[sdiy] Transistor questions

[Roman Sowa]

I'm sure you can build gigahertz radio stuff on deadbug THT style if you 
do it on daily basis. Another thing is that kind of speed in a circuit 
designed like audio circuit. No disrespect to David, but I dare to guess 
he never made any radio circuit before.

Yes, I agree, seeing 3ns slope would be tough to without active 1GHz 
probe for $4k, let alone the scope itself. OTOH one does not really need 
to see it, but calculate based on frequency measurements and how it 
falls down from expected number at very high frequencies. In this case I 
think even 30ns would be more than enough to get a good tuning without 
HF compensation

There is no comparator in this oscillator, it's SCR reset, that's why it 
can be fast. Very fast, but not so fast as David wants it. And nobody 
cares about edges being rounded by next circuit, it's only needed for 
frequency control linearity

But anyway, here's a thought - arent PS3100 VCOs tuned at basicaly 
constant frequency, varied only in small amount by LFOs and pitchbend? 
Even if it's +/-2 octaves, I don't see any need for super precise tuning 
scale to start with.

Roman



W dniu 2025-03-03 o 13:17, mskala at northcoastsynthesis.com pisze:
> On Mon, 3 Mar 2025, Roman Sowa via Synth-diy wrote:
> 
>> I don't think there's a way to achieve 3ns rise time in this circuit,
>> expecially when done in THT.
> 
> Having anything happen in 3ns implies bandwidth of at least about 300 MHz,
> i.e. the bottom of the UHF range, which is certainly possible on a
> carefully designed through-hole board.  I've seen ham radio projects do
> it, and note that ordinary FM broadcast recievers routinely handle(d) 108
> MHz signals on through-hole boards, let alone UHF analog television
> receivers.  Conventional wisdom today includes exaggerated ideas of the
> disadvantages of through-hole.  But the hams would usually use dead-bug
> construction instead of a through-hole board for this range.  Getting it
> to work is far from a gimmie.
> 
> A related issue is that if you achieved it, *would you know*?  You quite
> possibly don't have an oscilloscope fast enough to see such an edge.  The
> comparator in your oscillator which needs to respond to the edge is likely
> not fast enough to care, and if the signal ever has to pass through an
> amplifier, or even a cable, designed primarily for audio, it'll round off
> the 3ns edge.
> 
> I don't understand the calculation that goes from "6kHz fundamental" to "I
> need 3ns rise time" but I think that's the step to look at.
>