Driving 'Fiddy's Flybacks' in flyback mode. / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Driving 'Fiddy's Flybacks' in flyback mode.

first previous 8 9 10 11 next last

Move Thread

LAN_403

Ash Small

Wed Nov 19 2014, 08:26PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

I've been giving the board layout quite a bit of thought, and, as I mentioned above, it really does seem that using as many SMD components as possible is by far the best way to reduce parasitic inductance, as this facilitates the shortest distances between other conponents, thus keeping wire and track lengths as short as possible.

Having read this article,

along with the comments below it, it would appear that 'reflow soldering' isn't beyond my ability, so the current plan is to use Maplin stripbpard, with SMD components mounted on the side with the copper strips, and all the non SMD components mounted on the reverse side of the board.

Then I can mount the SMD components in one go, and then flip the board over and mount the non-SMD components, soldering these by hand.

It could take a while to sort out the board layout especially as I haven't finalised all the components yet. The other thing I still need to give some thought to is the foil primary. My initial thought on this are aluminium foil with paper insulation and crimped connections, with twisted wires from the board to the flyback to reduce inductance.

Shrad

Thu Nov 20 2014, 10:25AM

Registered Member #3215 Joined: Sun Sept 19 2010, 08:42PM
Location:
Posts: 780

what about a custom lamination tool to flatten classic copper wire?

I guess it would be relatively easy to construct a set of successive steel rollers which would be geared to a motor... no?

Ash Small

Thu Nov 20 2014, 12:25PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Shrad wrote ...

what about a custom lamination tool to flatten classic copper wire?

I guess it would be relatively easy to construct a set of successive steel rollers which would be geared to a motor... no?

Copper foil is easy enough to obtain. The simplest 'custom tool to flatten wire' is a hammer and anvil

Shrad

Thu Nov 20 2014, 12:45PM

Registered Member #3215 Joined: Sun Sept 19 2010, 08:42PM
Location:
Posts: 780

yeah, but I meant "in a reliable, repeatable and consistent way" ;)

Ash Small

Sat Nov 22 2014, 07:55PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Well, I'e started thinking about board layout, and come up with the sketch below:

(This is 'looking down onto the 556', with pin 1 at top left, although the SMD components are on the reverse of the board. It's just drawn this way, to show the relative positions of everything.)

The IC socket and MOSFET (not shown) are on one side of the board, the SMD components are on the other, with the copper strips.

Blue for SMD ceramic capacitors, red for thick film SMD resistors and green for SMD Shottky diodes.

There are currently ten x 10uF ceramic capacitors between Vcc and GND, equalling 100uF right at the 556. I've gone with the concensus and used Shottkys for ground bounce protection, with just a gate resistor between the OUT and GATE pins.

I've not added the potentiometers or switched capacitors yet and I've not added the components in red on the schematic.

I've included the schematic in the sketch, so it's easier to follow.

Once again, any comments/suggestions will be welcome

EDIT: I've just noticed I've left a capacitor off pin 11, and maybe I should add capacitors to pins 4 and 10, with one side connected to Vcc and the other connected to GND.

Sulaiman

Sun Nov 23 2014, 01:55PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

Although schottky diodes are excellent, high voltage high current units aren't cheap,
at the frequencies used for flybacks you can use UF400x, UF540x etc. (x = 8 for 1kV)

I still think that a bunch of TVSs is better than a resistor for the snubber
(try TVSs for MkII)

Where does the +Vcc come from?
if you are using transfromer/bridge rectifier/reservoir capacitor/regulator
you could take the +ve supply for the transformer primary from the reservoir capacitor
this allows a simple low power regulator (7812 etc.) for the 555/556
and max power available to the primary.
Higher primary dc supply usually gives better efficiency
due to 'lost' voltages across transistor, diodes etc. and lower current for a given power.

Last but not least, be sure to use a traditional 555/556 and not the low power CMOS types,
that often do not have the required current output.

Ash Small

Sun Nov 23 2014, 03:46PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Sulaiman wrote ...

Although schottky diodes are excellent, high voltage high current units aren't cheap,
at the frequencies used for flybacks you can use UF400x, UF540x etc. (x = 8 for 1kV)

Do you mean the MOSFET protection Shottkys or the 'ground bounce' protection Shottkys on the output of the 556?

I understood that the ground bounce Shottkys don't need to be anywhere near 1kV. I was actually planning to use 600V devices for the MOSFET protection.

Sulaiman wrote ...

I still think that a bunch of TVSs is better than a resistor for the snubber
(try TVSs for MkII)

I probably will add some TVS's, I've not given the snubber circuit a lot of thought yet.

Sulaiman wrote ...

Where does the +Vcc come from?
if you are using transfromer/bridge rectifier/reservoir capacitor/regulator
you could take the +ve supply for the transformer primary from the reservoir capacitor
this allows a simple low power regulator (7812 etc.) for the 555/556
and max power available to the primary.
Higher primary dc supply usually gives better efficiency
due to 'lost' voltages across transistor, diodes etc. and lower current for a given power.

Initially I'll run everything from 12V, although it will be fairly simple to increase coil voltage at a later date. As I understand it, the 'reflected voltage' is a product of 'turns ratio' and output voltage, so I should be able to increase primary voltage without needing higher voltage components, this should just result in shorter 'on' times, in order to maintain the same 'volt seconds'.

Erlend^SE

Sun Nov 23 2014, 07:08PM

Registered Member #1565 Joined: Wed Jun 25 2008, 09:08PM
Location: Norway
Posts: 159

Quick idea: add a extra winding to return the excess power to the supply. (winding + diode).

Have less windings than the primary, so it's still possible to build a bigger flyback voltage.
Non-delivered energy is returned to the supply, for use on the next pulse)

It may able be possible to have a winding to make the controller cut back the duty when it starts to generate more than a given voltage (feedback)

None of this apply, if a SMPS controller is used, given it's used with secondary feedback.

Ash Small

Sun Nov 23 2014, 07:17PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Erlend^SE wrote ...

Quick idea: add a extra winding to return the excess power to the supply. (winding + diode).

Have less windings than the primary, so it's still possible to build a bigger flyback voltage.
Non-delivered energy is returned to the supply, for use on the next pulse)

It may able be possible to have a winding to make the controller cut back the duty when it starts to generate more than a given voltage (feedback)

None of this apply, if a SMPS controller is used, given it's used with secondary feedback.

Yes, I've seen this idea before, thanks for reminding me about it. It's certainly something I intend to experiment with once it's up and running.

Ash Small

Mon Nov 24 2014, 08:33PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Ok, so this is the board layout for all the components on the schematic except those in red.

SMD capacitors are blue, SMD Schottky diodes are green and SMD thick film resistors are red.

Potentiometers are 22 turn cermet that I happened to have lying around. The switches in the schematic have become 'jumpers' in the board layout, as these take up less space (I think) and are cheap.

I'm currently working on the layout of the MOSFET and schottky diodes, all of which need separate heatsinks, I think. (EDIT: I think one Schottky diode can go on the same heatsink as the MOSFET, so three heatsinks in all, I suppose.)

I'll then look at snubber placement.

The number of smoothing capacitors has increased to 18 at the moment, which is 180uF. I expect it will end up with more. I'm just adding them where it's convenient, can't do any harm, can it?

first previous 8 9 10 11 next last

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob