Recommended mid and tweeter

[Loren42]

If one has a 2235H they wish to use for a woofer, what might be a good recommended midrange and tweeter combination to compliment the 2235H?

This is for HiFi and driven by a tube amp,

[Robh3606]

Why are you asking?? I thought you had a 3 way built already. If you are looking at a clasic JBL system then a L300 clone is the obvious choice. If you want more complexity that a 4344 fits the bill. Depends on what you are looking for. A two way the 4430 is the obvious choice.

Rob:)

[Loren42]

Why are you asking?? I thought you had a 3 way built already. If you are looking at a clasic JBL system then a L300 clone is the obvious choice. If you want more complexity that a 4344 fits the bill. Depends on what you are looking for. A two way the 4430 is the obvious choice.

Rob:)

Yes, but it is problematic. I used an Audax PR170M0 for the mid at someone else's recommendation and had someone turn a crossover design based on that.

Turns out the Audax has an SPL of about 100 and the crossover uses a lot of resistors to pad down the mid and it is still almost 4 to 6 dB too loud. When the crossover calls for a 45 Watt resistor I start to cringe thinking about all the power that needs to be bled off.

So I am just looking at all the alternatives (including further padding of the Audax).

[Ruediger]

Yes, but it is problematic. I used an Audax PR170M0 for the mid at someone else's recommendation and had someone turn a crossover design based on that.

Turns out the Audax has an SPL of about 100 and the crossover uses a lot of resistors to pad down the mid and it is still almost 4 to 6 dB too loud. When the crossover calls for a 45 Watt resistor I start to cringe thinking about all the power that needs to be bled off.

So I am just looking at all the alternatives (including further padding of the Audax).

The real problem when padding down a loudspeaker is not the wasted power but rather the loss of damping by the amplifier.

If You had horn drivers You would need to pad them down as well.

The best solution is an active one. Have a look at dbx driveRack and at EV DC-one. That are affordable digital crossovers.

Ruediger

[Loren42]

The real problem when padding down a loudspeaker is not the wasted power but rather the loss of damping by the amplifier.

If You had horn drivers You would need to pad them down as well.

The best solution is an active one. Have a look at dbx driveRack and at EV DC-one. That are affordable digital crossovers.

Ruediger

That is an option. I have a Ashley XR2001.

However, I just got done building a splendid tube amp and I really don't want to build one or two more, plus the tangle of extra wires right now.

It would be helpful if I could model the crossover in HT's Bass Box/X-Over Pro, but the crossover was designed by someone using LEAP and I can't seem to replicate the circuit topology in my program.

Next step is to drag one of the beasts outside into the yard and perform a sweep. That should give me a better idea of what factors are room artifacts and what factors are really attributes of the speaker.

I am thinking of laying the cabinet face up so that I minimize reflections from trees, fence, and the house in my yard. Unfortunately, it is going to be a bit windy here for a few days, but that is better than the forcast for Washington, DC. :D

[Flaesh]

..Audax has an SPL of about 100 ..
can you use 4*2235 on channel? ;)

[Loren42]

can you use 4*2235 on channel? ;)

Might as well get started punching eight 15" holes into the concrete wall. :D

[Robh3606]

Post the schematic

Rob:)

[Loren42]

Post the schematic

Rob:)

Here is the complete LEAP document that I received. Note, the plot for the 2235H with the 10" long ports was for reference only. The cabinet is currently tuned to 28 Hz. The target was 32 Hz and that was what the crossover was designed for.

http://www.mdbq.net/pyramid/LEAP.pdf

[Ruediger]

I had overlooked the link to the LEAP document, sorry.

What really fascinates me is that such a simple crossover can change a loudspeakers characteristic completely.

Take the 2235H data sheet. Page 1 of the LEAP document does not show the irregularities beyond 1 kHz and the steep descent beyond 2 kHz. Page 4 shows an irregularity at 500 Hz which is not present in the 2235H data sheet.

