Ah...the schematics on page 1 of this thread. I now see the confusion.
When in doubt, I always go back to the schematics in the F4 manual. See page 13 here: https://www.firstwatt.com/wp-content/uploads/2023/08/prod_f4_man.pdf
As you can see, 220uF is the actual part. In my earlier response I thought perhaps you didn't have 220uF but did have 470uF caps.
When in doubt, I always go back to the schematics in the F4 manual. See page 13 here: https://www.firstwatt.com/wp-content/uploads/2023/08/prod_f4_man.pdf
As you can see, 220uF is the actual part. In my earlier response I thought perhaps you didn't have 220uF but did have 470uF caps.
TLDR - The F4 is a great buffer amp! Build it!
Beware that biasing it requires proper TL341s and maybe tweaking R5 or R8
Longer version
Hi, after building several Pass amps (ACA, F5, Aleph J) I decided to go for a F4. I had been playing with several iterations of the MoFo (great stuff, for the price and simplicity it is wonderful!) and wanted the same concept but a bit more powerful and transparent. The F4 is been playing since this morning and it sounds VERY good indeed.
For now, I'm using it with a custom tube preamp, designed and built by a friend and ex-partner in a small tube amp and preamp company we owned in the past. The pre is based on the PCC85 tube (I know, it's for UHF work but it sounds really good here), and is capable of driving a buffer amp to very healthy levels.
Although still playing with a modest DAC, USB-connected to a Volumio Raspberry Pi, and driving a pair of home-made Juliets (Scott Moose DCR for the Markaudio 12p), I'm really impressed with everything about the amp, especially timbre and soundstage (incredibly focused but super-wide, way outside the speakers). Speed is first-rate, on all the audio band, although no hint of hardness.
Even if nothing else, building the F4 serves to show how good the preamp is, I had only tested it with the MoFo. It has too much gain to all Pass amps I own, especially Aleph J or X150.8, which I use in my main system, balanced, direct from DirectStream dac (and I'm using the 20dB attenuator on the DAC for some of the recordings, or the digital volume control will be around 50 - in 100).
As much as I love the MoFo, all it takes is 5 minutes of the F4 to understand that they play on different leagues. I'm really curious about the match with my Martin-Logan Montis and PS Audio DirectStream. I also built and love the BA-3 FE as a preamp and it should be really good with this amp. A lot to try out, even some SET amps that finally will be able to drive some demanding speakers...
Anyway...
I essentially posted to report on some troubleshooting I had to go through. When finished, I could bias the MOSFETs only to around 70 mV on the source resistors (0.47 ohm). I had to lower R5 to 5k AND replace the TL431 for the pair that came with the MOSFETs to achieve proper bias.
I bought the boards and semiconductors pack from diyaudio store and forgot that it brought along a couple of TL431, so I ordered some from Mouser and used these. Now I understand why the package includes the TL431...
Thank you very much to all, especially Nelson Pass (incredible designs and generosity!), 6L6 (for the great guide and advice on the R5) and ZenMod (for the tip to replace TL431).
BTW, even with matched devices from the store, voltage across the several source resistors is not the same, dispersion rgd the average is around 10% for the "worse" case, at least with 200mV. No sign of current hogging, but definitely some deviation. Since I believe store devices were correctly matched (no numbers on them, though), I guess some of the variation may come from the resistors (3W, 5%, which are the same I always used in the past, for the other amps anyway). No problem with offset though, easy to get it to virtually zero and it is stable (for now, amp is still open, no covers).
I'm going to let it "cook", finetune final bias/offset, install definitive speaker terminals (still using what I had at home) and then maybe I'll get some better matched resistors and change them.
Again, thank you!
Getting back to hear some more music...)
Beware that biasing it requires proper TL341s and maybe tweaking R5 or R8
Longer version
Hi, after building several Pass amps (ACA, F5, Aleph J) I decided to go for a F4. I had been playing with several iterations of the MoFo (great stuff, for the price and simplicity it is wonderful!) and wanted the same concept but a bit more powerful and transparent. The F4 is been playing since this morning and it sounds VERY good indeed.
