UK Vintage Radio Repair and Restoration Discussion Forum

I'm a sucker for a neat looking valve amp especially one that's crying out for attention. I picked up this DULCI at a recent Audiojumble. Its a pretty exact copy of the Mullard 10 watt amplifier circuit. It looked rather forlorn and was minus all the valves. I was in 2 minds ,in case either of the transformers was faulty, but the seller generously allowed me to remove the baseplate and test these with a conveniently placed multimeter. All was well so I took the plunge. After a quick clean, and some basic cold tests, the only problem areas to be sorted before power up were

1. open circuit main electrolytic ( both halves )
2. Leaky coupling caps to the EL84's ( already replaced at some point but 'past it ' )
3. Remaining HUNTS caps ( 2) to be replaced for good measure.
4. Cathode bypass caps missing on the EL84s ( 270 ohm cathode resistors had been replaced before, so perhaps 'forgotten' )
5. New mains lead ( an 8 inch lead is little inconvenient ! )

The electrolytic was a dual 50+50, lug mounted into a paxolin plate, riveted to the chassis. After careful extraction and a rummage in the parts bin, a 'Radiospares' 50+50+50 was found and tested fine on the ESR meter. Taller than the original, but reasonably 'period'
The new Caps were some 100nF /800 V orange drop which I found in a long forgotten drawer. The others being normal VISHAY type yellow jobs.

The 100K anode loads on the phase splitter were out of spec. so these were also changed. Mullard state that the mismatch caused by not reducing the value of one of these ( as per the Leak TL12 + ) was insignificant - more on this later.
A set of valves was fitted and power applied slowly. Up came a nice reassuring hiss ( over sensitive input, as designed, alas ) and the generator, power meter and scope attached to see what was what.

All seemed to be well and it looked clean on the scope at the rated 10 watts. The frequency response checked at 1W was virtually flat down to 10Hz and extends upwards to 30Khz so not bad at all A slight 'droop' is noticeable after 10Khz but this seems common with a lot of these designs. THD at 1Khz came in at 0.4 % ( spec is 0.3) at 10 watts. There was a problem with the ECC83 phase splitter, which would crackle as soon as the valve was touched. Having checked the base and all the soldered connections and applied some DE-OXIT, this turned out unusually to be the valve itself ( ostensibly a NOS Mullard ) - that's a new one on me. At first, I could not quite believe it, but fitting another ECC83 effected a cure - Weird !

As small diversion, I decided to check the balance of the drives to the EL84s and found they were not quite exact. To see if this really mattered, I attached a 1Meg pot across one of the 100K Anode loads and adjusted this until the drives were equal. This was then removed, measured and padding resistors of the same value ( 630K ) were wired across the anode load. Re-testing showed this had made an improvement and the THD was now down to 0.25% at 10 W.

As my bench isolated AC supply is slightly low at around 225 volts, I then repeated the test with 'full mains' This produced marginally more output approaching 14 W and the THD was still less than 1%. For a 'budget' amp. this is really rather excellent.
Just to test for true hum and noise the input was grounded and I could then only hear a very faint 100Hz hum with my ear right up to the bench test speaker. Providing the source impedance can be lowered ( with a suitable resistive pad ) then first stage noise won't be a problem.

All I need to do now is find another one !!

A.