Boost converter cuts off

[Kyrk]

I am trying to power a small class d amp with a boost converter like this https://www.amazon.eg/-/en/Yeeco-Converter-Adjustable-Transformer-Stabilizer/dp/B074J9D278or this https://www.ebay.co.uk/itm/252676342631?hash=item3ad4af0367:g:prIAAOSwJ7RYSpuR  they both use the same chip, xl6009, and they have a very similar design. They should be able to accept up to 4A and output 2.5A according to spec. The amp module is very small, i am not asking more than 2-3watts from it.

I've tried a number of configs. I want to supply 12v to the amp using a 2s battery (so from 8.4 to 12v conversion), the issue is that both of the modules are unable to handle the current spikes while playing bass at moderate power making the amp to cut off.
Powering the amp directly from the batteries works fine. I tried using 3 batteries in series (Around 11.4V) and boost to 12v, the converter still couldn't handle it.
In a previous project it could supply constant 5v to a smaller amp using a single cell battery.

I've also tried the same with another boost converter a more beefy one https://www.amazon.co.uk/LAOMAO-Converter-10-32V-12-35V-Adjustable/dp/B00HV43UOG, this worked fine but it's much bigger than i would like. What is the issue with the aforementioned converters can I fix it somehow, is there a better option at the same size?

[CountChocula]

I would guess that the problem is the PSU's transient response—when the amplifier needs to play a loud sound, it suddenly draws a higher amount of power that the supply cannot provide fast enough, causing the voltage to droop and the sound to cut off. This could be caused either by the batteries not being able to supply enough power, or by the boost converter being too slow to respond to the sudden load increase.

I'm not sure that you can really do much… you could try adding a largish tank capacitor on the output of the boost converter (even if just to validate that this is the problem), but that will likely create other problems, and would probably just end up taking up more space than a better power supply.

[Kyrk]

Since the batteries on their own can power the amp just fine, and the beefier boost converter also doesn't have an issue, I guess we can conclude its the boost converts fault. I just thought they would be fine supplying the transients because we still talk for less than 1A. What kind of cap could i add on the output? Do you have in mind a better boost converter in the same footprint?

At the end of the day i can just change the battery to 3s and not use a boost converter, but I am really limited in space.

[CountChocula]

Since you're building a one-off product (I assume), you can just try to experimentally find a cap value that works. Right now, you have 220µF on the output; you can add more capacitance in parallel until either you get the behaviour you want, the thing gets bigger than you're willing to deal with, or the circuit goes unstable and starts oscillating out of control.