Just a thought...how about some Single Ended Gain modules

[clintrubber]

[quote author="Gus"]However PRR? or someone else posted something like maybe the circuit drawn as shown "sounded better" I am to lazy to look for the thread.[/quote]
I'm fairly sure it was PRR, but in the vein of: why not try it first as drawn & listen, before starting tweaking right away.
I searched for his words for a minute, but couldn't find it back.

I guess the technically more healthy versions can be made to sound good as well (of course), but maybe they'll be 'too good'... as in 'updating' a fuzz-box to get rid of all those nasty distortions :wink:

 

[clintrubber]

[quote author="Wavebourn"]gentlemen;

let's switch to my schemo since it is more adequate?

However, if someone already patented such combination of transistors, resistors, capacitors... we may say him "thank you and go away from DIY community"[/quote]
Your schematic looks like a logical progression based on the various already existing circuits...
...so because of that it shouldn't be too hard to get it patented with todays level of competence at the patent offices :green:

 

[skipwave]

The oft-cited PRR post:
http://groupdiy.com/index.php?topic=1504&&start=59


Now, with that out of the way, this thread has set out a 48V supply as a design parameter. As Analag points out, we can make use of that.

The more ideas that are contributed, the closer we seem to get to the FETBloke. That is not a bad thing, maybe we need something slightly simpler (lower parts count) than a Bloke and intended for a single supply rail.

 

[Gus]

with 48VDC one needs to think about the fet.

 

[clintrubber]

[quote author="Gus"]with 48VDC one needs to think about the fet.[/quote]
All-BJT then, the FET must go :wink:

But 48 & the JFET can co-exist of course, some extra parts.

 

[Sorr]

But I like the FET on the front end and no input cap!

 

[clintrubber]

[quote author="Sorr"]But I like the FET on the front end and no input cap![/quote]
Circuit could be 'flipped vertically' so that the first BJT is a PNP and then you could still do without an input-cap :wink:

But w.r.t. the JFET: indeed easy way to get Hi-Z.
But BJTs can do that as well (look for instance at the 990-DOA: BJT-input & high input impedance) & have the benefit of losing that 'generous' Vth-spread.

 

[skipwave]

[quote author="Gus"]Vgs?[/quote]

Indeed, the 48V rail makes the 2n5457 unsafe, with its Vgs of 25. Mike recognized this and switched to 2sk170, which handles 40V. Can we implement some simple protection measures without over complicating the circuit? Some middle ground between the dropping resistor and the extra bjt.

Maybe Analag's method is the simplest, with the fet riding atop a high volt bjt. Couldn't the BC550 go there? Not as wide a safety margin as the 546, at 50V vs. 80V, but still usable.

 

[tk@halmi]

Cascode the input. Then you can use most JFETs and get into the FM reception range at the same time.

 

[bcarso]

If the source Z is high consider carefully the gate leakage dependence on Vdg. 40V for the SK170 is a limit, not so much a recommended operating point.

See page 4 of this: http://www.ortodoxism.ro/datasheets/toshiba/1027.pdf

For a low-pinchoff device like this you get most of the performance by Vdg ~> 5V.

 

[skipwave]

Someone here (I think it was Thomas Holley) posted a cascode input discrete opamp design with 2sk389 and BC546. This was awhile back.

 

[triodemike]

here is the latest out of simiulation using that topology, subed high gain transistor for darlington.

 

[bcarso]

Too busy to sim myself but it looks like the 5210 gets a mite too hot.

 

[Gus]

Maybe part of the sound of that simple stage is the use of the 24V supply think gain and headroom with 24V and 48V.

People gave hints with possible 48V problems with that min parts count circuit that a sim might not "see".

Now how good is that cc fragment with a Zener vs two 1n4401s when you are clipping on the ground side?

 

[JohnRoberts]

Thank you for addressing the output stage current source. :grin:

I ASSume the LF pole is in the source?

JR

PS: If you don't have a specific design brief, how will you know when you're finished? :?:

 

[Guest]

I ASSume the LF pole is in the source?

...or PS filter? 390 Ohm / 100 uF, what's the frequency?

 

[JohnRoberts]

The 100 uf pole is something like 4 Hz but even if the cap was open it shouldn't affect the gain 3dB. Luckily I don't really care but the plot is inconsistent with my meat based SIM.


JR

 

[triodemike]

 

[JohnRoberts]

Thank you for confirming my meat SIM :grin: , some day I need to figure out that computer jazz...

If the answer doesn't look right, maybe it isn't right. That's why I never embraced SIM, you need to know the answer to use it well, but I protest too much. Slide rules required you knowing the answer first too and were damn useful...

We just need not confuse precision with accuracy.

JR

 

[Guest]

[quote author="JohnRoberts"]Thank you for confirming my meat SIM :grin: , some day I need to figure out that computer jazz...

If the answer doesn't look right, maybe it isn't right. That's why I never embraced SIM, you need to know the answer to use it well, but I protest too much. Slide rules required you knowing the answer first too and were damn useful...

We just need not confuse precision with accuracy.

[/quote]

Yes, your "meat SIM" underestimated a PS filter.