Burnt a costly drivers coil with a pricey amp../Followup and prevention of costly mistakes..

[jay michael]

Good timing on this discussion, this is has been my focus over the last couple days.  Starting to feel better about my Danley limiter settings with the help from the guys on the Danley FB page.... but now taking a harder look at my 21sw152's in the skrams.  Taking cues from some points found in the bassboss post, the danley white papers and some comments made here, I have come up with the following numbers that I am considering to use with my 21sw152 4 ohm Skrams.  I am running 2 drivers per channel at 2 ohms loads on a powersoft k10.  I plan to use the peak limiting function on the k10 and set up my venu360 to operate in rms limiting mode.

Using the recommendations listed by Paul for techno trance he's saying use 1/3 to 1/2 aes power (2000 watts) Using 1/3 to be conservative would give me about 666 watts and 51 volts.  The Danley article suggests at least 100ms but up to 2-3 seconds for attack, which falls in line similar to what Paul is saying plus adding 2-4 seconds release.  

What do you guys think, seem like a reasonable place to start?

 

[jay michael]

Also, to echo what sme has been saying, I have been chatting with a guy on fb who uses a few skhorns in high stress club environments.  He mentions that he gets noticeable heat build up in his cabinets when he is working them real hard, to a point where he cycles them out for periods of time to give them breaks.  Perhaps what sme mentioned in regards to this style of cabinet might have some merit, a little extra safety buffer may not be a bad idea?  

[SME]

You have a K10 which has a "true power" limiter, which I would definitely use.  I would configure it to use a fairly long attack time (10 seconds?) and an even longer release time (maybe a minute if possible?).  I would also choose a quite conservative power value, like 350 W per driver.  The idea is that you shouldn't normally hit this limiter.  It's there to protect the equipment against worst case scenarios of long continuous tones at full blast or a DJ pushing the sound into severe clipping.  Couldn't these drivers take more like 700W a piece?  Maybe.  But if one is frequently pushing long-term true power beyond 350W, adding more cabs might be the better solution.  The drivers are certainly likely to last longer if they aren't being driven so hard as to heat the cabs up.

 

[radulescu_paul_mircea]

He's K10 doesn't have DSP I suppose, otherwise why wouldn't he use the internal DSP?

In a SKhorn or Skram subwoofer, with 8 or 4 ohm drivers , one wouldn't have to use a peak limiter with any amp except K20. 

Engage clip limiter in K/X series or soft clipping on any other amp and then use a really long term limiter. If the limiter is not a gain reduction system, but a normal limiter, take care with the voltage values , because if there is more distortion than it should, it will increase the average power and decrease the crest factor. 

A 21SW152 coil without moving can get burned even with 200 watts ( meaning continous current or sines at frequencies where excursion is really really low.

In use, you can use 500 watts as an excellent protection strategy if you don't stay in limit permanently. 

[radulescu_paul_mircea]

11 hours ago, jay michael said:

He mentions that he gets noticeable heat build up in his cabinets when he is working them real hard, to a point where he cycles them out for periods of time to give them breaks.  

This cabinet is tricky. The sound is absolutely stunning, but the thermal behavior is not best. It is best used in a low bandwidth and techno is to be avoided.

Even though a dual 18TBX100 sub is from -6 to-10 dB in maximum output compared to my dual 221 Ipal Skhorn XL subs, half the dimensions, on techno the are as loud when limiters intervenes. They do not sound nowhere near as good, but they are quite close in output. This is because the impredance minimum from 45 Hz to 85 Hz is dead centered in the band where that kind of music is centered 

[Ricci]

Excellent posts all. 

Thanks for posting some of your limiter findings Paul. 

To reiterate for anyone reading along...The very short version is that the modern long excursion pro woofers in cabs tuned above about 25Hz really shouldn't need to worry too much about short term peak voltages and setting these may not help protect the drivers in most cases. Over excursion isn't nearly as likely as it used to be with older short stroke drivers and that is mainly what this protects from. The enemy of modern drivers is heat and long term buildup in the voice coil. To that end you need to employ a limiter that protects from too much average power over a longer time period and this will need to be set MUCH more conservatively than the driver power ratings. If the limiter is constantly engaged you need "more rig for the gig" or to turn down the volume some. 

[Ricci]

On 8/8/2019 at 9:02 AM, radulescu_paul_mircea said:

This cabinet is tricky. The sound is absolutely stunning, but the thermal behavior is not best. It is best used in a low bandwidth and techno is to be avoided.