The Audax midrange shows heavy irregularities from 3 kHz on (Audax data sheet). They are cured on page 1 of the LEAP document.

The tweeter response has a notch at 15 kHz in the tweeters data sheet, this is gone as well.

Need to go now, more tomorrow.

Ruediger

[Loren42]

I had overlooked the link to the LEAP document, sorry.

What really fascinates me is that such a simple crossover can change a loudspeakers characteristic completely.

Take the 2235H data sheet. Page 1 of the LEAP document does not show the irregularities beyond 1 kHz and the steep descent beyond 2 kHz. Page 4 shows an irregularity at 500 Hz which is not present in the 2235H data sheet.

The Audax midrange shows heavy irregularities from 3 kHz on (Audax data sheet). They are cured on page 1 of the LEAP document.

The tweeter response has a notch at 15 kHz in the tweeters data sheet, this is gone as well.

Need to go now, more tomorrow.

Ruediger

Good catches!

I am beginning to feel there are some real cracks in this design. :(

[Ruediger]

I looked a bit around at linearx' s website. One ***CAN*** do loudspeaker measurements in a prototype enclosure, and one ***CAN*** import such measurement files into the crossover shop, but obviously this has not been done in Your case.

Another oddity which jumped into my eyes is that the bass coil's DC resistance is not specified and thus cannot have been taken into account for the bass enclosure. One could at least have specified an upper limit.

The Audax midrange seems to be usable from 400 Hz to 2.5 kHz, and the tweeter could cover the range from 2.5 kHz up.

Frankly: I think it is a waste of time to continue the current approach with the current components.

There are plenty of alternatives, two seem to be very attractive to me.

No. 1 is a 4430 clone.

No. 2 is a two way system with a 2" driver and a beryllium diaphragma (later refinement). There is a company offering be diaphragmas for JBLs.

Ruediger

[Loren42]

I looked a bit around at linearx' s website. One ***CAN*** do loudspeaker measurements in a prototype enclosure, and one ***CAN*** import such measurement files into the crossover shop, but obviously this has not been done in Your case.

Another oddity which jumped into my eyes is that the bass coil's DC resistance is not specified and thus cannot have been taken into account for the bass enclosure. One could at least have specified an upper limit.

The Audax midrange seems to be usable from 400 Hz to 2.5 kHz, and the tweeter could cover the range from 2.5 kHz up.

Frankly: I think it is a waste of time to continue the current approach with the current components.

There are plenty of alternatives, two seem to be very attractive to me.

No. 1 is a 4430 clone.

No. 2 is a two way system with a 2" driver and a beryllium diaphragma (later refinement). There is a company offering be diaphragmas for JBLs.

Ruediger

Poop. Okay. I can see where the LEAP analysis and crossover I got was not a good plan.

I am exploring all routes. Seems a shame to ditch the cabinets after all that work (at least I did not put the koa veneer on them), but if it turns out I am going down a dead end street, then the sooner I stop, the better.

Can you tell me more about option #2? Specifically, about what horn components I need? If there are part numbers or something to help me figure out what I need.

The number 1 option I think I can simply do some research on to see what components are needed there. The number 2 option is a little more nebulous.

Thanks, again.

[Ruediger]

Poop. Okay. I can see where the LEAP analysis and crossover I got was not a good plan.

I am exploring all routes. Seems a shame to ditch the cabinets after all that work (at least I did not put the koa veneer on them), but if it turns out I am going down a dead end street, then the sooner I stop, the better.

Can you tell me more about option #2? Specifically, about what horn components I need? If there are part numbers or something to help me figure out what I need.

The number 1 option I think I can simply do some research on to see what components are needed there. The number 2 option is a little more nebulous.

Thanks, again.

