From: Joe on
On Sat, 12 Jun 2010 07:57:35 -0700, John Larkin wrote:
> On Sat, 12 Jun 2010 08:53:20 +0100, John Devereux wrote:
>>John Larkin writes:
>>> On Fri, 11 Jun 2010 17:43:39 -0300, YD <ydtechHAT(a)techie.com> wrote:
>>[...]
>>> That's nice. It's running pretty hard class C.
>>>
>>> Try this:
>>>
>>> C2 = 10 uF
>>> L2 = 0.25 uH
>>>
>>> That will move it to class A, with about 50 millivolts p-p drive at
>>> the emitter. That's more like what I had in mind originally. Vb is
>>> stiff at about +0.6, and Vc dips down to about -0.1, a little below
>>> Ve, sucking a brief blip of current out of the base cap. The sine wave
>>> is just slightly flattened on the bottom.

Yes, those changes to C2 and L2 cut the p-p signal at emitter from about
525 mV to 50, but the currents at c, b, and e change only a little -
eg peak currents (mA) Ib 1.22 vs 1.20, Ic -5.14 vs -5.34, Ie 3.91 vs 4.15.

The FFT with YD's C2, L2 values is fairly clean, but with the C2, L2
that you suggest above, there are a lot of signal components only
20dB down from the 1.581kHz fundamental, vs at least 44dB down from
the 1.578kHz fundamental in YD's.

>>> It's useful to add a small resistor, like 1 milliohm, in the emitter
>>> so that you can probe the current. It's interesting.

>>You don't need to add the the resistor. You can just measure the emitter
>>current by clicking on it!
>>
> I couldn't get that to work. Maybe my aim isn't good enough.

One can click the 'Select which traces are visible' icon (just
below the help button; looks like a miniature graph), or
right-click in a Plot window and click 'Visible traces',
and then click Ie(Q1).
From: YD on
Late at night, by candle light, John Larkin
<jjlarkin(a)highNOTlandTHIStechnologyPART.com> penned this immortal
opus:

>On Sat, 12 Jun 2010 08:53:20 +0100, John Devereux
><john(a)devereux.me.uk> wrote:
>
>>John Larkin <jjlarkin(a)highNOTlandTHIStechnologyPART.com> writes:
>>
>>> On Fri, 11 Jun 2010 17:43:39 -0300, YD <ydtechHAT(a)techie.com> wrote:
>>>
>>>>Late at night, by candle light, John Larkin
>>>><jjlarkin(a)highNOTlandTHIStechnologyPART.com> penned this immortal
>>>>opus:
>>
>>[...]
>>
>>>
>>> That's nice. It's running pretty hard class C.
>>>
>>> Try this:
>>>
>>> C2 = 10 uF
>>> L2 = 0.25 uH
>>>
>>> That will move it to class A, with about 50 millivolts p-p drive at
>>> the emitter. That's more like what I had in mind originally. Vb is
>>> stiff at about +0.6, and Vc dips down to about -0.1, a little below
>>> Ve, sucking a brief blip of current out of the base cap. The sine wave
>>> is just slightly flattened on the bottom.
>>>
>>> It's useful to add a small resistor, like 1 milliohm, in the emitter
>>> so that you can probe the current. It's interesting.
>>
>>Hi John,
>>
>>You don't need to add the the resistor. You can just measure the emitter
>>current by clicking on it!
>>
>
>I couldn't get that to work. Maybe my aim isn't good enough.
>
>John
>

Just hover the pointer over the terminal until it looks like a
clampmeter. One of those things I found quite by accident, and nicely
surprised.

- YD.
--
File corruption detected. Select option:
1 - Call the cops
2 - Call the press
3 - Bribe it

Remove HAT if replying by mail.
From: Capt. Cave Man on
On Sat, 12 Jun 2010 19:38:42 -0300, YD <ydtechHAT(a)techie.com> wrote:

>Late at night, by candle light, John Larkin
><jjlarkin(a)highNOTlandTHIStechnologyPART.com> penned this immortal
>opus:
>
>>On Sat, 12 Jun 2010 08:53:20 +0100, John Devereux
>><john(a)devereux.me.uk> wrote:
>>
>>>John Larkin <jjlarkin(a)highNOTlandTHIStechnologyPART.com> writes:
>>>
>>>> On Fri, 11 Jun 2010 17:43:39 -0300, YD <ydtechHAT(a)techie.com> wrote:
>>>>
>>>>>Late at night, by candle light, John Larkin
>>>>><jjlarkin(a)highNOTlandTHIStechnologyPART.com> penned this immortal
>>>>>opus:
>>>
>>>[...]
>>>
>>>>
>>>> That's nice. It's running pretty hard class C.
>>>>
>>>> Try this:
>>>>
>>>> C2 = 10 uF
>>>> L2 = 0.25 uH
>>>>
>>>> That will move it to class A, with about 50 millivolts p-p drive at
>>>> the emitter. That's more like what I had in mind originally. Vb is
>>>> stiff at about +0.6, and Vc dips down to about -0.1, a little below
>>>> Ve, sucking a brief blip of current out of the base cap. The sine wave
>>>> is just slightly flattened on the bottom.
>>>>
>>>> It's useful to add a small resistor, like 1 milliohm, in the emitter
>>>> so that you can probe the current. It's interesting.
>>>
>>>Hi John,
>>>
>>>You don't need to add the the resistor. You can just measure the emitter
>>>current by clicking on it!
>>>
>>
>>I couldn't get that to work. Maybe my aim isn't good enough.
>>
>>John
>>
>
>Just hover the pointer over the terminal until it looks like a
>clampmeter. One of those things I found quite by accident, and nicely
>surprised.
>
>- YD.

