Vifa C17WH-

Vifa C17-II, Monacor DT300+waveguide
Copyright 2010 © Troels Gravesen

Monacor SPH176 midbass and DT300 + waveguide. The SP176 is an interesting driver with a vented, stamped metal chassis but some other time. Here we stay with the Vifa C17. The waveguide needs 4 pcs. 4 x 12 mm standard metric screws to be held in place. They don't follow the kit.

I've been looking at the Monacor DT300 tweeter for some time and finally decided to buy a pair. This is a 30 mm soft dome and how would this driver perform compared to my much loved ScanSpeak D3806/8200 tweeter? The latter has a 38 mm voice coil and really needs a supertweeter to make it up to 20 kHz.

The DT300 may be a suitable compromise for two-way systems providing a possible low point of crossover. What's even more interesting is a new waveguide made especially for the DT300 tweeter. This very short plastic horn increases sensitivity by some 6 dB at 2 kHz and a very simple 1st order crossover is possible with the Vifa C17 drivers. I never thought I'd make a 1st order crossover work properly, but this tweeter seems to make it possible, helped by the gentle roll-off from the C17 drivers.

It may be difficult to distinguish between a horn and a waveguide, but usually a horn takes a compression driver and a longer horn, where a waveguide in principle can be applied to any driver (think of PA bass speakers), from small domes to large cones. A waveguide is usually a very short and shallow horn, in this case only some 40 mm deep.

Individual response of drivers. With the waveguide in front of the dome, the sensitivity is raised to 100 dB/2.8 V at 2 kHz and the tweeter can be equalised by a very small capacitor, e.g. 2.7 uF to flatten the response. This also means that distortion may be significantly reduced in the crossover region as the dome really doesn't move much due to the acoustic amplification from the horn. 2.7 uF in series and a LCR circuit to flatten the driver's impedance is all that's needed to make a point of crossover around 2.5 kHz.

What's seen above is what happens if the magnetic oil is removed from the tweeter. Actually I've never seen a tweeter "submerged" in magnetic oil like this one. There was oil all over the place, the pole piece being covered with this magnetic sauce.
Removing the oil increases response down towards 1 kHz, making the crossover construction much easier. I'll skip the details but running the LspCAD on the un-modified tweeter wasn't that easy.
The red graph was from before the tweeter mods were fully completed.

Above the impedance profile of tweeter after removing the magnetic oil (red) + mods.
The DT300 needs a lot of burn-in. Not what I usually find on tweeters, but this one seems to be the exception.

Routing new front panels for the C17 and DT300/waveguide drivers.

To get the drivers as close as possible it's necessary to do the routing as seen on these photos. The position of the drivers is shown on the drawing below. I was afraid the waveguide was made from some brittle plastic that would fracture from the router but I was able to route very slowly into the plastic without trouble. The plastic turns a bit hot and starts melting, but take it easy and only do a fraction of a millimetre at a time. The waveguids are cheap - less than 10 US $) so don't panic if you break one. If you have a router with variable speed you may try running at low rpm. My router hasn't.

Router plan for the drivers.
What I also did this time was lifting the vent to be completely symmetrical to
the two Vifa drivers. I used a Monacor MBR75 vent.
I'll try reducing the vent length to 50-75 mm to increase vent tuning to 45-50 Hz.

The crossover:

Above the crossover as presented by the LspCAD after fine-tuning of the 4th crossover. LspCAD seldom gets it completely correct. You can always make a nice looking performance on LspCAD, but after hard-wired fine-tuning and inserting the real values into the LspCAD it doesn't always look so good.

Tuning the frequency response is minor trouble due to the C17 and DT300 drivers having minor notches around 4 kHz, thus the LCR circuit for the tweeter at 4 kHz (15R+1.5uF+1mH). The other LCR is part of the high-pass section giving a smooth 1st order rolloff.

The reverse null test was bit confusing as a sole 1 mH to the C17 drivers made the best measuring performance but transparency was seriously reduced from this simple approach. So back in went the RC circuit consisting of 4R7 + 6.8 uF, which basically flattens the impedance of the C17 drivers. Very elementary..

I often get the question on the order of components in a LCR circuit, but you can have any order of components in an LCR circuit. It doesn't matter. It can be LCR. RCL, CLR, etc. No problem. It does the same thing.

Crossover components

Having two 8 ohms speakers in parallel takes a low-ohm series coil, thus the cored 1 mH/0.11 ohm.
Both coils for the tweeter notch filters can be high-ohm.

Measurements

The red graph is the response of the two C17 drivers from only a 1 mH series inductor. The blue is when we add the impedance correcting RC circuit. Where the red graph very much seems to resemble a true 1st order roll-off, the blue seems to have a roll-off of 6 dB/octave from 1-2 kHz. From 2-4 kHz the roll-off appears to be more like 9 dB/octave and from 4-8 kHz we have a 12-18 dB roll-off.

Impedance of drivers without crossover attached. The red is the tweeter with all modifications as described later.
There's an annoying notch at 160 Hz. Some cabinet resonance from my test cabs. The final cabs will be added further bracing. Read first article, C17.

Above the response from drivers and summed response of the finished crossover with 2R2 to the tweeter. Close to what the LspCAD simulation predicted. Maybe even a little better. Overall we have a rather flat response and eventually I raised the tweeter series resistor from 1R5 to 2R2 to reduce treble response. With 1R5 the sound became a little too forward.

