I have one of these desktop lamps from Home Depot that has a transformer and a 20w incandescent bulb in it. I put a 4004 diode in series with bulb to calm down the light. The intensity is very nice now but the base which contains the transformer seems to be a bit warm. Is there a reason the transformer should run warm because I am pulling current from the transformer in only one direction? It's been running for 2 days down on my workbench and hasn't set the place on fire so I assume it's probably ok.
Quite possibly. And if it does, I would bet your modification would void any insurance you have to cover such an event.
You now have a DC load on a transformer and it is driving the core into saturation, which causes it to produce loots of extra heat as the primary winding stops acting like a high value inductor and just like a piece of wire across the part of the cycle remaining after saturation.
Other than the saturation part, the transformer has little problem with the DC load.
I wrote out a description of a solution to the problem made only with more diodes, but decided that, since it might involve burning down your house, I had better keep it to myself.
You can always keep the lamp outside so it doesn't do any damage if it catches fire. To bring the light to your workbench, use a thick bundle of fiber optics. Drill a big hole in the wall for the bundle. This also lets you smell the smoke from outside quicker. Win-win all around, I think.
Are you sure it did not get hot before you added the diode? I can't see how putting a diode in series would do anything but decrease the load on the transformer and hence run it cooler.
What you did was essentially make it a half wave rectifier and therefor you cut the power to the bulb in half.
How bout this, remove the diode and let it run for a day or so and then check and see if its warmer or not. (maybe use a thermometer for more exactness)
If your worried about a fire then put a smoke alarm near it. If your not around the lamp much then don't leave it on when your gone.
Get a lower wattage bulb with the same voltage rating and base.
--
Paul Hovnanian paul@hovnanian.com
-----------------------------------------------------------------------
Procrastinators: The leaders for tomorrow.
This is a somewhat arcane bit of transformer-ology, but the core can become magnetized in one direction and run closer to saturation. If it was close to saturation already, it may be saturating now.
What I'd do is this:
(1) Take out the diode and run the lamp for a while. This will demagnetize the transformer and also let you measure the temperature in normal operation.
(2) Put the diode back in and see if the transformer gets hotter. Bear in mind that it may take a while to magnetize and start saturating.
I would say that is putting things mildly, to the point of politely to extent of a little euphemistically. Usual practice is to push the core into saturation already as much as is allowable without specific temperature or temperature rise limits being exceeded at maximum rated input voltage, or maybe without samples known to burn up during testing at maximum rated input voltage plus whatever margin of safety the manufacturer deems appropriate. Keep in mind that the manufacturer may be a low bidder.
The magnetizing and saturation should mostly kick in within a few cycles. However, the core's magnetic properties probably get slightly worse with increasing temperature, so things may compound a bit after several minutes. Also, higher primary current increases primary winding temperature, and higher winding temperature increases I^2*R losses in the winding, leading to more primary winding temperature rise, and it could take a while for this positive feedback process to either run out into a stabilized elevated temperature or proceed to failure of primary winding insulation.
Keep in mind that if the transformer temperature does stabilize before the transformer blows, it can easily do so at a level that dangerously compromises the integrity of the winding insulation.
It's more like the bulb getting 60% rather than 50% of full power, given reduced resistance of the bulb's filament at lower filament temperature. (Along with light output close to 30% of full due to light production efficiency being less at the lower filament temperature.)
60% of normal average current broken into a 50% duty cycle has RMS current being about 84% of normal (I^2*R loss of load-related-current 60% of full, with other losses not decreased), and DC secondary current makes things a lot worse for the core which is probably already being pushed!
If I was going to put a diode in series with the bulb, I would plan on the transformer overheating badly and going blooey! As in only have this running while I was around to unplug the line cord and shovel the thing into a bucket of water or a toilet as soon as it blows! I would not want to wait for a smoke detector to tell me such an arrangement went blooey, since at that moment an air current could be taking smoke away from the detector until things progressed so much that the smoke detector would get hit with smoke no matter which way air currents were moving. By then, there could be a roomful of smoke and maybe an outright fire.
I think the idea someone posted of using a lower wattage bulb makes a lot more sense!
Possible problems with a higher voltage bulb: 12 volt 20 watt strikes me as likely being halogen. One problem is finding a version designed for a moderately higher voltage and with the same style base.
Another problem is that when a halogen bulb is underpowered, the halogen cycle does not always work. Theoretically, evaporation of the filament is supposed to slow more than the halogen cycle slows if the bulb is underpowered, but it does not always work that way. One problem that sometimes occurs from underpowering a halogen bulb is that the halogen cycle slows more than a "reverse halogen cycle" performed by contaminants, and the bulb could get blackened with filament material. This is more likely with el-cheapo halogens from manufacturers other than GE, Philips, Osram/Sylvania and Ushio.
One fix is to use a non-halogen 28V bulb, but then you will need a different socket, which may refuse to fit and may not be nearly as easy to get as a 10 watt 12V halogen bulb.
yeah thats true but lets get one thing right about the transformer....
putting a diode in series with the secondary does NOT push the core towards saturation.. The PRIMARY still has AC on it and the magnitizing current that flows in the primary keeps the core centered. Whatever current that flows in the secondary is canceled by an additional flow in the primary. The magnitization swing of the core is determined by the primary voltage and not by the secondary or primary current.
Putting a diode in series with the primary of course is a big problem.
Putting a diode in series with the SECONDARY is NOT a problem.
I think you are being extraordinarly charitable to the engineering and safety involved. This is Home Depot: I'm 100% sure that every cost- cutting measure has been taken at every step in the supply chain, and that every component was made by a low bidder.
If the primary winding has zero resistance, then the voltage might remain pure AC, but this is not the case, especially for small transformers.
Loading the secondary with DC also loads the primary with DC, and that produces a voltage drop on half of the half cycles that is not there during the other half. That current drops voltage across the primary resistance, and that drop is subtracted from from the line voltage for those half cycles. This effect definitely unbalances the AC applied to the winding, and walks the core toward saturation.
Now, if you arrange another voltage drop for the unused half cycles that averages the same as the resistive drop during the loaded half cycles, and put that drop in series with the primary, you can bring the net voltage applied back to much closer to zero, and the transformer will be happy, once again. At least until the lamp burns out.
Hmmm OK that's a good point. Let me think about it.
It's interesting therefore to note that the core is actually stressed more (and may approch saturation) during the 1/2 cycle that the diode is NOT conducting and the primary voltage is therefore higher.
So I'm thinking that if the core starts to saturate when the diode is off, the primary current will increase during the "off" 1/2 cycle, until the primary voltage drops to reaches some equalibrium.....it's hard to predict which combination of conditions ends up with more heat in the transformer.....
Very interesting question... maybe time for someone to make some measurments...
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.