Poor build quality of HF

Topics relating to lighting control gear, including magnetic & high frequency ballast and ancilliary devices such as starters, ignitors and switchgear.
User avatar
Jack_T
Posts: 273
Joined: Tue Jul 12, 2011 7:56 pm
Favourite Manufacturer: Thorn
Area of interest: Fluorescent lighting
Location: Guiseley
Contact:

Poor build quality of HF

Postby Jack_T » Fri Aug 31, 2012 12:48 pm

I've had a lot of ballasts fail recently due to just lazy manufacturing. I find it stupid that the ballasts are hardly even able to be rewired as the terminals tend to break! I don't know how common this is for them to be either dead on arrival or just poorly made though! What experiences have you guys had?

User avatar
LinearSLI-H
Posts: 1023
Joined: Wed Jun 29, 2011 10:20 am
Favourite Manufacturer: PHILIPS and VOSSLOH SCHWABE
Area of interest: HID and Associated control gear
Oldest item: Philips 160W MBT/U 1951
Location: Cheshire

Re: Poor build quality of HF

Postby LinearSLI-H » Sat Sep 08, 2012 8:15 pm

I have never used electronig gear myself, I do have some Philips matchbox ones and they seem to be ok, but we have a set of Crompton battens at work that are electronic, their a good 10 years old now and still running well, see here.
FEED THE MOOMIN!!

Ash
Posts: 137
Joined: Tue Jun 12, 2012 8:28 pm
Area of interest: Any electricity and computer tech

Re: Poor build quality of HF

Postby Ash » Mon Sep 10, 2012 3:35 pm

How about putting a resistor of too low wattage ? It works, sorta, and it also emits that hot electronics smell. Couldn;t put in a proper resistor that costs like 10 cent more tops

User avatar
SuperSix
Posts: 660
Joined: Wed Jan 05, 2011 7:01 pm
Favourite Manufacturer: Fitzgerald
Area of interest: Fluorescent and Mercury Discharge Lighting
Oldest item: BTH 200W Carbon Filament Lamp
Location: Cambridgeshire
Contact:

Re: Poor build quality of HF

Postby SuperSix » Mon Sep 10, 2012 6:02 pm

The new place at work is completely HF, with the exception of my 40 year old SRS Pop Pack and it's absolutely *****! Three of the brand new 28W T5 Dextra battens had failed before we even moved in, a further two have failed in the last week! One of the Thorn PL-L wall washers on the shop floor failed today! It's just bollocks!

User avatar
Danny
Posts: 829
Joined: Thu Apr 21, 2011 10:24 pm
Favourite Manufacturer: GEC
Area of interest: Street lights fluorescent discharge
Oldest item: 120v 60w carbon lamp 1905
Location: North east england

Re: Poor build quality of HF

Postby Danny » Mon Sep 10, 2012 8:59 pm

@ Eliot maybe fit switch start 36W Fittings in place :D

User avatar
Tom
ATL Admin
Posts: 881
Joined: Mon Jan 03, 2011 5:39 pm
Favourite Manufacturer: Pre 1993 Thorn
Area of interest: Fluorescent and Discharge Lighting
Oldest item: One of a few pip top lamps!
Location: Oxfordshire
Contact:

Re: Poor build quality of HF

Postby Tom » Mon Sep 10, 2012 10:28 pm

It's a shame there isn't a suitable magnetic ballast for 28w T5! I think there'd be a lot of love for T5 if they were ballast and starter switch. Upon switch-on, the starter closes and preheats the tube's cathode filaments. Upon opening, the magnetic field within the ballast collapses, causing a voltage pulse which strikes the tube. This is often repeated twice or more, causing the distinctive blinking at startup.">switchstart! Only success I've known is 14w on 13w SS and the 80w on Transtar 8' 85w units!
I LOVE LAMP!

User avatar
LinearSLI-H
Posts: 1023
Joined: Wed Jun 29, 2011 10:20 am
Favourite Manufacturer: PHILIPS and VOSSLOH SCHWABE
Area of interest: HID and Associated control gear
Oldest item: Philips 160W MBT/U 1951
Location: Cheshire

Re: Poor build quality of HF

Postby LinearSLI-H » Tue Sep 11, 2012 7:34 pm

They have a magnetic ballast for the 28watt 2D lamp? :roll: , I have one in my bathroom! :D
FEED THE MOOMIN!!

Ash
Posts: 137
Joined: Tue Jun 12, 2012 8:28 pm
Area of interest: Any electricity and computer tech

Re: Poor build quality of HF

Postby Ash » Wed Sep 12, 2012 7:28 pm

The 28 2D can work on 18W T8 choke. Our Eltam 18W T8/20W T12 chokes are actually rated for 18 and 24W PL-L, 26W PL-D and PL-T , and 28W 2D

Worth trying also 30/32W T8/T12/circline choke too

And how about the 54W T5 ? Perhaps it can be underpowered somewhat (say at ~40W) without damage ?