Even though a dual 18TBX100 sub is from -6 to-10 dB in maximum output compared to my dual 221 Ipal Skhorn XL subs, half the dimensions, on techno the are as loud when limiters intervenes. They do not sound nowhere near as good, but they are quite close in output. This is because the impredance minimum from 45 Hz to 85 Hz is dead centered in the band where that kind of music is centered 

 It's not just low impedance. It's low impedance in an area of minimum excursion. The bad news is this is an issue with almost any type of cabinet that can be built. 

Sealed:  the impedance remains relatively high and there is no minimum excursion frequency at all. Unfortunately the driver is trapped in a chamber which can heat up quicker, the driver uses a lot of excursion to produce output increasing distortion, output is lower and so is overall efficiency and sensitivity. Also the drivers are exposed to the elements and direct radiating which carries some sound signature with it. Smaller than vented.

BP4: This type of system has a sealed chamber which controls driver excursion below tune but it does carry a minimum excursion point in the middle of its bandwidth which also corresponds closely to an impedance minimum. Driver still enclosed in a sealed chamber as well. Can have very high sensitivity but only over a very limited bandwidth. Most of the 1 note BP mythology comes from this type of box due to limited bandwidth, vent noise and compression, bad group delay, etc. Driver can be better protected. Can be made about the same size as vented. 

Vented: Vented offers better overall sensitivity and efficiency than sealed over a large bandwidth. Similar to sealed excursion profile through most of the upper bandwidth which means lots of excursion needed for high output. Near vent tune the driver excursion drops to almost nothing and the impedance is at a minimum. The only thing saving the drivers from cooking in most cases is that the tuning is low enough that the content at tune is rare. Tunings at >40Hz could be problematic. Drawbacks are limited upper range output, lowish sensitivity, direct radiating sound, distortion, drivers exposed, vent noises and compression. Good size vs extension.

Passive Radiator: See vented. Add extra cost, completely sealed cabinet, more cones and moving parts to potentially fail or get damaged. No vent noise though. 

TH: Tapped Horns have a series of minimum excursion frequencies with usually at least 2 in the used bandwidth. Both of them will correspond to low impedance areas as well. A positive with TH's is that the driver is not enclosed and there should be some air exchange from the enclosure and past the driver. A negative is that the driver motor is usually partially exposed which gives a "half" direct radiating sound and gives greater chance of environmental damage. Also generally these have a limited bandwidth of about 2 octaves if the design is good. Slightly bigger than vented for the same corner frequency.

BP6: These are similar to TH's. These have at least 2 minimum excursion points in the bandwidth. Both of them will correspond to low impedance areas. One is at the lower port tuning. The other will be mid band corresponding to the upper vent tuning or path length depending on which type of design it is. These can have a reasonable bandwidth and there is some air exchange in the enclosure since the driver is not enclosed on either phase. No direct radiating sounds from driver operation. Drivers are somewhat better protected than with direct radiators. Slightly bigger than vented for the same corner frequency but generally can be made smaller than TH's. Drawbacks of vent compression / noise. Possible group delay or response issues if not designed well. 

BP8: These have at least 3 minimum excursion points. As with every other cab design they correspond to areas of low impedance and 2 if not 3 of them will fall in the primary bandwidth. See BP6 notes as the rest is quite similar. Can have a very wide effective bandwidth. Can be made to sound very clean. Very difficult to design and very sensitive to small changes in driver parameters or port/slot/horn loading or configuration. Issues with shifting of vent or radiator loading at war volume. Much bigger than vented and usually bigger than a TH for the same corner frequency. 

FLH: These have a series of minimum excursion frequencies that correspond to impedance minimums. Usually at least 2 will fall in the primary bandwidth. Additionally the driver motor  is usually enclosed in a very small sealed chamber. Before someone attempts to correct me on this I am not talking about full size bass horns, which for all practical purposes do not exist in pro audio, home, car or really any other audio systems, except in a handful of fixed installs. Yes if you get them near full size the impedance and excursion profile changes drastically but no one uses subs the size of a room for 1 driver.  Almost all of these that people actually use are greatly undersized. In that case they will have pronounce excursion minimums and impedance minimums corresponding closely with them. Anyway...These can have stellar response, efficiency, sensitivity and distortion with no direct radiating sound and no vent noise. The issues are that the drivers can be cooked easily on the undersized variants, these keep getting better the larger you make them but they need to be downright HUGE at which point they can be bettered by stacks of smaller options. 

Compound horns: See BP6...regardless of what exact mutation or spin -off there's at-least 2 excursion minimums and the corresponding impedance minimums, See a trend here? 

