Want to kill RF oscillation from audio amplifier IC

[Northe]

I bought a Joyo JA02-II 5 W battery guitar amplifier. My intent is to use it with crystal radios and to make probes to use it as a signal tracer.  It works fine, and using it's "overdrive" function gives me a lot of gain. I actually sounds pretty good, too.

However I discovered that the amplifier has a free-running oscillation from the CS81385 IC amplifiers in the mf broadcast band. If I put my finger on the IC it moves the frequency, may even stop it, but take my finder away and it drifts back to approximately the same frequency.  I tried wrapping aluminum foil around the amplifiercase, which substantially reduced the radiation.

I'm wondering if someone can offer suggestions to stopping the oscillation without a a lot of work on the teensy-weensy PC board with macroscopic surface-mount components.   I figure worst case I could shield the inside of the plastic case with copper foil tape or carbon paint, but would really like to be able to actually kill the oscillation.

Thanks in advance for any suggestions.

[BrokenYugo]

CS3138S (not 5) looks to be a class D amplifier, it's supposed to run at a high frequency, just not let that out. You probably need more filtering on the output and maybe power input, but that's hard to say with no pictures of the board or schematic. Looks like the datasheet is Chinese only. Switching frequency is stated at 500kHz which would put the second and third harmonics right in the AM broadcast band, so that's probably what's going on.

https://datasheetspdf.com/mobile/978910/CHIPSTAR/CS8138S/1

Note the couple mentions of ferrite beads and 1nF caps to ground on the two output pins, are those present?

[Northe]

Thank you.
You have reminded me of my ignorance concerning class D audio amplifiers.

After peering at the mostly Chinese data sheet using Google Translate it finally occurred to me that it was a class D amplifier. And, yes, the interference is more or less around 1500 kHz.  So far I have not seen a schematic, for the amplifier board but I'd place a bet that the design is probably standardized and used in a number of entertainment applications. I'll keep searching for one.  And yes, it really seems that the RF is getting out primarily through the speaker/output pins

As far as the filtering on the output/speaker pins, there are lots of surface-mount chips on, but I'll have to peer even closer to see if there is any evidence of ferrite beads.  So far it appears that adding a 0.22 or 0.33 uF capacitor right across the output pins may be killing the signal emission with the amplifier still functioning. It's difficult to tell because I'm using 400 V caps and touching the pins with the leads. I'll probably see what sort of small low voltage caps I can order so I can tack them on the board. 

Thanks again!
Northe