Transmission line bass cabs

Discussion in 'Amps and Cabs [BG]' started by Bass Unique, Nov 3, 2011.


  1. steve_rolfeca

    steve_rolfeca Supporting Member

    I think that you're overstating BFM's position, and possibly missing an important point. I don't think that Bill is arguing for testing without listening at all. I think he's simply making the valid point that testing and listening should go hand in hand.

    As Bill points out, a correctly-interpreted impedance plot can point out issues like cabinet leaks. What if you have a problem that isn't severe enough to sound obnoxious, but does compromise performance? Running an impedance plot can serve as a confidence check, to ensure that your cab is ready for critical listening.

    Otherwise, a build flaw might lead your ears into believing that you needed to adjust your port tuning, when in fact it was right in the first place, and some other factor was at play...
     
  2. Bass Unique

    Bass Unique

    Nov 3, 2011
    Wiltshire
    Below is the impedance plot I have taken this morning. I used two multimeters. One to make the voltage stay at 2.83 V (+/- 0.02) and the other to measure current. So, I have a plot showing the impedance for a constant voltage input (I adjusted for each point to maintain voltage).

    I used a 5 Hz resolution up to 100 Hz and then 10 Hz above 100 Hz. In order to accurately locate the peak, I used 1 Hz resolution between 50 and 60 Hz. Peak is at 56 Hz.

    Impedance 1210T.jpg

    I have to say I was not expecting a single peak at 56 Hz as this was not how it was designed. I thought it would have a broader much smaller peak. There appears to be a broad dip around 30 Hz (I assume this to be the reflex tuning as the line should not be tuned this low). The line tuning could be lower than I expected but I was expecting another dip around 55 Hz.... not the case!

    I leave you to look at the curve and make comments....
     
  3. Bass Unique

    Bass Unique

    Nov 3, 2011
    Wiltshire
    Just used Win ISD to try and recreate the curve. looks like a reflex with 110 litre box and tuned to 26 Hz (or something like a sealed box 86 litres which actually equates to the actual volume of the whole box....)

    According to my model, it cant be tuned to 26 Hz through either a reflex or line mode. Any ideas?

    Impedance model.jpg
     
  4. billfitzmaurice

    billfitzmaurice Commercial User

    Sep 15, 2004
    New Hampshire
    Owner, Bill Fitzmaurice Loudspeaker Design
    You need to take the chart down to 5Hz and up to 200Hz, and have a standardized half-space measurement to be sure of what you have. Otherwise you might end up like Karlson, who went to his grave thinking the design bearing his name was a horn, when in fact it was a 6th order series tuned bandpass. Marketed as the former it was at best considered an oddity. Marketed and perfected as the latter it might have been successful.
    +1. You can't fine tune a design if you don't know what it's doing, and there's no worse tool than the ear for that purpose. Not that you can't use the ear alone, but it takes far longer and sacrifices many trees in the process.
     
  5. Bass Unique

    Bass Unique

    Nov 3, 2011
    Wiltshire
    Right - according to my vented only Win ISD model, if I consider it to be a 39 litre enclosure (rear chamber only) with a long low tuned port, 370mm wide by 80mm high on average over its length, then this produces a tuning frequency to be 39 Hz at 1.35m long port/line.

    According to Win ISD, this should give me an impedance peak at just over 80Hz (not 56 Hz actual) and a rising peak around and below 20 Hz. This model has good control of the cone as well for power handling and does not approach 9mm down to the B string.

    For the model to work as a reflex to explain the impedance peak at 56 Hz, the box has to be the full 86 litres (the volume of the chamber plus the line) with the same 1.35m long port. However, this model is supposed to consider the port volume separately from the chamber. According to Win ISD, the box is appearing larger than it really is so that I am gaining the volume in the port AND it acts as a port. So, I seem to be getting the benefit of the full box size plus the volume in the port. Port tuning is then 26 Hz. According to Win ISD, this tells me my speaker will fart out at 225w around 42 Hz. However, the speaker is better controlled than this with nothing like 9mm excursion with my amp flat out at 300W 8 Ohms.

    So, comparing the modeled outputs, if it is considered the larger box for tuning and impedance and the smaller box for cone excursion (which the Win ISD model cannot support), around 32 Hz the smaller reflex and sealed box are similar (with reflex adding more in the 35 to 120 Hz band as you would expect). The impedance peak appears to look like a sealed box (as BFM predicted) but an output 5 to 6 dB higher at the low end.

    Plenty for you to digest there!
     
  6. Bass Unique

    Bass Unique

    Nov 3, 2011
    Wiltshire
    Bill, I did not display that, I could add it to my graph. Interestingly, my only other explaination which was it may appear as a 6th bandpass would require a peak aroudn 130 Hz. it was pretty flat from 100 to 200 Hz with no sigificant ripples.