A paper about the #1 option: http://audioheritage.org/vbulletin/showpost.php?p=75815&postcount=1

You need 2425H or 2426H drivers (8 Ohm) and 2344 horns. Do ***NOT*** buy J drivers (16 Ohm), because the xover is sophisticated and cannot easily be adapted to 16 Ohm drivers.

Info about aftermarket beryllium diaphragmas can be found here: http://audioheritage.org/vbulletin/showpost.php?p=276077&postcount=1

A very good 2" horn is the 2397 Smith horn. It requires an adapter. A Smith horn can also be built with relative ease. See: http://audioheritage.org/vbulletin/showpost.php?p=276077&postcount=1

There are several 2" drivers of interest. That is a topic of its own. Maybe somebody else can explain?

I do have a 4430 clone with J-drivers and an active xover. I do also have 2445 and 2450 drivers, and 2390 and 2395 lenses. The differences are small. The 2" drivers sound effortless, and the 2395 is my favourite.

Ruediger

[Robh3606]

Before you throw in the towel try padding the midrange down. You should also call Madisound and tell them your not happy and explain that your mid is too hot. You did pay them right?? Try using their customer service. They can easilly adjust the pad values for you. If they are receptive you can send them a text file of your measurements they can import and use as a reference curve in LEAP.

They used heavily smoothed canned curves. Leap Enclosure Shop has a huge library of them. You need measuements of you drivers in your box for Leap to work unless the canned curves are close to your in box measurements.

Rob:)

[jerv]

Of course, building clones of classic JBL systems and replicating their crossover is the safest and probably the best path. On the other hand - if you want to continue the present approach, these are my recommendations:

- The Audax PR17M0 is a very good sounding mid. In my experience, it is best used between 400 Hz and 1,8-2 kHz. 4kHz is too high: see the resonances (ripples in frequency response from 3kHz), and it will also beam too much. It is far from 100dB/W (that's a gross excaggeration in the product sheet) - more like 95-96dB /1W. You will still need to pad it down, of course, but 45W resistors :screwy: is total overkill. 10W is more than enough.
- I have no experience with the Morel MDT-37. From the spec sheet, it seems that the sensitivity is maybe a tad low to be used with the Audax (only 93dB).
- The midrange is mounted in a separate, narrower box. This makes the bafflestep higher in frequency, and lowers the efficiency in the 400-700 Hz area.
- The crossover seems wrong, with too much attenuation of the tweeter, and probably too much emphasis in the 700-1,5 kHz area. And the 45W resistors is total nonsense.
- My advise: Dump the tweeter, replace with 1" compression driver on small waveguide/horn, for example the XT120, available from usspeaker.com. Sounds very good, with Beyma CP380, B&C DE250 (or even the cheap DE10), or JBL 2425/26.
- Get all drivers measured separately in the actual box (without crossovers, of course). Use less than 20 mS gating and distance of at least 2 meters. Measure on axis, and make several measurements off axis (+/- 15 degrees up/down, +/- up to 30 degrees left/right). If possible, measure outdoors as well.
- Get someone with good software and good working knowledge of loudspeaker filters to model the network for you.

These are complex tasks. If you don't feel up to it, the safest route to a good system is to follow a more well-trodden path. As have been mentioned before in this thread, 4430 or 4344 are obvious candidates, since both have the 2235 bass unit. Numerous threads is to be found in this forum on both systems.

[Ruediger]

You can leave the 45 W resistors. If they can handle that power AND IF THEY ARE PROPERLY MOUNTED they can dissipate the heat more easily and don't get so hot.

Ruediger

[Loren42]

A paper about the #1 option: http://audioheritage.org/vbulletin/showpost.php?p=75815&postcount=1

You need 2425H or 2426H drivers (8 Ohm) and 2344 horns. Do ***NOT*** buy J drivers (16 Ohm), because the xover is sophisticated and cannot easily be adapted to 16 Ohm drivers.

