The reason I’m putting this on is because it’s so weird. I had never heard of this amp before this one walked in for repair/service, and I can’t find any reference to it at all anywhere.
The real mystery is that it is, internally, very very, like a musicman 212 combo, and the reason that is so peculiar is that the musicman combos were of an extremely unusual design, mainly in the way the output 6L6′s were interfaced to the front end, which was entirely designed using discrete chips (TL072′s if I remember right).
The drive to the output stage is applied via the cathodes of the 6L6′s by two fairly robust transistors which have the case pattern of T0 122, and the grids are held common. Very unusual, so was the proamp a rip-off of Music Man, or some kind of development project? Don’t know, except that the workmanship isn’t on the same planet as Music Man. I’d be interested to hear if you know anything.
Sound City were designed by Dave Reeves, who later developed Hiwatt amplification. They were produced for the Dallas distribution company from the early 1960′s.
As with many amps from the beginnings of the sixties, it had no master volume control, which meant that the background noise level wasn’t that great. Reeves got around this in an interesting way, by fitting a two position toggle switch on the back of the amp. This is not a power limiting switch; it drops the overall gain of the amp, so is equivalent to a two level master gain control, and can make a big difference to the noise level value.
Unfortunately, there was a preset pot associated with this switch, (VR6) which set the gain difference between full gain and the reduced value, and this was in the cathode circuit of V3b (the last preamp valve before the phase splitter). This caused a lot of trouble, because it’s a dc path, and presets don’t generally like dc. The stock answer was to disconnect the switch and short out the pot, but this means that the amp runs at full gain (and also full noise). A more useful alternative is to reinstate the switch, and replace the preset for a fixed resistor, say around 24K.
Another problem that this amp suffers with and which was also present when new, is associated with the tone network situated behind the front plate. It’s a frequency selective ladder network, (or series of pi-type filters if you like) and is a high impedance network, which will therefore pick up hum badly. To clean this up, you get a piece of cardboard and stick some baking foil to it. You can secure this by one of the front bolts and it fits under the ladder network. It cuts the hum a lot. There is no screening under this circuit in the original. Make sure that you have a good connection to ground on the foil; and be very wary of voltages floating around in the amp. Some Sound City’s had close to 800volts on the anodes. Switch it off, disconnect, and leave it for a few minutes. Then test around for anything you might not want to put your fingers on!
Make no mistake, although these amps never had the kudos of Hiwatt, Marshall or Orange, they were very nicely built and generally well-designed amps that go very cheap at the moment. No bad buy, then.
According to the customer, this model amp has had a fair number of problems, mainly with switching noise, but also reverb hum. In the prescribed mods from Fender, Bruce Zinki (the designer) blames much of this switching noise on the relays of the earlier models, which is fair comment, but the real problem is not that the relays are bad, but they are high voltage (24volt) which means that switching the coil gives a considerable spike. Later these were modified to 5volt higher impedence coil devices. Even that’s not too bad, except that the feed to the coils is through about a foot of unscreened cable within the chasis which happens to drift underneath the PCB, past various sensitive preamp passives. Not a great idea.
In order to limit this spiking problem, Fender recommend that R19 be changed to a value of 9.1K. This actually drops the operating voltage to the relay, and as the one I had was operating with twenty volts to a 24 volt coil already it seemed not a good idea. I shunted that resistor with an electrolytic cap of about 10uF, which took the edge off the spike and did a fair bit for the noise. After that I replaced the long relay/footswitch supply cable for a piece of screened lead and that also improved matters no end. The cap slows the action of the relay a bit, but it’s only a few miliseconds, and nothing you would be aware of in use.
The R13 that is recommended to be reduced is a grid leak resistor and If you halve this it will drop the gain by around 6db, which will also cut the switching noise, but only because the overall gain is that much less. So I left that as it was, as the other mods had cleaned the amp up a lot. The reverb (which wasn’t buzzy on this one) did exaggerate the switching noise, and I couldn’t get round this. I think it’s associated with the layout of the PCB, which there’s not much you can do about that.
This particular amp had another problem. The treble control acted as a volume. This turned out to be a dry joint on the grid of V3b. Also watch out for the missing coupling cap from the anode of V2a, it is there but not in the schematic.