Scoops: These have the primary notch at the tuning frequency from the "scoop" and usually a smaller less severe one in the middle of the bandwidth.  As usual the minimum impedances are very near these minimum excursion points. These are sort of between a TH and a vented cab. These can have very high sensitivity and efficiency but only over a relatively narrow bandwidth and the response usually turns quite nasty above that point. Group delay and resonances can be a big problem. Drivers are exposed and these start to have direct radiator sound when pushed. These are generally not known for high SQ due to the response shape and other behaviors. 

I could keep going with TL's etc.... There are a ton of slight variations on the major design types, each trying to have their own special name or reinvention of the wheel. 

 

The take away of the above is this...Low impedance = more power applied. Low excursion = no cooling going on. Sealed rear chamber = trapped air/heat. So we want high excursion for cooling and low power? From a driver survival standpoint yes. But drivers run out of linear excursion quickly and it greatly increases distortion? That's also true. The 2 are at odds with each other. 

ALL of the cabinets that are not sealed/ free-air / IB are going to have at least 1 "Kryptonite" frequency band where low impedance (BIG POWER!) and low excursion (NO COOLING!) happen simultaneously. 

Bass reflex and BP4 will only have 1 Kryptonite frequency band. Bass reflex will be at tune. 

Every other design you can think of: BP6, BP8, Compound Horn, FLH, Scoops, TH, etc... Will have at least 2 Kryptonite frequency bands in their primary bandwidth. Some will have 3. 

If we look at it as only a function of handling power and self cooling the alignment that will take the most real average power is free-air, followed by sealed most likely. The next tier down is probably vented. Everything else would be a step down from there due to the multiple "Kryptonite" bandwidths. In the real world we have to consider the output vs the power input though. Some of the other alignments are capable of producing equivalent output with far less real power applied. It then becomes a complicated issue of how much power will each cab really handle vs the output that amount of input provides. A large sealed cab might be able to handle 500w broadband long term but a TH with the same driver may only handle 250w due to much less excursion and cooling being developed. However if the TH is 6dB louder on the 250w...

 

[peniku8]

This was my idea with the F6, which can be put into a 0-latency chain via a WSG compatible console.

I just found that the maximum attack time of 500ms might be too fast in some (most?) cases, but this example shows how you could protect cabs with an excursion minimum around 32Hz from overheating due to sustained low crest factor signals (sine waves in this case), without affecting any content but that specific frequency.

 

[Samps]

Is the Pi cooling plug effective?   Haven’t seen it used anywhere else which leads me to believe it’s not very effective.  

[Ricci]

On 8/10/2019 at 5:54 PM, Samps said:

Is the Pi cooling plug effective?   Haven’t seen it used anywhere else which leads me to believe it’s not very effective.  

It may help with long term heat soak effects but it won't help with rapid voice coil heating at tuning or other WCS frequencies. Something like that takes quite a bit of time to remove heat. Voice coils may be extremely close to the pole and top plate steel in the motor but there is still a small air gap. Even very large coils are still relatively small so they can get extremely hot very quickly. 

[menace]

Awesome thread! Ill be back for a reread(and no doubt some questions) soon when I get some box's together.

I will say that a 5hz sine wave through a bridged lab 6400 into 21sw152-4 created a smell very quickly! VC still measures fine but it was a good illustration to self how fragile these beasts can be!

[Ricci]

5 hours ago, menace said:

Awesome thread! Ill be back for a reread(and no doubt some questions) soon when I get some box's together.

I will say that a 5hz sine wave through a bridged lab 6400 into 21sw152-4 created a smell very quickly! VC still measures fine but it was a good illustration to self how fragile these beasts can be!

Single digit sine waves are some of the worst. Impedance is very low so current input is high.  Technically there is a lot of cone excursion but it is a very slow oscillation that produces low velocity air movement across the coil and through the motor structure. Be careful with that stuff!

[kipman725]

I seem to have burnt a load of drivers using the same amp this weekend.  4*15" per channel (only 2.5" VC though) @ 8 ohm total load per channel, sealed boxes.  Hitting the built in limiter a lot and for sustained periods of time.  I just didn't have enough output to match my tops and the required SPL after loosing a sub due to transit damage.  Noticed that one of the drivers seemed to be stuck in place at the end of the night... (totally melted) I presume there are more.

[Ricci]

On 8/20/2019 at 12:34 PM, kipman725 said:

I seem to have burnt a load of drivers using the same amp this weekend.  4*15" per channel (only 2.5" VC though) @ 8 ohm total load per channel, sealed boxes.  Hitting the built in limiter a lot and for sustained periods of time.  I just didn't have enough output to match my tops and the required SPL after loosing a sub due to transit damage.  Noticed that one of the drivers seemed to be stuck in place at the end of the night... (totally melted) I presume there are more.

