This is a quick post to record a problem and fix I had with my old Teac BX330 Audio Amplifier.
They are basically a good old amplifier. Their primary feature is that they are DC coupled, meaning there is no bottom end cutoff frequency for those audio buffs so focused on purity of their bass notes.
You will need a service manual to follow this. They are available on-line. I paid for mine but I notice they are now available for free download.
The symptom was that the speaker protection circuit refused to engage relay K301. This relay activates the audio to the output.
The protection circuit was doing it's job.
The problem turned out to be a failed JFET Q203 in the front end of the DC coupled amplifier. This JFET was leaking +15V to it's gate The right channel was the fault trigger but the left channel also displayed the same fault to a lesser degree.
The JFETS are PN 2SK109C. I was unable to locate a source or any detail on them and with no serviceable device from which to determine characteristics, finding an alternative was going to be something of a guessing game.
Going by the circuitry and voltages present, I came up with Fairchild JFET PN J112 as a suitable alternative.
The original JFET is a matched pair in one case. The alternative devices are separate so you need two for each channel and you need to source them from the same batch. I was able to source them on eBay.
You will need to bend the pins so that the Gate and Source are in the correct location to match the Q103/Q203 footprints. The Drains are at either end of the original device, Gate next, then Source. Note the order on the J112 is Drain, Source Gate, hence the requirement for some pin bending.
The adjustment procedure on page 5 of the service manual effectively sets the DC voltage on the speakers to 0VDC. This cannot be achieved when the faulty JFETS are in circuit. On the worst (RH) channel the minimum I could get was 16VDC. That would destroy any connected speaker if it weren't for the protection circuit. If you find the adjustment is range limited it's time to make some minor modifications.
I found the left channel adjusted to zero as is but the right channel needed some slight circuit modification. Namely R222 paralleled with 10K and R223 bypassed with a shorting link.
