I can hold them indefinitely. My laser thermometer is out of service but a meat thermometer rested precariously on each heat sink reads about 45c.
Almost precisely 45C, now corroborated by an IR temp gun. Offset crept up to ~25mV on both channels so I popped the lid and twiddled it back down a bit. Thanks, everybody.
Ok it turns out that my bias current has been a bit higher than I thought, around .65v. I was still hooked up to the variac and forgot to double check it again plugged straight into mains. Either way, the temps I reported were with the higher current so I pushed things up a tiny bit more (.680v) and I’m still not hitting 50c. Should I be concerned? I’m using a 4U chassis and keratherm insulators.
Sounds good! Check your DC offset, adjust as necessary and leave the bias where you have it. If your heatsinks are handling it , no worries!
Final Bias Steps
"WOW"
I have had my newly constructed F5 Ver. 3.0 running for 4 days. What an amazing amp. I have all of my CD's ripped to FLAC, playing through an Allo Boss Dac on top of a Raspberry Pi.
Three Dog Night, AC-DC, Robin Trower, Led Zepelin, Steely Dan, Willie Nelson, Johnny Paycheck, Zac Brown, don't forget Areatha, John Lee Hooker, Muddy Waters, Little Richard, and
Holly Catfish Batman "Ella Fitzgerald". I have been listening to MP3's so long I forgot what a good source and a Great Amp sounded like. Thanks Nelson.
Don't try to guess my age by the music I listen too.
Now my question: I biased the amp to 485mv, one channel at a time. I am now ready to to bias the amp to 600mv. Should I do this "One Channel at a time" or can I do this with both
channels connected to power. I will disconnect V+ V- and Grnd if advised to bias one channel at a time. I will also adhere to Nelsons Temp recomedation of "GRIKY"
Time to listen to some "Al Green"
"WOW"
I have had my newly constructed F5 Ver. 3.0 running for 4 days. What an amazing amp. I have all of my CD's ripped to FLAC, playing through an Allo Boss Dac on top of a Raspberry Pi.
Three Dog Night, AC-DC, Robin Trower, Led Zepelin, Steely Dan, Willie Nelson, Johnny Paycheck, Zac Brown, don't forget Areatha, John Lee Hooker, Muddy Waters, Little Richard, and
Holly Catfish Batman "Ella Fitzgerald". I have been listening to MP3's so long I forgot what a good source and a Great Amp sounded like. Thanks Nelson.
Don't try to guess my age by the music I listen too.
Now my question: I biased the amp to 485mv, one channel at a time. I am now ready to to bias the amp to 600mv. Should I do this "One Channel at a time" or can I do this with both
channels connected to power. I will disconnect V+ V- and Grnd if advised to bias one channel at a time. I will also adhere to Nelsons Temp recomedation of "GRIKY"
Time to listen to some "Al Green"
Keep both channels running. Get one up to the desired bias, then the other one. check back with the first one. You need to get the entire chassis up to temperature.
Also, make sure you keep the lid on the case so it warms up to the final operating
condition. Only take it off to make adjustments and then
put it back on and let the temp and bias measurements stabilize. Go back and
forth between the channels. You'll likely have to tweak this a few times.
condition. Only take it off to make adjustments and then
put it back on and let the temp and bias measurements stabilize. Go back and
forth between the channels. You'll likely have to tweak this a few times.
F5 Simulation with KiCAD/ngspice
For those of you interested in F5 simulation and using KiCAD, I have made some setups for KiCAD/Eeschema/ngspice. Electro-thermal simulation of the output stage (transistors, heatsinks, thermistors) is included, as well as load changes.
Installing and using ngspice - an opensource simulator
Installing and using ngspice - an opensource simulator
For those of you interested in F5 simulation and using KiCAD, I have made some setups for KiCAD/Eeschema/ngspice. Electro-thermal simulation of the output stage (transistors, heatsinks, thermistors) is included, as well as load changes.
Installing and using ngspice - an opensource simulator
Installing and using ngspice - an opensource simulator
What is a safe minimum current rating for a power switch for an F5 with a soft start board? I'm in Europe, so 240V mains. My current fuses are 1.25A, but I'm guessing that since they are slow blow they may actually be handling a higher current during switch on.
The reason I ask is I'm looking for a long (~20mm) flatted toggle switch (DPDT or DPST) for the front panel of my F5 which currently only has the switch on the back. I've found a few candidates, but one is out of stock everywhere I've looked (APEM 5646AB-8), one seems like it might have too low of current rating (C&K 7201P1Y4ZQE, 2A@250VAC).
My best bet at the moment seems to be the NKK M2022E4S1W01, rated for 3A@250VAC:
I'd prefer not to have to add another transformer to power a relay board, even though I know that would open a lot more options for power switches...
