Hello everyone! It has been a while.
Last time I was here, I had just replaced the (2x) EL-70 drivers in my Planet10 microTowers with (2x) Pluvia 7.2HD. At the time, I had trouble deciding whether to go with series or parallel wiring. The EL-70s were nominal 4ohm, so I originally went with series wiring for nominal 8 ohm total. However, the Pluvias are nominal 6 ohm, so I had the conundrum of choosing 12 ohms (series) or 3 ohms (parallel).
Ultimately I chose the lazy way (12ohms, series) because the speakers were already wired that way and my Tubelab SSE has 4, 8 and 16 ohm taps (went with the 16 ohm taps and it sounds excellent).
Trouble is, now these speakers do not get along at all with my "Wiener" TPA 3118 Amp. When I ordered it, I selected the 4 ohm output filter configuration and it is fairly well-known that the results can be pretty terrible if these amps are played with inappropriate speaker loads, especially such a massive mismatch of 4 ohms vs. 12 ohms and especially when the rising impedance curve of these "naked" drivers makes matters in the top end exponentially worse for the Wiener (not so much for the SET!). This isn't just theoretical - it sounds horrific.
gmarsh, designer and builder of the "Wiener", concluded that a Zobel is a fantastic addition to flatten the impedance and help get the best from the Wiener. I grabbed his code from that post and started playing with it in GNU Octave. (Thanks Gary!) The only way forward to get these speakers to work with the Wiener Amp is to go for parallel wiring, and I might as well add the Zobel while I am at it. The simulations I have done in GNU Octave, with some help from an online Zobel calculator have made this all pretty straightforward.
Question #1: The Pluvia 7HDs wired in parallel will have an ReVc of 2.7 ohms (nominal 3 ohms impedance). This will not be ideal for the 4 ohm tap on my SET, nor the 4 ohm target load of the Wiener Amp. Should I add an extra 1 ohm of resisance? (could be accomplished by adding a 2 ohm 10W resistor in series with each driver).
Question #2: Right now the naked speakers running in series (12 ohms nominal, no Zobel) are a very nice match with my SET. Will rewiring as proposed and adding a Zobel screw that all up?
Any feedback will be very much appreciated! Thanks.
Last time I was here, I had just replaced the (2x) EL-70 drivers in my Planet10 microTowers with (2x) Pluvia 7.2HD. At the time, I had trouble deciding whether to go with series or parallel wiring. The EL-70s were nominal 4ohm, so I originally went with series wiring for nominal 8 ohm total. However, the Pluvias are nominal 6 ohm, so I had the conundrum of choosing 12 ohms (series) or 3 ohms (parallel).
Ultimately I chose the lazy way (12ohms, series) because the speakers were already wired that way and my Tubelab SSE has 4, 8 and 16 ohm taps (went with the 16 ohm taps and it sounds excellent).
Trouble is, now these speakers do not get along at all with my "Wiener" TPA 3118 Amp. When I ordered it, I selected the 4 ohm output filter configuration and it is fairly well-known that the results can be pretty terrible if these amps are played with inappropriate speaker loads, especially such a massive mismatch of 4 ohms vs. 12 ohms and especially when the rising impedance curve of these "naked" drivers makes matters in the top end exponentially worse for the Wiener (not so much for the SET!). This isn't just theoretical - it sounds horrific.
gmarsh, designer and builder of the "Wiener", concluded that a Zobel is a fantastic addition to flatten the impedance and help get the best from the Wiener. I grabbed his code from that post and started playing with it in GNU Octave. (Thanks Gary!) The only way forward to get these speakers to work with the Wiener Amp is to go for parallel wiring, and I might as well add the Zobel while I am at it. The simulations I have done in GNU Octave, with some help from an online Zobel calculator have made this all pretty straightforward.
Question #1: The Pluvia 7HDs wired in parallel will have an ReVc of 2.7 ohms (nominal 3 ohms impedance). This will not be ideal for the 4 ohm tap on my SET, nor the 4 ohm target load of the Wiener Amp. Should I add an extra 1 ohm of resisance? (could be accomplished by adding a 2 ohm 10W resistor in series with each driver).
Question #2: Right now the naked speakers running in series (12 ohms nominal, no Zobel) are a very nice match with my SET. Will rewiring as proposed and adding a Zobel screw that all up?
Any feedback will be very much appreciated! Thanks.
Last edited:
Apologies for the incoherence here as thanks to my dentist, I've been stuck on a mixture of codeine, amitriptyline and ibuprofen & I'm a bit light headed.
