Here are some more pictures of the Prem-4 having removed the covers.



Power supply transformer, btw all the blue caps are for power supply smoothing and filtering, each is 1300uF 350V. They are wired in series pairs with 100K/2W resisters in parallel with each one. The plate voltage is 500V.



Output transformer, these are supposed to be special, don't know the details.



Labels are peeling off due to age.



Right, I have 30 years experience in electronics related fields, have a circuit diagram and I am still crapping myself at taking the covers of this thing. What I am about to do could kill.

Underneath now.

The PCB is double sided but not plated. No solder resist so hand soldered. No silkscreen. Not impressed, especially as I notice what might be a wiring error (white/red wires go to a green wire and not what appears to be the pin intended for connection).

Most of the wiring on the right is from the output transformers. The amplifiers is ultralinear so 5 wires per transformer.

the potentiomters are for setting the bias for each output tube. I think Kudos for using decent size pots for this and not presets.

First thing check the power supplies.

The iron items on the right are 320mH, 600mA DC chokes, these are wired between the 500V smoothing caps (2) and a seperate pair of filter caps for each channel. The "regulated" or rather "isolated" supplies come from the latter with each input tube having its own "isolated" output.

The isolated voltages are 20V higher than they should be. will need to look into this.

The heater supply is top left near the orange cap, This was not working correctly, half the rectifier has blown up (seems a common problem for me), replaced it on this side of the PCB. Now the input tubs glow nicely.



here is the dead rectifier



One of the channels had blown its plate fuse, most likely culprit is a cathode resistor which is 20R wirewound. By measuring the cathode to ground resistance with the tubes removed should find it, and sure enough we do. But to replace it means getting at the other side of the board and that means de-soldering some wires. I desoldered the left hand side ones after making lots of photos. This is quite scary. Before messing with this always check the big caps are discharged before approaching. Fortunately the resistors across the caps discharge them fast and the resistors are in good condition.



Crikey, theres a lot of expensive caps in here.



I am using an I-pod for source, it works well. It works fabulously well with my friends I-phone, but I don't want to blow that up.



Hopefully you can see the blown cathode resistor. near to it you can see why it went bang. Things have a habit of falling through the holes onto circuit boards in my house. This is one bad thing about circuit boards without a solder resist. Going to examine the rest of the board for FOD.



These presets are to match the gain in the phase splitter. No need to touch them yet.



Near the bottom is the circuit that allows for the bias setting. There is 1M between the cathode and the LF353 input and a 6V2 Zener to ground. I confess I don't really understand how this is supposed to work. Presumably the zener is there to prevent a big excursion taking out the op-amp.



Power supply isolator previously discussed. The high value brown resistors tend to drift, and these 560K resistors actually measure close to 700K. I will remove them and place them on the other side of the board for easy replacement later, maybe when this warms up the will be correct. The other resisters measure OK. This might explain the elevated voltages I read on the isolated supplies.




I think the 50uF caps are there to keep the input tube heaters alive when the power goes off long enough for the HT supply to discharge.



20R W/W resistors are also used to connect the screen grids to the ultralinear tappings.



Input valves go here



I have removed the Shakeproof washer and the blown resistor at this stage



Ok after re-assembly, I am using only 4 of the 8 output tubes. Each transformer tapping is driven by two tubes in parallel. We should be ok with just one tube so we have 50W instead of 100W, at least to check the amplifier is working.

At this stage I have not replaced the plate fuse (they are Very Expensive, only have 3 spare). The amp will actually work with a very small current (there is an LED across the fuse that tells you it is blown, the current through the LED is sufficient for the amplifier to give a very low output when it works. Of course the bias is then wrong so the LEDs are lit.



Plate fuse is in now, amp is running.



Lights out



Start to put some covers back on and audition, while thinking about getting new parts. Those W/W resisitors are outside spec and maybe should be replaced. The high value Allen Bradleys are easy enough to find.



It sounds very good, not sure replacing the Vishay W/W for KOA Film type is worth the risk, open to suggestions.