ScanSpeak Discovery 3-Way Classic
Copyright 2014-20 © Troels Gravesen


Before we start: You cannot use any other drivers than those used here without needing a new crossover and you cannot use any other baffle design (e.g. all flat) without needing a new crossover, and I can't help you here. Please read this file before writing. Should you for some reason dislike the ring-radiator you may use the D2604/830000 by increasing attenuation a little due to 2 dB higher sensitivity.

This speaker has been on my to-do list for a long time! So long that the ScanSpeak Discovery series wasn't developed yet. What I had in mind was the ScanSpeak Classic series from 22W, 13M and maybe the 9500 tweeter. However, the 13M drivers are no longer available and with current range of Discovery midrange drivers I don't miss them.
When I launched the SEAS 3-Way Classic I never expected this speaker to be built in any major number due to the classic Seventies design. However, builders response proved me wrong.
Please DO NOT ask about the sonic differences between the two. It's been quite some years since I made the SEAS version.

What the ScanSpeak 22W/8534G00 bass driver has the SEAS CA22RNX doesn't, is a higher mechanical Q due to the fiber glass voice coil former, delivering better low-level detail and transient response. The 10F/8414G10 is a magnificent midrange driver bridging the gab between the 22W bass and classic R2604 ring radiator. This R2604 tweeter is already a classic and I don't see why it shouldn't be around even longer than some of the ScanSpeak Classic drivers, which are well into their third decade.

This ScanSpeak 3-way Discovery features an all-LR2 filter from a stepped baffle and delivers an excellent level of transparency. The midrange and tweeter levels can be adjusted to personal taste by attenuation resistors. Small 4" midrange drivers always sounds different from the common "6+1" set-up due to better dispersion in upper-mid, thus some adjustments can be necessary depending on personal taste, your room and not least source of music and amplifiers.

For both midrange and tweeter series capacitor I have used Jantzen Audio Superior Z-caps due to enhanced transparency and this option and one with Standard Z-caps will be shown. Superior-Z caps don't come cheap, but if you want to hear what these Discovery drivers can really do, there's no way around super caps.

System sensitivity: 88 dB/2.8V, 1 meter. Impedance: 8 Ohms. Minimum impedance: 5.5 Ohms. Points of crossover: 800 and 3000 Hz, 2nd order LR. Time-aligned tweeter for improved transient response and phase integration.
Any change to front panel design and you need and new crossover and I can't help.


Click images to view large
View left image at 3200 pixel for high-res screen


Useful links:
Please read these files before e-mailing:

You cannot change cabinet front panel dimensions and drivers' placement without needing a new crossover - and I cannot help.
You cannot use any other drivers with the crossover shown here.
Please read these files before e-mailing:



Click image to view large

Download driver specs: R2604/832000    10F/8414G10    22W/8534G00

I decided to go for the 8 Ohms version of the 22W driver making a better partner for tube friends. The 4 Ohms version might have added another 3 dB to system sensitivity, but I reckon impedance more important here. These Discovery bass drivers feature high mechanical Q, a promise for dynamic bass that doesn't "die" at low levels. These Discovery 8" drivers are quite large, actually making some 240 cm^2 membrane area from my calculation. Some 10% more than the average 8" driver.
The 10F middriver was tested here. This is one smooth driver and I'd like to use it also in a small lap-top micro system with one of the small 3/4" neo domes from ScanSpeak, e.g. D2004/602000. Some day.
For treble the R2604/832000. I know this tweeter so well having used it numerous times over the years. Unbeatable for the price to my thinking.


Just because the 10F can do it all up to 17-18 kHz doesn't mean we can take the point of crossover between mid and treble up above the main treble range, e.g. 6-8 kHz. Technically we can, but it doesn't sound as good as going down to around 3 kHz. Dispersion is the thing here. The 10F cone has a diameter of 68 mm meaning it will start beaming around 5 kHz, thus we go a little below that point. Like the SEAS 3-Way Classic, this 3-way ScanSpeak features an even simpler crossover due to the stepped front panel allowing true LR2 filter to be realised (2nd order to tweeter). It can be revealed that C2021 is 22 uF and C1021 is 8.2 uF ands I strongly suggest these two caps being Superior-Z caps. These are the two most critical components in the crossover delivering all he transparency we can have from these drivers -and there's a lot to be had!