This impulse response is rather unusual as the impulse from the tweeter and two midbass drivers appear to start simultaneously due to the tweeter's voice coil being some 50 mm behind front panel level.

The step response displays the tweeter connected with inverted polarity (first sharp peak) and midbass drivers with positive polarity. This seems to be a time-coherent system. Due to the wave-guide, the tweeter voice coil is almost in the same plane as the C17 voice coils, but due to the first order filters producing a 90° phase shift, the overall phase shift between tweeter and bas will be 180°, thus the inverted polarity of the tweeter. We might try inserting an all-pass filter before the tweeter to get all drivers connected with same polarity, but I haven't tried this.

Cumulative spectral decay from final crossover. Actually this tweeter seems to have a very clean response with no apparent resonances except for a small dip at 15 kHz. It doesn't disturb my peace and it's tricky business to work around 15 kHz but I'll see if I can further improve the tweeter mods to smooth the response for the sake of good order.

Impedance plot of system. A significant rise in impedance around the point of crossover. However, nothing serious to my Copland valve power amp. Should you worry about this from using low-wattage SET amps, read below:

Impact on impedance from adding a LCR circuit across terminals: 0.27 mH + 6R8 + 22 uF. A very easy load to any amplifier that can run a 4 ohms speaker.

Impact on FR from tweeter LCR (15R + 1.5 uF + 1.0 mH). The minor bump at 4 kHz is removed. And it's audible.

The sound?

Well, this DT300-waveguide certainly changes the sound of the C17 set-up. A waveguide in front of a dome tweeter sounds different compared to standard tweeter and transients appear faster than ever. It's difficult to compare to the first C17+27TFFC set-up because a lot of things has changed. Where the C17/27TFFC had a 2nd/3rd order filter we're here dealing with a 1st order filter. I tried a similar 2nd/3rd order approach to this set-up but the sound didn't seem to "come out of the box". I all sounded very nice, really nice, smooth and clean, but it didn't come to life the way this first order filter does. First order filters also have their drawbacks due to the huge overlap between drivers and the mutual colouration they may introduce, but again, I think the reduced distortion of the tweeter due to the waveguide plays an important role here. A good sign is that I generally play louder with the C17+DT300 compared to the C17+27TFFC. In the beginning I thought the DT300 had some distortion at high levels until I realised how loud I was playing. And the DT300 needed time to break in. The current set-up does have limitations in terms of overall loudness, but they can play loud enough. No doubt about it!

It's rare I sit in front of a pair of new speakers, pulling out old CDs and vinyls to hear what I missed before. But this time I did. I've never heard such depth from live recordings as here. Driver integration is better and one explanation may be that the waveguide will enlarge the soundwave coming from the small dome to a size almost equal to the size of the C17 driver.
Did I hear someone say: "Yeah, right! The guy's got a new speaker and hear things he never heard before." I very much quote myself, because we'll always hear things we never noticed before when we connect another pair of speakers - because no two speakers are alike. Nevertheless, this speaker will be on the agenda for some time before the final cabs are made and they will be in my permanent collection of speakers - as representatives of particular interesting d'Appolitos. They can do things none of my other speakers can do.

The bass from these C17 drivers is excellent and despite the poor measured performance in terms of bass extension, the bass appear deep and powerful. It's very dry and anything but one-note or mushy. Mind you, the radiating area of the two C17 drivers is around 280 cm^2, close to a 10" driver, thus we have the bass and mids coming from a huge area compared to most other small speakers. And I still consider the C17 a relatively small speaker with an internal volume of approx. 40 litres.

The C17/DT300 is very much a "window to the music" rather than a speaker that "creates music in a room" like the PMS with its wide dispersion and ease of placement. The C17/DT300 takes more care in placement and toeing in and it may be more picky on what drives them. The best I've heard was from a MingDa 300B push-pull integrated, 20 wpc.
Did I say I never enjoyed the audience's handclapping so much as from these speakers? My vinyl live recordings are on the turntable all the time.

Tweeter mods:

Remove the faceplate and the voice coil as seen to the left. The vented pole piece is covered with some heavy felt material and the hole through the polepiece is not damped at all. A bit strange. Seems like a rather simplistic approach. Why have a vented pole piece and a damped rear chamber - and then block the vent? One reason may be that the felt pad works as an acoustic vent to the rear chamber.

Remove the magnetic oil with some small sheets of paper - and there's a lot of oil. Pic to the right.

The voice coil needs thorough cleaning. Use soft paper (Kleenex) to remove residual magnetic oil. Make sure to remove all the oil as it adds weight to the voice coil.

A ring of felt material (1 mm thich) is glued to the polepiece and 200 milligrams of Monacor MDM3 (2/3 wool + 1/3 polyester foam) is stuffed in the pole piece vent.

If you think this is too much, you may get away with only removing the magnetic oil. And this must be done to make the crossover work as intended.

After modifying the tweeter I had both tweeter running fullrange on the MLS signal for some time to speed up burn-in. I suggest you let them play full-range music for some time. These tweeters are quite rugged, so don't be afraid to ad moderate power.

Go to main page

Go to Vifa C17-I