SMPT power supply LY6806 based

Hello all, I am looking for a schematic or a data-sheet for the LY6806 circuit.

On the website **broken link removed** (which seems abandoned since 2014) nothing available.

This circuit is used in a variable 3V-24V 3A cheap power supply.

Thank you for any information.

it may be difficult to find the schematic. if the purpose is to repair it, you can do it without schematic. most of the time one of the tiny electrolyte capacitor (the one across the power source of the driver IC) become faulty (less value). you can change it and check.

Hello and thank you mbarazeen, no it's not down, to understand and make changes is better when have all the documentation!

It's nt unrealistic to create such a schematic yourself. Exact name of diodes and values of smd capacitors may be not feasible to figure, but it should be enough to try to locate the fault.

Here, part of the schematic made by myself:



Once again, the PSM is not down!

I just want to try to change it to reduce the minimum voltage to 3V or less and keep up to 24V.

One way to see, is to replace the shunt controller TL431A with an LMV431 instead, the VREF changes from 2.5 V to 1.24 V, but not sure that is enough?

Try linking out R12, that should allow it to go slightly lower.

Hello and thanks @rjenkinsgb,

R12 & R15 are just to limit RV1 action but in all case not able to reduce drastically the Vout voltage.
Du to the fact LM431 is not linear the curve transfer looks like this :

A significant change in the output voltage was possible with the replacement of R9 with a value of 150 ohms.
But not enough to reduce Vout under 3V.
One solution I see is to replace the actual TL431 by a LVM431 the differences between both are the Vref, 2.50V for the first, 1.24V for the second.
But not sure for stability ?
Any ideas on this ?
Thanks.

Looking at it again, I've realised that the minimum will be limited by the TL431 minimum voltage, plus the optocoupler LED voltage drop - likely 1.8V or more.


The LVM431 may work OK as far as that goes, with over a volt less drop?
The output voltage range will be halved, though. You should be able to compensate for that by reducing the value of R15, eg. 2K7 or a 4K7 in parallel with the 5K should get it back to somewhere near the full range.

Another option would be re-arrange the optocoupler drive circuit, so it can work at a rather lower voltage:
In essence, add a PNP transistor with emitter to positive, then an emitter - base resistor and a resistor from base to the TL430 / C5 junction.

Then connect the opto LED and its series resistor between the transistor collector and 0V.

That should give operation to at least a volt lower output, but is more likely to cause stability problems.
Try the LVM431 first?

vidalv said:

Hello and thanks @rjenkinsgb,

R12 & R15 are just to limit RV1 action but in all case not able to reduce drastically the Vout voltage.
Du to the fact LM431 is not linear the curve transfer looks like this :

Click to expand...

The TL431 is just a 2.5V adjustable 'super' zener, linear or non-linear doesn't really apply.

R10 and R11 are what makes it non-linear, removing those will make it much more linear.

With those removed you should be able to get closer to the 2.5V minimum voltage of the device, shorting out R12 will make a slight reduction, but with R10 and R11 removed the decrease will be slight.

But you need to bear in mind that the feedback is via an opto-coupler - so the output voltage must be high enough to turn the LED on - so you won't be able to go too low anyway.