Awesome 👍In my bench tests, I have used around 20 Vdc for the power supply (due to the limited heatsink), so I can measure around 12 Vpp over the load without appreciable clipping. I have reached close to 16 Vpp, but with obvious cuts.
The table I uploaded includes an additional margin of safety so that the distortion profiles remain adequate for what it was designed for (single-ended character throughout its excursion, with predetermined cadence in its distortion profile and predominance of H2 over the higher ones). ). It is a more conservative table in terms of power expectations.
The voltage drop that I considered for the LM317 plus the voltage drop in one of the power resistors was 3.75 V, while for the mosfet (plus the voltage drop in one of the power resistors) was 5.3 V.
Thanks for the nice PCB layout. I have a few comments if I may: To be considered or not, your call.
1. Was J4 added so we can put a small value quality cap across the output cap ?
2. Could you add a resistor, ex 330 ohm , 3W across J2.
3. Could the input cap C2 made much bigger to accommodate Polypropylene cap. and add multiple holes to fit many types.
4. Should you add a gate resistor for Q1, ex 220 ohm ?
5. Should the width of the traces carrying the most power by enlarged?
Thanks
Eric
1. Was J4 added so we can put a small value quality cap across the output cap ?
2. Could you add a resistor, ex 330 ohm , 3W across J2.
3. Could the input cap C2 made much bigger to accommodate Polypropylene cap. and add multiple holes to fit many types.
4. Should you add a gate resistor for Q1, ex 220 ohm ?
5. Should the width of the traces carrying the most power by enlarged?
Thanks
Eric
1 yes. I will add the text 'bypass cap'.
2 yes
3 yes
4 yes
5 not a problem to make wider
I've recently been learning how to use Kicad. As a 75 year old, I find it difficult to hold a soldering iron steady, but I don't seem to have a problem with PCBs. I have found these youtube videos to be an excellent introduction.
It may take a day or so to make the changes.
ray
2 yes
3 yes
4 yes
5 not a problem to make wider
I've recently been learning how to use Kicad. As a 75 year old, I find it difficult to hold a soldering iron steady, but I don't seem to have a problem with PCBs. I have found these youtube videos to be an excellent introduction.
It may take a day or so to make the changes.
ray
Thanks for everything Ray.
Learning how to use Kicad at 75 is simply amazing 👍 BRAVO !
Thanks for the video links.
No rush for the PCB changes, much appreciated.
Eric
Learning how to use Kicad at 75 is simply amazing 👍 BRAVO !
Thanks for the video links.
No rush for the PCB changes, much appreciated.
Eric
Before making changes 2 (330 ohm across J2) and 4 (gate resistor), I think it would be courteous to ask permission from diegomj1973.
C2 now allows for P of 5 mm 7.5mm 10mm 15mm.
Designing PCBs is a wonderful exercise for the mind. I am still very much in the learning stage.
ray
C2 now allows for P of 5 mm 7.5mm 10mm 15mm.
Designing PCBs is a wonderful exercise for the mind. I am still very much in the learning stage.
ray
A simple scheme that includes a transformer, rectifier and filter will be more than enough. For the filtering capacitor I recommend a value of 6800 uF up to a power of approximately 6 W per channel (one 6800 uF capacitor for each channel), so that the ripple does not exceed 2.44 Vpp. This requirement will be necessary to be aligned with the minimum resolutions for 16-bit content.
Best regards
Best regards
On the PCB, bigger pads and wider traces at R1 and R2 will help dissipate heat. Wider trace to C1 would be good, too.
Add decoupling capacitor to power input?
Add decoupling capacitor to power input?
Lovely design!
My only suggestion is to increase the feedback resistor values. Currently the input impedance is 330 ohms. A little low. You can easily raise it to 4-5k, scale the 2.2k accordingly and then most SS preamps will drive it fine.
My only suggestion is to increase the feedback resistor values. Currently the input impedance is 330 ohms. A little low. You can easily raise it to 4-5k, scale the 2.2k accordingly and then most SS preamps will drive it fine.
Gate on IRFP150 is pin 1 (not pin 2), a gate stopper resistor might be a good idea as well as trying to make the traces from input to gate as short as possible
and yes increasing input impedance and feedback resistor accordingly
and yes increasing input impedance and feedback resistor accordingly
Last edited:
Thanks for the various corrections and suggestions.
In addition, after looking at the Zenductor, I propose adding the following:
1 - an alternative rcj-046 rca input in addition to J1
2 - providing mounting holes for a HS380 heatsink 2.2 C/W
3 - an alternative CP-037AH DC supply in addition to J3
Hopefully, it will all fit on the 100x100 mm board.
