Sir David Wheeler Esq . . . . .
CONSIDERATIONS . . .
So the main switch is definitely OK and you definitely were shorting the battery connector box out. If that is using Dee size cells . . .vice Cee's . . . and a sum of 6 of them, a dead shorting across them, would have each of them pumping out fom 5-6 amps / 30 w for short durations. BUT the consideration was that you were shutting off power to the main system, alhough, be it in a quite spectacular way.
I see a ver r r r y l o o o o o w probability of ANY service info being forthcoming or being found for that unit at this time..
Now, with you having the main power switch at its on position at all times.
Confirm for me, does 1 jump across, accomplish that desired effect or might it take multiple times to finally end up with the unit coming on, working normal and renaining on until you finally turn it off with the main power switch ?
So lets blaze ahead and see what facts we can develop from a close in photo examination of the boards topside and the bottom foil trace interconnects.
We see that the closest terminal to the corner will be the DC adapter power plug then the battery connector plug and the final power switch plug, with the two internal speaker plugs being of no interest to the problem.
Looking at the foil side of the board we see all negative leads soldered to the ground plane foil, the positive leads float, but they are being within TINY foil islands.
I see the corner plug + connection traveling down the edge of the board until I can see it no longer.
I can not see either of the two other plugs trace paths since they are on the other unviewable foil side of the board.
The most power used in that unit would seen to be with that AUDIO amp IC that you see right beside the speaker plug connectors. Read its part number on its top and pass to us so that we can see if 9 volts is being at its lower parameters, that the unit will work at.
Otherwise, I could see that, being hidden back in the left quadrant of the circuit board, that they are using a switch mode boost converter to hop up the supply voltage .
Your disrupting the power with a short . . . . and its resultant loss of power . . . . could KICK the inoperative converter into operation, as the power comes back.
That being considered, and not shorting the battery connector, instead , how about a more direct approach by using just battery power and then you just turn on the power switch and unplug and plug in the battery connector to blip-blip-blip-blip-blip-blip the power on and off several times to see if the unit jumps into operation just after one of those power plug in / turn on blips.
That replicates your shorting action in turning the unit on and off but the batteries will be happier.
Also you might just try flipping the main power switch on and off in the same manner.
You say . . .
Update: I have also had similar success with a perfectly timed jamming in and out of the charger in the back of the device.
Which means that this below is then confirmed . . . .
Stepping ahead, with not yet knowing your prior results, then leave the power switch in its on position and move down to the AC wall warts end connector and accomplish the same fast on off effect with the unplugging of THAT connector into its receptacle to see if that makes it work . . . . if your timing is right . . . . and you are holding your tongue, now pressed into the side of your jaw . . . just in the right position.
Leave it plugged in if the unit comes to life at some attempt.
Until I see a good photo that will let me see the foil path interconnects between the power switch, the battery holder, the wall warts connector plug AND their actual power connection being eventually made into the circuit board; I am currently seeing the wall wart connector plug directly having its connections co-joining with the battery connector.
Then the power switch connects their shared + connection into the + buss of the circuit board.
E- capacitors . . . . I see 4 of them across the board near the connectors 1 at the center of the board near an 8 pin IC . . .its numbering I.D. is being needed . . . . and it is of a skimpy/minimal 10VDC rating . . . used with a 9V power supply ? ? ? ?
The next two units sit either side of the AF power IC, I need the IC numbering and the voltage rating of those caps in case that chip is operating at > that the 9DC of the battery supply (via an on board power up converter circuit ).
Then we see the final E-cap in that row and it looks like a 10V rating on it also.
Then nearby, sits a dumbell shaped ferrite cored inductor with its magnet wire winding. That could be used in series with the incoming power for EMI / RFI /HASH AND CRAP filtering . Or it COULD be associated an earlier mentioned DC up converter supply, as being its ringing inductor, but NOT into that NEARBY cap with its mere 10VDC rating .
Seems like at least two E-caps are hidden from my viewing, at the other end of the board, can you supply their values and VOLTAGE ratings . . .plus positions .
Another photo of that area might answer all, as to there being an up- converter power supply hidden in there.
A deteriorated, marginally low value E-cap in those positions could account for exactly your needing of " kick starting " the unit .
Thaaaaaaasssssssssit . . . .fer now . . . ..
73's de Edd . . . . .
The things that come to those that wait, just may be the few things left by those who got there FIRST.
.