Man...You are having a time of it lately!

It sounds like you don't have enough cabs to provide the level and duration you are looking for. 

What drivers were the 15's? Amp is Behringer NU3000? I assume these were midbass modules?

[jay michael]

Really appreciating this thread, thanks for everyone's input. It seems every step of the way in this game there are so many new layers of learning that are required.    

[LSC]

Yeah so much to learn especially when it comes to limiting and how manufacturer specs are calculated.. I reckon they should also state true RMS figures besides the aes and program power ratings.. 

Found myself quite limited in regards to limiters as I'm using a DCX2496.. might need a proper upgrade to be able to limit in multiple ways

I'm also very pleased by all the input provided and will try to implement the limiting settings jay michael tried last weekend

Also very intriguing idea of an impedance measuring-limiter device, maybe that'd help lots of people.. 

[SME]

True RMS would be a very useful number to have from manufacturers, albeit still highly dependent on mounting and ventilation conditions.  I wonder what happens if you take the driver with the highest rated power on the market (assuming it's not too exaggerated) and install it into a barely-big-enough box stuff with insulation.  Then test it by passing DC current (so it doesn't move) for 2 hours on each power increment.  How much power is that thing gonna take?  Maybe we should pool together to buy one and take bets. 

[kipman725]

My sub situation is pretty marginal as its only a hobby for me.  Normally I'm running 12*15"low cost  drivers in undersized sealed boxes of an NU6000 which I equalise flat.  This has just about been enough for indoors.  However with only 8 drivers at an outdoor event I just didn't have enough sub; however to good news is that on closer examination to have only lost one driver due to thermal failure during the event and the one that failed before hand in transit so should be able to get up and running again soon.  Longer term I'm looking into changing my boxes into eight isobaric bass reflex (weak motors) which will increase output about +10dB according to hornresp and should be fairly compatible with other better drivers as an upgrade path.

[Ricci]

8 hours ago, SME said:

True RMS would be a very useful number to have from manufacturers, albeit still highly dependent on mounting and ventilation conditions.  I wonder what happens if you take the driver with the highest rated power on the market (assuming it's not too exaggerated) and install it into a barely-big-enough box stuff with insulation.  Then test it by passing DC current (so it doesn't move) for 2 hours on each power increment.  How much power is that thing gonna take?  Maybe we should pool together to buy one and take bets. 

That would be interesting to see. I'd chip in if someone wanted to do it.

Too bad I don't still have the old AETechron 7560's / Crown M600's. DC coupled amps for running MRI and other medical equipment that would legit dump around 1000w into very low impedance loads and they could be ganged up in pairs. perfect amps for that. 

Wouldn't have to be the highest power handling sub on the market. Something reasonably cheap and beefy. Bonus points if it is easily factory repaired or re-coned. An Orion HCCA is relatively cheap, has a big 4" coil and recone kits are cheap and easy to find. I think they say it's a 2500w driver. If just running DC current there really isn't a need to have the whole driver even. Just run the coil of a re-cone assembly and monitor the temps and at what point it starts to burn. It won't be surrounded by the motor and frame structure but it also won't be moving through the air like it normally would. 

Start at 50w and increase by 50w at a time and let the temp stabilize in between. It won't last long. 

[SME]

5 hours ago, Ricci said:

That would be interesting to see. I'd chip in if someone wanted to do it.

Too bad I don't still have the old AETechron 7560's / Crown M600's. DC coupled amps for running MRI and other medical equipment that would legit dump around 1000w into very low impedance loads and they could be ganged up in pairs. perfect amps for that. 

Wouldn't have to be the highest power handling sub on the market. Something reasonably cheap and beefy. Bonus points if it is easily factory repaired or re-coned. An Orion HCCA is relatively cheap, has a big 4" coil and recone kits are cheap and easy to find. I think they say it's a 2500w driver. If just running DC current there really isn't a need to have the whole driver even. Just run the coil of a re-cone assembly and monitor the temps and at what point it starts to burn. It won't be surrounded by the motor and frame structure but it also won't be moving through the air like it normally would. 

Start at 50w and increase by 50w at a time and let the temp stabilize in between. It won't last long. 

I meant it as more of a thought experiment, but if someone with the right amp wants to do it...

Actually rather than doing 50W increments (which may be too big if the driver cooks at, say, 150W), a better idea might be to monitor the resistance of the coil in order to estimate likely thermal power compression at different true RMS levels and to inch closer to the actual limit.