For DC supply, I have been thinking about using the Dienoiser LM317 supply from the Power Supplies forum
https://www.diyaudio.com/community/...y-317-based-v-reg.331491/page-47#post-6255032
ray
In addition, after looking at the Zenductor, I propose adding the following:
1 - an alternative rcj-046 rca input in addition to J1
2 - providing mounting holes for a HS380 heatsink 2.2 C/W
3 - an alternative CP-037AH DC supply in addition to J3
Hopefully, it will all fit on the 100x100 mm board.
For DC supply, I have been thinking about using the Dienoiser LM317 supply from the Power Supplies forum
https://www.diyaudio.com/community/...y-317-based-v-reg.331491/page-47#post-6255032
ray
Hi Ray
You could add your 3 suggestions, it’s your PCB design after all. It doesn’t affect the amplifier design.
I would add these 3 items for a monobloc that dissipate less heat. According to Diego in post no.15 with a 24V supply one channel will dissipate a total of about 15W of heat. With a 2.2C/W heatsink your theoretical temp rise will be 33C so in a 25C it’s a heatsink temp of 58C, very hot. If the mosfet and the LM317 each have their own heatsink it’s doable but I wonder if they are physically too close.
BR,
Eric
You could add your 3 suggestions, it’s your PCB design after all. It doesn’t affect the amplifier design.
I would add these 3 items for a monobloc that dissipate less heat. According to Diego in post no.15 with a 24V supply one channel will dissipate a total of about 15W of heat. With a 2.2C/W heatsink your theoretical temp rise will be 33C so in a 25C it’s a heatsink temp of 58C, very hot. If the mosfet and the LM317 each have their own heatsink it’s doable but I wonder if they are physically too close.
BR,
Eric
Last edited:
This is Very Cool...............................
This type of configuration is/was actually on my list one of my next things to try as well, I am currently using Resistors, and then BJT's as CCS's in a design concept I have been working on !! 🙂
The amp I have built works well, based on my sim schematic Found Here,
https://www.diyaudio.com/community/...jects-someone-could-share.288899/post-5195755
I am glad that you posted this and some results, looks nice, Keep up the Great work !!
I was thinking of using use it in place of load resistor for the bjt driver on my FET gate. 😉
In its current form it has a 2N2222 (2x or 3X is even better), but it does get really hot and it is not very stable with temp, don't need 12v of swing on the gate either (maybe 5v or 6v max on the gate even on an extreme output design) so I was thinking about using the 3 terminal regulator current source configuration.
I have always been a fan of Elvee's Chip regulator amp, but never made one yet, I bought a bunch of LM337 to do so a while back and they were all fake.........
So I never made that one yet. :/
Good to see that it may work well and will give it a try. 🙂
Cheers !!

jer 🙂
This type of configuration is/was actually on my list one of my next things to try as well, I am currently using Resistors, and then BJT's as CCS's in a design concept I have been working on !! 🙂
The amp I have built works well, based on my sim schematic Found Here,
https://www.diyaudio.com/community/...jects-someone-could-share.288899/post-5195755
I am glad that you posted this and some results, looks nice, Keep up the Great work !!
I was thinking of using use it in place of load resistor for the bjt driver on my FET gate. 😉
In its current form it has a 2N2222 (2x or 3X is even better), but it does get really hot and it is not very stable with temp, don't need 12v of swing on the gate either (maybe 5v or 6v max on the gate even on an extreme output design) so I was thinking about using the 3 terminal regulator current source configuration.
I have always been a fan of Elvee's Chip regulator amp, but never made one yet, I bought a bunch of LM337 to do so a while back and they were all fake.........
So I never made that one yet. :/
Good to see that it may work well and will give it a try. 🙂
Cheers !!

jer 🙂
Latest version. I decided not to add the heatsink since I was planning to attach the mosfet to a much larger heatsink.
The board size is now 100mm x 100mm so there is a lot more room. Suggestions to take advantage of the extra room are welcome. Again, I see I haven't improved since my long ago programming days when mistakes were numerous, so please point them out.
I mostly build tube amps and would use carbon composition (non-inductive) resisters for grid stoppers. Is that the case in transistor designs?
ray
The board size is now 100mm x 100mm so there is a lot more room. Suggestions to take advantage of the extra room are welcome. Again, I see I haven't improved since my long ago programming days when mistakes were numerous, so please point them out.
I mostly build tube amps and would use carbon composition (non-inductive) resisters for grid stoppers. Is that the case in transistor designs?
ray
- Home
- Amplifiers
- Pass Labs
- Hybrid ZEN Amplifier + LM317 = Efficient and simple