User avatar
XmasLightGuy
Posts: 81
Joined: Tue Mar 08, 2011 1:23 am
Area of interest: Linear Fluorescent
Location: Colorado
Contact:

Re: Poor build quality of HF

Postby XmasLightGuy » Sun Nov 11, 2012 10:44 pm

I got a F54T5 HO to fire up on a F40T12 HPF magnetic rapid-start ballast but...
It was slow to come on
It overdrives the ballast (total wattage with the T5 is 66w where total with a T12 is 50w)

Interesting test... might work better with a F32T8 HPF rapid-start (rare item thoug)

--------
On the original subject:
Yeah I've had some junk quality HF ballasts that didn't even last a full year...

Ash
Posts: 137
Joined: Tue Jun 12, 2012 8:28 pm
Area of interest: Any electricity and computer tech

Re: Poor build quality of HF

Postby Ash » Tue Nov 13, 2012 3:28 pm

You used 2 lamp ballast ?

User avatar
XmasLightGuy
Posts: 81
Joined: Tue Mar 08, 2011 1:23 am
Area of interest: Linear Fluorescent
Location: Colorado
Contact:

Re: Poor build quality of HF

Postby XmasLightGuy » Tue Nov 13, 2012 5:08 pm

No, it was the 1-lamp version.

I have also posted a pic in the gallery now too.

Ash
Posts: 137
Joined: Tue Jun 12, 2012 8:28 pm
Area of interest: Any electricity and computer tech

Re: Poor build quality of HF

Postby Ash » Wed Nov 14, 2012 3:55 pm

May be better with a 2 lamp ballast when you wire it for 1 lamp (red and blue wires if using US standard colors)

User avatar
funkylamp
Posts: 69
Joined: Thu Aug 09, 2012 10:58 pm
Favourite Manufacturer: Westinghouse Lamps
Area of interest: fluorescent-MV lighting
Location: U.S.A.

Re: Poor build quality of HF

Postby funkylamp » Fri Nov 23, 2012 3:39 am

I think the arc voltage and current is all wrong for T5 for F40T12 rapid start.
F32 two lamp would prolly do better shunted

User avatar
XmasLightGuy
Posts: 81
Joined: Tue Mar 08, 2011 1:23 am
Area of interest: Linear Fluorescent
Location: Colorado
Contact:

Re: Poor build quality of HF

Postby XmasLightGuy » Thu Nov 29, 2012 1:29 am

@Ash: Yep a 2-lamp ballast (wired in 1-lamp mode) should do better since i believe its a bit higher OCV (plus the ballast wouldn't be overdriven)..I have some extra 2x F40 ballasts laying around and do plan to try that sometime

@funkylamp: Yep I'm sure arc voltage / current is probably wrong...but it is just a 'test' after all. Interestingly though the ballast (with lamp attached, but before it lights) is around 10w, then 66w when lit.. so something less than 56w going to the lamp?
I could try a F32 ballast...but with the rapid-start ballasts work 1-lamp or 2-lamp(wired for 1-lamp) should give the same result once lit (in either case it will "see" a 4-foot T5 and adjust for that)
1-lamp F32 RS ballasts exist, but I've never come across one..I do have a few 2-lamp'ers

Ash
Posts: 137
Joined: Tue Jun 12, 2012 8:28 pm
Area of interest: Any electricity and computer tech

Re: Poor build quality of HF

Postby Ash » Thu Nov 29, 2012 4:55 pm

If yu can measure temps i think you can even have a guess whether it is ok to leave the lamp with that ballast. In HF ballast the interesting temps are the electrolitic capacitor (most sensitive component) and switches (their tem can tell about the stress on them)

User avatar
XmasLightGuy
Posts: 81
Joined: Tue Mar 08, 2011 1:23 am
Area of interest: Linear Fluorescent
Location: Colorado
Contact:

Re: Poor build quality of HF

Postby XmasLightGuy » Sun Dec 02, 2012 5:09 pm

I can't measure temps - other than feel the over-all temp (those old magnetic RS ballasts are basically a sealed unit). And it doesn't feel overly hot or anything. The RS ballasts here also have a thermal protector so if they overheat they simply shut down.

------
Interestingly HF ballasts don't seem to have that thermal protection. LOL here's what's supposed to be newer/better technology... :doh

Ash
Posts: 137
Joined: Tue Jun 12, 2012 8:28 pm
Area of interest: Any electricity and computer tech

Re: Poor build quality of HF

Postby Ash » Mon Dec 03, 2012 4:14 pm

Thats plain laziness not not include that

But there is the problem :

The component that is most sensitive to temperature is the electrolitic capacitor. Thel life of such capacitor is rated at its max allowed temp and is very short (as per the rating) : 1000-2000 hour at 105C. Now a rule applies, that states that every 10C the capacitor is cooler, it will have double the life. So lets take a capacitor rated to 2000h at 105C, it''l do 4000h at 95C, 8000h at 85C, 16000h at 75C. But wait, 75C is often about the realistic temp the ballast is working at!! so there we go with the life of the capacitor.... and the ballast lasts about 2 lamps. Take a 1000h rated capacitor and you have the lamps outlast the ballast - and not by premature failure of a component, but the expected EOL of the capacitor as can be estimated from its ratings.....

