Vintage Scanspeak drivers, modern crossover, trials and tribulations

[DrBoar]

It all started with a fleamarket Ortofon P20 and
http://www.troelsgravesen.dk/SS_vintage.htm.
Making a modernized version of the P20 with a more elaborate crossover and just using the larger tweeter went fine. Installed in a box everything sounded good. At the time I had everything patched together, to tidy things up I among other things resoldered the wires on the tweeters. In this case the usual tags are mechanically very sturdy metal rods, about 5mm in diameter and 40mm long. For some reason this mechanical sturdyness was not linked to thermal robustness, and my soldering killed one of the tweeters....

Well I had an other pair of 4 Ohm tweeters just as the D3804 dome. So I just swapped it in and it did not sound that good. The replacement a vintage Luxor dome not that dissimilar to Peerless domes of the time with a high Q resonance at 1 kHz is quite different from the D3804 that has a very well damped resonanse at 700 Hz. Blue and green curves respectively.



With the crossover at about 2 kHz bad things happens with that high Q resonance. I hastily made a LCR trap of what I had at home. Orange= driver, pink and green= LCR trap. Blue= driver and LCR combined.



I guess that I could work with components values to increase the 1.5-5 kHz one ohm or so. System impedance is not far from Troeds design.



Frequency response from the individual drivers (close) and system 1 m on axis looks quite good. I used the the XTZ Room Analyser II, a setup aimed at analysing rooms so the measurements are quite basic and is affected by reflections from the surroundings as well as the foot of the microphone. I might revisit this later with better measurement gear.







The box is a rather squat onken style box tuned to about 55 Hz. With removable baffle it has proven to be a very useful tool for testing various drivers. After decades of very slim speakers they do look quite odd especially on stands.