Oh hang on, the Wilson can actually have a lower parts count than the two-transistor mirror: I don't need degeneration with the Wilson, do I?
I still would use a little degeneration lets say around 30 ohms......for thermal stabilization of the mirror. These resistors are probably not required in ic implementations of the mirror because the transistors are built on the same substrate.
Jam
Jam
Last edited:
Now it's a JamJar
You might want to reserve a space on the pcb for a cap in parallel with R7 just in case someone wants to operate it unregulated.
You might want to reserve a space on the pcb for a cap in parallel with R7 just in case someone wants to operate it unregulated.
Not quite yet 😉........ but I suggest this offset adjustment solution as being less evasive ( a 5k - 1K pot should work) leave out the trim pot on the first differential, otherwise it's complete in my humble opinion, but it is Jeff's call.
solution........
Jam
solution........
Jam
Attachments
Last edited:
I like that offset adjustment. If nothing else, at least it allows my LTP to look like an LTP again.
So here are the contestants: BuzzBomb, SixPack and Beefcake:
So here are the contestants: BuzzBomb, SixPack and Beefcake:
Regarding beefcake, I have been seriously intending on replacing TL431 with just an appropriate value Zener/led combination as a voltage reference and using a trim pot, to dial up the bias voltage, then you can use small value polypropylene caps instead of large electrolytics as used with TL431.
For the purpose of biasing up a lateral fet source follower output stage
Maybe 3 or 4 green leds or appropriate Zener with trimpot.
Or just keep throwing leds at the bias circuit till near enough is good enough (ie anything that achieves around 1.5A to 2A) and just have dc offset.
For the purpose of biasing up a lateral fet source follower output stage
Maybe 3 or 4 green leds or appropriate Zener with trimpot.
Or just keep throwing leds at the bias circuit till near enough is good enough (ie anything that achieves around 1.5A to 2A) and just have dc offset.
Last edited:
Regarding beefcake you definitely want to add the feature of being able to load the VAS with a resistor to ground.
You could replace 2sk2013 with a lateral fet too, without any degeneration (or maybe 1 Ohm), nice and thermally stable at 50mA.
You could replace 2sk2013 with a lateral fet too, without any degeneration (or maybe 1 Ohm), nice and thermally stable at 50mA.
Hi Mark,
Oops! As drawn neither makes any sense. The compensation resistor needs to be outside the bias spreader, not inside it. (Then it will be clear that the capacitor just goes across the spreader.)
Thanks for catching that!
Cheers,
Jeff.
Been scratching my head about that one.
My current TL431 bias setup for ref.
Idea A, GLED431 replaces TL431, or B? (neither have been built). My rambling thoughts lead on to should I use the LTP regulator? OK It's getting complicated again, do you really need all this? then thought, what's wrong with the amplified diode if a thermistor is included? Can't be that good if people are moving away, need a low noise Vref......... Go to Line 3 🙂
Attachments
Something like this.
R5/R6 is a trimpot for DC Offset.
Just find the right led/Zener combination to get the bias you need.
No thermistors required for laterals.
R5/R6 is a trimpot for DC Offset.
Just find the right led/Zener combination to get the bias you need.
No thermistors required for laterals.
Attachments
Last edited:
Just reading green leds have a negative temperature coefficient of between -3.6mV/K to -5.2mV/K. That could actually be quite useful for hockey pucks too.
I like dumb stuff that just works.
https://www.onsemi.com/pub/Collateral/TND328-D.PDF
I like dumb stuff that just works.
https://www.onsemi.com/pub/Collateral/TND328-D.PDF
I think that might be a little difficult, the jump between 4 & 5 LED's would be not enough bias current and Oh bugger, I need bigger heatsinks! 😀