ETD29 size transformer for a 'scope clock – design tool or procedures?

[Chris56000]

Hi!

Having spent five hours searching through "dead", "sold out" and "contact me" web pages that haven't been updated in 15 years, as well as trying to order three times on evil bay for a scope clock kit, (listings promptly disappear the minute I place an order with messages from sellers claiming they're sold out!) can anyone furnish me with a design tool or winding procedures/specification for an ETD29 flyback–design transformer  that will have the following outputs please, for a scope clock based on the SC200 design:–

a) Primary – suitable for 9–24V input;
b) Efficency 80% or better;
C) Sec 1 – 5–0–5V at 1.0A;
d) Sec 2 – 125V at 50mA;
e) Sec 3 – 6.3V at 0.6A, 2500V insulation;
f) Sec 4 – 15–0–15V at 0.1A

The approximate total power output needed is about 30VA.

Switching device is LM2587T–adj. or similar in TO–220 style package. Note – the circuit attached suggests the 3A rated LM2586 is suitable, but for an extra safety margin I propose to use the 5A rated LM2587T!

Will a horizontal format ETD29 transformer core be suitable for this?

Chris Williams

PS!

I did email David Forbes of Cathode Corner asking whether he could supply the transformer or furnish details of how to make it, and to date I've not had a reply!

[mag_therm]

Probably will have to wind it yourself.

First calculate how many primary turns to get inductance, then how many turns on sec 3.
I think it will be feasible, if a little time consuming, to wind it.
The ETD29 is a bit big for 30 VA, but that makes it easier to hand wind.
Maybe condicors can be   0.5 mm (24 AWG) for pri  and 0.25mm (30 AWG) for the secondaries.

[T3sl4co1l]

I don't like how many doublers there are... the LM3586 is a boost/flyback controller, with peak current mode control.  It'll choke, driving much of a doubler.

Easy enough to change it to all-flyback, and also use fewer supplies, like the uh, U17 supply could be a zener shunt regulator from C55, and C54 and associated parts can go away.  (If you have a PCB and kit, easy enough to DNP these parts, and strap on new parts, between pads wherever.)  Would think POS5A and POS5GPS can be supplied from the input instead, whatever that is (7-12V I suppose).  That leaves NEG5, POS250 + HV multiplier chain, Heater, and Blanking, which are all unavoidable.  Of those, only the multiplier needs to draw forward-switched load current, which is annoying, but at least it's better than, well...all of them.

Shouldn't need many turns on a core that size, and Bmax = 0.1T or even higher should be reasonable.  The big deal is: getting the air gap right for the required magnetizing inductance, and getting the windup right for the required (low enough) leakage inductance, for good efficiency.

Probably the primary can be a single layer per section (instead of two series sections as shown), and made in two or three sections, all of which get wired in parallel for the total winding.  This is important for leakage.  The other windings go inbetween these sections, with enough layers of tape to meet the insulation requirement.  Probably, the HV winding should be by itself, then the little secondaries can be grouped together, and blanking maybe should be on its own as well just because of the isolation rating.

The blanking winding could also use triple-insulated wire, which can sometimes be salvaged from things, or maybe you can find some on eBay.

Tim