Return to KLUBNL.PL main page

rsgb_lf_group
[Top] [All Lists]

Re: LF: W1VD amp help - more waveforms

To: LineOne <[email protected]>
Subject: Re: LF: W1VD amp help - more waveforms
From: Andy Talbot <[email protected]>
Date: Thu, 12 Apr 2018 22:34:00 +0100
Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=mime-version:in-reply-to:references:from:date:message-id:subject:to; bh=kXFh5hb8wGHCNJVFiHu23y0k0eZkYbeKiWeFbmPJh1Q=; b=Oz7bAN08QkioC9QFIUSt/PbGvOkPajbbQYi0YeE2wZWf7xQDdgTLUDNAbStLbEPJay I4LuepszQWLq+4/lt8VSQlU/zc+bWCYpeR4uKHzqPFMm1E6QuMRqrLXcNnndaoSMzCif hOswzjdyjl2j/dGjQZmKrO3O69f8h0Q7KSvGvmhbRhLYDmRsSR3fR1nFqkXnGtTVqXSM L7nVMDiXEt7CHrX8ekXAlQezAx7jWRBA0LYFFxZqEObBCC7pfkv2W3U4oN1hyJbUzCtw v7GU4EoaT97UrLPCCDJAxkaF3kdfnLvZDHGvlgSx/PqOe5oXC8EJG/BIbF8mxMR268g7 /N7Q==
In-reply-to: <8D5E7536EB86447B920A6B0C99E2BA88@gnat>
References: <1UQRRBrZ33.5N62o0UWd72@optiplex980-pc> <[email protected]> <[email protected]> <[email protected]> <[email protected]> <[email protected]> <[email protected]> <BFE1E3A433664900AE95DE6D9EE733BC@StevePC> <[email protected]> <7950E881690348FB8A48C42DD2735F8D@StevePC> <1UQRSH98Tx.2IrTZ1ue7L6@optiplex980-pc> <[email protected]> <7FBAA3A0B6C84158BA005CFF09A5A7EA@DELL4> <[email protected]> <[email protected]> <8D5E7536EB86447B920A6B0C99E2BA88@gnat>
Reply-to: [email protected]
Sender: [email protected]
Yes, I agree with Alan and fail completely to see why the driver should have any effect on the ringing. It just cannot.

 As I keep saying, such ringing is commonplace in SMPSUs and is due to leakage inductance in the transformer.   I don't like torroids as they don't allow you to enclose the winding nearly as much as as ETD or pot core type of construction does and I'm sure contribute more leakage L than more enclosed core types.   There is more wire facing outer-space than there is facing ferrite when you have windings of only a couple or turns.  .

But everyone else just seems to have accepted torroids are the best thing, so carry on.   Someone must have originally said they're the way forward but I'd love to know why no one just uses standard SMPSU cores.  

I'll just tootle along with switching power amps built with tank circuits and only then into a transformer,  and let everyone else go their own way.   It just seems such a simple and obvious route, I cannot understand the need to generate a square wave, force it though an unwilling transformer then use a complicated low pass filter to clean up.

The only LF amp I've built where push-pull FETs go straight into a transformer is a linear (well, almost linear) one - just to see if it could be done.   About 40 - 50 Watts from a 28V supply, 60 - 70 watts saturated with scope for improving it with better FETs   http://www.g4jnt.com/Linear_LF_PA3.pdf   .    The output transformer for that was just an RM10 sized pot core of F44 material.   I suspect that if pushed hard into saturation, and the feedback CR removed from each drain-gate it would also show ringing on the edges.

But there's little call for linear amps on the LF bands.

Andy G4JNT

On 12 April 2018 at 21:43, Alan Melia <[email protected]> wrote:
Paul, Chris, Jay et al
Just a wild thought....Jays LPF is not inside the PA box. The filter is giving a frequency dependent load on the PA. Could it be possible that some reflected impedance is exagerating the ringing. As Andy said unwanted resonance somewhere. For accurate efficiency measurement you need to know what you are actually measuring but from the stability point of view the filter is not necessary. Is it possible (and worth trying) taking the filter out of circuit and putting the transformer straight into the load ??  Then see if the ringing is more controlled??

I honestly cant see why the driver would affect this. As far as I can see there is no sign of a ring on the gate waveform. As it goes off it drops straight to ground, starting before the other phase switches on. My naive though is that the filter may not be damping the ring but the wide band load might. Filters need treating with care theycan behave oddly when not terminated with resistive loads. Then there is the coupling that might be occuring inside the case.

Probably rubbish but I think its an angle we have not considered yet. :-))

Alan
G3NYK

----- Original Message ----- From: "Chris Wilson" <[email protected]>
To: "N1BUG" <[email protected]>
Sent: Thursday, April 12, 2018 7:19 PM
Subject: Re: LF: W1VD amp help - more waveforms




Hello Paul,

I  am looking at your photo of the LPF. The specs say L1 and L3 should
have  52  turns,  and  L2, the centre one, should have 73 turns. i was
unable to get 73 turns on the toroid without winding on top of a lower
layer,  and  insulated  with  Kaptan tape. This made the overall depth
quite  a  bit  deeper  than  L1 and L3. In the photo it looks like the
centre  toroid  overall is no deeper than the end ones, and the amount
of bolt thread protruding *LOOKS* the same as the end ones. Your build
is so painstakingly neat though, it may be that you have aesthetically
made all the toroids look identical.

http://www.chriswilson.tv/paul-lpf.jpg

Thursday, April 12, 2018, 5:42:13 PM, you wrote:

Jay,

When the package arrives I will swap out the caps and the IR2110. If
that doesn't do it I will order a new 74F74 and swap out the
resistors. If all of that fails I don't see anything else I can do.
Realistically at my level of expertise if it doesn't work at that
point I will have reached the end of the road. :-(

The caps I have are KEMET part # R82EC1100Z350J

Here is a link to the datasheet:

https://www.mouser.com/datasheet/2/212/KEM_F3101_R82-1103738.pdf

Paul





--
Best regards,
Chris                            mailto:[email protected]





<Prev in Thread] Current Thread [Next in Thread>