I've been studying the F5T in prep for a build, ran it through Spice, etc. and I want to build my own PCBs, bin my own parts, etc. strictly for the educational value. I'm even planning on building my own chassis. I'm at the point where I need to acquire parts, but after searching (and searching...and searching...) and reading this thread, among others, I see a continuing problem with parts availability. I went on a search / rampage looking for substitutes and can't find any. I'll admit I'm new to these parts and amp design in general, so I'm still learning what to look for, but I have come away wondering why several manufacturers make seemingly good low-noise N channel jfets up the yahzoo, but no one seems to make p-channels. So that brings me to say "uncle" and ask you guys... What is the current understanding with regard to substitutes for the fets (IPS and OPS) in this project, keeping in mind I want parts that I can buy from a distributor (e.g. Mouser, etc.) in good quantity. I would prefer to stay away from ebay/paypal sellers. Tried to buy something the other day from an seller via paypal and my CC company locked my account for suspected fraud. More time wasted getting that unlocked. Just don't want the hassle anymore. I want to call Mouser and be done with it. I realize many want to stick with the TO-92 (myself included), and I personally hate soldering surface mount stuff, but does anyone have any ideas for surface mount equivalents of the 2SK170/2SJ74? I'm willing to change my PCBs if I can avoid changes that would destroy the quality and intent of the F5 circuit. Thanks for any help.
tricky
I wouldnt be so sure about that
the 400mm is made with split heatsinks
which demands for certain designs to use that
or else the 300mm would be more effective....IMO
longer fins would be more effective
that depends on the fet spacing. With the store Boards, 400mm will be the best way. or even 500mm.
yes...just dont use 'heat-spreader-bar', unless all parts have been machined, and are 100% plane
If I could work out a method of reading the peak charging current in the secondary circuit using a DMM, I would have done it already.
But I can't, I'm not clever enough.
Total dissipation is around 300W roughly calculated in my head here as I sit and type. The outputs are arranged on 2 separate output boards, 3 per board (there are some photos of my setup on the KSA50 thread's last few pages). There are not any hotspots on my sinks so the heat is spread evenly. Ta when I did the biasing was about 30 deg C.
What tool or device would I need to be able to make the measurement? Fast blow fuse is too slow?
thats about 60W more then F5t v1/v2. if that is for 2 ch. and ambient temp around 5-10c more then most have. that is a huge difference
and what clearing do you have under the sinks?
In my test setup there were not any clearing under the sinks. When the case is finished there should be 30mm clearance.
What is the efficiency of the F5T V2? According to the calculations for my KSA50 it's efficiency is about 40%. Is the F5 so much more efficient? Please note that I assume most F5T builders will probably build the V2 variant biased to provide 50W, hence the comparison to the KSA50. Or don't you think it is valid?
What is the efficiency of the F5T V2? According to the calculations for my KSA50 it's efficiency is about 40%. Is the F5 so much more efficient? Please note that I assume most F5T builders will probably build the V2 variant biased to provide 50W, hence the comparison to the KSA50. Or don't you think it is valid?
F5t v2 runs on +/32V and around 1.8A bias. wich gives it a dissipation of 115W pr ch.
krell often run a very high bias. and then the dissipation goes up.
the 0 clearing also play a role in Your high sink temp.
so with those sinks, 35W less dissipation pr ch, 10c lower ambient, and 40mm+ clearing under the sinks. then the heat will be nothing like the heat you get it will be atleast 15c lower.
krell often run a very high bias. and then the dissipation goes up.
the 0 clearing also play a role in Your high sink temp.
so with those sinks, 35W less dissipation pr ch, 10c lower ambient, and 40mm+ clearing under the sinks. then the heat will be nothing like the heat you get
it will be atleast 15c lower.
maybe also focus on good venting of the box itself
it would be possible to make it work in unity with heatsink convection
difficult with a standard box design
but might work well if lifting the lid up using print board stands
imo...not sure I would like to mess with anything that wouldnt work with 'normal' heatsinking....whatever normal is
it would be possible to make it work in unity with heatsink convection
difficult with a standard box design
but might work well if lifting the lid up using print board stands
imo...not sure I would like to mess with anything that wouldnt work with 'normal' heatsinking....whatever normal is