1.   Small gauge inductors

2.  More baffle step compensation

3.  The Sonic Cap

None of these three items are monumental.  The combination of all three is not monumental.  However, they do contribute to a somewhat significant change noteworthy of the 1801b designation.

Small Gauge Inductors

This is the least significant change to the 1801 crossover.  I experimented with small gauge inductors in the 1801 crossover circuit where a large gauge inductor + resistor were placed. There are a couple places in the 1801a crossover where resistors are directly in series with inductors. Both of these locations are in parallel with the driver. The values of the the inductors made a resistor substitution viable. Rather than use a 14ga inductor and resistor, I use a small gauge (very small gauge) inductor and no auxiliary resistor. This works because the very small gauge inductor has the resistance built-in. This implementation is very similar to the resistor wire found in older automotive ignition systems. The resistance is built into the wire.  After implementation, I measured the frequency response of the woofer, tweeter, and overall. I also measured the phase. Then I listened to the results. The results were quite interesting. The measurements were virtually identical. There was a smooth .5 db change across the MLS sweep. This change in db is not audible.  It might have even been my jig shifting slightly.  The listening experience was a little different.

When using the small inductor in the woofer circuit I noticed a slight audible improvement. I was puzzled slightly by this. I decided to check it again. Yep, it was slightly better. I am not totally certain why it was better, but I can speculate. First, it could be that the DCR (DC resistance) of the small inductor works better in the circuit. When Dennis Murphy designed the initial crossover for the 1801 he modeled 1 ohm of resistance in this circuit. The large inductor + resistor combination yielded about 1.15 ohms of resistance. The small inductor yielded .86 ohms of resistance. It could be that the speaker liked .86 ohms better. Second, it could be that the small gauge inductor was a theoretically better resistor. The inductor has more surface area than the resistor and should (??) dissipate heat better. Third, it could be that I eliminated a connection when eliminating the connection between the inductor and resistor. I believe the first reason is the most likely cause for the improvement, but this is only a guess. I could easily create a good marketing campaign from any of these reasons, but I won’t. I really don’t know why it sounded slihglty better.  Nonetheless, it did.

With the small inductor substitution in the tweeter circuit I noticed no audible change. I really can't explain why, but it is nonetheless the case.  I have listened to the small gauge inductors in my own living room for several months now, and am very happy with them.  Upon mentioning the use of small gauge inductors to Dave Elledge, he commented,

Another gentlemen of very solid knowledge affirmed my thoughts about using small gauge inductors when resistance is desired.

The overall result is that the small gauge inductors are here to stay. Despite their very wimpy appearance, they perform wonderfully. They provide a slight sonic improvement in some places. They are also less expensive. It is a win-win situation. I can only speculate about why these small inductors work equal/better than a much more expensive component selection. There are good reasons for many aspects of this issue. I am satisfied that the smaller components will remain in my crossover as the design(er) dictates a need for higher DCR in that circuit.

More Baffle Step Compensation

A few folks commented that the 1801a crossover sounded slightly lean in the midbass region.  Dennis and I took this to heart, and agreed that this was a viable complaint.  We heard it too.  Getting this region of response correct is a little tricky though.  It requires a fair amount of room integration. 

Baffle loading takes place on a speaker much like horn loading.  The difference is that most speakers don't have horns, they have flat baffles.  The baffle loading is less significant, but present.  There is a 6db rise in response from drivers mounted on a flat baffle.  The db rise related to the size of the baffle and the frequencies involved. 

As the frequency length approaches 1/4th the baffle width, a loading will take place.  The onset is gradual, and generally begins at about 200hz.  By 800 hz on most baffles, the loading is fully present.  This means that electronic means must be used to eliminate this loading.  This is fairly easy to accomplish in the crossover, but there is another concern.  It is the room.

The room walls, especially the back wall,  will theoretically and practically extend the size of the loudspeaker baffle.  This effectively lowers the onset frequency of baffle loading.  For ease of discussion I will only address the back wall.  A larger theoretical baffle lowers the frequency of the baffle loading and increases mid-bass response.  The significance of coupling between the loudspeaker baffle and the back wall depends on the distance between the baffle and the back wall.  If the loudspeaker is closer to the back wall, there will be more coupling.  If the loudspeaker is further from the back wall there will be less coupling. 

Adding more baffle step compensation electronically in the 1801b is probably the most significant change for the 1801.  The result is that the 1801b could be placed further into the room without the need for back wall coupling to increase mid-bass energy.   It obviously sounds slightly warmer in the mid bass.  A few folks tried the 1801a placed well into the room.  The consensus was that the mid-bass suffered an intolerable serious drop in performance.    The 1801a was slightly lean in the mid-bass, and placing it well into the room exacerbated the lean condition.  The 1801b chops about 2db off the midrange & high frequencies.  This brings the mid-bass into a richer balance.  I find that male vocals sound slightly warmer.  This is because the fundamental frequency of their voice is more projected.  This is most obvious with Andrea Bocelli.

This change is not monumental, but it is significant.

The SonicCap

The last change is an easy one, but deserves mention too.  Henceforth, the SonicCap will be used in all 1801 crossover locations.  There are a few reasons for this.  The most significant is that the SonicCap is that good.  It is clean, quiet, and simply makes music.  Since I generally don't make own colorful comments, I'll interject some from Ed Majewitz.  Ed has many years in the hifi business and copious experience.  Ed's remarks pertain to an application in the MBOW1.

Upgrading from the 1801a to the 1801b

I have commented on this issue with a few folks offline.  I will be sending some crossovers to Jack Giuffre for his comments.  I am not sure that a crossover swap to the 1801b is money well spent.  IMO, if it cost $50, the upgrade would be a no-brainer - do it!  IMO, if it cost $250, the upgrade would also be a no-brainer  - don't do it!  Jack will hopefully provide his comments in by the end of March 2003.

The summary/cost will be thus:

There are only 2 capacitors retained from the 1801a crossover.

An upgrade from the old crossover to the new crossover will cost $115+$10s/h.  This retains 2 capacitors from the previous crossover.

If you don't have any SonicCaps and want them, it will cost $145+10s/h. This retains only 1 capacitor from the previous crossover.

The cost reflects the retail cost of components + $15 for development.

Jack eventually provided feedback and his sentiment is stronger than mine.  I thought the 1801b change/improvement was relatively subtle.  He didn't.  Jack conveyed a notable improvement in character and quality with the 1801b crossover.