ballast trouble in home built uv light box

[matt miller]

Friday night I built a uv light source with the plans here: Dead Link Removed . I bought the same ballasts that are listed in the parts list. I bought them locally at Grainger: http://www.grainger.com/Grainger/searchresults.jsp .

All of the lights don't come on every time. They've all come on a couple of times, then when I plug it in a few minutes later only a few will come on. The rest are glowing dimly on the ends, but do not fully come on. There are 8 bulbs/4 ballasts. Wrong ballasts??

My darkroom is fairly cold (basement, winter, Iowa). It's about 58-62 in my darkroom right now.

Any ideas

 

[Canuck]

Have you tried juggling or twisting each paired tube that doesn't come on? I gather its a pair of tubes at a time that doesn't light up 100%? I found that for me and fluorescent fixtures, this does the trick, due to lack of 100% clean contacts with the socket.

 

[matt miller]

I have jiggled & adjusted the tubes a couple of times. I have also checked the wires to the lampholders. I plugged it in a while ago & all of the lights came on. I let it run for about a minute and then unplugged it. I waited about a minute & plugged it back it. Just two lights on the end came on. I unplugged it & plugged it back in. None of the lights came on, but all were glowing dim on the ends. You're right about the lights in pairs. 2 per ballast.

 

[Canuck]

Hmm, just a few thoughts, just in case you haven't tried it yet, but I have a feeling you have . Please bear with my questions, just thinking out loud .

Do the ones that light up (or not light up for that matter), light up all the time on a consistent basis? If so, can you just swap them with the ones that do not light up. If they light up there, could mean the tubes are the culprits. If they don't kight up, could be the ballast being wrong for the app. May not have enough oomph to get them started.

When I suggested you try unplugging the tubes, did you try it while things are plugged into the power? Sometimes that does the trick and then for some magically reason, all it nice in the universe again ... for awhile ..

 

[photomc]

Matt, sure you have done this, but I had some problems when I did my first unit as well (only a 4 bulb unit) so let me ask the following - there are two red wires going to one socket, then a pair of blue wires going to the 2nd socket of the pair -right? Then you have two yellow wires going to the first socket - opposite end, then two jumpers that go the the 2nd socket - does this sound right? If not that may be the problem, with the diagram like it is it can be hard to find the correct placement..

Just looked at the ballast you used..if it was the #3X964 from Grainger and you are using the 24 inch BL or BLB bulbs I think that is the problem. The bulbs are F20T12's and the ballast is for F15T8's which are like 15 or 18 inch bulbs..so my guess is that the ballast does not have enough ump! (sorry don't know the correct term) to light the bulbs up and keep them there.

Good luck..and let us see something once you have it working.

 

[rogein]

Matt,

I don't see a grounded metal reflector in the url plans. I didn't use one in my first 'box' and had problems with the tubes coming on - once I added it the bulbs all came on without problems.

 

[Jeremy]

Matt, I had this problem when the Gra-Lab my box was plugged into was plugged into a powerstrip, but once I plugged the Gra-Lab into the wall it worked better.

Re-reading the above sentence it is kind of confusing so here is what I am trying to say:

Box--->Gra-Lab Timer--->Powerstrip--->Wall = bulbs not coming on

Box--->Gra-Lab Timer--->Wall = bulbs coming on no problem

 

[Canuck]

Hmm, I think a quick check of the grounds on the unit is in order ...

 

[Michael Mutmansky]

Matt,

Three possible problems (I think they were all mentioned aove).

1. Ballast - Lamp mismatch. Look on the ballast and see what the ballast is designed for. A T-8 ballast will drive a T-12 lanp most of the time, but it won't be an optimal situation. This could cause a problem starting in a cold environment. They also won't be 'driven' at the appropriate level, which will cause premature lamp failure, and possibly premature ballast failure.

2. Lack of grounding behind lamps. It is important to have a piece of metal behind the lamps that is tied into the system ground. This is an integral part of starting a lamp with many ballasts. I use a piece of aluminum flashing that was painted white behind the lamps on my home-built unit. I've never had any starting problems to speak of.

3. Carefully check the wiring diagrams. They will be on the ballast label. Not all ballasts are wired the same way, so read the diagram carefully and make sure that you actually have them wired properly.

I doubt the temperature is the primary problem, because most ballasts are designed for 50 degree starting. However, if you have another problem, like the lack of grounding, the temperature will make the problem worse.


---Michael

 

[matt miller]

Just looked at the ballast you used..if it was the #3X964 from Grainger and you are using the 24 inch BL or BLB bulbs I think that is the problem. The bulbs are F20T12's and the ballast is for F15T8's which are like 15 or 18 inch bulbs..so my guess is that the ballast does not have enough ump! (sorry don't know the correct term) to light the bulbs up and keep them there.

You're right, the Grainger site does call these ballasts out for F15T8. These are the ballasts I used, but on the ballast itself the first item listed is "(2)F20T12" so I though it would work for these bulbs. I'm using the 24" F20T12BL. Maybe I should take these back & get something different?

I don't see a grounded metal reflector in the url plans. I didn't use one in my first 'box' and had problems with the tubes coming on - once I added it the bulbs all came on without problems.

I don't have a metal reflector. The lampholders I used are Grainger 4YA99. I mounted them in galvanized angle on each end. I grounded this angle, as well as the ballasts.

Box--->Gra-Lab Timer--->Wall = bulbs coming on no problem

I had the unit plugged directly into the outlet, but will remember what you said when I hook it up to a timer. In fact, maybe I should be testing it with the timer.

2. Lack of grounding behind lamps. It is important to have a piece of metal behind the lamps that is tied into the system ground. This is an integral part of starting a lamp with many ballasts. I use a piece of aluminum flashing that was painted white behind the lamps on my home-built unit. I've never had any starting problems to speak of.