Adding series R will alter the LF load alignment so generally is best avoided. If you're happy with the current configuration, I'd be inclined to stick with that, but that could also just be bone-idleness (and a lot of prescribed drugs) on my part. You can try the Zobel if you like -it's unlikely to do any harm. Unless it's a massively oversized type though, it's probably not going to do a huge amount, as the Pluvias have a reasonable amount of copper in the motor & inductance isn't massively rising with frequency. Since they're cheap to implement, might as well give it a shot.
FWIW, providing you have power to burn, you could try shunting the series-wired drivers with a power-resistor or several paralleled power-resistors so you have, say, 50w power-handling. Call it 22ohms starting point & see what gives from there. Inefficient, but so are series resistors, & providing you go into it for a good reason rather than a wild shot in the dark, no harm in it, as it's completely reversible and costs little to try out.
Adding series R will alter the LF load alignment so generally is best avoided. If you're happy with the current configuration, I'd be inclined to stick with that, but that could also just be bone-idleness (and a lot of prescribed drugs) on my part. You can try the Zobel if you like -it's unlikely to do any harm. Unless it's a massively oversized type though, it's probably not going to do a huge amount, as the Pluvias have a reasonable amount of copper in the motor & inductance isn't massively rising with frequency. Since they're cheap to implement, might as well give it a shot.
FWIW, providing you have power to burn, you could try shunting the series-wired drivers with a power-resistor or several paralleled power-resistors so you have, say, 50w power-handling. Call it 22ohms starting point & see what gives from there. Inefficient, but so are series resistors, & providing you go into it for a good reason rather than a wild shot in the dark, no harm in it, as it's completely reversible and costs little to try out.
Thanks Scott!
Okay, so no series resistance - check.
I'm curious to know more about this shunt idea - both the theory and the practical implementation. Sadly, I only know enough about electronics to be dangerous. With the SET, I do not really have power to burn.
What about the idea of rewiring to parallel and adding the zobel? This should play relatively nicely with both amps? No? (Nominal 3 ohms). While I am happy with the current speaker configuration + SET, the same speakers with the Wiener are abysmal. I'd like the speakers to play nicely with both amps, and perhaps others.
P.S. I hope you feel better soon!
Okay, so no series resistance - check.
I'm curious to know more about this shunt idea - both the theory and the practical implementation. Sadly, I only know enough about electronics to be dangerous. With the SET, I do not really have power to burn.
What about the idea of rewiring to parallel and adding the zobel? This should play relatively nicely with both amps? No? (Nominal 3 ohms). While I am happy with the current speaker configuration + SET, the same speakers with the Wiener are abysmal. I'd like the speakers to play nicely with both amps, and perhaps others.
P.S. I hope you feel better soon!
Last edited:
IO would have no issue swiring P7HD in parallel. Its DCR is consistent with a nominal 8Ω speaker, so that wil give the optimum Weiner load.
A Zobel on the loudspeaker is to help the amplifier. P7HD has pole piece cap (AFAIK) and does not exibit a hige rise in the HF.
dave
A Zobel on the loudspeaker is to help the amplifier. P7HD has pole piece cap (AFAIK) and does not exibit a hige rise in the HF.
dave
It doesn't have to seem this way. Your amplifier with various loads on different taps will either be Voltage limited or current limited. Don't be afraid to experiment to find the optimum tap.
Thanks gents!
Parallel it will be. Zobel... probably. Heck, I could easily make that switchable, or even just bridge it across the terminals.
Parallel it will be. Zobel... probably. Heck, I could easily make that switchable, or even just bridge it across the terminals.
Reactance annul zobel 😉
Yeah, normally used cheap pots to dial in each speaker, then sometimes replaced with an adjustable power resistor or created small resistor 'banks' if there were any inquisitive types around.
Mysterious isn't it how the 'old ways' have a habit of working. 😉
You made me a bit of a convert to resistor banks a few years back -I went a bit OTT and subsituted a regular padding resistor with 8 (I think it was 8) series-parallel high precision metal films. Not only easy to fine-tune -a surprising difference for the better. Shows engineering counts, as usual (if we eject some of the mad audiophile verbiage used to sell some components).
I do like diode-biasing film caps too (prefer it to using batteries actually).
You made me a bit of a convert to resistor banks a few years back -I went a bit OTT and subsituted a regular padding resistor with 8 (I think it was 8) series-parallel high precision metal films. Not only easy to fine-tune -a surprising difference for the better. Shows engineering counts, as usual (if we eject some of the mad audiophile verbiage used to sell some components).