Info about aftermarket beryllium diaphragmas can be found here: http://audioheritage.org/vbulletin/showpost.php?p=276077&postcount=1

A very good 2" horn is the 2397 Smith horn. It requires an adapter. A Smith horn can also be built with relative ease. See: http://audioheritage.org/vbulletin/showpost.php?p=276077&postcount=1

There are several 2" drivers of interest. That is a topic of its own. Maybe somebody else can explain?

I do have a 4430 clone with J-drivers and an active xover. I do also have 2445 and 2450 drivers, and 2390 and 2395 lenses. The differences are small. The 2" drivers sound effortless, and the 2395 is my favourite.

Ruediger

Link #1 appears to require a subscription to AES Library that is beyond my reach.

Links 2 and 3 appear to be the same and lead to a dead post that was pulled.

[Ruediger]

Link #1 appears to require a subscription to AES Library that is beyond my reach.

Links 2 and 3 appear to be the same and lead to a dead post that was pulled.

Aaaargh!!! Sorry, I don't know how to insert proper links into a post. Some userguide somewhere?

You can make the links work: click the link which I supplied. In the upper right corner of the target page there is another link, "Thread: ...link...".

Still one of the links is wrong, it should point to the Smith horn thread but does not. I'll try again: http://audioheritage.org/vbulletin/showthread.php?t=27649

Ruediger

[Loren42]

Before you throw in the towel try padding the midrange down. You should also call Madisound and tell them your not happy and explain that your mid is too hot. You did pay them right?? Try using their customer service. They can easilly adjust the pad values for you. If they are receptive you can send them a text file of your measurements they can import and use as a reference curve in LEAP.

They used heavily smoothed canned curves. Leap Enclosure Shop has a huge library of them. You need measuements of you drivers in your box for Leap to work unless the canned curves are close to your in box measurements.

Rob:)

That is always an option since I paid for the crossover design and the two Audax mids as per their recommendation.

I find that I want to gather as much information as possible and also try to quantify my cabinet's response so that I can ask intelligent questions.

Given the product I got from them the first time I am not confident that they will apply the required diligence the second time. Sort of like asking the locals for directions. They will get you just far enough so you won't come back and ask them again.

I guess this is one of the big reasons that drive us to DIY. I have Bass Box Pro and X-Over Pro. Might be time to learn better how to use them if they can do an acceptable job. The problem is that the circuit topology of the LEAP design is not compatible with the approach and limitations of X-Over Pro, so I can't easily just tweak the LEAP design with the tools I have.

I also do not have the actual driver response curves in X-Over Pro's library. I assume I will need to measure them with something like WT3 Tester or is there a better tool or method?

You mentioned measuring the response in the box without crossovers. My concern is that the measurement itself is flawed because of nasty room acoustics. I really don't have facilities (anachronic chamber) at hand to do accurate measurements without room artifact. I need a clever way around that.

[Robh3606]

You can't get the LEAP crossover into here?? The way they show the Leap schematic is all the Series and Parallel components in different legs. For example the 2 sereis resistors in the tweeter could be combined into a lump value for RP1. You may even be able to fool it by using inductor as resistors using very low inductance values and the resistor value as the DCR as another example.

Rob:)

[jerv]

You mentioned measuring the response in the box without crossovers. My concern is that the measurement itself is flawed because of nasty room acoustics. I really don't have facilities (anachronic chamber) at hand to do accurate measurements without room artifact. I need a clever way around that.

Very few have access to an anechoic chamber, but you can get ok measurements in ordinary living rooms. To get around the room reflections interacting with the driver, you must gate the measurements. A very good freeware software package for this, is http://www.diyaudio.com/forums/multi-way/144984-holmimpulse-measuring-frequency-impulse-response.html

[Mr. Widget]

I really don't have facilities (anachronic chamber)Can an anachronic chamber be used for time travel? Perhaps Ian needs one of these? :D


Back to the point at hand, I have had very good customer service from Madisound. I would explain your situation and see how they might be able to help you.


Widget