For now, I'm using it with a custom tube preamp, designed and built by a friend and ex-partner in a small tube amp and preamp company we owned in the past. The pre is based on the PCC85 tube (I know, it's for UHF work but it sounds really good here), and is capable of driving a buffer amp to very healthy levels.
Although still playing with a modest DAC, USB-connected to a Volumio Raspberry Pi, and driving a pair of home-made Juliets (Scott Moose DCR for the Markaudio 12p), I'm really impressed with everything about the amp, especially timbre and soundstage (incredibly focused but super-wide, way outside the speakers). Speed is first-rate, on all the audio band, although no hint of hardness.
Even if nothing else, building the F4 serves to show how good the preamp is, I had only tested it with the MoFo. It has too much gain to all Pass amps I own, especially Aleph J or X150.8, which I use in my main system, balanced, direct from DirectStream dac (and I'm using the 20dB attenuator on the DAC for some of the recordings, or the digital volume control will be around 50 - in 100).
As much as I love the MoFo, all it takes is 5 minutes of the F4 to understand that they play on different leagues. I'm really curious about the match with my Martin-Logan Montis and PS Audio DirectStream. I also built and love the BA-3 FE as a preamp and it should be really good with this amp. A lot to try out, even some SET amps that finally will be able to drive some demanding speakers...
Anyway...
I essentially posted to report on some troubleshooting I had to go through. When finished, I could bias the MOSFETs only to around 70 mV on the source resistors (0.47 ohm). I had to lower R5 to 5k AND replace the TL431 for the pair that came with the MOSFETs to achieve proper bias.
I bought the boards and semiconductors pack from diyaudio store and forgot that it brought along a couple of TL431, so I ordered some from Mouser and used these. Now I understand why the package includes the TL431...
Thank you very much to all, especially Nelson Pass (incredible designs and generosity!), 6L6 (for the great guide and advice on the R5) and ZenMod (for the tip to replace TL431).
BTW, even with matched devices from the store, voltage across the several source resistors is not the same, dispersion rgd the average is around 10% for the "worse" case, at least with 200mV. No sign of current hogging, but definitely some deviation. Since I believe store devices were correctly matched (no numbers on them, though), I guess some of the variation may come from the resistors (3W, 5%, which are the same I always used in the past, for the other amps anyway). No problem with offset though, easy to get it to virtually zero and it is stable (for now, amp is still open, no covers).
I'm going to let it "cook", finetune final bias/offset, install definitive speaker terminals (still using what I had at home) and then maybe I'll get some better matched resistors and change them.
Again, thank you!
Getting back to hear some more music...
)
Balanced F4
I've put two F4's in the same case - I didn't want to deal with two cases. It can also be run paralleled if jumpers are used. The devices and boards are from the DIY Audio Store, and the transformers from Tiger Toroids. The inrush current limiters are bypassed with a relay after 20 seconds or so. I used the cheapomodo to figure snubbers. Otherwise there's nothing innovative or unusual here. It sounds as you might expect - mighty fine. Please excuse the crappy camera, but I think the pics at least give the idea. The only disappointment - the lovely blue switches I got from Amazon have red LEDs in them, if you can believe it. Ruins the color scheme.
Thanks to Woodworker for his help in the planning stages. The amp runs about 52º on one side and 54º on the other (after running a break-in track overnight) and the biases are pretty close to spec, so I think I'll be OK heat-wise.
I've put two F4's in the same case - I didn't want to deal with two cases. It can also be run paralleled if jumpers are used. The devices and boards are from the DIY Audio Store, and the transformers from Tiger Toroids. The inrush current limiters are bypassed with a relay after 20 seconds or so. I used the cheapomodo to figure snubbers. Otherwise there's nothing innovative or unusual here. It sounds as you might expect - mighty fine. Please excuse the crappy camera, but I think the pics at least give the idea. The only disappointment - the lovely blue switches I got from Amazon have red LEDs in them, if you can believe it. Ruins the color scheme.
Thanks to Woodworker for his help in the planning stages. The amp runs about 52º on one side and 54º on the other (after running a break-in track overnight) and the biases are pretty close to spec, so I think I'll be OK heat-wise.
It's 5U/500 - I wanted to be sure to have enough heat sink. In its current position (next to an A/C vent) it stays below 50º after several hours of idle. Drawback - it weighs at least a half a pound per watt.