Speaker/Amp Impedance Issue

Gerry E. · 3627

0 Members and 1 Guest are viewing this topic.

Offline Gerry E.

  • Full Member
  • ***
    • Posts: 190
on: April 25, 2017, 12:08:46 PM
I’ll try to keep this as brief as possible.  Last year I moved my vintage ALTEC 756B/Jensen RP302 speaker drivers from 3-foot wide JE Labs style open baffles to Pure Audio Project (PAP) OBs.  I did this for two reasons.  First, the PAP OBs are half as wide as the JE Labs.  Second, the PAP OBs have built-in 15 inch woofers.  Two per side, so I no longer needed a separate powered sub-woofer.

I wasn’t sure if my Bottlehead Paramount 300Bs were powerful enough to drive the 15” woofers, plus the ALTEC/Jensen combo, so I used a Crown DriveCore 2 amp with built-in DSP crossover to power the woofers.  This worked very well and also let me dial in the level of the woofers using the Crown’s built-in level controls.
 
I recently got the idea to see what would happen if I tried powering the speakers full-range off the Paramounts.  In order to do this, I purchased a pair of 80hz 4 ohm 12db/octave low-pass filters from Parts Express.  The reason the filters are 4 ohms is because the two 8 ohm woofers per side are wired in parallel.

OK, so how did it sound? Not so great at first! The low bass seemed to be missing in action. Then I remembered reading that 12db/octave filters are 180 degrees out-of-phase. I assumed this was causing significant cancellation between the PAP woofers and ALTEC 756Bs. 

The simple fix was to wire the woofers out-of-phase (reverse the +/- input connections) and that made all the difference! Now the low bass was back in business. See the attached photos for the system setup.  I sacrificed some bass response below 40hz but the overall sound is now more integrated (coherent).  The Paramounts were up to the challenge!     

My only concern now is the load impedance presented to the Paramounts.  My Paramounts have the upgraded Magnequest BH-5 output transformers which only have 8 and 16 ohm seconday taps.  I posted something about this on the Audio (Hi-Efficiency) Asylum and with some help, determined that the speaker impedance is below 4 ohms.  Is this a problem?  Here are my options:
 
1.  Don’t worry about it because it sounds good and the amps seem to be working just fine.
2.  Re-wire the woofers in series which will bring the speaker impedance up to the 8 – 16 ohm range.  If I do this, I will need 16 ohm low pass filters.                                         
3.  Get different output transformers for my Paramounts.  One’s with 4-ohm output taps.
4.  Get different woofers.  One’s with a 16 ohm impedance so wired in parallel will be 8 ohms.

Any comments or suggestions?  Thanks!

Gerry     



Offline Zimmer64

  • Full Member
  • ***
    • Posts: 210
    • Blog
Reply #1 on: April 26, 2017, 08:47:30 AM
If you like what you hear, leave it as it is. The impedance mismatch results mainly in a low damping factor. I don't think you hurt or overly stress the amp.

Michael

Nelson Pass F5 Turbo V2, Quickie (mod), S.E.X. 2.1, Tubes4hifi SP14, Dynaco VTA ST 70, Tubelab SSE, Vroemen Diva Superiore ER4, Jordan JX92S VTL, 47 labs 0647 CD, Aqvox DAC, Rowen Absolute pre / psu / power amps, BG Neo3 / Betsy / Eminence A15 open baffles


Offline Paul Birkeland

  • Global Moderator
  • Hero Member
  • *****
    • Posts: 19365
Reply #2 on: April 26, 2017, 10:23:36 AM
2.  Re-wire the woofers in series which will bring the speaker impedance up to the 8 – 16 ohm range.  If I do this, I will need 16 ohm low pass filters.               
I would try this first.  If you bought a prefabricated crossover from PE, you may want to calculate your own values and just but the pair of inductors and caps without the PC board.  You could leave the amp set at 8 Ohms for the 16 Ohm woofer array on each channel.

Paul "PB" Birkeland

Bottlehead Grunt & The Repro Man


Offline Gerry E.

  • Full Member
  • ***
    • Posts: 190
Reply #3 on: April 27, 2017, 04:07:01 AM
Thanks guys!