3. Carefully check the wiring diagrams. They will be on the ballast label. Not all ballasts are wired the same way, so read the diagram carefully and make sure that you actually have them wired properly.

I don't have metal flashing, but I do have the lampholders mounted to galvanized angle, and the angle is grounded at both ends.

All of the lights come on sometimes & then the next time only 2 will come on. I will recheck everything tomorrow. If I still have no luck, I'll replace the ballasts with a different type. Does anyone have a suggestion?

Thanks for all of the help, I really appreciate it.

 

[Michael Mutmansky]

Matt,

The ballast sounds fine from your description.

The grounding you have may not be sufficient. For one thing, the entire system needs to be at the same ground potential to work optimally. Additionally, the ground is used to help strike an arc in the lamps, and it needs to be continuous the entire length of the lamp. I recommend trying this...

Take a piece of bare copper wire, and run it parallel to the lamps behind each lamp, and connect it to the metal brackets on each end. You can staple it in place, but make sure it does not get crossed with any of the lead wires, and connect it electrically to the metal on both ends.

If that doesn't do it, you may want to try running it in a marmer room, and see if that solves the problem.

---Michael

 

[matt miller]

The grounding you have may not be sufficient. For one thing, the entire system needs to be at the same ground potential to work optimally. Additionally, the ground is used to help strike an arc in the lamps, and it needs to be continuous the entire length of the lamp. I recommend trying this...

Take a piece of bare copper wire, and run it parallel to the lamps behind each lamp, and connect it to the metal brackets on each end. You can staple it in place, but make sure it does not get crossed with any of the lead wires, and connect it electrically to the metal on both ends.

If that doesn't do it, you may want to try running it in a marmer room, and see if that solves the problem.

I'll try this.


I was confused about how to ground the ballasts & if I even need to. The wiring drawing on the ubildit page shows a grounding wire connected to a screw in the ballast flange (where I screw it to the board). I've grounded each ballast like this as well. Is this the correct way to ground the ballast? Or do I need to have a ground on them?

 

[photomc]

This has been very helpfull...as the electronic ballast for F20T12's are Very Hard to find. Also, the ground is most interesting..Michael are you saying that a seperate ground should be used for the bulbs/sockets in addition to the ground that you use for the ballast? Or just make sure that the ballast have proper grounding, and that should work OK?

 

[Michael Mutmansky]

Matt,

Look on the label for the ballasts. They probably show the housing going to ground.

If so, then you should use the grounding wire in the flange, but make sure there is a good electrical connection there, make sure there isn't paint on the flange that will interfere with the electrical continuity.

Good luck on the exposure unit. I hope you quickly get the problems worked out, and move on to the fun stuff.


---Michael

 

[matt miller]

I got it working. I ended up cutting a piece of sheet metal to line the top of the exposure chamber behind the bulbs. I tied the sheet metal into the ground and walla, the lights came on. IMO the Edwards ubildit plans are flawed by not including this. Now I guess I need to paint the metal white?

Thanks for the help everyone. I can't wait to use it now.

 

[andrewfrith]

Matt..i too built a UV box from the eepjon plans and have had an issue where some of the tubes dont turn on immediately (i have to jiggle a few of them around to get them to turn on)...if i do like you and put a piece of sheet metal behind the bulbs, how does it get wired into the ground? do u mean the ground for the ballasts or some connection to the bi-pole sockets?

thanks
-andrew

 

[matt miller]

I just wrapped the stripped end of a ground wire around a screw & screwed it into the metal, 1 at each end. The other ends of the 2 ground wires were tied into the power cord ground wire. Seems to work, but there may be a better way.

 

[andrewfrith]

i'll give that a try..thanks matt

-andrew

 

[Shinnya]

Again, again, and again

Hi everyone:

I am almost there to finish the UV box. There is one more thing about it that I want to make sure before I do it. It is about the ground.

I have read this thread, and there is a couple of things that are not quite clear to me. Here's the images of my ballasts with other wires.

1) If I were to put bare copper from one end to the other, I will do it in the direction of A. Am I right on this. Or can it be in the direction of B as well? Or does it matter? When I do this, I am connecting these screws, right? Also, is it a problem if the wire touches ballasts?

2) If I were to put a sheet of metal on the back of lamps, what exactly am I connecting to? It is those screws which hold the sockets? Or do I need to connect it to the screws holding ballasts?

Thank you again for your help as always.

Warmly,
tsuyoshi

 

[photomc]

*bump* can anyone answer this...

 

[Jon King]

The direction (A, or B) won't not matter. The wire should be grounded - the third, green wire in the U.S., not the 120V return line. The ballast exterior should be designed to be grounded, so connecting them with the wire won't be a problem. If the wire does not work, you may need a metal sheet behind the lights, as some people earlier in this thread were using.

 

[Murray@uptowngallery]

The metal surface behind the lamp requirement is critical for magnetic ballasts because the start isn't as aggressive as with electronic ballasts...at least the older vintage (self resonant) e-ballasts I worked with in the 'old days' didn't care. The voltage would rise until the lamp lit, even at insanely cold temps where the lamps weren't supoosed to light...lamp life probably suffered. I don't know the physics, not the same as flash tubes, but I believe a capacitive effect, maybe something to do with electric field before the lamp ignited.

My mother-in-law had an old house with ungrounded outlets & a handyman-installed ceiling mount fluor. fixture usually didn't light, just the ends glowed. It always freaked her out when I waved my hand mysteriously near the fixture and the lamp would light.

Murray

 

[Murray@uptowngallery]

Newer generation ballasts have a lot of care put into lamp failure detection and lamp life, so they may not tolerate non-spec installation as well.