Take a look at this and the posts that come after it:
https://www.diyaudio.com/community/threads/the-f6-revisited.405444/post-7560367
https://www.diyaudio.com/community/threads/the-f6-revisited.405444/post-7560367
I have been listening to the Rev. 1 boards, built with my BOM and IRFP240 outputs now for several weeks. also using the Jensen transformer, since I had them. The sound is WONDERFULL! Dead silent at idle, No distortion to my ears when playing music. Thanks for puting together this project.
I am planning to build a high-powered version soon, and am considering a few things. Due to the blue LED having a slightly positive thermal coefficient, as opposed to the green's negative, I fixtured up to place 4 green LEDs in series for use with the IXYS outputs. Also, I'll use the 3.6K resistors. My question is: How high a power supply voltage would be practical before the outputs can't take the heat from the bias required for higher output power? I'm thinking of using a 24 or 25V transformer for the PSU, which may require 50V caps.
Any thoughts?
I am planning to build a high-powered version soon, and am considering a few things. Due to the blue LED having a slightly positive thermal coefficient, as opposed to the green's negative, I fixtured up to place 4 green LEDs in series for use with the IXYS outputs. Also, I'll use the 3.6K resistors. My question is: How high a power supply voltage would be practical before the outputs can't take the heat from the bias required for higher output power? I'm thinking of using a 24 or 25V transformer for the PSU, which may require 50V caps.
Any thoughts?
The rate limiting step in this instance will be the size of your heatsinks, and whether the output IRFP 240 MOSFETS can handle the dissipation (the maximum is 150W for this device, but generally you want to be at no more than 75watts for longevity) . So what size enclosure and heatsinks do you have? A 24 or 25V secondary will give you roughly +/- 30V DC after rectification and resistance losses. 50V caps would be a good choice here. If you are at +/- 30V, then you should limit your bias to 1.2A = 72 watts. But IRFP240 Mosfets are plentiful so YMMV.
If you want a higher bias, then you will need to choose Mosfets that can handle a higher dissipation, such as the IXTQ75N10P (360 watt dissipation!). Read post 114 for info.
In summary, I will quote 6L6 who wrote this in the original F6 illustrated build guide:
Best,
Anand.
Last edited:
Thanks for the reply. I'm planning to use the IXTQ75N10P outputs with Keratherm pads. I have the standard 4U 300mm chassis. I expect heatsink capacity will be the limiting factor. I guess I'll find out when I build it how much power can be dissipated before getting too hot. 55C is very hot. That IR thermometer in the garage will come in handy.
I wonder how much power this will make into 8 ohms...
I wonder how much power this will make into 8 ohms...
Indeed, fans can help a lot. I use simple add on fans blowing into the sinks of my class A amps, most of which have 3U300 sinks. Convenience is king, so I have this product. External USB powered, and operationally silent when using the lowest setting. My sinks that are up to 55C without the fans are just barely warm to the touch with the fans. Easy.
Yeow!!
Indeed, with the diyaudio F6 using IRFP240’s, using 4U/300 I noticed it was biased at just about 1.05A which is ~50 watts. And at 50 watts, the heatsinks were about 45 degrees C.
Best,
Anand.
The IXTQ75N10P is an excellent device for the F6. It is best kept at or below 26 Vds. Check out the SOA graph in the datatsheet. At this voltage bias current can range from 1.6A to 1.8A depending on the heatsinks. This will be good for deep Class A operation to around 30 Watts.
I happen to be running mine just beyond the upper edge of the SOA at 26.5V and 1.75A. They do sound great that way. I added extra L brackets inside my chassis to boost the thermal capacity. Heat is spread to the baseplate, frontplate and backplate. Less than 50W per device w/ No fans.
I happen to be running mine just beyond the upper edge of the SOA at 26.5V and 1.75A. They do sound great that way. I added extra L brackets inside my chassis to boost the thermal capacity. Heat is spread to the baseplate, frontplate and backplate. Less than 50W per device w/ No fans.
Thanks for the correction ZM, my brain conked out (fatigue) and forgot overall dissipation is shared amongst multiple output devices!
Best,
Anand.
The IXTQ75N10P parts need an extra 1V of Vgs to bias correctly. With the new PCBs that is easily accomplished by substituting a blue LED such as the LTL1CHTBK5 for one of the green ones in each string. R1 and R2 don't need to change, though I like the sound of splitting R1 between the R1* and R1 locations. R11 and R12 can be from 68Ω to 100Ω.
That is good advice. However, I have broken all three of those rules... 😬 I do think they are terrific parameters for the new builder... Between the Zenductor and this screenshot below, it should be okay to run your FETs hot so long as you have some experience. In other words, if your heatsinks get to 60C on a hot day and your FET cases at 70, I wouldn't worry too much. The component that seems to stress the most are the caps but even those seem pretty rugged compared to the ones from back in the day.
In the summer my house has been 80F-85F inside. I still use my amps. I wouldn't be surprised if my heatsinks have been as high as 65C... Not a single problem so far.
Here I hope I've uploaded the updated BOM with the IXYS outputs, and options for the blue LEDs and hight voltage 3.6K bias resistors.
I've completed my hi-voltage version... Antek AS-3224 X-former, Bridge Diode blocks, 88,000 mFd per phase power supply filtering. Installed one extra .47ohm resistor in the CRC, though it may not make a significant difference. I used the Cinemag CMOQ-4LPC with the ZM adapters.
On the workbench, Voltage at clipping into my 8.2 ohm dummy load was 53.6V P-P. Calculates to about 43+ Watts! Input was in excess of 4.5 volts P-P from the HP signal generator. Hooked up to the system, driving B&W 801s, as good as the baseline build but with more...
This amp will rock with Rachmaninov or Led Zep. Dire Straits never sounded better. If you build it be aware that the higher voltage power supply will be pushing the output transistors just out of their SOA per their datasheet! I biased mine to about .685 volts across the 0.47 ohm source resistor and have not had any trouble yet in at least 8 hrs of on time continuous. Temp at transistor mounting screw was 55 degrees C. Heatsinks were nice and warm, like cozy up to them on a cold day...
I've completed my hi-voltage version... Antek AS-3224 X-former, Bridge Diode blocks, 88,000 mFd per phase power supply filtering. Installed one extra .47ohm resistor in the CRC, though it may not make a significant difference. I used the Cinemag CMOQ-4LPC with the ZM adapters.
On the workbench, Voltage at clipping into my 8.2 ohm dummy load was 53.6V P-P. Calculates to about 43+ Watts! Input was in excess of 4.5 volts P-P from the HP signal generator. Hooked up to the system, driving B&W 801s, as good as the baseline build but with more...
This amp will rock with Rachmaninov or Led Zep. Dire Straits never sounded better. If you build it be aware that the higher voltage power supply will be pushing the output transistors just out of their SOA per their datasheet! I biased mine to about .685 volts across the 0.47 ohm source resistor and have not had any trouble yet in at least 8 hrs of on time continuous. Temp at transistor mounting screw was 55 degrees C. Heatsinks were nice and warm, like cozy up to them on a cold day...