Korg Forums

Hi all,
I am a tech who received a Triton LE 61 key for repair. The unit had experienced power issues, maybe some type of spike... anyway when I got it I found the internal voltage regulator had failed and it was sending an excess voltage into the 5 volt rail. I replaced the regulator and fet and come caps and got the unit to turn on and work... almost. Some of the patches, primarily the piano and strings, it seems, are highly distorted, while other patches are perfect and clear as a bell. I tried reinitializing the unit, rewriting all the patches in global mode, and even updated the OS with no success. I entered the diagnostic mode and found "PCM rom error, S1". I have deduced that at least one of those 4 PCM roms has been corrupted or is blown out. I have reflowed the solder around them to no avail. From the service manual, they seem to be parts
320040038, IC MX23C6410MC-10 BK4-H (S)
,320040039 IC MX23C6410MC-10 BK4-L (S),
320040040 IC MX23C6410MC-10 BK5-H
320040041 IC MX23C6410MC-10 BK5-L (S),
or on the schematic ic's 7, 8, 48, 45.

I did call the korg parts supplier and found they are discontinued.

I also checked on Ebay and found a whole broken mainboard for 150, however I am hesistant to buy it since it's possible those ROMS are bad too.

Just wondering if anybody has any thoughts on my diagnosis and any possible workaround or solution for finding parts.
Thanks very much!

Hi Chris, I saw your blog post about the power failures on Korg TR Le also I let a comment about that.

http://fixingelectronics.blogspot.com.a ... 7450738099

Researching the schematics on TR the same switching power supply is in use. The main problem is when voltage on AC transformer goes down below 9V AC. This situation leads to the increase of the current on P-channel MOSFET CHP6302 on 5V output rail which supplies 90% of the overall consumption. If AC voltage drops on transformer the switching current on the MOSFET device increase to sustain 5V regulation... but this can lead to a short circuit on drain-source MOSFET junction terminals. A short circuit on MOSFET fries the mainboard logic ICs due the "sudden unregulated voltage on 5V rails". In other words the mainboard chips are fried because the 5V voltage rail is unregulated => same voltage as external power supply rectified due the short circuit on MOSFET junction.

The switching power design on TR and Le is not smart from engineering perspective because AC power drops frequently happens due socket transformer failures or mains 110/220v power fluctuations on stage. The TR/Le series has not protection on that technical scenario, unfortunately. I never opened my TR 88 but after your experience repairing the Triton Le I will do the changes replacing factory MOSFET with a 252 DPACK case for improved heat dissipation with more source-drain voltage (VDS) and current capabilities. The factory MOSFET case size SOT23-5 is suitable for portable devices such cell phones (marked as F1 on the pic below measures only 3x3mm in size), and here the power dissipation is to weak when MOSFET drain-source current and temperature goes up due transformer AC fluctuations.



All this happens due the simplicity of the PWM (pulse with modulation) switching controller S-8520F50, the device using a similar SOT23-5 case has no power shut down or protections when the unregulated input voltage goes to low or to high. To let clear the question the PWM controller has an input operational range from 2,5 to 16 volts. When AC power adapter voltages drops extremely low for example to 6V AC the rail regulation is sustained increasing the switching current on MOSFET device which can lead to a sudden short circuit on the drain-source junction. The other scenario is the fusing of the MOSFET itself remaining as an open circuit.

On a linear regulator for example a LM7805 the most used regulator this situation would be avoided because the device starts to drop the output voltage when input voltage goes down, at the expense of power efficiency and heat generation due linear regulation theory. I prefer that after all because a linear regulator is safe when 110/220v power adapter fluctuates over a wide range. A PWM switching controller should have protections to avoid the possibilities of damages on MOSFETs stages.

Another issue on MOSFETs is the dynamic degradation of the RDS (resistance drain source) switching feature due device aging. Here the MOSFET heat dissipation starts to increase due leakages currents and the poor switching efficiency, leading over time to a sudden "short circuit/melt down or fusing" on MOSFET junction. Because the PWM switching controller on TR/Le don´t sense input current and voltage there are no protection on that fatal scenario.

What I don´t understand is the use of SOT23-5 mini components on the switching power supply. The SOT23-5 case is suitable for portable devices for example cell phones or tablets. When space is not a restriction the use of conventional ones are the best recommendation. On miniature MOSFETs the specifications on power dissipation, voltages and current capabilities are lower as standard ones, affecting the endurance and durability over time.