Re: LF: Progress? with the PA

[N1BUG <paul@n1bug.com>]

Thanks Andy.

I need more time to get my head around that equation. I'm still struggling with a near vertical learning curve on all of this. I never worked with ferrite (except in RFI suppression) until very recently. To start with I don't know what figure to use for N turns on this 16:1 transformer with 4 x 10 turn windings. There must be beginner level articles somewhere that walk one through this stuff, but I haven't yet found them.

For that matter I'm not sure how to calculate Vrms on something that isn't a sine wave.

All I can say at the moment is the core is a EPCOS / TDK B64290L674X830

N30 material with a ui of 4300, 36 x 23 x 15 mm.

I'm still interested in understanding the issues with this PA as a learning exercise but am preparing to move on to something else for on air use.

Paul N1BUG

On 01/07/2018 12:34 PM, Andy Talbot wrote:

What core ?

What  dimensions ?

Does it obey the fundamental equation for saturation in a core ?

Vrms = 4.44 .F . N . A . B

F Hz,   N turns,  A core area area in m^2   B Tesla

Keep B < 0.1 and  ferrite saturation should be fine

And if it is saturating, you need more turns, not less

Andy  G4JNT

On 7 January 2018 at 17:26, N1BUG <paul@n1bug.com> wrote:

http://n1bug.com/n1debug/ASB_LF_Amp-20171223.jpg

Recall that last week I had increased turns on the output transformer from 3x5 to 3x10. That reduced heating of the FET but did not help to stabilize the PA. It would put 12 to 14W into the antenna before oscillation started.

Today I changed it to 4x10 turns so it is now a 16:1 transformer. This has made a big change! The PA will now put more than 30 watts into the antenna before oscillation starts with 13V on it. For the first time I noticed some slight warming of the transformer. I think it is the core heating up, not the wire.

If I put 19V on it, I briefly see almost 50 watts into the antenna but the transformer becomes hot very quickly and within just 3 or 4 seconds, power begins to drop rapidly.

At this point I guess either:

The core is not suitable material or not big enough...

or

Maybe it needs fewer turns (but this may lead to too much heat in the FET again?) At 30W output with 13V, that FET is already running quite warm with 4x10 turns.

At least I finally found something that makes a change!

73,
Paul N1BUG