nitpicking no more…
Thank you for the replys.
I won't attempt to reinvent the wheel, even if it's tempting to declare a round rounder than round 😛
I'd have to become a genius in math/physics, get a thermal camera and lots of gear to conduct experiments on my own. Alas, this won't happen anytime soon.
It's often just sort of a feeling of something's complexity without really understanding the things behind. And since thermal dynamics seems, well, more accessible than electronics...
But I must say it would be nice to find a way to increase the thermal transfer/heatsink efficiency (aka reinvent the wheel
), so that they could be reduced in size. I like small components...
Thank you for the replys.
I won't attempt to reinvent the wheel, even if it's tempting to declare a round rounder than round 😛
I'd have to become a genius in math/physics, get a thermal camera and lots of gear to conduct experiments on my own. Alas, this won't happen anytime soon.
It's often just sort of a feeling of something's complexity without really understanding the things behind. And since thermal dynamics seems, well, more accessible than electronics...
But I must say it would be nice to find a way to increase the thermal transfer/heatsink efficiency (aka reinvent the wheel
), so that they could be reduced in size. I like small components...
What changes are suggested to optimize the F4 for a 4 ohm load? Additional bias current and an increase in cap filter size?
Thanks Mighty ZM!
Just ordered the F4 PCB boards and MOSFET Kits from DIYAudio Store and Antek 18v+18v 500VA transformers to build 2 F4's running balanced. Already have the matched JFets and Universal PS boards.
First project and thanks 6L6 for pointing me this direction!
Happy New Year!
Rob
Just ordered the F4 PCB boards and MOSFET Kits from DIYAudio Store and Antek 18v+18v 500VA transformers to build 2 F4's running balanced. Already have the matched JFets and Universal PS boards.
First project and thanks 6L6 for pointing me this direction!
Happy New Year!
Rob
Trying to understand amp design... at least the F4 schematic. Adding another pair of output MOSFets would add ~10 watts to class A output to 4 ohms and ~20 watts in AB? More heat to worry about and need for additional drive current?
Rob
Rob
Last edited:
which speakers?
I don't like over-general answers ..... specific are much more useful
with 8R load , voltage envelope (read - rails voltage) is limiting factor for power
with 4R load , current envelope (read - Iq) is limiting factor for A Class , while for rest of power, amp is going to Class B and again rails are limiting factor
will it work better if you increase Iq (increasing A Class window), not easy to say without knowing more about load .....
if you have enough hardware (heatsink , Donuts, caps) you can easily increase Iq with present number of mosfets in output ........ say that 30W per piece is still safe and sane, if htsnkng is proper
don't use force , use a bigger hammer
I don't like over-general answers ..... specific are much more useful
with 8R load , voltage envelope (read - rails voltage) is limiting factor for power
with 4R load , current envelope (read - Iq) is limiting factor for A Class , while for rest of power, amp is going to Class B and again rails are limiting factor
will it work better if you increase Iq (increasing A Class window), not easy to say without knowing more about load .....
if you have enough hardware (heatsink , Donuts, caps) you can easily increase Iq with present number of mosfets in output ........ say that 30W per piece is still safe and sane, if htsnkng is proper
don't use force , use a bigger hammer
Magneplan MG-20.1's which seem to like power. I have the impedance vs freq somewhere... I think they go to 3 ohms at higher freq.
Rob
Rob
I take that back. The Maggies are 4 ohms across the range except at crossover points which peaks to ~10 ohms at 200 Hz and 1500 Hz.
I also drive a REL powered sub from the amp outputs if that makes a difference.
Rob
I also drive a REL powered sub from the amp outputs if that makes a difference.
Rob
well , bridged/balanced F4 will cope well with Maggies
will it be enough for your expectations , only you can decide, afterwards
if not , be prepared for much more complicated build than F4 is ......... I mean - natural extension is BA3 Bal. , simply going up in level of everything - rails and Iq , number of outputs , kilos of copper and aluminium .......
will it be enough for your expectations , only you can decide, afterwards
if not , be prepared for much more complicated build than F4 is ......... I mean - natural extension is BA3 Bal. , simply going up in level of everything - rails and Iq , number of outputs , kilos of copper and aluminium .......
Thanks ZM.
I'm not planning to modify this F4 build. I am also trying to understand the engineering behind this that many on this forum know so well. I am an engineer... just deal with fluid flow instead of electron flow! I know enough to be dangerous...
So bear with me for a moment.. If I was to add one more MOSFET stage (and the associated local circuit - essentially copying the Q5 and Q10 sections in the F4 schematic), the JFets would need to drive 33% more voltage (or is that current?) to the output stage.
I'm not planning to modify this F4 build. I am also trying to understand the engineering behind this that many on this forum know so well. I am an engineer... just deal with fluid flow instead of electron flow! I know enough to be dangerous...
