Saibot,
Thanks for the diagram and measurements. It makes it a hell of a lot easier for me, or others to debug your circuit without having to go there and actually poke it with our own multimeters like geeks in heat
Heres my thoughts on whats going on.
Firstly:
I dont know the code for the schottky diodes that i got. I bought two, but one of them broke (like the glass just shattered.)
Glass diodes are almost always designed for signals, not power. Therefore their rated current is much lower, usually about 200mA maximum. Now, I'm pretty sure that a single charger circuit would take about 200mA, maybe more. In the configuration you have (2 parallel groups of boards, in series) then because you have two paralleled, combined they take 400mA. This, if not enough to blow your shottky diode is most certainly giving it greif, thus its ridiculous voltage drop
Alternatives are: 1) Get some power shotky diodes, rated at 1A+ such as the 1N5817, 1N5819. OR 2) Use Standard power diodes, non shottky, just plain old silicon power ones. Aka the 1N400X (X = 1 to 7). The only reason shittky's were suggested in the first place was because we assumed (wrongly) that you wanted 3v from a 3.3v supply. And shotkys have a smaller voltage drop.
Secondly:
The confugration you've used, using both paralle and series configurations, in an ideal world should be fine. If all went to plan, they would each have 1.5 across them, and drawing the same current. In the practical world things rarely operate that way.
For a start, these circuits will not 'drop' a certain voltage. That depends purely on how much current they get, so you must regulate either one or the other and its FAR easier to regulate their voltage, and let them draw whatever current they need. Also, the problem with series configurations is...the output of each is referenced to 0V. That means that usually (may not be) the negative of the output (going to the - of the cap) is connected to the GND of the input. In series you have two 'GND's, which are at different voltages. This means that current can flow from one circuit to the other, when you connected them up. The ONLY exception to this is if you are using each 'paralleled' group to charge a seperate cap bank. If you are charging one big bank, then you can't have them in series.
So, sorry to tear up your circuit but here's what I would do
Put all boards in parallel. So that means all the negative inputs (GND, 0v) are connected. Do the same with the power input. I understand you want them to have 1.5V across them, and you have a 3.3v supply, but you will have to use diodes to drop this, as putting boards in series, at least for YOUR purpose, isn't wise.
Refer to the diagram I provided in a previous response. In that simple configuration, each board gets the same input voltage (1.2V to 1.5V) but will draw its own current. If they are all the same type them they should all draw the same current. Also use standard pwoer diodes, a LOT easier to get hold of. Perpas its not that efficient, but its safe, and should be reliable. So at least its a start so you can get things working. You could change the circuit to convert 3.3V to 1.5V later on. First, get things working without blowing stuff up
(took me years to learn that)
If you don't connect the outputs together, you can measure every individual boards output. Should be more or less the same, but without a capacitor connected, its pretty muich meaningless, as the CAP determines the output voltage. It will rise as the cap charges, then level off at around 300-320V. Unconnected could be up to 400V+. You can now connect up all the outputs in parallel. NOT series. In parallel everything has the same voltage, but is capable of providing more current. In series, everythign gets the same current but has different voltage drops.
Now, to explain why you measured the voltages you did:
These units are designed to operate from a single AA cell. When you draw a lot of current form a cell, their voltage drops. So even a new AA battery, when powering sometihng that takes 200mA, its voltage will drop to about 1.3V. Also, as the battery wears down, its voltage drops, so anything designed to be powered by them usually can operate down to 1.0-1.1V. So the units that had 0.9V across them, were in fact working quite well (400v output). Where as the units that had 0.6V across them weren't, because thats below the voltage of a 'dead AA battery'.
As I keep saying when you measure the output of these boards,
without anything connected to them, you'll measure huge voltages. This is because, as you draw more current from them, their voltage drops. Without anything to draw current (multimeters don't really draw anything) theirs nothing to hold the voltage down so it could reach 500V.
When you connect a capacitor across it, its voltage will drop, and, as the cap starts to charge, it'll rise slowly, settling at a voltage UNDER the 330v which the caps are rated at. So, don't worry about 400V, it won't ever get there when charging your capacitors...providing you don't give the units more than 1.5V each.
Sorry for the long post. I hope that explains all.
Blueteeth