Now the most heat is generated not on the caacitor, but on the switching transistors and HF transformer. Those often actually work at temps near 70-80C, and they can go up without damage (but ith damage to the PCB - more on that later). Now if the space in the fixture around the ballast is cool, the ballast will dissipate the heat and the capacitor in it won;t be subjected to such temps. If te ballast is in a tight place (or additionally heated by the lamps etc) then the capacitor will be overheated and die

All that is as long as al components are good quality. If they are not then it will blow way faster

Now there is also damage to the PCB. Take for example the rectifier bridge on the input. It does not heat much so is often overlooked by the designers completely. In many ballasts it is just 4 SMD diodes behind the PCB. Now they do get hot when working. Not enough to exceed their own rating (and therefore they dont burn out), but they discolor and slowly disintegrate the material of the PCB (and perhaps compromise its electrical isolation....), can detach the tracks from it (well now the real fireworks can begin), and the capacitor is often right there on the other side of the board - not good for it either

Now think of this : The thermal design margins there are tight as they are, so perhaps manufacturers prefer the ballast to overheat a bit and keep going instead of having it switch off (at least without protection it will last for a while, with protection it will be cycling right away and they'd be flooded with RMA's.....). Cooler ballast = need better design and construction = more expensive and/or larger size

Also example from computers : CPU's have thermal protection in them. You can lift the heatsink off a working CPU (the newer ones, the thermal protection did not allways exist) and nothing will happen to it. Problem begins when the CPU is working at high temp but somewhat below its thermal shutdown temp. It is hot and it conducts heat to stuff around it - can blow up some capacitors on the board for example. In laptops they usually have even tighter thermal design limits, however they use better capacitrors - so the capacitors are usually not issue there, but the chipset/gpu (which are soldered t the board) can just break off... Green RoHS soldering even makes it easier as it is more fragile than lead containing soldering

User avatar
XmasLightGuy
Posts: 81
Joined: Tue Mar 08, 2011 1:23 am
Area of interest: Linear Fluorescent
Location: Colorado
Contact:

Re: Poor build quality of HF

Postby XmasLightGuy » Sat Dec 08, 2012 2:42 am

Ash wrote: All that is as long as al components are good quality. If they are not then it will blow way faster

This is the key phrase "good quality"...
I think these days all electronics ( -everything- not just ballasts) are made with one thing in mind: "lowest possible manufacturing cost" quality doesn't matter. LOL

Ash
Posts: 137
Joined: Tue Jun 12, 2012 8:28 pm
Area of interest: Any electricity and computer tech

Re: Poor build quality of HF

Postby Ash » Sat Dec 08, 2012 7:50 am

Something like this



Elna is one of the highest quality elecrolitic capacitors out there, if not the one. They also make a lot of optimized capacitors for specific applications. And they have some capacitors specially for ballasts :

http://www.elna.co.jp/en/capacitor/pdf/ ... 0_11_e.pdf

look at pages 98-103 in the book (pages 94-99 in the file). Actually look at the one on page 102 whih is the best. It is not 2K hours like most cheapies. It is rated to 10K hours at 105C. This means 20K 95C, 40K 85C, 80K 75C. Most electronic ballasts are rated around Tcase = 70C so it is to be expected that the capacitor inside is at like 75C. So it will last 80K hours. At 8 hours/day use it'll make 27 years - not bad



But have you seen much ballasts that lived so long ? Well i have one that was in use (permanently powered) from 92 to 2011 and is still ok. But its not too common

User avatar
MetalHalideHater
Posts: 162
Joined: Sat Jul 02, 2011 6:31 pm
Contact:

Re: Poor build quality of HF

Postby MetalHalideHater » Sat Dec 08, 2012 9:36 am

rubicon also make good capacitors. You know when something is wrong when there are 85C capacitors within the ballast.

Ash
Posts: 137
Joined: Tue Jun 12, 2012 8:28 pm
Area of interest: Any electricity and computer tech

Re: Poor build quality of HF

Postby Ash » Sat Dec 08, 2012 10:11 am

There are some good ones i know of : Elna, Sanyo, Rubycon, Nippon Chemicon, Nichicon

Want to give a try to Vishay and Illinois Capacitor - they look like high possibility to be good (look high quality on the outside) but i dont find any installed anywhere to get conclusions (well, the capacitors are made in USA, most electronics manufacturers won't ship caapcitors from USA to China to install in their boards....)

Ost are cheaper and widely used in mainboards, and appear to be good

Panasonic are considered good but i dont like them as much, seen them baloon up more than others in computer mainboards

And beware of fakes there are loads going around. Gotta look at them to tell (different shape "X" in the top, different fonts in the label etc), and only sometimes they have actual fake-looking names (where a letter is different). For example you said Rubicon, if they are really written with i and not y then they are fakes


Return to “Control Gear”

Who is online

Users browsing this forum: No registered users and 1 guest