    I took the impedance down to 10 Hz as well (thinking the amp may not like much lower than this). The voltage drops off fairly significantly below 20 Hz so I had to make significant adjustments - this may be a Markbass design intent to protect the speakers from subsonic excursions when slapping. These was a slight rise to around 20 Ohms by 10 Hz (unlike a sealed box).
     
  7. billfitzmaurice

    billfitzmaurice Commercial User

    Sep 15, 2004
    New Hampshire
    Owner, Bill Fitzmaurice Loudspeaker Design
    I would not use an instrument amp for critical measurements. I calibrated my HolmImpulse with the testing amplifier in the calibration loop to be sure that there's no coloration of the result from the amp.
     
  8. Bass Unique

    Bass Unique

    Nov 3, 2011
    Wiltshire
    I would agree for a lab test, it would be easier to use a lab amplifier which should drive down to 5 Hz easily. However, by operating on discrete tones one at a time and then adjusting the voltage and monitoring current at the same time, there cannot be "colouration". There is more likely to be errors in my multimeters. However, we are looking for relative changes only. The absolute value appears to be approximately 2 Ohms higher than I would have expected and I cannot believe the crossover accounts for all of this.

    So, we have a small rise around 10 Hz dropping to the system resonance around 25 to 35 Hz and then a peak at 56 Hz (coinciding with more movement of the cone) dropping to a flat area between 100 and 200 Hz.

    I need my response graph now to see if it matches the 86 litre low tuned prediction. This will have to be at the end of the week (other work demands my time).
     
  9. Bass Unique

    Bass Unique

    Nov 3, 2011
    Wiltshire
    I hate to say this but it would have to be a massive leak to significantly change a port tuning... I always block the port and see if the cone then behaves as a sealed box to check for leaks (cone returns slowly in a sealed box when pressed - in a vented, it immediately returns to its rest position). Its a simple test to prove there is no leak.

    If there is a small leak such as a tiny crack or badly fitted terminal block which only just produces an audible chuffing, I would be surprised if you could find it on an impedance plot for a vented box.... it would have to be significant in terms of cross section compared to the port (ie about 1/10th of the cross section and above may start changing tuning slightly).
     
  10. im no bill fitzy but i hope ill do ok
     
  11. will33

    will33

    May 22, 2006
    austin,tx
    Following along here...it's interesting. Some of the finer bits of the design tweaks here are a bit above my knowledge base to have much for helpful suggestions. I do notice that while it seems to have one impedance peak, more sealed cab style, that peak looks pretty high to be sort of "tube amp friendly" if that's a consideration.

    Cool work, I'm curious to see how it turns out.
     
  12. billfitzmaurice

    billfitzmaurice Commercial User

    Sep 15, 2004
    New Hampshire
    Owner, Bill Fitzmaurice Loudspeaker Design
    I modeled something similar in HornResp and you get a similar impedance value at 56Hz, with a lower peak around 12Hz. If your gear won't allow an accurate measurement below 20Hz that could be why you're not seeing it.
     
  13. Petebass

    Petebass

    Dec 22, 2002
    QLD Australia
    12Hz? What's going on down there if you're not using an octave down pedal on a a note slightly lower than a low B? :)

    I jest. I realise this is a theoretical conversation and I'm actually quite enjoying it.
     
  14. billfitzmaurice

    billfitzmaurice Commercial User

    Sep 15, 2004
    New Hampshire
    Owner, Bill Fitzmaurice Loudspeaker Design
    Nothing's going on down there, it's just where you'll see an impedance peak with a VB that has an Fb in the vicinity of 25Hz, which is one possibility for what the OP has here.
     
  15. Bass Unique

    Bass Unique

    Nov 3, 2011
    Wiltshire
    Bill, yes there is a small hump but I would not call it a "peak" as such as it was pretty smooth and flat. It seems to start to rise around 20 Hz and is just tailing off by 10Hz. We can see more than half the "hump" which matches your model so by 10 Hz and below, it should be dropping further.

    I think if I need to go lower than 10 Hz, I would use our Crown amp at work which should go down to around 2 Hz pretty flat. I am sure the MB would do it but it is requiring a large voltage change by 10 Hz to keep things the same. Without an ocilloscope here at home I cant see if the waveform is skewed so would wait for a work trip to do this one.

    Anyway, it closely matches your model which is good news. Did you model show a hump/peak around 130 to 150 Hz? I was expecting one but there was only a flat zone up to 200 Hz.

    I think 25 Hz is a little low for a bass guitar as a tuning frequency. This could probably be raised to the low 30's without a major change to the low end and it may increase output efficiency further for more fundamentals from the bass.

    This could be done in a number of ways but if it is possible to increase efficiency and reduce size then this cannot be a bad thing.

    1) The line length could be different - If this is a reflex effect, changing the length slightly would probably not be a bad thing to reduce size.

