Has anyone built one of Martin King's ML TL's to use with an FE206 and a SI t-amp? What BSC values did you use?
I started out with a 10ohm resistor and a 3mH inductor, along with his suggested 15ohm Mills resistor and 3uF cap in parallel with the driver, but that seems to kill my mid range a bit much.
Any suggestions? I'd rather not mess around _too_ much with the circuit, since I mounted them inside the cabinets (my room is too cramped at 10x12' to put the circuits outside). How would it sound with the BSC totally removed?
-Pat
I started out with a 10ohm resistor and a 3mH inductor, along with his suggested 15ohm Mills resistor and 3uF cap in parallel with the driver, but that seems to kill my mid range a bit much.
Any suggestions? I'd rather not mess around _too_ much with the circuit, since I mounted them inside the cabinets (my room is too cramped at 10x12' to put the circuits outside). How would it sound with the BSC totally removed?
-Pat
I would say that the only real way would be to try it and find out. How long has the speaker been breaking in for?
Nate
Nate
Experiment with different series resistor values. The parallel RC is a Zobel network to flatten out driver impedance, don't mess with that. And the inductor value determines the frequency where BSC comes in, it is based on the baffle width.
The series resistor serves two functions, first to increase driver Qts to make the FE206E suitable for the vented cabinet, second to determine the amount (dB) of BSC correction.
A cute implementation is using a rheostat (i.e. variable resistor) to give you on-the-fly adjustability:
http://www.quarter-wave.com/General/BSC_Variable.pdf
The series resistor serves two functions, first to increase driver Qts to make the FE206E suitable for the vented cabinet, second to determine the amount (dB) of BSC correction.
A cute implementation is using a rheostat (i.e. variable resistor) to give you on-the-fly adjustability:
http://www.quarter-wave.com/General/BSC_Variable.pdf
I don't like to be the bringer of bad tidings, but this might be tricky. The Tamp is not overly powerful, and while it sounds sweet enough when being driven within its own very moderate limits, push it hard (into clipping) and it quickly starts to sound vile. Now, with solid state, you're probably going to need close to the upper end of the values Martin specifies to bring the naturally rising response of this driver down, and this is going to kill efficiency and require more grunt from the amp: not power so much as current -the FE207E might have been a better bet in these circumstances. How long have you had the drivers? They'll need at least 250 hours on them before the start to give of their best.
With the circuit removed, and using a normal solid state amp (like the Tamp) you will have pretty much 0 bass, a rapidly rising mid-range, and a harsh treble. A few ideas; unfortunately, few are cheap fixes.
1) Junk the Tamp (lovely though they are) in favour of something else with more power and current-pumping ability. Cheap idea: a 2nd hand NAD 3020 would work wonders, or you could build one yourself if you feel up to it -the circuit diagram's are lurking in many places on the web. Another alternative would be to try going down the current-drive approach Nelson Pass has done with his First Watt amps, though I personally think that the 5% THD of the F1 at full output is way too much for my taste -it's a 0 feedback design however, and the circuit could probably be modified to include some, which might help with that. This sort of amp will minimise the values for the circuit. Final thought -a couple of gainclone monoblocks. I don't normally like monoblocks, but this route would again allow better current driving from the chip amps, and reduce the risk of them going into clipping (same problem as the Tamp, but they're usually more powerful and versitile.) You can circumvent this problem by paralleling a couple of these power-op-amps, which is cheap and easy enough and will give all the grunt you'd need.
2) Replace the 206E with the 207E
3) Feel like a wide baffle cabinet? Retain the existing cabinet, but extend the sides. Quick experiment for you to try: tape a couple of pieces of MDF the same width as the cabinet to the sides of the front baffle, perhaps angled slightly backward, so the front baffle the driver 'sees' has in effect been tripled in width. You'll find you need far less BSC than before. I've tried this, and it works very well. Have a look here; it's a multi-way, but that's not important, and anyway, it makes interesting reading. The pictures of the experimental setups will give you a few other ideas. http://www.troelsgravesen.dk/diy_loudspeaker/IBL.pdf
Hope some of this helps
Best
Scott
With the circuit removed, and using a normal solid state amp (like the Tamp) you will have pretty much 0 bass, a rapidly rising mid-range, and a harsh treble. A few ideas; unfortunately, few are cheap fixes.
1) Junk the Tamp (lovely though they are) in favour of something else with more power and current-pumping ability. Cheap idea: a 2nd hand NAD 3020 would work wonders, or you could build one yourself if you feel up to it -the circuit diagram's are lurking in many places on the web. Another alternative would be to try going down the current-drive approach Nelson Pass has done with his First Watt amps, though I personally think that the 5% THD of the F1 at full output is way too much for my taste -it's a 0 feedback design however, and the circuit could probably be modified to include some, which might help with that. This sort of amp will minimise the values for the circuit. Final thought -a couple of gainclone monoblocks. I don't normally like monoblocks, but this route would again allow better current driving from the chip amps, and reduce the risk of them going into clipping (same problem as the Tamp, but they're usually more powerful and versitile.) You can circumvent this problem by paralleling a couple of these power-op-amps, which is cheap and easy enough and will give all the grunt you'd need.
