Announcement

Collapse
No announcement yet.

Just caught my speaker on fire!

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • John L. Murphy
    replied
    Re: Just caught my speaker on fire!

    We owe it to those sharing our homes to conduct a basic power handling test before putting a DIY loudspeaker system into service. To do this operate the system at the loudest level it will ever be used for an extended period of time (at least 2 hours) and then evaluate the temperature rise of all the components. Any component that is too hot to touch should be upgraded to handle the dissipated heat without becoming "too hot to touch". Even the most carefully calculated predictions of power dissipation are no substitute for actual verification of temperature rise when it comes to the fire safety of our DIY loudspeakers.

    Consider this, a 10 Watt resistor is only rated for 10 Watts of dissipation as long as it is held at room temperature. At elevated temperatures it must be "derated" to a lower power rating. Resistors and electrolytic capacitors are definitely the weak points in the system when it comes to temperature rise so check them first.

    While it is an interesting exercise to predict the power dissipation of a resistor based on assumptions about energy distributions in program material and such, no professional engineer would ever accept those estimates as assurance that his loudspeaker would be safe when operated at its rated power for long periods of time.

    Play Safely!

    John

    Leave a comment:


  • evilskillit
    replied
    Re: Just caught my speaker on fire!

    Lookin sharp man. If you manage to catch those on fire then thats just bad voodoo. I'd check to make sure nobody has hexed you.

    Leave a comment:


  • davepellegrene
    replied
    Re: Just caught my speaker on fire!

    Finally got a chance to get these repaired. Started at least. I ended up using two 8 ohm 20 watt resisters paralleled to replace the 4 ohm 10 watt that got hot. I lifted them up off the board a half inch and spaced them a quarter inch apart. Here is a pic how I remounted the crossover boards for better circulation. I mounted three boards with grooves cut in them to slide the xo boards into. One on the bottom, one on top and one in the back. They hold the xo boards in vertically so I get air flow from bottom to top.These are also directly behind the upper woofer so it should provide plenty of air flow for cooling. This way I can get my rigid fiberglass around the xo boards on the cabinet walls and not cover them. I may move the xo board on the left over into the next groove to the right to get it away from the wall a little more.
    So far I like this set up. I think I am going to do my Rock It speakers this way and put extra grooves so I have more choices on placement.

    Dave

    Leave a comment:


  • philiparcario
    replied
    Re: Just caught my speaker on fire!

    yes they are and they are heavy enough for a lot more. 3000 watts no problem. why solder when you can use them!

    Leave a comment:


  • rtkelly
    replied
    Re: Just caught my speaker on fire!

    Mr. Arcario;

    Are those SQ.D busses you used? Are they heavy enough?;)

    Rollie

    Leave a comment:


  • philiparcario
    replied
    Re: Just caught my speaker on fire!

    found my 300 watt 1 ohm 2500uf sub filter pictures. the resistor alters qts a bit and the cap acts as a 15hz high pass. this unit never gets hot period. my dayton rss390 is a great IB sub with this filter addition. those resistors are great
    Attached Files

    Leave a comment:


  • markk
    replied
    Re: Just caught my speaker on fire!

    Originally posted by whatatrip View Post
    I am fairly new to all of this but have always wondered about what wattage is required for a given driver. There have been questions in the past but mainly with reguard to 2-way designs. I've been doing 3-ways and have had problems with resistors getting hot. Paper and wood ignite at roughtly 450 F so Dave's resistors must have been gotten close to 400 F. That just can't be safe. I would be cautious about that kind of heating even with a heat sink.

    What I have found is that the wattage of the resistor has to be close to double the RMS capacity of the driver. So far that has been the case. I have some DC130BS-8 drivers that are rated at 30 watts and I have 2-25 watt resistors that get warm when under hard load. I believe the mid range drivers in the minis are rated at 25 watts. It looks like the resistors that were used are 10 watt resistors. I hope Dave will do some testing and report his results.

    The enclosures were primed yesterday and today they get painted but when I get done I will do some more serious testing of the resistors.

    If I put a multitester in line with the driver, and set it for amps and take AC voltage readings, would I be able to estimate what the actual wattage is going to the driver?

    You can't really make assumptions based on the driver wattage. Those are somewhat dubious.

    I think a couple rules are in order.

    First, what is your system sensitivity and how loud are you going to play. If you have a 3 way that is 86dB/2.83v and is 4 ohm, and you are going to play at 95dB average (pretty loud, but not extreme), you're going to average 16 watts. Not much to worry about.

    If the speaker is only 82dB sensitive, and you want 100dB, now you need 128 watts into the speaker

    Again, the amount converted to acoustic energy is 0.5-2%, so basically 98%, or, essentially all of it, will be converted to heat.

    If a 2/3 of the music power (defined however you want to) falls into the midrange band, then you need to dissapate 85 watts into the mid circuit resistors and VC

    Say you have a 1 ohm series, and the 4 ohm shunt in parallel with a midrange unit. For math convenience, let's just say that the impedance of the driver is relatively flat at 4 ohms in the passband.