The reason I ask is I'm looking for a long (~20mm) flatted toggle switch (DPDT or DPST) for the front panel of my F5 which currently only has the switch on the back. I've found a few candidates, but one is out of stock everywhere I've looked (APEM 5646AB-8), one seems like it might have too low of current rating (C&K 7201P1Y4ZQE, 2A@250VAC).
My best bet at the moment seems to be the NKK M2022E4S1W01, rated for 3A@250VAC:
I'd prefer not to have to add another transformer to power a relay board, even though I know that would open a lot more options for power switches...
Last edited:
Back when I biased my F5 I had a problem where I needed to set uneven biases on one channel in order to get 0mV at the output (e.g. +543mV and -557mV, where my target was ~550mV). The other channel was almost perfect (+550mV and -551V).
I just recently received another set of parts (M2X boards and matched quad of JFETs), and it made me wonder about the quad matching scheme.
For a matched quad (NP), are they all matched to each other, or in pairs? I guess what I am asking is that if you built the amp, set it up and biased it, and then swapped LSJ74s (leaving the LSK170s in their original positions), would you expect to see a difference at the outputs?
I may try swapping them, or even putting in my new matched quad, just to see if I can minimize the bias spread. I'm assuming it's not actually a problem to have non-matching +/- voltages, but it would be a learning opportunity to see what parts affect what measurements.
I just recently received another set of parts (M2X boards and matched quad of JFETs), and it made me wonder about the quad matching scheme.
For a matched quad (NP), are they all matched to each other, or in pairs? I guess what I am asking is that if you built the amp, set it up and biased it, and then swapped LSJ74s (leaving the LSK170s in their original positions), would you expect to see a difference at the outputs?
I may try swapping them, or even putting in my new matched quad, just to see if I can minimize the bias spread. I'm assuming it's not actually a problem to have non-matching +/- voltages, but it would be a learning opportunity to see what parts affect what measurements.
Last edited:
@booja30
the pros here suggest that everything within 10% is fine
you're within 1.3% - so nothing to be worried about
just keep offset small.
cheers
A.
the pros here suggest that everything within 10% is fine
you're within 1.3% - so nothing to be worried about
just keep offset small.
cheers
A.
Hello booja30
Check this one 16A/250 V AC:
C1710HO BULGIN - Switch: toggle | Pos: 2; SPDT; ON-ON; 16A/250VAC; -25/70degC; IP40; AE-C1710HOAAA | TME - Electronic components
Best regards
Carlos
Sorry
That one is ON-ON....
This one is more apropriated:
C1750HO BULGIN - Switch: toggle | Pos: 2; DPST; OFF-ON; 16A/250VAC; 28A/28VDC; IP40; AE-C1750HOAAA | TME - Electronic components
Check this one 16A/250 V AC:
C1710HO BULGIN - Switch: toggle | Pos: 2; SPDT; ON-ON; 16A/250VAC; -25/70degC; IP40; AE-C1710HOAAA | TME - Electronic components
Best regards
Carlos
Sorry
That one is ON-ON....
This one is more apropriated:
C1750HO BULGIN - Switch: toggle | Pos: 2; DPST; OFF-ON; 16A/250VAC; 28A/28VDC; IP40; AE-C1750HOAAA | TME - Electronic components
Last edited:
I agree. Those deviations are smaller than the tolerances of the parts you're using. Those resistors you're measuring over aren't 0% tolerance parts and that's what determines your voltage given a current.
Thanks guys. I had a feeling it was ok but the build guide implies they should be the same voltages. Anyway, I just placed a big Mouser order today so I can have some more fun building M2X boards, and a BA2018 preamp with two VRDN PSUs. I also bought a few power switch options to try (NKK 20mm DPDT flat toggle). Now to wait for the BA2018 board to come back in stock...
Hello everyone, some time ago I finished the Pearl 2, B1 and F5 chain, I like it a lot. I also implanted Mundorf audiophile condensers in B1. It was worth it! don't believe anyone who says something different.
But I want the helicopter in "The Wall" by Pink Floyd to land IN my room and not outside.
When I think of driving my JBL Ti6K via bi-amping. Before I start getting material, the community would like to ask if F5 and F5Tv2 can play correctly in phases?
I wanted to build the F5 Turbo anyway, it would be the right opportunity.
But I want the helicopter in "The Wall" by Pink Floyd to land IN my room and not outside.
When I think of driving my JBL Ti6K via bi-amping. Before I start getting material, the community would like to ask if F5 and F5Tv2 can play correctly in phases?
I wanted to build the F5 Turbo anyway, it would be the right opportunity.