I caught it right away. I'm a 'prober' by nature.
From: John Larkin on
On Sat, 12 Jun 2010 19:38:42 -0300, YD <ydtechHAT(a)techie.com> wrote:

>Late at night, by candle light, John Larkin
><jjlarkin(a)highNOTlandTHIStechnologyPART.com> penned this immortal
>opus:
>
>>On Sat, 12 Jun 2010 08:53:20 +0100, John Devereux
>><john(a)devereux.me.uk> wrote:
>>
>>>John Larkin <jjlarkin(a)highNOTlandTHIStechnologyPART.com> writes:
>>>
>>>> On Fri, 11 Jun 2010 17:43:39 -0300, YD <ydtechHAT(a)techie.com> wrote:
>>>>
>>>>>Late at night, by candle light, John Larkin
>>>>><jjlarkin(a)highNOTlandTHIStechnologyPART.com> penned this immortal
>>>>>opus:
>>>
>>>[...]
>>>
>>>>
>>>> That's nice. It's running pretty hard class C.
>>>>
>>>> Try this:
>>>>
>>>> C2 = 10 uF
>>>> L2 = 0.25 uH
>>>>
>>>> That will move it to class A, with about 50 millivolts p-p drive at
>>>> the emitter. That's more like what I had in mind originally. Vb is
>>>> stiff at about +0.6, and Vc dips down to about -0.1, a little below
>>>> Ve, sucking a brief blip of current out of the base cap. The sine wave
>>>> is just slightly flattened on the bottom.
>>>>
>>>> It's useful to add a small resistor, like 1 milliohm, in the emitter
>>>> so that you can probe the current. It's interesting.
>>>
>>>Hi John,
>>>
>>>You don't need to add the the resistor. You can just measure the emitter
>>>current by clicking on it!
>>>
>>
>>I couldn't get that to work. Maybe my aim isn't good enough.
>>
>>John
>>
>
>Just hover the pointer over the terminal until it looks like a
>clampmeter. One of those things I found quite by accident, and nicely
>surprised.
>

I know that, but I had a hard time hitting the emitter current. The
resistor is a bigger target. Maybe I should buy more pixels or
something.

John

From: John Larkin on
On Sat, 12 Jun 2010 15:10:45 -0700, John Larkin
<jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote:

>On Sat, 12 Jun 2010 15:28:44 -0500, John Fields
><jfields(a)austininstruments.com> wrote:
>
>>On Fri, 11 Jun 2010 20:04:42 -0700, John Larkin
>><jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote:
>>
>>>On Fri, 11 Jun 2010 08:16:29 -0500, John Fields
>>><jfields(a)austininstruments.com> wrote:
>>>
>>
>>>The .wav thing is cute. But the ringdown is being caused by the LC's
>>>Q, not by the transistor oscillation. The R2-C4 thing dies out pretty
>>>soon, and then the tank rings on its own. That why you need R4; at
>>>infinite Q, it would ring forever.
>>>
>>>If you can get an LC with a Q that high, you can simulate a nice bell
>>>noise by just whacking it and letting it ring... like a real brass
>>>bell. Much simpler.
>>
>>---
>>
>>Speaking of real world parts and more-or-less realistic bong
>>frequencies, L1 here is a Caddell-Burns 5.6mH 2.9 ohm choke (P/N
>>7200-34), C1 is 10�F with an ESR of an ohm or less, and the thing
>>oscillates at about 675Hz.
>>
>
>Xl of 5.6 mH at 675 Hz is 23.8 ohms. So, ignoring core losses, the
>inductor's Q is just over 8. Your model, with the 1K across L2, has an
>inductor Q of about 470.
>
>If you model with a Q of 8, it's pretty awful. And if you play with
>the base bias resistor, you can make something that looks fairly
>chaotic, sort of ratty like a superregen.

Oh, I see the 2.9 ohms in the box off to the side. I must be pulling
the Q down to about 4. But even at 8, the oscillation is loaded pretty
bad. It's hard to get good inductors at these frequencies.

So what's R4 for?

John

First  |  Prev  |  Next  |  Last
Pages: 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
Prev: Twin T circuit wanted
Next: And blocking oscillators