2) Replace the 206E with the 207E
3) Feel like a wide baffle cabinet? Retain the existing cabinet, but extend the sides. Quick experiment for you to try: tape a couple of pieces of MDF the same width as the cabinet to the sides of the front baffle, perhaps angled slightly backward, so the front baffle the driver 'sees' has in effect been tripled in width. You'll find you need far less BSC than before. I've tried this, and it works very well. Have a look here; it's a multi-way, but that's not important, and anyway, it makes interesting reading. The pictures of the experimental setups will give you a few other ideas. http://www.troelsgravesen.dk/diy_loudspeaker/IBL.pdf
Hope some of this helps
Best
Scott
paanta said:
Everything that Scottmoose said. I personally thing that all of the Fostex FExx6E drivers are unsuitable for resonant cabinets. That these drivers shouldn't be forced into MLTL's is not news, but the mystique of the FE206E seems overpowering. If you can afford the cost, get a pair of FE207E's.
But. now you have the FE206E's. You need to play with the values of the series resistor (you didn't specify what you used here) and the bypass resistor. You need to do this with the circuit outside the cabinets. Temporarily remove the filters and wire the drivers directly to the binding posts. Then wire the filters between the amp and the binding posts.
The easiest way to test resistors is to wire the low side of an L-pad in place of each resistor. This gives you a 0-8 ohm variable resistor. If you need more, pad up the L-pad with a fixed resistor.
Dumbass said:
Well, no.
There are two resistors involved here, a series resistor ahead of the filter and the resistor in parallel with the inductor. The series resistor sets the effective Qte of the driver at low frequencies and therefore sets the bass end of the FR curve. The parallel resistor is out of the circuit at low frequencies and the only series resistance offered by the filter is the DCr of the the inductor. At high frequencies, once the inductor rolls out, the effective resistance of the filter is the sum of the two resistors. This is the resistance that sets the high end of the FR curve.
Bob
Martin actually only has a single resistor in parallel with the inductor, doesn't matter if the series resistance goes there or in front of (or behind) the BSC, everything is happening below the BS cutoff anyway.Bob Brines said:
I'll add :
..... as long as you account for it when sizing the BSC components.
If you place the series resistance after the BSC circuit, you need to account for it when calculating the size of the inductor and parallel resistor. I have been including it in with the parallel resistor to simplify the design and reduce the part count. But what ever works best for the designer is the right answer.
If you have not already, take a look at my latest version of the BSC Sizing article on my site. It was improved this past summer.
..... as long as you account for it when sizing the BSC components.
If you place the series resistance after the BSC circuit, you need to account for it when calculating the size of the inductor and parallel resistor. I have been including it in with the parallel resistor to simplify the design and reduce the part count. But what ever works best for the designer is the right answer.
If you have not already, take a look at my latest version of the BSC Sizing article on my site. It was improved this past summer.
Dumbass said:
Well, you picked your moniker!!
Draw out the circuit -- Amp, series resistor, series inductor paralleled with a resistor, driver.
Now assume the series resistor is 2 ohm and the bypass resistor is 6 ohm. At DC, the inductor is a short -- 0 ohm -- so the driver sees 2 ohm of series resistance.This shapes the bottom end of the FR curve. At 20k Hz, the inductor is an open -- infinite resistance -- so the driver sees 8 ohms. This shapes the top end of the FR curve. The value of the inductor sets where the transition occurs.
Bob
Now, now, no need to get snarky.Bob Brines said:
You and Martin are correct, I didn't account for the need to change inductor value if resistance is between BSC and driver. Nevertheless, as Martin pointed out, he only has a single series resistor in his design, and it can be varied to yield different low-end response.
Just a word about Zobel network.
All the MJK's designs are great and it's rally great he is sharing all this knowledge with us. But I just found Fostex to sound much better without recommended Zobel network than with it. Zobel kills the openes, spacialness and directness of the sound, but that might be just mine opinion.
By the way, have anyone compared FE 206E with FE 207E? I suppose 207 is more likely to have good bass from these enclosures.
Bartek
All the MJK's designs are great and it's rally great he is sharing all this knowledge with us. But I just found Fostex to sound much better without recommended Zobel network than with it. Zobel kills the openes, spacialness and directness of the sound, but that might be just mine opinion.
By the way, have anyone compared FE 206E with FE 207E? I suppose 207 is more likely to have good bass from these enclosures.
Bartek
- Status
- Not open for further replies.