ATC SCM20 upgrade

Hi everybody
The knowledge about ATC loudspeakers on this forum is impressive, and I would very much appreciate to receive any input from you.
I own an ATC SCM20.
Recently I have upgraded the Xover components (to Duelund resistors, Auricap Caps, Alphacore inductors), replaced the tweeter (Vifa -> ScanSpeak 9800) with excellent results overall. I consider upgrading to the SL version of the mid/bass driver.
(1) How much better is the SL driver compared to the normal one?
(2) Does the Xover need to be adapted?
(3) Is there a way to purchase those drivers and install them myself?
Thank you very much! Bruno

 

The most importand part of the 20 is the bass/mid driver and in my experience the SL version is significantly better/clearer, especially at lower volume levels.

I didn't think the existing crossover was badly done at all and the components were amply rated for many years of good quality and often hard use, although I respect your changing the tweeter for something better. Are you able to measure what the Scan is doing level wise compared to the mid, as it has a wider bandwidth as I remember?

The only way you are going to get a pair of SL drivers I think is to buy a pair of 20SL's and modify those. You'll get a bi-wiring option then as well IIRC. Selling your modified speakers on may be a problem though. Why not contact ATC, give them the serial numbers of your speakers and see if they'll sell you a pair of SL drive units as they do with active owners.

 

Thanks for the reply! I do not want to sell it. Anyway, besides the fact it is not original, I guess most potential buyers of used ones would prefer the SL .
Actually, I did hear significant improvement after the Xover parts upgrade.
To account for the different efficiency between the old and new tweeter (went for the specs as published by the manufacturer) I adjusted the values of the resistors (L-pad). The sound is shining more (which I like better) after the tweeter change but I attribute that to the different characteristics, not sensitivity of the tweeters.
I do listen at relatively moderate levels most of the time. Therefore, your comments motivate to go for the upgrade.
There is an upgrade option for 1700$ a pair. However, the distributor (Switzerland) mentioned the Xover needs to be changed which (1) I doubt based on published Xover data found on the net, and (2) I could imagine that ATC is reluctant to work with "tinkered" stuff. I will contact them directly whether any Xover modifications are required.

 

There is more to swapping a tweeter in a passive speaker than sensitivity. Is the ScanSpeak a known drop-in replacement? If not, you have just completely changed your speaker.

 

Impedance is the same, frequency response is similar to the Vifa tweeter and suitable for the Xover point at 2.8 kHz. Anything else?

 

Which tweeters are we talking about exactly? The ScanSpeak D2904/980000 and the Vifa DX25 thing? What is the exact Vifa unit?

 

Actually, a Scanspeak D2905/9700 has replaced a Vifa D26TG-35.

Specs:
Vifa
* Power handling: 100 watts RMS/140 watts max * Nominal impedance: 6 ohms * DC resistance: 4.6 ohms * Frequency range: 2,000-30,000 Hz * Magnet weight: 8.5 oz. * Fs: 940 Hz * SPL: 90 dB 1W/1m * Net weight: 1-1/4 lbs. * Dimensions: Overall Diameter: 4-1/16", Cutout Diameter: 2-7/8", Mounting Depth: 1-11/16", Magnet Diameter: 2-7/8", Magnet Height: 1-1/4".
Scan Speak
* Power handling: 225 watts RMS/320 watts max * Nominal Impedance: 6 ohms * DC resistance: 4.7 ohms * Frequency range: 2,000-30,000 Hz * Fs: 500 Hz * SPL: 89.5 dB 2.83V/1m * Net weight: 1.5 lbs. * Dimensions: A: 4-3/32", B: 2-7/8", C: 1-3/4", D: 2-7/8".

 

Hi, the tweeter you have replaced has a differnt impedance curve, the peak is in a differnt place, so the crossover would need to be modified to maintain the same acoustic roll-off at the xover point. You might like to read my posts in this thread to understand why.

https://www.audio-forums.com/as-rediect/showthread.php?t=19049

Still, if you like the results great, but its not the same speaker it was and the response will be different.

 

Thanks!

