SpeakerTalk

[speakerguru]

[audiolabtower]

Thanks, that explains everything. So the new former helped with dissipation and allowed the spec to be increased to 100W, do you remember if it was concurrent with the launch of the aB version, or just accidental around the same time so some original 104s might have the uprated driver?

[speakerguru]

Just around the same time. The vc change was a result of too many warranty returns and pressure from Sales & Marketing. The aB was the first result of the target function approach to xo design following the arrival of the R & D computer, a HP Fourier Analyser. The aB also addressed the issue of low impedance at crossover. The previous Model 104 dipped below 4 Ohms around crossover, which was "naughty" for a nominally 8 Ohm speaker, and caused premature current limiting with some amplifiers. Quad's Peter Walker would have complained and Laurie Fincham always took their comments seriously despite initially making g light of it .

[audiolabtower]

Interesting, as an aside I built the original Quad 405 when it first came out (the only parts difficult to source in those days were the output transistors, I bought a set direct from Quad) and it was fine with the 104 but better with the 104aB. It peaked at 6 ohm and was well on the way down at 4 ohm, which the 405-2 addressed. I was going to modify the boards to take the new protection module which allowed high current peaks into low impedances but by then had moved on to more transparent amps.

https://lh3.googleusercontent.com/pw/ACtC-3dQvY_JzOlnOCQI3CU_r4M1J-VCAfQoQxaVOznwBUZXqrwvB4H787HUCOJcciuNkPQAIo4Wt8e4dvN7XoIFlhRYtIln38AEb5hzRJ6tR4uHEFL5GecoKoCAA20Ho5RteivAFeoN5nxfTZdwcKmMM_rU=w1326-h994-no?authuser=0

https://lh3.googleusercontent.com/pw/ACtC-3e1LYjMRpAMPP43DuijYdIp6ZtZEgvv2Xlc6xqQ3K1w1ap5IfRQTmpikp0GiORh1wLEbJZashvvCkon9qtHUcOD945429KvcFAgtOvPmCIe4t4wYvtC-Du4LDYsiPEg7JNmU-QnZn44O0h_1BuCGpyf=w1326-h994-no?authuser=0

That computer was ground breaking. It allowed Kef to mass produce stuff to the tolerances Spendor still had to hand select.

[speakerguru]

I'll see your aside and raise it . When the Q405 came out, the boot was on the other foot. Their current limiter was far too aggressive and could be activated with 8 Ohms speakers which were well within BS and IEC limits for impedance.

What was far worse was that when they current limited the amplifier effectively had a negative output impedance which, if bigger than the speaker Z magnitude at any frequency, formed an oscillator at that frequency which swung the whole supply (100V p-p clipped) across the speaker. Sounded horrendous rather than just bad, as per regular output clipping. Quad changed the limiter to a more friendly and non-oscillating design with the 405-2, mostly as a result of my testing at KEF.

[audiolabtower]

Yes, the 150 watts at 6 ohm was probably resistive and large or dynamic reactance was not handled as well, surprising really from the electrostatic people. I added their thyristor circuit across the outputs.This was described as a dc shutoff in case bias diodes went open circuit, but had a time constant, I wonder if it was there to be also triggered under your described conditions. I must say I never played the 104s loud enough to have any bad experiences, but the aB was definitely better.

[speakerguru]

We used 8 x 405 cards in the KM1 amplifier but without the thyristor crowbar circuit d.c. protection. It was not needed and it was always on the verge of triggering with high LF content. IIRC if you tested with 20Hz sine wave you could see it trying to cut in every half cycle.

[audiolabtower]

Wow, yes, of course, you used it in your KM1!