So bear with me for a moment.. If I was to add one more MOSFET stage (and the associated local circuit - essentially copying the Q5 and Q10 sections in the F4 schematic), the JFets would need to drive 33% more voltage (or is that current?) to the output stage.
input JFets will cope with additional pair of Mosfets (them being parts driven with voltage , not current) and final result will be "just" increased A Class window , while you'll still have rails voltage as limiting factor for max power on 4R load
as I said - you have enough space to increase Iq with present numbers of outputs, before resorting to adding them more
as I said - you have enough space to increase Iq with present numbers of outputs, before resorting to adding them more
unnecessary questions...
Hi friends, happy new Year!
me again, with some more questions...
After the lengthy capacitor-selection-process (many thanks for the patient and thoroughful answers!), here's my next obstacles...
...which are the PI Resistors R1-R8, and, respectively, R_Optional1 – R_Optional6. Are they improving things?
My next conceptual question is: 6L6 built his F4 / PSU with bridges and did not use the diodes-section... I understand it's perfectly working, but why not building the diodes-section? (Is it better or easier, or both?) Since both great 6L6 and mighty ZM use and propagate them, I'm quite sure it is at least equally good.
But the question-marks remain, as I haven't yet understood the function of the diode-board, I need some brain-food to understand it? (I already read over the link to a "beginner's guide to building a PSU", provided by 6L6. Will read it again. Is there more?)
Many thanks!
david
Hi friends, happy new Year!
me again, with some more questions...
After the lengthy capacitor-selection-process (many thanks for the patient and thoroughful answers!), here's my next obstacles...
...which are the PI Resistors R1-R8, and, respectively, R_Optional1 – R_Optional6. Are they improving things?
My next conceptual question is: 6L6 built his F4 / PSU with bridges and did not use the diodes-section... I understand it's perfectly working, but why not building the diodes-section? (Is it better or easier, or both?) Since both great 6L6 and mighty ZM use and propagate them, I'm quite sure it is at least equally good.
But the question-marks remain, as I haven't yet understood the function of the diode-board, I need some brain-food to understand it? (I already read over the link to a "beginner's guide to building a PSU", provided by 6L6. Will read it again. Is there more?)
Many thanks!
david
not being familiar with pcbs in question , I'll better pass answering that
regarding Graetz Bridges - that's matter of personal choice - which one you'll use and how much you're gonna to lily-guild them
regarding Graetz Bridges - that's matter of personal choice - which one you'll use and how much you're gonna to lily-guild them
Happy New Year!
They are part of the CRC... (Capacitor, Resistor, Capacitor...) Filter. More educated minds will need to chime in, but I think pi filters are traditionally CLC... Capacitor, Inductor, Capacitor.
Anyway - You need the resistors. Populate all but the optional. If you decide to build another amp and/or later if you decide you need the optional resistors, it's easy-peasy to add them.
Let's start with the purpose. All together, your "PSU" needs to turn AC into nice, smooth, linear DC. In this case bipolar - positive and negative supplies. Here's the first Google result for some decent brain food.
Rectifier - Wikipedia
Section 1 - Rectification. Monolithic bridges vs. independent components (the diode board) may involve a few choices... in somewhat of an order, but it may not be perfect.
Cost - Advantage monolithic
Space - Advantage monolithic
Time / Ease of assembly - Advantage monolithic.
Versatility - Ever want to see how different diodes perform? Does your brain get all tingly when you hear "fast switching / soft recovery"? - Advantage discrete.
Performance - Advantage diodes if your intent is to eke out every last bit of performance, but then... what diodes?
There are also regulated supplies out there etc. etc. etc. The possibilities are endless.
Section 2 - smoothing - see above re: filters. That's your caps and resistors. I had this web page in my notes.
DC Power Supply design
I like pictures... 😀 It shows only the cap effect, but it's one I refer to as I learn. IMO, his whole section on PSUs was well-written and easy to digest.
Hope that gets you rolling in the right direction. Don't agonize. Unless you really want to get into the minutiae, it will come down probably to cost, space, and time. I've seen people claim that the discrete diodes perform better, but I've never seen anyone claim they could hear a difference. I still built the diode sections... Oh well, I never claimed to follow logic. 😀
Thank you, itsallinmyhead! Appreciate a lot!
The link to google was, ahem, obvious. I got so lost in the diodes-jungle that I forgot to look after the forest.
(And, although I see the logic, I'm not sure yet wether I should follow it or not. 🙂 — so I'm back contemplating about the paradoxon between logic and solder 😀 )
david
This illustrates what the diodes are doing. The source at the left side is the AC from the transformer.
Full-Wave Rectifier w/ Filter
Full-Wave Rectifier w/ Filter