Power supply should be 60-70V. I prefer 70V. Total current draw per channel would 20-25 mA, let’s say 30 mA to be safe.
I posted a power supply PCB somewhere that works well.
Congrats! Of course, would be happy to help! Though you really want George, he’s the master. It sounds very interesting. One challenge is to make the output impedance low enough to properly drive 8 ohms. In this case the output impedance would likely be dominated by the choke impedance. What values are you using for the feedback resistors?
Probably not anymore. I built a perf board based FETSET amp back in early 2023 on a giant heat sink. Posts #787, 872, 873, 877 and others. I used that amp at about 25 WPC off and on until it died late last year. I have planned to rebuild it with a pair of CPU coolers for heat sinks and maybe lay out a real PCB for it, but that has not happened yet. All of my work with SE mosfet amps used high voltage parts running on about 300 volta of B+ into a 600 ohm OPT which was intended for vacuum tubes. All used the IXTH6N100D2 DEPLETION MODE mosfets. In some early experiments I blew up a couple of these mosfets from high frequency oscillation. I have zero DIY experience with any mosfet design that directly drives a loudspeaker. I still use my nearly 40 year old Yamaha NS10M Studio monitors as reference speakers and don't want to do anything that could kill them as they are not replaceable.
I did look up the IXTH20N50D part and found mention that it is discontinued or only available in large quantity buys. I don't know how true this is, but check before you choose this part, Digikey recommends the IXTH16N50D2 as a substitute. I have never seen either part.
No question that board does a fine job and its usefulness is not limited to Rahul's line stage. Not sure how one would be able to adjust gain better than simply finding the zener string that gives you what you want. A little insert and solder never hurt anyone!
Rahul, a question for you, please.
I need to add gain - something I had mentioned to you before. I have been using the FE2022 before the SCG from my MEH woofers. I do not hear any ill effects but I thought I should ask if feeding them the full X10 gain of the FE2022 OK? I had them set for X5 but need more on one channel.
Would SCG care what signal level it receives?
Hope all is well for you.
Take care,
Rahul, a question for you, please.
I need to add gain - something I had mentioned to you before. I have been using the FE2022 before the SCG from my MEH woofers. I do not hear any ill effects but I thought I should ask if feeding them the full X10 gain of the FE2022 OK? I had them set for X5 but need more on one channel.
Would SCG care what signal level it receives?
Hope all is well for you.
Take care,
So nice to hear from you, Rick! You pose an interesting question and I have thought about this question a lot. The SCG topology naturally allows a wide input swing. With a gain of 10X, the input voltage is multiplied by 10. If we know the max output swing, we can calculate the max input swing before clipping.
Let’s say the supply is 70V and Vd is set to 40V. It could then theoretically swing 30V up and 30V down before clipping, i.e., 60Vpp. Divide that by 10 and you get 6Vpp at the input before the output clips. If the gain is lowered, then the input can take more voltage before the output clips.
You can totally stitch two preamps together. I did that for a while when I was using DSP to do room correction and was losing gain in the computer. Use the volume pot before the SCG, that’s the right place for it. Incidentally, I too used the FE2022 together with the SCG. Since then, I am using a lighter touch on the room correction and don’t need the extra gain anymore. My goal is to be DSP-free before BAF 2025
Let’s say the supply is 70V and Vd is set to 40V. It could then theoretically swing 30V up and 30V down before clipping, i.e., 60Vpp. Divide that by 10 and you get 6Vpp at the input before the output clips. If the gain is lowered, then the input can take more voltage before the output clips.
You can totally stitch two preamps together. I did that for a while when I was using DSP to do room correction and was losing gain in the computer. Use the volume pot before the SCG, that’s the right place for it. Incidentally, I too used the FE2022 together with the SCG. Since then, I am using a lighter touch on the room correction and don’t need the extra gain anymore. My goal is to be DSP-free before BAF 2025
Thanks. I have used IIR DSP for years. I bet you will never get what you received from DSP any other way. I am a purist at heart but have had to accept the reality of rooms versus loudspeakers. The room always wins. But we are given a chance to subvert the room with good EQ.
Yes, that is the same reason I am needing the gain.
Working with ACOURATE FIR and since it cannot fix dips everything gets lowered in level which is better, I guess, but it does require additional gain. So it is not like the voltage level will be appreciably higher than before - just starting at a lower level.
I am only in the learning stage but I can tell you I am hearing infinitesimal details never heard before and I have much learning ahead of me.
I gave been using FE2022 and SCG for the woofer channel of the MEHs but am starting to think i might need to add a half gain FE2022 ahead of the tweeter channel.
