Greetings...
I'm in the process of building a pair of bi-amped bookshelves with Seas L15RLY/P and 27TDC. I'm going to use a pair of bridged Symasym's for each woofer, and a single Symasym for each tweeter.
Now, they'll be actively crossed over at 2KHz, but i also intend on using a Zobel network on the woofer. The part i'm concerned about is the voltage rating for the cap in the RC network. The amps will have +-35 to +-40V rails, and because they'll be bridged, if i understood correctly, the voltage swing doubles, so that gives me a max of 160V peak-to-peak on the woofer.
Currently, i have a small stock of 63V ERO (Vishay-Roederstein) film caps from which to put together the 15uF i need. Given the situation, would it be too risky to use these 63V caps, or would 160V (or more) caps be necesary?
Many thanks in advance
Chris
I'm in the process of building a pair of bi-amped bookshelves with Seas L15RLY/P and 27TDC. I'm going to use a pair of bridged Symasym's for each woofer, and a single Symasym for each tweeter.
Now, they'll be actively crossed over at 2KHz, but i also intend on using a Zobel network on the woofer. The part i'm concerned about is the voltage rating for the cap in the RC network. The amps will have +-35 to +-40V rails, and because they'll be bridged, if i understood correctly, the voltage swing doubles, so that gives me a max of 160V peak-to-peak on the woofer.
Currently, i have a small stock of 63V ERO (Vishay-Roederstein) film caps from which to put together the 15uF i need. Given the situation, would it be too risky to use these 63V caps, or would 160V (or more) caps be necesary?
Many thanks in advance
Chris
Hi,
the normal maximum voltage across the speaker terminals is |Vrail|
For a bridged design this becomes |2*Vrail|
Your bridged amp running on +-40Vdc could send 80Vpk to the speaker.
But, there is a complication.
If the amplifier output suddenly switches off, the back emf from the speaker can exceed the driving voltage.
It's all about inductors trying to oppose the change.
I would use 80V times a suitable factor.
100V (1.25*) seems a little too close, 160V (2*) would be a lot better. 250V (3*) better still.
15uF is just in reach of motor run/start capacitors. Using a PP version of one of those, although unusual, would be a safe start (they're not too expensive).
the normal maximum voltage across the speaker terminals is |Vrail|
For a bridged design this becomes |2*Vrail|
Your bridged amp running on +-40Vdc could send 80Vpk to the speaker.
But, there is a complication.
If the amplifier output suddenly switches off, the back emf from the speaker can exceed the driving voltage.
It's all about inductors trying to oppose the change.
I would use 80V times a suitable factor.
100V (1.25*) seems a little too close, 160V (2*) would be a lot better. 250V (3*) better still.
15uF is just in reach of motor run/start capacitors. Using a PP version of one of those, although unusual, would be a safe start (they're not too expensive).
The very nature of this heavy duty cap, predisposes it more to being certain types... which are also good for mains filtering... paper and film etc, as oposed to the normal selfhealing stuff... whish I could remember the publication now... One of the reasons are that other types can fail as closed circuit...
but the speakers are never driven to see the Vpp signal.
The Vpk is the worst case.
Take one amp with output terminal at -40V and the other amp sending +40V to the other output terminal. This is what bridging allows the terminals to do.
Now, the speaker sees +40pk - (-40Vpk) =80Vpk across it's input terminals. The only time the speaker can ever see more than the Vrail to rail voltage is that sudden cut off inductive spike.
Hi Richie,
each amp has supply rails of +-40Vdc.
each amp can send a maximum voltage to the output terminal of either +40Vpk or -40Vpk.
wire them in bridge mode and the maximum across BOTH terminals becomes +40Vpk to -40Vpk or -40Vpk to +40Vpk.
In both cases the maximum voltage at the speaker terminals is twice what a single amplifier can deliver. i.e. 40Vpk in non bridged mode and 80Vpk in bridged mode.
each amp has supply rails of +-40Vdc.
each amp can send a maximum voltage to the output terminal of either +40Vpk or -40Vpk.
wire them in bridge mode and the maximum across BOTH terminals becomes +40Vpk to -40Vpk or -40Vpk to +40Vpk.
In both cases the maximum voltage at the speaker terminals is twice what a single amplifier can deliver. i.e. 40Vpk in non bridged mode and 80Vpk in bridged mode.
You don't have to worry about the inductive spike unless the amp is misdesigned and does not have output clamping diodes. If it doesn't have output clamping diodes and also has VI-limiting it will blow up itself when the protection circuit activates, which is bad design. 
