mfortuna wrote:The data sheet shows a universal circuit that includes a 400V cap after the bridge and connected to the 1200P60. It also mentions a 450V maximum on the HV input.
It does indeed, but the circuit must be configured to accommodate 450Vdc applied to pin-8. More below.
It seems Irobot could have designed a universal charger.
Of course it could. It could also have made the Fast Charger
function on 110-250VAC mains. The Company did that with its Rapid Charger product,and it had a chance to see what it cost. Thereafter, it refrained from doing more than was necessary!
...Can you explain the damage scenarios?
As I alluded, the circuit must be first rev-engineered to see what is there. W/o knowing the high-voltage section's wiring, a whole lot of guessing is required. Notice in the (On-Semiconductor ...) Figure 1. "Typical Application" schematic there is a TBD-value resistor connecting the rectified-mains voltage to pin-8, And from there an asterisk takes you to this note below the diagram: "
* Please refer to the application information section". You can get a pretty good idea of the trouble a designer must endure to make this device work well by reading that application information section.
Suffice it to say: a) VCC (pin-6) for the 1200P60 appears to be developed by passing the HV on pin-8 (see On-Semi. Fig. 2) through an internal constant-current section, and through undefined impedance to GND, b) VCC_max = 16Vdc, c) by doubling HV, the developed VCC should double, and d) if VCC was anywhere near 10Vdc for the 120VAC design, it would go over the top when 240VAC mains powered the HV-section. If exceeding that "16Vdc MAX" means the IC is fried, or simply won't work over 16V, I don't know.
But, notice these other things. Figure-2 shows there is an overload function. Maybe that has saved the 1200P60! By reading (in the application section below the graphs) about "Dynamic Self−Supply" (DSS) you may see there is another way of handling a large HV value by using a "Dstart" diode.
You will also see words like "with an auxiliary winding. It will automatically disconnect the internal startup source and the IC will be fully self−supplied from this winding." as another means for accommodating a high supply voltage. I am doubting that iRbt followed this path because omitting the tertiary winding from the xfnr is one oi the apparent cost-savings the Company would use.
Is the DC output voltage determined by the zener voltage plus the drop through the opto-isolator {LED}?
Yep, that method is quite typical. Roomba's FC ckt is not that simple, and I expect the SPSU's is also more complex, but basically that's the idea.
