On 3/17/2017 5:35 AM, tabbypurr@gmail.com wrote:
> On Friday, 17 March 2017 03:56:37 UTC, David Eather  wrote:
>> On Fri, 17 Mar 2017 13:56:13 +1000, David Eather <eather@tpg.com.au> wrote:
>>> On Wed, 15 Mar 2017 22:39:48 +1000, <tabbypurr> wrote:
>>>> On Wednesday, 15 March 2017 12:10:36 UTC, Martin Brown  wrote:
>>>>> On 14/03/2017 13:58, tabbypurr wrote:
>>>>>> On Tuesday, 14 March 2017 09:22:05 UTC, Martin Brown  wrote:
>>>>>>
>>>>>>> There used to be old school analogue function generator chips that
>>>>>>> made a triangle wave and then applied diode shaping to get a
>>>>>>> pseudo-sine wave. HP made one design implementation that was
>>>>>>> surprisingly good. Intersils 8038 was the poor mans alternative for
>>>>>>> DIY.
>>>>>>>
>>>>>>>
>>>>> http://www.intersil.com/content/dam/intersil/documents/icl8/icl8038.pdf
>>>>>>
>>>>>>>
>>>>>> I built one of those decades ago. What a car crash. The wave forms
>>>>>> were hopeless. I don't remember the details to know why, I presume
>>>>>> the problem was the 8038 though.
>>>>>
>>>>> It was never anything like as good as a Wein bridge sine wave but it
>>>>> was
>>>>> good for about 0.5% THD if you trimmed it properly. I suspect
>>>>> manufacturing tolerances made it inconsistent batch to batch.
>>>>>
>>>>> Cute chip in its day, but that was a long time ago.
>>>>
>>>> I doubt it managed 50%, let alone 0.5%. It had 3 outputs, sine square &
>>>> triangle. At some frequencies one output looked more like one of the
>>>> others should, and the others were just a mess. It was dire, and yes I
>>>> followed the advised circuit. It might manage 0.5% at some frequency,
>>>> but as a sig gen it was a real failure. If I ever get the time I'll
>>>> look at it again one day, it's on a shelf somewhere.
>>>>
>>>>
>>>> NT
>>>
>>> there were a few kits designed with it - they really did manage 0.5%
>
> Presumably they used only part of its frequency range.

Yeah, the audio portion.

-- 

Rick C

On 17/03/2017 09:35, tabbypurr@gmail.com wrote:
> On Friday, 17 March 2017 03:56:37 UTC, David Eather  wrote:
>> On Fri, 17 Mar 2017 13:56:13 +1000, David Eather <eather@tpg.com.au> wrote:
>>> On Wed, 15 Mar 2017 22:39:48 +1000, <tabbypurr> wrote:
>>>> On Wednesday, 15 March 2017 12:10:36 UTC, Martin Brown  wrote:
>>>>> On 14/03/2017 13:58, tabbypurr wrote:
>>>>>> On Tuesday, 14 March 2017 09:22:05 UTC, Martin Brown  wrote:
>>>>>>
>>>>>>> There used to be old school analogue function generator chips that
>>>>>>> made a triangle wave and then applied diode shaping to get a
>>>>>>> pseudo-sine wave. HP made one design implementation that was
>>>>>>> surprisingly good. Intersils 8038 was the poor mans alternative for
>>>>>>> DIY.
>>>>>>>
>>>>>>>
>>>>> http://www.intersil.com/content/dam/intersil/documents/icl8/icl8038.pdf
>>>>>>
>>>>>>>
>>>>>> I built one of those decades ago. What a car crash. The wave forms
>>>>>> were hopeless. I don't remember the details to know why, I presume
>>>>>> the problem was the 8038 though.
>>>>>
>>>>> It was never anything like as good as a Wein bridge sine wave but it
>>>>> was
>>>>> good for about 0.5% THD if you trimmed it properly. I suspect
>>>>> manufacturing tolerances made it inconsistent batch to batch.
>>>>>
>>>>> Cute chip in its day, but that was a long time ago.
>>>>
>>>> I doubt it managed 50%, let alone 0.5%. It had 3 outputs, sine square &
>>>> triangle. At some frequencies one output looked more like one of the
>>>> others should, and the others were just a mess. It was dire, and yes I
>>>> followed the advised circuit. It might manage 0.5% at some frequency,
>>>> but as a sig gen it was a real failure. If I ever get the time I'll
>>>> look at it again one day, it's on a shelf somewhere.
>>>
>>> there were a few kits designed with it - they really did manage 0.5%
>
> Presumably they used only part of its frequency range.

Yes. It was fine up to around 20kHz but at higher frequencies the 
triangle wave linearity degraded taking down the sine wave with it.

You had to work at the waveform symmetry to get 0.5% but <1% was easy.