80V is the maximum voltage the cap can see. A 100V cap would be perfect.
80V is the maximum voltage the cap can see. A 100V cap would be perfect.
A speaker cone driven by a sine wave will move one way then the other, reflecting the change of polarity of it's input stimulus. True, at any one time the max voltage across the speaker will be limited to the peak voltage, but the charge stored in the capacitor will be of one polarity and in the extreme case of a fast reversal of drive, could see the full pk-pk.
Hmm.. So, either way, i have to use some 100V caps minimum, to (also) have some safety margin.
Series-paralleling caps i have would be (at least) cumbersome (currently 2x6.8u + 1.5u @ 63V, that would mean 8x6.8u + 4x1.5u).
Guess i'll have to resort to ordering new 100V caps... And i was hoping i could use the caps i managed to source (for free)...
PS: i tried to do some calculations, to see how the voltage would be split among the resistor and the cap in a Zobel network, and i ended up with the cap seeing about 90% of the applied voltage. Is that right?
Series-paralleling caps i have would be (at least) cumbersome (currently 2x6.8u + 1.5u @ 63V, that would mean 8x6.8u + 4x1.5u).
Guess i'll have to resort to ordering new 100V caps...
And i was hoping i could use the caps i managed to source (for free)...PS: i tried to do some calculations, to see how the voltage would be split among the resistor and the cap in a Zobel network, and i ended up with the cap seeing about 90% of the applied voltage. Is that right?
This is odd...
I just searched, found and used three different online Zobel calculators, each giving me a different result.
http://www.carstereo.com/help/Articles.cfm?id=36 gave me 17.8uF and 6.9 ohms (Re=5.5ohms, Le=0.84uH)
http://www.apicsllc.com/apics/Misc/filter2.html#zobel gave 27.8uF and 6.9ohms (for the same parameters)
http://www.selectproducts.com/calculator_zobel.html gave 4uF and 10 ohms (for Znom=8 ohms and frequency 5KHz, as per the graph at http://www.seas.no/Prestige Basser PDF/L15RLY_P-H1141-08.pdf )
I just searched, found and used three different online Zobel calculators, each giving me a different result.
http://www.carstereo.com/help/Articles.cfm?id=36 gave me 17.8uF and 6.9 ohms (Re=5.5ohms, Le=0.84uH)
http://www.apicsllc.com/apics/Misc/filter2.html#zobel gave 27.8uF and 6.9ohms (for the same parameters)
http://www.selectproducts.com/calculator_zobel.html gave 4uF and 10 ohms (for Znom=8 ohms and frequency 5KHz, as per the graph at http://www.seas.no/Prestige Basser PDF/L15RLY_P-H1141-08.pdf )
I'm curious why you think a zoebel is needed with a active crossover.
Zoebels are useful with passive crossovers to stabilize the filter load, but an active crossover doesn't have that issue. Surely the Symasym can handle your drivers (you probably don't really need to bridge the woofer amp)
My
is avoid the expense and do without a zoebel for your woofer. I have been doing it happily for years.
Zoebels are useful with passive crossovers to stabilize the filter load, but an active crossover doesn't have that issue. Surely the Symasym can handle your drivers (you probably don't really need to bridge the woofer amp)
My
is avoid the expense and do without a zoebel for your woofer. I have been doing it happily for years. I was considering applying a zobel after the recommendation of the good sir i bought the drivers from, a quite well-reputed custom speaker designer in our "neck of the woods". He said it may very well / aid in the quality of the midrange produced by the woofer, and i had no reason to think otherwise.
The zoebel will flatten the load the amp sees and make it less reactive. If your amp doesn't respond well to reactive loads then a zoebel might make a difference. Tube amps are a problem this way.
I haven't heard the Symasym (yet), but I suspect it will handle any reasonable load just fine. But, if you want to see if it does make a difference, why not use standard crossover cap usually rated 250V+?
If you don't hear a difference with a zoebel, you can add it to your tweeter protection caps. I learned the hard way that amp turn on thumps can fry a tweeter in short order. I use 30 uf or more (3x 10 uf in the current model) in series with the tweeter. That's enough to keep that filter from affecting the response too much.
I haven't heard the Symasym (yet), but I suspect it will handle any reasonable load just fine. But, if you want to see if it does make a difference, why not use standard crossover cap usually rated 250V+?
If you don't hear a difference with a zoebel, you can add it to your tweeter protection caps. I learned the hard way that amp turn on thumps can fry a tweeter in short order. I use 30 uf or more (3x 10 uf in the current model) in series with the tweeter. That's enough to keep that filter from affecting the response too much.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.