[Funk Audio]

That actually becomes a calculable thing at that point. You can call the coil a flat heat sink, there are online calculators to tell you how many watts will reach what temp in ambient air. I put in the area for a 4" coil 2" long, at 50watts it hits 225c so that's where it would melt, high temp coil insulation is rated at 220.

[lilmike]

Why use an amp? Why not use a welder? A sine-wave is a sine wave. My TIG welder can handle short circuits and since it is a TIG, I have a foot pedal for controlling current. 

Just a thought.

[SME]

2 hours ago, Funk Audio said:

That actually becomes a calculable thing at that point. You can call the coil a flat heat sink, there are online calculators to tell you how many watts will reach what temp in ambient air. I put in the area for a 4" coil 2" long, at 50watts it hits 225c so that's where it would melt, high temp coil insulation is rated at 220.

Do you know what assumptions are involved?  Particularly concerning the air?  The situation is very different if the air is stagnant than if it is allowed to convect freely.  My guess is that a heat-sink calculator would assume free (natural) convection under some particular mounting configuration (say installed onto an infinite slab).  A sealed box stuffed with insulation may not allow much if any natural convection, so the situation may be worse than this calculation suggests.

On the up side, if the coil is mounted in very close physical proximity (on the order of one millimeter or less, by my rough calc) to other metal parts of the driver, enough heat could travel them through the air to make a difference.

[Funk Audio]

11 hours ago, SME said:

Do you know what assumptions are involved?  Particularly concerning the air?  The situation is very different if the air is stagnant than if it is allowed to convect freely.  My guess is that a heat-sink calculator would assume free (natural) convection under some particular mounting configuration (say installed onto an infinite slab).  A sealed box stuffed with insulation may not allow much if any natural convection, so the situation may be worse than this calculation suggests.

On the up side, if the coil is mounted in very close physical proximity (on the order of one millimeter or less, by my rough calc) to other metal parts of the driver, enough heat could travel them through the air to make a difference.

The calculator I found allows orientation and ambient temp settings, if I set it to 20c and face down the power drops to 42watts. I am sure one could calculate the thermal resistance of the air in the box and the wood as well and input the degrees/watt resistance to the calculator as well to get a more accurate sim. If I change the ambient temp to 50c degrees, to assume the inside of the enclosure might stabilize at that then the power to get the coil over 220c drops to 35watts. Adding steel helps but even setting the final surface area to what a huge motor would have keeping the source(coil) the same size and setting 1mm air between the coil and steel brings us up to ~55watts at 50c ambient, ~65watts at 20c.  My main point is no matter what you do its usually going to be a dismally low amount of true RMS power a driver can take for an infinite period of time.

[SME]

3 hours ago, Funk Audio said:

The calculator I found allows orientation and ambient temp settings, if I set it to 20c and face down the power drops to 42watts. I am sure one could calculate the thermal resistance of the air in the box and the wood as well and input the degrees/watt resistance to the calculator as well to get a more accurate sim. If I change the ambient temp to 50c degrees, to assume the inside of the enclosure might stabilize at that then the power to get the coil over 220c drops to 35watts. Adding steel helps but even setting the final surface area to what a huge motor would have keeping the source(coil) the same size and setting 1mm air between the coil and steel brings us up to ~55watts at 50c ambient, ~65watts at 20c.  My main point is no matter what you do its usually going to be a dismally low amount of true RMS power a driver can take for an infinite period of time.

Thanks for the clarification.

The obvious lesson here is that the worst-case scenario for long-term power with no ventilation is *very low*, and potentially lower than the numbers above because the calculator is very likely assuming free convection with an unlimited source of air at the given temperature.  For those following along, free convention occurs when warmer, more buoyant air begins to rise and pulls cooler air in from the sides or below.  Convection makes air cooling *much more effective* than it otherwise would be because *pure conduction in air is very poor*.  This is something that occurs much more readily when there is room for the rising air column to develop sufficient velocity.  A tiny sealed box fill with stuffing that increases air resistance, won't anywhere near as much convection compared to an open-air installation.

I'm thinking the temperature monitoring may be a very good idea for subs used in pro settings with EDM with potentially high duty cycle signals near tune.  One might need a lot more capacity than anticipated if reliable long-term operation at high levels is required.  Another thing that could help would be to purposely tune a bit lower than the lowest anticipated content.  A 30 Hz tune might be OK for music that's mostly 40 Hz and up, but stuff that hits at 30 Hz hard, it may not be a bad idea to tune a bit lower, to try to avoid the deadly combination of high level high duty cycle signals at minimum impedance and low excursion.