CMOS-FET input switch-- a bad idea?

In writing to Hafler authorized repair technicians about my aging 915 preamp and 9300 power amp, I have recieved the following comment

"One thing that you need to know is that the 915 has a solid state input selector switch that is operated by the front panel control. That is a sonic "bottleneck" but the tape input does not pass through that switch. The tape input is relay controlled and consequently will provide cleaner sound than the other inputs on the 915."

The Hafler manual states that this CMOS-FET electronic switch is a sonic advantage because it eliminates signal degradation due to mechanical switch contacts and allows the electronic switches to be located near the rear panel inputs for a shorter signal path and reduced interchannel crosstalk.

The tape input, of course, has a mechanical switch as stated by the technician.

What are the relative sonic advantages and disadvantages of the CMOS-FET switch compared to the mechanical switch? Do you agree with the technician that a mechanical switch must deliver cleaner sound?

Is the CMOS-FET switch, by its nature, cause for concern? Yes, I am going to connect my Slee phono preamp to the tape input and try to hear the difference, if any.

Assuming that the mechanical switch is in Good condition, I believe the sound would be Streets ahead than an FET transistor made to fully conduct, and act as a switch.

Infact if you do not need all the multiple input, go for dedicated hardwiring one input.

The FET 'switches' also assume that the signal passing through it is MUCH lower than the supply voltage fed to the switch. I dont know if in the Haffler its a 5 Volt or 12 Volt supply to the FET switches.

A 5 Volt supply is rather close to tghe output of CD players, infact some high output CD players could exceed this level, and their peak to peak even approach the 12 Volt suppy.

In the days when the Haffler was designed, 500 milli Volt inputs were the norm.

CD Players today often put out 2 V RMS ie 2x2.8 Volts peak to peak !

Well designed FET switches can deliver good sound quality ( the THD is impecable but not necessirily the sound quality )... but the early ones like in the HAFFLER were Probably too rudimentary by today's audiophile standards...

Just my 2 cents ...
I did a listenging test to compare the tape monitor relay input to the selector switch CMOS-FET input. Graham Slee, Janis Ian, Paul Simon and Barbirolli conducting Sibelius 2nd provided the short program.

If there is a difference between these two inputs, it is not much. Certainly a tiny amount compared to my recent switch from MAS Gray interconnect to MAS Blue. If there is a difference, the selector switch CMOS-FET connection is better. Maybe a little more low level detail and a little better rendering of space. Maybe not.

With the system as it is now, the MAS Blue is much more pleasing on the Slee phono preamp. The Gray is too forward in the upper midrange and lower treble. I guess I attribute some of this to the Hafler 9300 power amp. I did the testing with the Gray interconnect, which does etch the details and control the bass better than the Blue.

If this selector switch is "cheap and cheerful" as the Brits say, then cheers for Hafler using it to increase profit without hurting the sound. They STILL had to drop out of the hi-fi hype market shortly after making this preamp and are now kaput.

I also asked John Hillig about the selector switch, since I still plan to have this preamp modified by him. He wrote back as follows.

"Modified 915s sound pretty good. I too was worried about the FET switching but I have a better view of it now. "
IMO, all CMOS switches sound nasty, they're shrill and harsh plus they degrade THD+N measurements. Mechanical switches are my favorites, no electronics, no electronic artifacts. You just need to be sure to keep them clean.