    2) Indrease the line cross section or reduce the taper to increase terminus area - this would tend to increase the tuning frequency.

    3) I could reduce the volume at the rear of the driver. However, my initial calculations suggest that if I reduce the volume, I also have to reduce throat area to maintain tuning which would further reduce efficiency.

    It would be an advantage to reduce the height of the unit slightly and fit a smaller mid so a preferred size of material can be used to make it cheaper to build. I would therefore probably adjust the line length and taper to raise the tuning frequency to around 31 Hz. This should maintain an extended bass shelf without a significant loss of efficiency (1 to 2 dB around 30 Hz but may increase the 40 to 60 Hz range). I may slightly increase the cross section of the line as well.

    Thanks for the discussion everyone. I hope to post a 1/3 octave output response and an updated impedance plot Friday.
     
  16. billfitzmaurice

    billfitzmaurice Commercial User

    Sep 15, 2004
    New Hampshire
    Owner, Bill Fitzmaurice Loudspeaker Design
    It does, but that would probably have been tamed by your damping. HornResp doesn't allow for modeling damping in anything other than an IB. It also shows that your measured results are likely boundary assisted.
     
  17. Bass Unique

    Bass Unique

    Nov 3, 2011
    Wiltshire
    Do you mean the impedance curve was boundary assisted? What aspect of the curve do you believe shows this?

    Having read some other threads on TB which we have both posted in, it is good to know that my damping was sucessful in attenuating the low 100 to 200 Hz region without significantly affecting my low end output ! :D
     
  18. billfitzmaurice

    billfitzmaurice Commercial User

    Sep 15, 2004
    New Hampshire
    Owner, Bill Fitzmaurice Loudspeaker Design
    No, the response curve, although boundaries can show up on an impedance sweep too. As for the midbass spike, it's very narrow, and a manual sweep might have trouble finding it. But suppressing it shouldn't take more than the damping you have anyway.
     
  19. Bass Unique

    Bass Unique

    Nov 3, 2011
    Wiltshire
    I see - the response curve. We know that the first curve was not a true half space as it would have included the floor and the baffle step. Watch this space and one will be presented soon.

    There was a very small hump on the impedance curve at around 140 Hz which I focussed on at 1 Hz intervals but there was still no peak. Most likely damping.
     
  20. Bass Unique

    Bass Unique

    Nov 3, 2011
    Wiltshire
    Ok - This is the final one - used a proper measuring amplifier, small signal level, and the two microphone method suggested by BFM outdoors 30m+ to the nearest building.

    Measured three times and then averaged the results (just incase of something spurious). The results are third octave data which I then normalised to a 1 watt (2.83 volt) signal using a 100 Hz tone.

    Actual measured response outside third octave plot.jpg

    Very interesting and not quite what I expected but confirms what I suspected from my experience of this design.

    The white plot is the actual measurement. I joined up the dots to create a "curve". Remember this is for guidance only and for interest using a hybrid TL/reflex.

    Based on my 90 litre box (if this were conventional), I have modelled a EBS reflex tuning and obtained the red trace. This shows an extended bass shelf response as you would expect for a large box for this speaker. The green trace was taken from Greenboys 12 sub trace on the website and then normalised to 1 watt (for reference only as a good reflex).

    The actual output level is lower than I expected around 30 Hz but by 40 Hz, it matches the model. This is probably about right according to the impedance plot showing a resonance arond 25 Hz in this system so more bottom end. However, a reflex model of an extended bass shelf should show reduced output compared to the 12sub between 50 and 100 Hz. It would also show less cone control and a risk of over excursion.

    The actual response seems to show similar output to Greenboys 12sub in this region and from what I can see, similar cone control. This means that the system responds at the low end somewhere between a EBS/sealed box and in the mid bass similar to a conventional reflex box!

    I also measured the port just inside the terminus at low levels. This shows a much broader response than I would expect from a typical "port" with strong output from 25 Hz gently rising to 80 Hz peak then slowly dropping and falls sharply away above 160 Hz.

    I had a quick look at the broadband emissions as well and it is not particularly sensitive (as you would expect from a single 3012LF), its around 95 dB/W up to 1800 Hz and then due to the large mid response, there is a rising response up to around 3 kHz (this region is around 97/99 dB/W) which offsets the reduced output around 30 degrees to the axis. The tweeter is around 96 dB/W (the manufacturer of this item suggests 97).

    So, I will modify the mid next to smooth the 2 to 3 kHz region a little more and then the design is pretty much ready if anyone wants a go at building one. (no charge for this design as you have all seen it develop but don't all PM me at once... I will create a PDF to e-mail out).
     
  21. Primary

    Primary TB Assistant

    Here are some related products that TB members are talking about. Clicking on a product will take you to TB’s partner, Primary, where you can find links to TB discussions about these products.

     
    Jul 23, 2021

Share This Page

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.