concerned about the reliability of Eltam EHID electronic ballasts for HID lamps « on: January 24, 2013, 04:44:04 PM » Author: dor123
As you have seen from this video , the EHID electronic ballast have a one unusual feature that isn't exist in other HID electronic ballasts: After the factory engineer switched the power on, the ballast ignited the lamp after 2.5 secs from the moment of the turning on, while it checked what the lamp it is and if it is 100W or 150W.
These 2.5 secs, where the ballast still didn't struck the lamp, are actually a ballast POST (Power On Self Tests)(Much like the POST of your average computer, tablet, smartphone, gaming consoles, computer hardware, modems, routers, electronic appliaces, etc...).
The presence of a POST in this ballast, gives me the concerns, that the EHID electronic ballast, actually have a CPU (Processor), or at least a similar component the BIOS of the computer motherboard, meaning that this isn't just an electronic ballast, this is actually a computerized ballast (There is a computer inside the ballast: More components to fail).
So this ballast is have much more components than usual, and it is actually a small computer that controls the lamp (It can even be remote controlled by a special controller).
This gives me the conserns that this ballast, have more chances to fail prematurely.
Logged

I"m don't speak English well, and rely on online translating to write in this site.
Please forgive me if my choice of my words looks like offensive, while that isn't my intention.

I only working with the international date format (dd.mm.yyyy).

I lives in Israel, which is a 220-240V, 50hz country.

marcopete87
Member
*****
Offline

Gender: Male
View Posts
View Gallery

Re: Concerned about the reliability of Eltam EHID electronic ballasts for HID lamps « Reply #1 on: January 24, 2013, 05:01:00 PM » Author: marcopete87
These 2.5 secs, where the ballast still didn't struck the lamp, are actually a ballast POST (Power On Self Tests)(Much like the POST of your average computer, tablet, smartphone, gaming consoles, computer hardware, modems, routers, electronic appliaces, etc...).
The presence of a POST in this ballast, gives me the concerns, that the EHID electronic ballast, actually have a CPU (Processor), or at least a similar component the BIOS of the computer motherboard, meaning that this isn't just an electronic ballast, this is actually a computerized ballast (There is a computer inside the ballast: More components to fail).
So this ballast is have much more components than usual, and it is actually a small computer that controls the lamp (It can even be remote controlled by a special controller).
This gives me the conserns that this ballast, have more chances to fail prematurely.

what about microcontrollers?
it can use an mcu, so more efficient lamp operation with same (or less) amount of components than electronic ballast (you may replace ballast ic with µC)
Logged
Medved
Member
*****
Offline

Gender: Male
View Posts
View Gallery

Re: Concerned about the reliability of Eltam EHID electronic ballasts for HID lamps « Reply #2 on: January 24, 2013, 05:34:53 PM » Author: Medved
@dor123:
It is the other way around:
The microcontroller (= simple complete computer on a single chip, so one, single IC) is there to make the hardware simple, as it allow to eliminate a lot of components from the board, because the programability make it so universal, it could within the one single IC with few passive components around handle many control and diagnostic functions, where with "classic" implementation each of them would require some separate IC and many components around.
What it doe after power up is settling the PFC (that take the longest time, I guess, because the PFC have to have slow response in order to maintain good THD), detect whether the ballast is correctly connected (e.g. if someone didn't swap the lamp and mains wires), if there is no short circuit on the output,...

The lamp wattage detection is usually done during warmup: It first use the maximum current of the smallest rated lamp and observe the voltage. If the voltage rise quickly, it is most likely the lowest wattage lamp, so it continue with corresponding setting. When the voltage rise too slow, it mean the lamp is bigger than correspond to the current, so the ballast switch to higher wattage setting.
This algorithm continue, till it see the correct voltage rampup rate (so arctube warmup).
If the ballast "overshoot" with this algorithm (so switch to higher wattage settings than the real lamp), the lamp overheats slightly, what cause the voltage to go too high. As the response, the ballast switch back to the lower wattage setting.