    So you have a 1 ohm series resistor with a 2 ohm (parallel circuit of 4 ohm R and 4 ohm VC). 2/3 of the 85 watts will go into the parallel circuit, and be split between the two. So 28 watts of power will be dissapated in each resistor and 28 watts in the driver. I sort of fudged the number to make it work this way. The drivers and resistors wouldn't last long.

    If the 3 way sensitivity goes up 6dB, to say, 88, the power requirements are cut to one quarter. So 7 watts dissapated in each. Even if you still want 100dB average levels.

    If the police come over and say, "turn it down," even if the sensitivity is the same, if you drop the level to 94dB average, again, only 7 watts per resistor and driver.

    If the series resistor values are different, say 3-5 ohms, the power dissapated gets cut in half.

    So you have to consider the spl level you're targeting as well as the exact topology and resistor values. You can see why pro units don't like passive xovers. The high spl levels required mean much beefier components. Something odd about a 1kW amp into a speaker with a 3 watt resistor...

    Also, high SPL designs in general would require less power to achieve a given spl target, and less power dissapated.

    Of course, an active design doesn't have any series or shunt resistors to overheat, and maximum use is made of available amplifier power, since no power is wasted. (Think about the above example where 85 watts is being dumped into the mid circuit. The mid driver only "sees" 28 watts. In an active setup, an amp of 28 watts would produce the same spl as passive circuit with an amp of 85 watts. 56 watts of the amp's power is just flushed down the toliet as heat...)

    Leave a comment:


  • exojam
    replied
    Re: Just caught my speaker on fire!

    When I had to replace the resistors in my speakers I picked up some of the Visahs to handle some more heat. I ended up taking the original resistors off of the main crossover board and put them on a small section of a car amplifier housing. Since the amp housing is basically a big heat sink, I cut small enough sections to fit the resistors on it with some CPU paste between the resistors and heat sink. I than mounted the piece vertically in the speaker with about a quarter inch spacer so none of the heat sink touched the internal bracing and it would have air flow all the way around.

    James

    Leave a comment:


  • Jim Holtz
    replied
    Re: Just caught my speaker on fire!

    There has been a lot of conversation about this and rightfully so. However, I think we've just witnessed the "perfect storm".

    3-ways have been around for 40+ years (?) and fires have never been a problem. The crossovers in commercial 3-ways most certainly don't use high-fi parts to say the least and I've never heard of a fire issue.

    Here's my thoughts for what they're worth:

    I seems to me that the combination of tight resistor attachment to the peg board, smothered in hot glue and then covered with foam retained enough heat from the extended super high volumes to cause things to start smoldering. The 1st thing I have to point out is, the Mini's weren't designed for that volume level. I'm glad they survived it. They're tougher than I thought. :D

    I think lessons learned here are; position the crossover board where it'll have some air movement near the port area. Solder the resistors so they can be positioned up and away from the board so air can flow around it. If super high volumes are anticipated, double the resistors to increase their ability to handle the load. Finally, take it easy on the volume knob. Your ears will thank you for it. :rolleyes:

    Dave: I'm really glad you caught the fire in time and I'm glad the Mini's survived the episode.

    Jim

    Leave a comment:


  • whatatrip
    replied
    Re: Just caught my speaker on fire!

    I am fairly new to all of this but have always wondered about what wattage is required for a given driver. There have been questions in the past but mainly with reguard to 2-way designs. I've been doing 3-ways and have had problems with resistors getting hot. Paper and wood ignite at roughtly 450 F so Dave's resistors must have been gotten close to 400 F. That just can't be safe. I would be cautious about that kind of heating even with a heat sink.

    What I have found is that the wattage of the resistor has to be close to double the RMS capacity of the driver. So far that has been the case. I have some DC130BS-8 drivers that are rated at 30 watts and I have 2-25 watt resistors that get warm when under hard load. I believe the mid range drivers in the minis are rated at 25 watts. It looks like the resistors that were used are 10 watt resistors. I hope Dave will do some testing and report his results.

    The enclosures were primed yesterday and today they get painted but when I get done I will do some more serious testing of the resistors.

    If I put a multitester in line with the driver, and set it for amps and take AC voltage readings, would I be able to estimate what the actual wattage is going to the driver?

    Leave a comment:


  • Wolf
    replied
    Re: Just caught my speaker on fire!

    Also be advised that it was a 4 ohm resistor. If you have a 25 ohm or 33 ohm resistor in a notch filter, it will actually not get as warm as the 4 ohm resistor would, and therefore would require less power handling. I will use 5W 25 ohm or 33 ohm resistors due to cost in some projects, and I've not had issues.

    Later,
    Wolf

    Leave a comment:


  • Brewski
    replied
    Re: Just caught my speaker on fire!