-- 
Regards,
Martin Brown

On Friday, 17 March 2017 03:56:37 UTC, David Eather  wrote:
> On Fri, 17 Mar 2017 13:56:13 +1000, David Eather <eather@tpg.com.au> wrote:
> > On Wed, 15 Mar 2017 22:39:48 +1000, <tabbypurr> wrote:
> >> On Wednesday, 15 March 2017 12:10:36 UTC, Martin Brown  wrote:
> >>> On 14/03/2017 13:58, tabbypurr wrote:
> >>> > On Tuesday, 14 March 2017 09:22:05 UTC, Martin Brown  wrote:
> >>> >
> >>> >> There used to be old school analogue function generator chips that
> >>> >> made a triangle wave and then applied diode shaping to get a
> >>> >> pseudo-sine wave. HP made one design implementation that was
> >>> >> surprisingly good. Intersils 8038 was the poor mans alternative for
> >>> >> DIY.
> >>> >>
> >>> >>  
> >>> http://www.intersil.com/content/dam/intersil/documents/icl8/icl8038.pdf
> >>> >
> >>> >>
> >>> > I built one of those decades ago. What a car crash. The wave forms
> >>> > were hopeless. I don't remember the details to know why, I presume
> >>> > the problem was the 8038 though.
> >>>
> >>> It was never anything like as good as a Wein bridge sine wave but it  
> >>> was
> >>> good for about 0.5% THD if you trimmed it properly. I suspect
> >>> manufacturing tolerances made it inconsistent batch to batch.
> >>>
> >>> Cute chip in its day, but that was a long time ago.
> >>
> >> I doubt it managed 50%, let alone 0.5%. It had 3 outputs, sine square &  
> >> triangle. At some frequencies one output looked more like one of the  
> >> others should, and the others were just a mess. It was dire, and yes I  
> >> followed the advised circuit. It might manage 0.5% at some frequency,  
> >> but as a sig gen it was a real failure. If I ever get the time I'll  
> >> look at it again one day, it's on a shelf somewhere.
> >>
> >>
> >> NT
> >
> > there were a few kits designed with it - they really did manage 0.5%

Presumably they used only part of its frequency range.


NT

David Eather wrote:
>
> >> NT
> >
> > there were a few kits designed with it - they really did manage 0.5%
> >
> 
>

** Of course they did.

 Nutcase Thornton is a lying, psychopathic moron. 




..... Phil

On Fri, 17 Mar 2017 13:56:13 +1000, David Eather <eather@tpg.com.au> wrote:

> On Wed, 15 Mar 2017 22:39:48 +1000, <tabbypurr@gmail.com> wrote:
>
>> On Wednesday, 15 March 2017 12:10:36 UTC, Martin Brown  wrote:
>>> On 14/03/2017 13:58, tabbypurr wrote:
>>> > On Tuesday, 14 March 2017 09:22:05 UTC, Martin Brown  wrote:
>>> >
>>> >> There used to be old school analogue function generator chips that
>>> >> made a triangle wave and then applied diode shaping to get a
>>> >> pseudo-sine wave. HP made one design implementation that was
>>> >> surprisingly good. Intersils 8038 was the poor mans alternative for
>>> >> DIY.
>>> >>
>>> >>  
>>> http://www.intersil.com/content/dam/intersil/documents/icl8/icl8038.pdf
>>> >
>>> >>
>>> > I built one of those decades ago. What a car crash. The wave forms
>>> > were hopeless. I don't remember the details to know why, I presume
>>> > the problem was the 8038 though.
>>>
>>> It was never anything like as good as a Wein bridge sine wave but it  
>>> was
>>> good for about 0.5% THD if you trimmed it properly. I suspect
>>> manufacturing tolerances made it inconsistent batch to batch.
>>>
>>> Cute chip in its day, but that was a long time ago.
>>
>> I doubt it managed 50%, let alone 0.5%. It had 3 outputs, sine square &  
>> triangle. At some frequencies one output looked more like one of the  
>> others should, and the others were just a mess. It was dire, and yes I  
>> followed the advised circuit. It might manage 0.5% at some frequency,  
>> but as a sig gen it was a real failure. If I ever get the time I'll  
>> look at it again one day, it's on a shelf somewhere.
>>
>>
>> NT
>
> there were a few kits designed with it - they really did manage 0.5%
>