The Sound

There were three surprises for me when I first fired up these speakers. They play classical music very well indeed and female opera singers are a delight!. Secondly it can seriously rock due to the generous size and quality of the bas driver. This high-Qm driver does very well indeed and goes lower than expected. Last, but not least, the mid-tweeter integration works so well that even seriously sibilant recordings come through well. I know the R2604 so well but I've never heard a better match than this 10F middriver. I have a suspicion the smooth roll-off of the 10F is responsible for this. The 10F is linear way above the point of crossover and in reality it could play all of the treble range although dispersion would be an issue if we did so.


Click image to view large

The front panels are screwed into place, ATC style. After seeing Mr. Wang's interpretation of the SEAS Classic 3-Way, I decided to give this a try - and for the very good reason that I'm going to make a 2nd pair of front panels for a SEAS 3-Way Classic mkII, this time featuring the SEAS CA22RNY driver. Please don't ask when.

Construction images

Left: Panels piling up, ready for the final cuts. Right: Cutting top, bottom and sides 45 deg.

The black front panels are made from through-coloured MDF. This black MDF is really like HDF, high-density fiberboard, weighing some 0.84 g/cm^3 compared to 0.64 g/cm^3 for the common brown MDF, although it depends on MDF thickness. Thick MDF sheets often have a rather soft center as producers may cut production cost by reducing glue content. Black MDF has an almost ceramic feel to it and comes at approx. twice the cost of standard MDF.

Left: All panels ready for test assembly. Right: For bracings I used left-overs glued to size.

Left: First time assembly by tape. Right: Fitting in the rear panel.

Left: Cutting the bracing panels. As can be seen I cut 2 mm deep grooves to guide the brace panels. Makes assembly much easier.
Right: Making the two panels for the midrange cabinet.

Left: Holes to be routed in two upper braces. Right: Braces ready for trimming edges.

Remember to make braces for the two cabinets mirrored!

Left: Lower brace markings. Right: All braces ready for mounting.

Left: Testing braces. Right: Gluing the first cabinet. These cabs are easy, two straps and a couple of clamps.

Left: Sanding is tedious and the problem with Baltic birch is to avoid edge ripping. I usually start - by hand - rounding the edges gently with grade 120 to make the rounding having the radius of the outer veneer thickness, i.e. less than 1 mm.
Right: I always try to make the rear panel rebates just a fraction of a mm deeper than necessary. After gluing I remove the surplus by edge trimmer.

Left: First coarse (grade 80) sanding finished. Later I use grade 120 and 180 before adding lacquer. Time to add all fillets supporting front panels.
Right: Fillets in place. Next is routing front panels.

Click images above and below for large view.

A few words on chamfering driver holes. Please first read here.
The amount of energy coming from the the rear of the driver obviously is the same as that coming from the front. Thus, we need to make sure we have as little obstruction to the oscillating air from the rear side of the cone. The 10F driver has a very small neo magnet, which already helps a lot. Due to the thickness of the front panel, do as seen on photos. This is the only driver that needs special attention. Chamfer outer panel 45 deg to a depth of 5 mm from the rebate. The inner front panel needs some chamfering too. Do as seen on photos and take care to mark where the middriver cabinet is. The hole in the lower front panel is the same size as the 10F outer diameter. As can be seen on the last photo, the 10F driver now has free air flow to its chamber.
The hole of the lower front panel for the bass driver is the same diameter as the driver, 225 mm.


Mounting the front panels

Trying out drivers and port. Port is Ø 68 x 175 mm. Cut the 220 mm length port to 175 mm.

Left: Small cut-out in fillet to allow tweeter magnet. Right: As can be seen, lots of rear space for the small middriver.

Left: Hole for vent and cut-out in fillet.
Right: Front panel attached with M6 umbraco screws into threads made in fillets. Works excellent! Click right image to view large.
The front panels were sanded with grade 120 and given a single oil treatment. The cabs had two coats of matt lacquer.
I like the rough monitor look!