I do like diode-biasing film caps too (prefer it to using batteries actually).
Thanks very much for the link! I've created a spreadsheet for those calculations. Will be very handy.
I plugged the new values into the Mattlab code, dumped it into GNU Octave, and all looks good.
I'll probably play with the values a bit to find common R and C component values (or combinations thereof) that are "close enough" (as per Martin J. King's comments in the same thread).
(digging through parts bin)
Okay, just to be sure, if someone could double-check my math (which has never been my forte) I'd very much appreciate it.
Let's start with the Pluvia 7:
Re = 5.4 ohms
Re2p (2 drivers parallel) = 2.7 ohms
F @ 2xRe = ~ 15,000 (eyeballed from Impedance vs Freq. graph)
So if I plug 2.7 ohms and 15,000 into my spreadsheet, I get Zobel values:
Rz = 3.375
Cz = 3.93
(digging through parts bin)
Well, I happen to have 4 x 5W 8.2 ohm Xicon metal oxide resistors, which could be paralleled for 4.1 ohms (each speaker).
And I have 2 x 2.2uF 630v Obbligato Film Oil (non-polar) "audio capacitors", which are each approximately the size of D-cell batteries.
The values deviate considerably from the calculated values, but as expected the GNU Octave simulation results are not "terrible". Response gently rises starting at 9KHz ending +1.8dB at 20KHz, and the impedance isn't "crazy". (see attachment)
Not ideal, and a really stupid use of the capacitors since component "audio quality" is basically moot in a Zobel. Thinking I might purchase some more appropriate caps. I guess while I am at it, I could do some math and buy some resistors to get closer to the calculated value.
I dunno. These Obbligato caps have been sitting in my bin for a long time and not been used. The DCR is 0.00 as measured by my component tester. Also, a small boost on the top end might add some "air" that might be welcome... (if I can even hear it).
Okay, just to be sure, if someone could double-check my math (which has never been my forte) I'd very much appreciate it.
Let's start with the Pluvia 7:
Re = 5.4 ohms
Re2p (2 drivers parallel) = 2.7 ohms
F @ 2xRe = ~ 15,000 (eyeballed from Impedance vs Freq. graph)
So if I plug 2.7 ohms and 15,000 into my spreadsheet, I get Zobel values:
Rz = 3.375
Cz = 3.93
(digging through parts bin)
Well, I happen to have 4 x 5W 8.2 ohm Xicon metal oxide resistors, which could be paralleled for 4.1 ohms (each speaker).
And I have 2 x 2.2uF 630v Obbligato Film Oil (non-polar) "audio capacitors", which are each approximately the size of D-cell batteries.
The values deviate considerably from the calculated values, but as expected the GNU Octave simulation results are not "terrible". Response gently rises starting at 9KHz ending +1.8dB at 20KHz, and the impedance isn't "crazy". (see attachment)
Not ideal, and a really stupid use of the capacitors since component "audio quality" is basically moot in a Zobel. Thinking I might purchase some more appropriate caps. I guess while I am at it, I could do some math and buy some resistors to get closer to the calculated value.
I dunno. These Obbligato caps have been sitting in my bin for a long time and not been used. The DCR is 0.00 as measured by my component tester. Also, a small boost on the top end might add some "air" that might be welcome... (if I can even hear it).
Last edited by a moderator:
For reference, same sim, no Zobel. The massive peak is far out of the audio band, but for reasons that I don't understand this can still be a very bad thing (from what I have read). Something about ringing? reflections? possible amp damage (Wiener).
Last edited by a moderator:
Today I finally had a chance to implement this change, and I have to say it was a huge success!
Wrap up: Speakers are wired in parallel and now include a Zobel (4.4uF--4R1) for ~3-4ohm nominal load. They sound excellent with the Wiener TPA3118 amp - all ringing and top end harshness is gone. To my surprise, I do believe they sound noticeably better with my Tubelab SSE now as well (using the 4ohm taps, of course). This was unexpected.
Thanks to all who offered help and advice!
Wrap up: Speakers are wired in parallel and now include a Zobel (4.4uF--4R1) for ~3-4ohm nominal load. They sound excellent with the Wiener TPA3118 amp - all ringing and top end harshness is gone. To my surprise, I do believe they sound noticeably better with my Tubelab SSE now as well (using the 4ohm taps, of course). This was unexpected.
Thanks to all who offered help and advice!