Michael - It's good to get confirmation that the mismatch will likely do no harm.

PB - Your suggestion was also my preference (wire the woofers in series and replace the filters with 16 ohm versions).  My only hesitation was that I wasn't sure if series wiring causes any sound anomalies.  As for the low-pass filter, Parts Express (or anyone else I'm aware of) do not offer 16 ohms versions, so I'm on my own.  That's fine because this way I can chose better caps and coils.  Of course I don't have to go too overboard with these parts because it's only for the two lowest octaves.

I'll report back after the changes, thanks again.

Gerry           



Offline Zimmer64

  • Full Member
  • ***
    • Posts: 210
    • Blog
Reply #4 on: April 27, 2017, 08:50:49 AM
This might help you if you want to design it yourself as CB suggested: http://www.apicsllc.com/apics/Misc/filter2.html

Nelson Pass F5 Turbo V2, Quickie (mod), S.E.X. 2.1, Tubes4hifi SP14, Dynaco VTA ST 70, Tubelab SSE, Vroemen Diva Superiore ER4, Jordan JX92S VTL, 47 labs 0647 CD, Aqvox DAC, Rowen Absolute pre / psu / power amps, BG Neo3 / Betsy / Eminence A15 open baffles


Offline Gerry E.

  • Full Member
  • ***
    • Posts: 190
Reply #5 on: April 27, 2017, 12:29:45 PM
This might help you if you want to design it yourself as CB suggested: http://www.apicsllc.com/apics/Misc/filter2.html


Hi Michael:

I actually used a similar calculator earlier today to get the values I needed.  I just ran it again using the calculator in your link to confirm those values.  I grabbed a snapshot of the results (use the link below). 

I'm only concerned with the low pass values and here's a problem - Inductor L2 is specified at 45.0199mH.  However, when I look for an inductor with that high a value, I don't find any at the usual sources (Parts Express, etc.). 

Most are in the < 1mH - 10mH range.  I realize that 80hz is a very low crossover frequency but still expected to see higher inductor values available.  Maybe I'm missing something or not reading this right?

Thanks!

Gerry               



Offline Paul Joppa

  • Global Moderator
  • Hero Member
  • *****
    • Posts: 5768
Reply #6 on: April 27, 2017, 05:17:33 PM
High inductance like that is very heavy and large, especially with an air core. Here's a calculator:

https://www.diyaudioandvideo.com/Calculator/AirCoreInductorDesigner/

for instance, you can get 45mH at 0.85 ohms with 26 pounds (!) of 10-gauge wire, on a form 5" diameter, 2.5" long, and about 9" outer diameter. This is very close to the optimum (least expensive) design. The wire alone will weigh around 25 pounds and cost around $300 or more.

This is why active line-level crossovers and separate deep-bass power amps are so often used.

Paul Joppa


Offline Gerry E.

  • Full Member
  • ***
    • Posts: 190
Reply #7 on: April 28, 2017, 04:13:29 AM
for instance, you can get 45mH at 0.85 ohms with 26 pounds (!) of 10-gauge wire, on a form 5" diameter, 2.5" long, and about 9" outer diameter.

Well there goes that idea!  The inductor in the 4 ohm, pre-built low-pass filter from Parts Express is not that big so I ran the values through the calculator.  4 ohm woofers are 11.25mH, 8 ohm woofers are 22.5mH and 16 ohm woofers are 45mH (12dB/octave, Butterworth filter).  So, for every doubling of the woofer impedance, the inductor must be twice the size.

Thanks Paul!

Gerry         



Offline Zimmer64

  • Full Member
  • ***
    • Posts: 210
    • Blog
Reply #8 on: April 28, 2017, 06:29:56 AM
You could add another amp and do a line level or digital XO in front of the power amp and bi-amp

Nelson Pass F5 Turbo V2, Quickie (mod), S.E.X. 2.1, Tubes4hifi SP14, Dynaco VTA ST 70, Tubelab SSE, Vroemen Diva Superiore ER4, Jordan JX92S VTL, 47 labs 0647 CD, Aqvox DAC, Rowen Absolute pre / psu / power amps, BG Neo3 / Betsy / Eminence A15 open baffles