Left: Making M6 threads. Right: One step back and time to add 4 mm bitumen pads on all internal panels, even the interior of the midrange cabinet.

Cabinet damping: 8 mm felt on all internal sides incl. mid cabinet. 30 mm acoustilux at top (image to thet left). 30 mm acoustilux behind mid cab. Space is limited below port and I placed a piece of 30 mm acoustilux next to the port as seen on photo. On top of the crossover (mid section) and half up the sides the final piece of acoustilux.
I cut a thin gasket for the midrange cabinet as seen on photo. See here for details: I never use any gaskets for the drivers. The mid cab must be absolutely acoustically isolated from the bass driver! Stuff the vent and push the bass membrane and see if you can make the middriver cone move. If yes, check the gasket again and make sure you have sealed the wire hole (e.g. silicone glue). If you glue the front panel you shouldn't encounter any problems.

Note middriver connected with inverted polarity. You can do this on the driver or at the crossover, but don't forget.

The crossover is mounted on the rear panel behind the bass driver.
This takes a front panel fastened with screws. If glued you may have to divide the crossover into separate sections, one for bass and tweeter and one for middriver. The full board is too large to pass through the bass driver hole. 

Before finishing the cabs I just had to set up the speakers and start modeling the crossover.

These 8" bass drivers are big 8" drivers having some 240 cm^2 membrane area from my measurement - and they can rock! More on sound here. The actual size of a speaker driver is controversial. Too many manufacturers claim e.g. 7" midrange drivers where reality would be more like a small 6". The problem is that there are no standards for driver size. Based on history a 6" driver is in the 130-140 cm^2 radiating area range. Some ScanSpeak 6" drivers with a membrane area of 155 cm^2 are more like 6½" drivers. A true 7" driver should be in the 170-175 cm^2 range and many 8" drivers display 220 cm^2 membrane area. This 8" Discovery is to 240 cm^2. ScanSpeak conservatively claim 230 cm^2. But big it is and rock it can. 


Measurements may give us an idea of tonal balance of a system, i.e. too much or too little energy in certain areas. Measurements may tell us about bass extension if far-field measurements are merged with near-field measurements. In addition to this ports may contribute to bass extension. Most of us diy'ers do not have access to an anechoic room for full-range measurements from 20-20000 Hz.  
What cannot be seen is what kind of bass performance we get in a given room. Bass performance is highly dependent on in-room placement of your speaker and the same speaker can be boomy in one place and lean in another. Actual SPL level at 1 meter distance and 2.8V input is useful for en estimate of system sensitivity and combined with the impedance profile may give an idea of how powerful an amplifier is needed to drive the speaker to adequate levels.
What measurements do not tell is the very sound of the speaker unless displaying serious linear distortion. The level of transparency, the ability to resolve micro-details, the "speed" of the bass, etc., cannot be derived from these data. Distortion measurements rarely tell anything unless seriously bad and most modern drivers display low distortion within their specified operating range. 
Many people put way too much into these graphs and my comments here are only meant as warning against over-interpretation. There are way more to good sound than what can be extracted from a few graphs. Every graph needs interpretation in terms of what it means sonically and how it impacts our choice of mating drivers, cabinet and crossover design.

Left: Let's start with the 22W driver. Now, this is one linear driver allowing to go as high as diameter (beaming) allows.
Right: Impedance of all drivers. Initially a low port tuning was tested, thus Fb = 30 Hz. Maybe I'll stay there... Green is the 10F in its small closed box. Blue = tweeter.

Left: SPL @ 2.8V, 1 meter from final crossover. Measurement merged with bass near-field response at 200 Hz.
Right: Final impedance of system with port tuning @ 30 Hz. Port 68 x 175 mm.

Left: Response of individual drivers driven from crossover and summed response (red).
Right: Changing R2041 gives the option of adjusting the important upper mid/lower treble range. Resistor values and default value incl. in kit.

Left: Just to show the significance of R3012-C3011. Listening to the bass driver alone driven from crossover the 3 kHz hizz is noticeable.
Right: Tweeter attenuation from 4 resistor values. Lowest value provide a flat response (too much to my ears but the choice is yours).


No longer available, source components locally.



Connect A to B with wire.