    Originally posted by davepellegrene View Post
    I have been thinking about making a board that is vented. Maybe like cutting grooves with the table saw so it looks like heat sinks then flip it on the other side turn it 90 degrees then groove it that way. That will leave square holes were the grooves over lap. Then I thought about hanging them 90 degrees from the back of the cabinet kind of how a wine glass slides into a wine rack, but with a support top and bottom with a simple groove to slide them into. This would give them plenty of air flow all the way around them and be easy to access and mount. Isn't that how the boards inside a computer tower are mounted?
    Yeah it would give you close to the same concept I'd probably look for away to have them run 90 degrees to top and bottom though rather than the 90 degrees to the case/side (EG. PCI board). You could have the notches be in two braces running across the sides and slide the crossover board(s) into the bracing. Or screw the crossover board into center running braces to get the board in the middle with more airflow around it.

    Take it easy
    Jay

    Leave a comment:


  • Pete Schumacher
    replied
    Re: Just caught my speaker on fire!

    Originally posted by davepellegrene View Post
    Isn't that how the boards inside a computer tower are mounted?

    Dave
    Sadly, computer towers rely on forced air cooling, and the boards are mounted horizontally, the worst possible way for natural convection to help. Your idea, with the board sticking out into the space, vertically aligned, is a quite valid approach.

    I don't think you need to go to that extreme though. Simply getting the resistors off the board a quarter inch and an unobstructed air flow should be all you need to do.

    Leave a comment:


  • davepellegrene
    replied
    Re: Just caught my speaker on fire!

    Originally posted by Brewski View Post
    Dave How close was the foam on top of the XO to the resistor? I'm wondering if the heat built up there and caused the foam to start to melt and possibly drip on to the resistor.

    Take it easy
    Jay
    Jay,
    If I had to guess what happened I would say the resistor was so well insulated from the foam and glue that it just got hot to the point that it melted or burned everything around it. I don't think the foam would catch the board on fire I think the resistor did that. I would say the resistor to the board is what did the damage. Even the cabinet behind the board is chard. This was my first build so I really had no idea they could get that hot. All the builds I have done since I put as much power to them as I could with the XO's out of the cabinet and nothing ever got hot. They were all smaller two ways and the woofer was the first to go. My daughters three way Nightmare's I may have to address but she only has a 35 watt Haffler. When I tested them they would almost take all my GX3 had but only on base light material. The woofer is crossed at 1800 Hz so I have it probably doing a little to much. I will probably rework them down the road when I have a little more knowledge and make OX's are right.
    All my other builds I have pulled the foam out of and replaced with rigid fiberglass which I think is much safer. The only other build I am worried a little about is my Continuums because the XO's are packed in fiberglass pretty tightly. I use them more for when I want to listen while sipping a glass of wine not rocken out with a case of beer anyway. :D So they should be fine.

    I've had the Mini's for about two years and were hooked up to my Yamaha amp that is rated at 100 watts per channel and it would start to clip first. I think the Mini's are just one of those speakers that you could feed them with a fire hose and they would take it. Just when you mix that with a noob you might burn down the house.
    I have been thinking about making a board that is vented. Maybe like cutting grooves with the table saw so it looks like heat sinks then flip it on the other side turn it 90 degrees then groove it that way. That will leave square holes were the grooves over lap. Then I thought about hanging them 90 degrees from the back of the cabinet kind of how a wine glass slides into a wine rack, but with a support top and bottom with a simple groove to slide them into. This would give them plenty of air flow all the way around them and be easy to access and mount. Isn't that how the boards inside a computer tower are mounted?

    Dave

    Leave a comment:


  • Pete Schumacher
    replied
    Re: Just caught my speaker on fire!

    Originally posted by markk View Post
    Exactly.

    I'm not saying anything bad about the ministatements in particular, just about the power handling issues with three ways in general.
    If care is taken with component selection, matching driver sensitivity mainly, then padding will be reduced to a minimum, and power will be mostly dropped in the VC, not in the passive XO.

    My Byzantium 3-way is crossed at 300 and 2000, 2nd order. I've got the Hafler 500 feeding a 2 Ohm resistor which is in series with the LP inductor and HP capacitor. There's a zobel, an inductor, and currently, a 12 Ohm resistor, across the mid terminals. I had Ozzie blasting for 20 minutes yesterday, at levels I personally would not listen to for long considering the spectral distribution (LOTS of mid energy). I've no doubt the peaks were nearing 150W, with RMS levels around 20W. It was LOUD! I touched both the shunt resistor and series resistor on the mid. The shunt was barely above room temperature. The series resistor was only slightly warm to the touch. Tweeter padding resistor was the same temperature as the board it was mounted to.

    I have no doubt that the Statement mid XO could be re-worked with a shunt inductor to help control the low end where resonance occurs, which would allow for a larger shunt resistor and consequently reduced power consumption. Of course, that raises cost due to the addition of a large value inductor. And since both watts, and resistors, are cheap, it makes sense to double up on the parts to keep temperatures down and accomplish the task at hand.

    As Dave is finding out, keeping the parts free of obstruction to air flow will likely be all that is needed to keep this from happening again in the future.

    Leave a comment:

Working...
X