Sunday, April 6, 2014

Korg Triton LE Power Supply issues; heating up, distorted sound

This Triton came to me completely dead. The first thing I checked was the power supply, a 9v AC 3 amp wall supply with a 4 pin DIN connector... very much the same as some Alesis products... the DMPRO for example, which mentioned elsewhere in this blog.

I found a blown 2 amp inline fuse in the power supply, which can be disassembled (with the help of a drill). I am sure quite a few of these units have found their way to a trash can, with their only issue being this blown fuse. 2 amp fastblow and slow blow Pico fuses are available all over the internet.

After replacing the fuse, I turned the unit on, until to find that it was heating up, and only produced a horrid distorted tone.

I checked the voltage rails first, and noticed the 9 volt rail was substantially low. Through a process of elimination, and disconnecting one rail at a time, I came to find out that the 5 volt regulator had actually failed, and was sending excess voltage into the 5 volt rail, and consequently dragging down the other rails. It wasn't an extremely high voltage, at about 6 volts, and but it was enough to drag down the other rails.

This posed a substantial problem as this particular Korg uses a tiny surface mount switching regulator, and a tiny surface mount FET in conjunction with it. I found these parts in China, on ebay.

Once installed, the new regulator, fet, and a few caps (for good measure I replaced the filter capacitors around the regulator), the power rail came up and the unit turned on and worked. I also added a small 5.6 volt zener diode, which would short the rail to ground in case the voltage ever exceeds the specs again, in the hopes of protecting the chips.But this was too late...

Apparently, one of the PCM roms was damaged... probably by the previous over voltage, when the regulator first failed. There are four of these PCM roms on the Triton LE, and they provide the basis for all of the preset sounds and programs.

On this particular unit, after its resurrection,  some of the sounds, predominantly the organs and some brass sounds, sounded great. But others were distorted... primarily all of the piano patches and string patches. These must be looking for those damaged PCM samples.

Sadly, Korg no longer supplies these PCM roms, and so we had to settle for a partial repair, until such a time as I can scavenge these chips.

On the bright side, the sampling feature was restored, and many of the cool organ patches.. and probably about 50% of all the other patches sound perfect. But there was nowhere I could obtain those PCM roms.

However, if someone reading this has an old triton LE mainboard for sale, from which we could scavenge these chips, please let me know!


  1. Hi Chris, thanks very much for the information I will do the changes on my Korg TR 88. I figured out the weak design on the schematics and possibilities of damages due MOSFET short circuit on 5V rail output. After reading your post with pics I will replace factory P channel MOSFET with a T-220 case type to avoid further damages. Regards.

  2. Another problem on Korg HI synthesis is the sound engine TGL-96. The chip integrates an interpolation sampling processor, DSP effects processor stage and DRAM-ROM memory controller for samples. On normal use overheats to much which degrades its logic, so a passive cooler is strongly recommended to cool it down. I do that for safety because logical failures on TGL-96 chip is one of the reasons about distorted sound, Regards.

  3. This comment has been removed by the author.

  4. The reason why the P-Channel MOSFET entered in short circuit is simple. In the case of AC power drops or fluctuations the PWM controller compensate that because has a wide input range from 2,5 to 16V. On the MOSFET the current goes up if the input voltage goes down. An improper or low transformer voltage (below 9V AC for example 5V AC) will trigger the damage on MOSFET because the switching current starts to increase to compensate the voltage drop on the AC input. The PWM controller due its simplicity has not protection for that meaning that the root cause problem resides on the use of a weak MOSFET. At the same time 90% of power consumption hangs on 5V rail output worsening the problem. Regards.

  5. Hi,

    I fixed many Triton LE and TR boards .. the only way to get the PCM ROMS is from a Triton Classic 1st Generation boards, it use the same 4 pcm roms, Laster Triton Classic use smaller pack and double capacity.

  6. Thank you for this information my korg le 61 the buttons combi or program is flashing.As when the light does not have power But in the part of the smart media card is hot and I opened it is very hot the coil and this distorted sounds please i need help  I want to know the value or how to order diode smd is 3132

  7. which terminal of the adapter carries 9v and 5v help please

  8. In my opinion korg or erred in the design of this power supply including both the external psu and the circuit that stays on the motherboard. This external power supply should be switched as it did on subsequent models or there should be an internal protection circuit that protects the motherboard against extreme AC variations. Then I conclude that either the engineer was negligent or incompetent or even was conniving with the korg's intention to create a trap for users with the intention of selling new keyboards to the unlucky ones that would cause them to be burned by plugging the psu into 220v. There are thousands of keyboards from these bottlenecks around the planet that should have been working to this day like many other older korgs. If they were honest and respectful of their users they should have made a recall so that the authorized repair shops installed internal protection on the boards or continued to manufacture new replacement boards already with the correction of the problem thus causing their eternal survival. If the keyboard is so good in all aspects, why did the power supply have to be bad? Unfortunately the users are not united and demanding because if they would have made a class action against korg to require at least receive a new mainboard