We hooked up a stock (unmodified) 600T and a stock KT-917 side-by-side, on the same antenna, and found that they were almost indistinguishable. The tuners could not be differentiated based on audio quality or quieting. The KT-917 was very, very slightly more selective, but this difference could easily be attributable to normal variation between pieces. The 600T has an excellent MPX filter that drastically reduces noise on weak signals, while the KT-917 calls its mono/auto/stereo knob a "quieting control" with the MPX filter built into the "auto" setting. Having them next to each other made the difference in size obvious: while both have a 17-3/8" x 5-1/2" front panel, the 600T is 15" from the front panel to the back of the antenna screws, while the KT-917 is 17" deep (16-1/4" without the back panel "feet"). Also, the 600T's cabinet is smaller than its front panel (i.e., the front panel protrudes), while the KT-917's larger cabinet is flush with the front panel on all sides.
Both tuners were outperformed by a Sansui TU-X1 that was hooked up to the same antenna. The TU-X1 sounded cleaner, with more powerful bass, and was much more selective. However, our 600T that is modified with narrow filters blows away the stock TU-X1 for selectivity.
Bob's 600T vs. KT-917 Comparison
EDITOR'S NOTE: This section is out of date. We now know a lot more about the 600T, and audio mods have been done on a couple of 600Ts with great results. We will probably have an update eventually, but it's a low priority.
Both have such an incredibly complex schematic, it is impossible for me to see which would be better. It appears that the 600T had a slightly better IF section for the wide and normal filters than the KT-917. They are totally different in circuit implementation, but share many common design features, if that makes any sense. The RF front ends are done completely differently as well, even though both have 8-9 gangs.
The KT-917 - antenna input, 2 gangs, FET amp, 2 gangs, FET amp, 3 gangs, balanced mixer. It has 2 independent gangs for the oscillator as well, for a total of 9 true gangs.
The 600T - antenna input, single gang, FET amp, triple gangs, FET amp, triple gangs, balanced mixer. It has 2 gangs tied together for the oscillator tuning, for a physical count of 9, but functional count of 8.
I don't see any new or innovative circuits in the KT-917, and most of the fancy circuit ideas introduced with the 600T are retained in the KT-917, just implemented with different parts and circuits. Both have incredibly advanced discriminator and MPX sections, and would have great potential for outstanding sound given some simple changes we have successfully implemented with the KT-7500. After comparing the schematics, I see that the 600T is a very different unit from the KT-917, so the common statement about them being the same except for cosmetics is incorrect. I'm not sure who started that one, but we all seem to have followed it.
I don't think anyone ever did audio section mods on a 600T [until recently - Editor], but I have seen things all over the map written about them for sound quality. It does not enjoy an audiophile reputation from what I can see. As they are usually pricey, and collectors would want stock units, it may *not* be a good one to modify if you think you might decide to sell it later. The 600T seems to be coveted more by collectors (like me) that want it for the technology, or FM DXers, rather than for the sound quality. I'm sure it could be brought to sound good, though. But so can many other less expensive units. As for parts, the Multiplex does appear to be a discrete design, based on a traditional chip to extract the 38 kHz PLL MPX signal from the 19 kHz pilot. The actual L-R channel extraction is done using discrete transistors, which is good. This is similar to the Sansui TU-9900 scheme. Very few tuners have discrete MPX, which is why I highlight the ones that do. The output section, though, contains many electrolytics caps, 4558 op-amps, relays, and switches, which we now know all kill the sound. Easy to fix, but, none of us has ever gone there.
Unlike the 600T which uses 2 parallel filter paths (wide/normal and narrow) the KT-917 uses a single serial IF filter circuit with taps for the 3 filter bands. It starts with a ceramic filter, followed by an LC filter, followed by 4 "Surface Acoustic Filters." As each filter is normally 2 stages, this would give the narrow mode 12 stages (or elements).
In the KT-917 and 600T, there are filters used to drive the signal meter and muting circuits that have nothing to do with the IF signal that goes to the detector. Of the 3 ceramic filters in the KT-917, only one is in the true signal path, and the others are used in metering. Same in the 600T - I worked on one in which the 4-pin filter that was driving the signal meter was off in center frequency, so the meter "peak" did not match where the tuner "was." That bad filter was replaced with two new 3-pin 280 kHz filters. See the DIY Mods page for more on the 600T's and KT-917's filters.
Bob's Quick 600T vs. KT-8300 Summary
They use many similar circuits, but the 600T goes further to make it better. 8 gangs versus 6. Balanced mixer. Pulse count detector. Calibrated meters. Much bigger and beefier power supply. But actually very similar IF stages. Narrow circuits are the same. Both use three 4-element Murata 4-pin filters, equivalent to six 3-pin 280 kHz filters. The 600T uses a 6-pole LC in wide + more 6-pole LC for 12-pole LC in normal position. I would say the 600T's "normal" circuit is just about perfect for all-around use, and modified narrow is used for getting ultra-weak stations with ultimate selectivity. A 600T with modified narrow filters in top alignment is an incredible tuning machine. The KT-8300 has two 4-pole LC filters for 8 poles in wide. Both are phase linear, and when the set is properly aligned, will yield measured distortion below 0.05%, which is my current limit of measurent capability.
Both use the same stereo circuit, based on the HA1156 IC. It uses the IC to run the PLL to extract the 38 kHz fromn the 19 kHz pilot, but that's about it. The stereo channels are formed with banks of discrete transistors in both units, in a switching circuit driven by the 38 kHz frequency. Stuff not normally seen in other tuners.
The above MPX circuit is called the DSDC or "double switched demodulator circuit," and uses both transistors and FETs to perform the mono-to-stereo conversion, with the HA1156 performing only a small percentage of the functionality that it has. After the DSDC MPX circuit, the KT-8300 and the 600T both use normal op-amps (4558s in T-05 type cans, but with normal op-amp pinouts) for gain, de-emphasis filtering and output buffering.
The KT-8300 has a 5 MHz wideband detector with balanced output. It is a very good one, but it is not a pulse count detector. The pulse count detector requires a second IF conversion, with another local oscillator, mixer, etc. to take the IF down to around 2 MHz.