-- 
Using Opera's mail client: http://www.opera.com/mail/

tabb...is a lying fuckwit and usenet criminal 



  > > > 
> > > > 
> > > > ** Must be a delusion - cos it contradicts the experience of all others.
> > > > 
> > > >   Or just plain old bullshit made up by valve freaks. 
> > > 
> > > 
> > > It's the collective experience of far more people than you.
> > >
> > 
> >  ** No it's not - you pathetic lying POS. 
> > 
> > 
> > 
> > > >   I told you to fuck off.
> > > > 
> > > >   Do so or I will hound you off this NG. 
> > > > 
> > > 
> > > 
> > > You aren't capable of it. 
> > >
> > 
> >  ** Fact is, I am very good at it and very successful. 
> > 
> > 
> > > 
> > > Anyone can look at that figure 20 and see the claimed 11% distortion 
> > > plain as day.
> > > 
> > 
> > ** And they can see you posted a MASSIVE LIE as well.
> > 
> >  You are a one sad & sorry fuckwit without a single shred of decency in you.  
> > 
> >  Lying on a technical NG is incredibly moronic.
> > 
> >  Folk like you are pox on the face of the earth. 
> > 
> >  Drop dead. 
> > 
>
> 
> Mr Allison is plainly nuttier than I expected.
> http://www.intersil.com/content/dam/intersil/documents/icl8/icl8038.pdf
> Figure 20 on p10 plain as day shows upto 11% distortion on sine waves,


** Big shame what the typical number is over the audio band

   You stinking, lying fuckwit. 

   Get off the NG. 


.....  Phil

On Monday, March 13, 2017 at 1:28:39 AM UTC-7, olds...@tubes.com wrote:
> I've been looking into buying an Audio Generator (Sine and Square Wave).
> I mainly want this to run thru an amplifier to listen to the frequency
> repsonse of the amp and speakers. Nothing very scientific, just to see
> what these amps and speakers can do.... And on occasion to inject an
> audio signal into amp sections to dee if the audio is passing that
> stage. 
> 
> Originally I was looking at some of the old tube gear, such as the Eico
> model 377. I also looked at the Heathkit IG-18 a transistorized version
> from the 70s. Because I like that old test gear, I'd be happy to get
> either of these, or another similar. Both of these are Sine and Suqare
> Wave. But what I am finding is the prices combined with high S+H fees,
> make these things very costly, and I have not found any of them which
> claim to work. Sold as-is, for parts or repair. So, I may end up paying
> $60 or more and getting something that dont work, likely needs new caps,
> and so on..... If they were $25 or less, I'd take a chance, but not for
> that kind of money.
> 
> My search on ebay lead me to something called a "Function Generator".
> What the heck is that??? And also called a DDS device. (I have no clue
> what DDS means). But I carefully looked at this NEW device, sold without
> a box (case), but only needs a power module and test leads. They sell
> for around $14 from China, but since I dont care to order from China,
> I'll probably have to hunt one down from the US, and pay around $5 or
> $10 more. But that's ok. These have both the Sine and Square wave, as
> well as Triangle wave (what does that sound like?). 
> 
> While I prefer the older stuff, for around $20 a cheap wal-wart, and a
> little work putting it into a box, I'd probably buy one of these, *IF*
> it will do what I need..... I have read all the info on the ebay pages,
> but I am not 100% sure that this is a modern replacement for the old
> audio generators. One nice thing, is that they will show the exact
> frequency on the digital readout. 
> 
> IS THIS WHAT I AM LOOKING FOR?
> Is something like this made to really be used, or is it just another toy
> that does little and fails in a few months?
> 
> And can something like this work to test tube amps, without blowing up
> from the high voltages in tube gear?
> 
> One last thing that has me puzzled. There are two BNC connectors. One is
> labeled DDS, tho other is HS. 
> 
> I found this in the description. 
> DDS frequency range: 1Hz-65534Hz, high-speed frequency (HS) output up to
> 8MHz. 
> 
> I dont understand how or what the high speed feature does. Why would an
> audio frequency be or need "high speed"? And what is the point for this
> device to go above 20,000 cps, since we cant hear that anyhow? Going up
> to 8MHZ seens senseless, unless this also serves as a RF signal
> generator.
> 
> Here is the ebay URL.
> 
> http://tinyurl.com/h5celuc

Wavetek 145 is a really nice oldschool signal generator.

Sine/sq/triangle, plus sawtooth and pulses.  Also freq and amplitude modulated (makes some crazy raygun sounds.)   Also frequency-sweep, and with a scope, you can see an amplifier's spectrum.

On 3/16/2017 1:07 PM, tabbypurr@gmail.com wrote:
>
> Mr Allison is plainly nuttier than I expected.

No argument there and I have no idea what you expected.

> http://www.intersil.com/content/dam/intersil/documents/icl8/icl8038.pdf
> Figure 20 on p10 plain as day shows upto 11% distortion on sine waves, and figure 18 admits upto 10% linearity on triangle waves. What sort of loon sits there lying on something so trivial for every reader to go see. What a - I don't know what he is. Just a nut.

I believe the context of this discussion was "audio" gear.  The high 
distortion of the ICL8038 is only in the frequency range above 20 kHz.

-- 

Rick C