Anyone used MAX1044 to produce V++ X 2 as in Okko Diablo?

[bordonbert]

Hi guys.  I'm building a Diablo Clone and putting in the 18V rail option.  The MAX1044 circuitry is simple enough but it has a confusing aspect.  I've tried to attach it below, I hope it comes through ok.

I have looked into the suggested applications in the Maxim datasheet which is very clear and helpful with a voltage doubler mode straight out of the Diablo schematic.  The 1044 works by alternatively applying the rail voltage and the ground to pin 2 the base of C18 and top of C20.  This allows C18 to charge to almost V++ (-Vdiode) before lifting it onto the rail voltage to dump its charge into C19 bringing that to around 2xV++.  C20 is irrelevant to all of this and while it is all going on, it's bottom end is alternatively linked to the ground and pin 5 which is not used.  Why are the two systems being used in the Diablo circuit?  For voltage doubler mode surely C20 is redundant.

Is there anyone who has experience with the MAX1044 or the Diablo who can point me at something I've missed?

[madbean]

C20 doesn't need to be there as you said. Perhaps it is just a design error.

[bordonbert]

Thanks Madbean.  Logic told me it was not needed but there is always the chance that I missed something that an experienced used of this chip would know.  If there is a genuine improvement in performance for the cost of one cap it's worthwhile.  Thanks for the confirmation, C20 is out!

[teknoman2]

I did a while ago a sch for the okko diablo on eagle,
but I never had the time to finish the brd.
You can use it if you like it.

Cheers,
Panos.

[bordonbert]

That's great Teknoman, thanks for that.  It clearly shows that the additional cap is not there.  That's good confirmation that what we are saying is correct and the cap is a mistake and not needed.

[bordonbert]

One little thing I've noticed is that your power line shows no smoothing right at the input, only the reverse polarity protection diode.  All of your 200uF smoothing is after the 9/18V selector switch.  This puts effectively 250uF across the MAX1044 setup on 18V selection.  It won't strain the chip but I'm not sure of the ability of the preceding 47uF cap to top up the 250uF in the same way.  Instinct tells me it should make no difference as the overall current draw from it is the same as the original.  It will simply take a little longer to come up to 18V and should smooth the line better.  My original schematic shows 100uF and 100nF across the input and 100uF after the switch.

[teknoman2]

Trully correct mate,
generally the powering filtering in my sch isn't very good cause it was one of my first designs and attemps on eagle.
Indeed if you use 100uF and a 100n box cap before the voltage doubler the smoothing is better.

[bordonbert]

Woah!  Your 2 caps on the MAX1044 are 10uF, both of mine show as 47uF.  Now that is a significant difference as it affects the amount of charge that can be dumped into the reservoir capacitor on each cycle.  The MAX1044 datasheet application notes show capacitors as 10uF for most of the other applications but for the voltage doubler circuit they are not given a value.  As long as the chip can handle it bigger is better here.  It passes more charge each cycle and gives more storage capacity to offer higher currents for the same ripple.  And we have to remember the ripple will be at much higher frequencies than a standard 50Hz PSU.