To implement this with HW circuits would mean about 20 individual IC's, I would not like to see the "reliability"...
And in real life of luxury butique display light, where complete rearrangements happen nearly each week and include shuffling with different lamp wattages in the same fixture, all done by display designers and not electricians, such feature become quite essential in preventing explosions of overloaded lamps...
Logged

No more selfballasted c***

dor123
Member
*****
Offline

Gender: Male
View Posts
View Gallery

Other loves are computers, office equipment, A/Cs


WWW
Re: Concerned about the reliability of Eltam EHID electronic ballasts for HID lamps « Reply #3 on: January 25, 2013, 02:28:58 AM » Author: dor123
Medved: Several components that exists in the "traditional" electronic ballasts, should remain in a microcontroller based ballast, such as the components that limits the current (Choke or capacitor), the inverter circuit (The one that account for converting the 230V, 50hz to HV DC to HF AC and than to square wave 80hz AC), the PFC components (Active in this case I think) and the component that supposed to make the HV pulses, to ignite the lamp.
The microcontroller, can only replace the additional components that more equivalent to the ones that exists in real programmed start fluorescent electronic ballasts (EOL protection, anti-cycling, detection of no lamp or a defect lamp, external arc, rectification, etc...)
If all of the "traditional EOL protection" components, are very prone to failure (As seen in programmed start fluorescent electronic ballasts, the microcontroller idea is good: Another things to be performed on the lamp, can be programmed into it, without short its life.
However, what about if it would stay in a very high temperatures (Such as in an IP65 ballast in a road lantern or a small floodlight in a hot summery night or day? (There are cooling fins, but I don't know how they can perform in a hot summery weather).
Logged

I"m don't speak English well, and rely on online translating to write in this site.
Please forgive me if my choice of my words looks like offensive, while that isn't my intention.

I only working with the international date format (dd.mm.yyyy).

I lives in Israel, which is a 220-240V, 50hz country.

Medved
Member
*****
Offline

Gender: Male
View Posts
View Gallery

Re: Concerned about the reliability of Eltam EHID electronic ballasts for HID lamps « Reply #4 on: January 25, 2013, 09:12:30 AM » Author: Medved
The "traditional" ballast components include control regulation loops, what need a frequency compensation components in order to be stable. These components are resistor and capacitor networks, frequently with rather high time constants (e.g. in the PFC). And it is the long time constant stuff, what is most prone to leakages of whatever origin (high temperature, humidity,...), so tend to make mystery faults (on the bench top everything work correctly, but in the field it cause problems, because the humidity effects can not be replicated on the bench).
When you use the microcontroller, these control functions, include the related frequency compensations could be implemented in the microcontroller, so you get rid of these pains. What remain is indeed the power path components (inductors, capacitors, transistors, diodes,...), but all the control stuff could be reduced to few voltage sensing dividers and current sensing shunts (they would have to be there anyway, even with the analog control).
With fluorescent ballasts all the functionality is already integrated within single chip, as the fluorescent ballast market is so huge and the ballasts so simple and similar, it become feasible to develop dedicated specialized control IC's integrating all the control stuff. So there the microcontroller would have no advantage, unless it is used as an interface for digital diming control (as the protocols are so complex and the processing have to be in high extend customized)
With HID's the lamp types differ way more and they need way different controls (depend even on the application, what is the most important feature and what is the second one), so it would lead to way more variants of very complex control IC's for way smaller market, so it won't be effective anymore. Because many general purpose microcontrollers have the computing power to handle all the fast response loops, while their programabiility allow very complex custom regulation algorithms tailored for given applications. So they usually need only gate predivers as external IC's (as the microcontroller technology can not handle the 600V shifts), but that is all.

The temperature is not any bigger problem than with any semiconductor device: The weakest point are the bondwires, what start to fail (in long term) at 200degC. But the microcontrollers do not handle any power, so their temperature is the same as the board, so does not exceed ~120degC.
And present devices are designed for 150degC, so still quite a good margin.

So there is only one technology aspect making the weakest point, way more prominent than the bondwires: The soldering reliability. And this is greatly improved, when you eliminate the number of interconnections and the board complexity. And that is exactly, what the microcontrollers are for.
Logged

No more selfballasted c***

Print 
© 2005-2024 Lighting-Gallery.net | SMF 2.0.19 | SMF © 2021, Simple Machines | Terms and Policies