Australian Technical Production Services

Site Navigation

2 way Passive crossover with Tweeter protection  

There are plenty of Passive crossover kits and PCBs around but this one addresses a problem that many do not touch - that is Tweeter protection. 
Many of you will have noticed that tweeter Failures are not uncommon particularly in Sound reinforcement systems.  
I have heard many theories attempting to explain this but really the reason for this is very simple:
    
Tweeters, particularly the horn loaded compression driver type are very sensitive, often 6 to 12dB more so than the lower frequency drivers.
High frequency power of 'normal music' expected from the Tweeter is often as low as 25% of the total acoustic power.
What all this means is that a tweeter only needs to have a fraction of the power handling of the speaker system as a whole and 50 Watt tweeters in 500W speakers are not entirely unheard of.

Now when everything is working as it should this is fine, but when something starts feeding back, and your Amplifier starts trying to deliver full power to your Tweeter, then its future life expectancy is often measurable in fractions of a second....

Having had a number of Tweeter Failures I decided it was time to investigate protection.
The speakers I had on hand with blown tweeters had been protected with poly switches yet had still suffered failure. I had heard that selecting the best poly switch was always a balance between adequate protection and nuisance tripping so I ran some tests on three Poly switches: RXE110 (Rated at 2.2A or 38.72W @ 8 Ohms) RXE090 (Rated at 1.8A or 25.9W @ 8 Ohms) and RXE075 (Rated at 1.5A or 18W @ 8 ohms)

To test the response times of the poly switches I used the following test rig:

I used a Tektronix TDS210 as the oscilloscope running as single shot at 1 second per division and a dummy load of 4 Ohms this dummy load has a 2:1 output to the oscilloscope.
This allowed me to measure with a reasonable degree of accuracy how long the poly switch took to open.
For the first round of tests I ran 4.1 Amps through the poly switch and measured how long the poly switch took to open:
The RXE110 took almost 4 Seconds to open:
The RXE090 took almost 3 Seconds to open and even the RXE075 took 2.5 Seconds to open.

Having tested several Tweeters to Destruction (by no means an exhaustive test as Tweeters are expensive, but more to get an indication of how long a tweeter could last) I had found that most nominal 50Watt tweeters would be lucky indeed to last 2 seconds at 128 Watts.
I had also found that many of the cheaper '50 Watt' tweeters were in fact lucky to survive several seconds at full power, although to give credit where credit is due the two more expensive tweeters I tested did handle rated power for several hours - despite being mounted inside an insulated wooden box (the test got rather loud otherwise).

Conclusion Poly switches are woefully inadequate as Speaker protection.

Lightbulbs likewise proved to be problematic - either getting consistent lightbulbs or lightbulbs that provided adequate protection without false tripping or failure.

What I needed was protection that tripped at a preset point, allowed a slight (but not significant delay) before tripping to avoid tripping on otherwise harmless signals, preferably a circuit that reset itself when the danger had passed and just to add to the challenge a circuit that required no external power....


This is the circuit I ended up going with:
C9 and L3 act as a high pass filter while R9 provides a constant/predictable load for the filter.

D1 to D4 are UF4001 high speed diodes that operate as a full wave rectifier and C6 - 1uF (May be vearied to suit application) acts as a Filter/Delay.

R6 (Set as required) and R7 (560 Ohm) act as a voltage divider for the trigger for the SCR SC1 and set the voltage at which the protector trips. R6 may be calculated by R6= (Tv-0.56)*1000 where Tv is the trip Voltage.

R1 allows the Input filter to be bypassed if/when the protection is triggered this will mean that the protection remains active until it is reset by everything going quiet - I have not used this option yet but I suggest R1 of around 39 Ohms (if used).

C8 (10uF) resets the SCR when the Relay drops out - resetting the protection circuit

Finally Q1(MJE340), R4 (85 Ohm), ZD1 (3V3) and R5 act as a current limit to stop the Relay burning out when the protector has tripped.

RLY1 is a Relay with a 400 Ohm coil and R9 is set at 150 Ohms, This sort of extra load any half decent power amplifier should be able to drive without even noticing and this does make the Circuit self powered.

Initially I did experiment with alternative methods of resetting the SCR, that did not involve an extra pair of Relay contacts however Relay I ended up going with only cost about $1 more than the SPST Relay and has 5 Amp contacts instead of 3.

The reason for having a high pass filter in front of this circuit rather than installing this circuit after the existing high pass filter in the crossover is that if the tweeter alone is disconnected, then in the case of a second order crossover, the high pass filter for the tweeter will act as a series resonant circuit which may cause stability issues for some Amplifiers.

Of course this is a Crossover so by the time I add all the other components for a second order crossover we get:
R3 is the Pad for the tweeter and C2, L1 act as the high pass filter for the tweeter R10 provides an option for the tweeter to operate at a lower level if the protection is tripped although I suggest that there would have to be a very good reason for using this option (I have not used it).
 
 L2 and C7 are the low pass filter for the woofer while C4 and R2 provide impedance correction for the woofer.
Woofers typically are largely inductive by the time you get to their crossover point and  may have an impedance of 40 Ohms or more, as a result simple 2nd order crossovers can end up providing far less than 12dB per octave rolloff.
This may not be an issue but if it is then impedance correction will help fix this

F1 provides 'Line of last resort' only protection for the woofer - by far the majority of failures I see are of the tweeters and by the time you fry a woofer things are going horribly wrong and a Relay that can handle the sorts of currents involved is going to be quite substantial.

If I am to be honest though, the fuse was really added as the PCB needed a link and I figured an appropriately sized fuse was unlikely to cause any harm.

Calculations for all these component Values are included along with worksheets in the Instructions for this project (PDF)

Note that you need not add all of these components if you wish (not even the tweeter protection if you do not want - although I suspect you could get a cheaper PCB if you did not want this option).

Kit costs $18 (Australian) and includes:
PCB
Assembly instructions
Crossover Calculation equations


Email  for more details or Paypal instructions (to order) - note at this stage I will have to work out mail costs on a case by case basis...


Website created by Richard Freeman 
with the aid of CoffeeCup HTML editor