Do you have access to the impedance vs frequency curves of different drivers?
What is the effect of a lower Fs, like in the new tweeter? Will there now be dent in the combined output of the drivers, or a peak?
How could I calculate what modifications are necessary?

 

At 2.8 kHz, both drivers have 6 Ohms Impedance (left Vifa, right Scan speak). The Xover is 2nd order. Can I assume that the impedance at 500Hz is sufficiently far away to play no role?

 

Hi,

The best thing to do is to find out exactly what the crossover components are for the old tweeter and model the response they give with that tweeter. Then you can model the response of the new tweeter and adjust the components until you meet the same acoustic response of the old tweeter with old components. If you tell me the old crossover components I could have a quick attempt for you.

 

That is a very kind offer. I'll report back soon. Thanks, Bruno

 

Xover

Hi Tenson
Here is the Xover schematics. Instead of drawing my own, I got it from Mark's page (http://www.mhennessy1.f9.co.uk/showpic.htm?atc/crossover.gif).
This is a slightly later model that uses a Vifa 27TG-46-06 tweeter.
The specs of 27TG are as follows:

Znom 6 ohm
Re 4.6 ohm
Le@1kHz - mH
fs 650 Hz
Qms 0.95
Qes 0.77
Qts 0.43
Mms .3 g
Sensitivity: 2.83V / 1m 91 dB

Could you really tell how to change the Xover for the ScanSpeak 9700? That would indeed be fantastic!
Thanks, Bruno

 

Hi,

This should work provided that ATC updated their xover to archive the correct slope for this other tweeter and did not change the xover on the woofer at all. Probably a safe bet. You don't fancy taking your speaker apart to check the values of the components? At lest those on the woofer to make sure the xover is the same.

 

Hi
The woofer part is the same, that I know for sure.
The tweeter part is different, so ATC did change the Xover for the new tweeter.
Actually I did take have a look at the Xover as it now and I was able to verify that the woofer part is the same, however, as I hot-glued some parts to the board with the label face down (capacitor, inductor), did not keep the old parts and did not take proper notes (I know.... I hoped I could get around admitting it...) I had a hard time figuring out the tweeter part.
I hope that it might work this way?
Thanks a lot, Bruno

 

Hi,

I just looked, and there are no measurements for the 27TG-46-06, as it is a customised tweeter made for ATC.

You either need to find accurate measurements (frequency response and impedance curve) for the 27TG-46-06. Or you need to look at your speaker and tell me the values of the filter components for the Vifa D26TG-35 tweeter yours use, as we can find the data for that.

You would need to change the components anyway, so it doesn't really matter if you have to take them off the board, does it?

 

Hi - I guess you are right.
So here is the original tweeter section of the cross-over:

 

Hi,

Thanks that is much more useful. I'm not 100% clear how the 4R7 and 10R resistors are wired up though. Its a strange arrangement. It looks like the first two 4R7 resistors are in parallel so thats effectively 2R35 series resistance, then 10R to ground, and another two 4R7 resistors giving 2R35 in series. Is this right?

These are your non-tweaked L-Pad values, from the old tweeter are they?

 

Hi,

Assuming I got it right with the resistor wiring, this is what I have modeled.

You can see that the xover with the Vifa tweeter is clearly aiming at a 4th order Butterworth slope at 2.8KHz. The reason the Vifa does not perfectly meet the target may well be due to the fact I am working with manufactures specs rather than a measurement of the tweeter in the ATC box. However, I'm a little unsure about that rising top end. Forced detail, anyone?

In this picture you can see the target ideal slope at 2.8KHz in red. The black trace is what the Vifa is doing. The white trace is what the ScanSpeak does with the same filter components. You can see its not actually all that far off, so you made a good choice, but its not perfect.



Below is a comparison between the Vifa (top graph) with the xover filter you gave me, and the ScanSpeak (lower graph) with the components I would suggest. I chose values you can easily get at an electronics store. As you can see its gets a lot closer to the target with the tweaked values.



P.S. this forum seems to shrink pictures so you can click to make them big.

 

Thank you Tenson!
The resistors are correct, and I have given the values of the unmodified xover.
I am done modifying with the new components, sounds very good!
Besides, using boutique Xover components does make a difference.
Thanks again, Bruno