Line 6 Lowdown LD175 interesting problem

[Mick Bailey]

I've just been working on a Line 6 LD175 that after about 15 minutes started to play distorted for a few seconds and then reset to its power-on condition. I traced this to the 3.3v supply, which slowly dropped to 2.7v and caused the processor to drop out. The amp uses an LM317, rather than using a fixed voltage regulator, which means using a pair of resistors to set the output voltage. The regulator tab is bolted to an area of copper on the PCB with a standoff used as an additional heatsink, but despite this the regulator got sizzlingly hot. Where are the resistors mounted? - on the back of the regulator tab on the other side of the PCB so they get cooked.

Now, the regulator should have an input of 8v according to the test point on the PCB, but it measured 11v. Just to be sure I checked all three LT supplies and they all read high by x1.414, despite having independent rectification for +8v,+16v and -16v. They do however all share the same secondary winding. I checked the transformer strapping to make sure it's configured correctly for UK 240v mains - there's a common (black) 0v, yellow (220v) and red (240v) primary configuration, selectable by plugging in either of two positions. My fix was to remove the resistors and replace them with 1/4W metal film leaded types mounted slightly off the board and connecting directly to the regulator pads. This is easily done on the reverse side of the board and there's plenty of room. I also made up a heatsink from sheet metal and used thermal compound. Now the regulator gets barely warm and the output is stable, so this should hold up.

[Jon Snell]

Just proving that training a budding engineer to check for correct power supply voltages and ground points, before 'jumping in at the deep end', is the easiest way to start the fault finding process.
Poor component placement is a common issue in Chinesium equipment I find.

[Helmholtz]

I checked the transformer strapping to make sure it's configured correctly for UK 240v mains - there's a common (black) 0v, yellow (220v) and red (240v) primary configuration, selectable by plugging in either of two positions.

So was the PT actually wired for 220V? If so, this would account for only around 10% increase in secondary voltage. But you measured over 40%. What's the reason for difference?
Dropped load current(s)? What did I miss?

[Mick Bailey]

The secondary is correctly wired for 240v according to the markings on the PCB and transformer wiring configuration. I measured the actual voltages applied to the transformer primary as well as checked the pinouts on the mains input board where the voltage is selected. The secondary output is a 9v AC loaded. This feeds a voltage doubler for the +/-16v supplies respectively, and the 8v supply is 1/2 wave rectified off the same feed. I concluded that the amp was made this way - I can't find any evidence of a fault with the transformer or its wiring. It's not uncommon to find amps running over-voltage here, but this seems to be too high to account for supply differences.


Edit,

I just found the transformer spec in the service manual and the output of 9v AC is correct according to the drawings. So this means the marked test points can't be right. Also the primary wiring colours are different between the drawing and the actual amp. According to the manual this particular amp has the wiring colours for a 110v/120v transformer. To get the test points to measure correctly the mains has to be at 175v on my variac.

https://manualzz.com/doc/24195168/ld...anual---guitar

[Jazz P Bass]

Thanks for the manual!

[Jon Snell]

More than the described voltage on the input of an LM217 is of no consequence, as its maximum input voltage is not exceeding the permitted value of 40volts.
As long as the output voltages are correct, the regulators are fine.

[Helmholtz]

But the dissipation of the LM317 will be higher by maybe 40%.

[Mick Bailey]

Exactly - the input voltage is within design spec, but the heatsinking is inadequate for the extra dissipation and there's excessive localized heating of the PCB.

[Jon Snell]

That is why my comment regarding Chinesium equipment; post 2. Good design but poor building with no thought to heat dissipation etc. One would have thought the makers would have thought about the extra dissipation ... maybe not, we will never know.
A little like the Fender Blues Juniors being used in the UK with 240volt mains voltage here; the output valves are over run as the biasing in set for 220volt mains tapping. They run at 100% sometimes and that is without signal! Change the 22k for 33k in the bias chain to cure that, after binning the old EL84s.