Path: ...!eternal-september.org!feeder2.eternal-september.org!news.eternal-september.org!.POSTED!not-for-mail
From: Bill Sloman <bill.sloman@ieee.org>
Newsgroups: sci.electronics.design
Subject: Re: 80dB now but still needs improvement at 1KHz
Date: Tue, 5 Nov 2024 13:17:23 +1100
Organization: A noiseless patient Spider
Lines: 64
Message-ID: <vgbv7l$177js$3@dont-email.me>
References: <vg9h7b$277o$1@nnrp.usenet.blueworldhosting.com>
 <vg9pnu$ekhh$1@solani.org> <vga6jh$suru$1@dont-email.me>
 <vgatph$46a$1@nnrp.usenet.blueworldhosting.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Injection-Date: Tue, 05 Nov 2024 03:17:26 +0100 (CET)
Injection-Info: dont-email.me; posting-host="195fb685e5cb46f5c705e43421dd297d";
	logging-data="1285756"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX1+3cObsx+aygY5Glojdav5ODyjDwmWuwhc="
User-Agent: Mozilla Thunderbird
Cancel-Lock: sha1:Fw4ccooRtLquI1fi+W36uwqKM1I=
X-Antivirus: Norton (VPS 241104-0, 4/11/2024), Outbound message
In-Reply-To: <vgatph$46a$1@nnrp.usenet.blueworldhosting.com>
X-Antivirus-Status: Clean
Content-Language: en-US
Bytes: 4325

On 5/11/2024 3:46 am, Edward Rawde wrote:
> "Bill Sloman" <bill.sloman@ieee.org> wrote in message news:vga6jh$suru$1@dont-email.me...
>> On 4/11/2024 5:31 pm, Jan Panteltje wrote:
>>> On a sunny day (Sun, 3 Nov 2024 23:06:04 -0500) it happened "Edward Rawde"
>>> <invalid@invalid.invalid> wrote in
>>> <vg9h7b$277o$1@nnrp.usenet.blueworldhosting.com>:
>>>
>>>> Simulation isn't fast but if you let it complete and do an FFT on the last 30 seconds, it's 80dB down at all unwanted
>>>> harmonics.
>>>> There do however seem to be unwanted sidebands close in either side of 1KHz
>>>> Any suggestions for improvement?
>>>
>>> Publish a circuit diagram.
>>> Not that hard to make !
>>
>> He did. The .asc file is a circuit diagram. I've looked at it and run the simulation - if only for 200 seconds rather than 300
>> seconds.
>>
>> The FFT says that sine wave is as good as he claims.
>>
>> The circuit diagram  doesn't make it clear what it's various parts are there to do - it makes sense to group the components in a
>> way that lets somebody looking at the circuit diagram get some feel for what the components are doing.
>>
>> And it you are using the LT1994 it makes sense to read the data sheet carefully enough to notice that the Icom pin 2 should be
>> bypassed with at least 100nF to ground.
>>
>> https://www.analog.com/media/en/technical-documentation/data-sheets/1994fb.pdf
> 
> Yes there's no way I would use a part without reading the data sheet.
> Pass 1 fount the 0.1uF but I hadn't got around to adding it.
> Pass 2 found that I couldn't go above +/- 6V
> Pass 3 which I haven't done yet would take a lot more time and I probably won't do it because I doubt I'll be able to build this for
> real.
> 
> Please see also my reply to Jan.
> 
>>
>> John May's circuit makes it clear that the second FET isn't strictly necessary. There are cheaper ways of getting rid of the
>> even-order harmonics.
> 
> Does his circuit have better performance?

The version I got privately certainly did.

> Just curious, it's not a competition.

Actually, I've decided that the second FET is a bad idea. Two different 
FETs need two different gate drive voltages to give the right channel 
resistance, and the only guide you've got to that is the amplitude of 
the single sine wave you are producing. A large number of different 
pairs of FET gate drive voltages could give you the same stable amplitude.

If you monitored the second harmonic component of the output you would 
have a second independent output which could let you optimise both gate 
dries at the same time, but that would be an unnecessary refinement.

>> I've managed to dig out the precision, full wave rectifier that I used in my circuit, which is a half-wave rectifier to which you
>> add just enough of the full sine wave to deliver both halves of the sine wave (one of them inverted) at it's output, which calls
>> for a couple 10k 0.1% thin-film precision resistors on a common substrate, which you can buy off the shelf.

-- 
Bill Sloman, Sydney