If anyone has the inclination Mr. Bruggemann is a rare fellow. Very patient and generous with his skills. One realizes quickly that the cost of software is minimal when you consider how much time he will spend to help you get things right. There is a more automated approach to this but I prefer getting a full understanding of what is being done to the signal. I am well beyond what I heard with the xilica Solara. A far more immersive sound.
THANKS and take care,
Yes, that is the same reason I am needing the gain.
Working with ACOURATE FIR and since it cannot fix dips everything gets lowered in level which is better, I guess, but it does require additional gain. So it is not like the voltage level will be appreciably higher than before - just starting at a lower level.
I am only in the learning stage but I can tell you I am hearing infinitesimal details never heard before and I have much learning ahead of me.
I gave been using FE2022 and SCG for the woofer channel of the MEHs but am starting to think i might need to add a half gain FE2022 ahead of the tweeter channel.
If anyone has the inclination Mr. Bruggemann is a rare fellow. Very patient and generous with his skills. One realizes quickly that the cost of software is minimal when you consider how much time he will spend to help you get things right. There is a more automated approach to this but I prefer getting a full understanding of what is being done to the signal. I am well beyond what I heard with the xilica Solara. A far more immersive sound.
THANKS and take care,
I've also been using Acourate filters (for at least 15 years). I understand wanting to be DSP free, but I've given up and now do the exact opposite - run my turntable into a Salas USFP which then runs into an ADI-2 Pro FS-R for ADC for applying the filters then back out to DAC. Heresy I know, but it sounds great.
DSP is a huge boon for crossover design and to improve speaker performance. No doubt about it. And Acourate FIR is a proper professional tool. I have used the trial version and it does its job very well.
I have been using DSP since about 2009. Was an early adopter of miniDSP. At that time, it didn’t have FIR filters. So I moved to doing the processing in the computer. Much more flexible. Being the cheapskate that I am, I used the free (but powerful) DRC engine. Also rePhase and REW by member pos here. Excellent tools.
Like @aljordan I ran the turntable into an RME ADI-Pro FS2 and then into the computer, then do processing, and back out to the ADI. I have since moved to a TDA1541 DAC. And am now on a passive crossover with light DSP. I really hate running the turntable into an ADC. It sounds awesome but could be better.
I have been using DSP since about 2009. Was an early adopter of miniDSP. At that time, it didn’t have FIR filters. So I moved to doing the processing in the computer. Much more flexible. Being the cheapskate that I am, I used the free (but powerful) DRC engine. Also rePhase and REW by member pos here. Excellent tools.
Like @aljordan I ran the turntable into an RME ADI-Pro FS2 and then into the computer, then do processing, and back out to the ADI. I have since moved to a TDA1541 DAC. And am now on a passive crossover with light DSP. I really hate running the turntable into an ADC. It sounds awesome but could be better.
That's how I started with room correction as well. It took me so long to learn how to configure the software that I decided to write my own software to help configure it for me. That was a long time ago though.
OMG! I can’t believe that you wrote that! DRC Designer is so helpful for using DRC. Amazing! And thank you for making it!
Using MEH speakers I am forced to use DSP. I run my phono through an RME ADC, also. I do not consider it heresy!
I find ACOURATE to be a boon for those of us using idler drive turntables - the sharp LF filter is the best solution I have heard.
Compromise is the one thing we can be sure of. I am certain whatever is lost with good DSP is overwhelmed by the good things it can do.
By the way, Rahul - what would you think of coaxing 24 db of gain from the FE2022? Yes, I am using the FE2022 to drive the SCG.
Thanks and take care,
I find ACOURATE to be a boon for those of us using idler drive turntables - the sharp LF filter is the best solution I have heard.
Compromise is the one thing we can be sure of. I am certain whatever is lost with good DSP is overwhelmed by the good things it can do.
By the way, Rahul - what would you think of coaxing 24 db of gain from the FE2022? Yes, I am using the FE2022 to drive the SCG.
Thanks and take care,
This gain stage sounds fantastic Rahul. I've had just a couple of long listening sessions so far but the first impression is very promising. Sweet, rich, a bit warm... but none of these things in excess or presented in any sort of distracting way. The music is layered, complex, natural, organic, and all the detail and clarity is still there. I can see why this thread has stayed near the front page of this forum for a very long time. This should also be really good in front of the Burning Amp output stages or an F4. Lynn's HGF certainly seems like a great fit..
Yes, it is single ended. Using 55 volts of B+. I asked the question at the FE2022 and received an answer from Ben Mah.
He was concerned about output voltage but (as you know) I am trying to get back total gain so the stage would not be working any harder than it did before FIR filters.
It is worth a try.
THANKS for your help (as always), Rahul
He was concerned about output voltage but (as you know) I am trying to get back total gain so the stage would not be working any harder than it did before FIR filters.
It is worth a try.
THANKS for your help (as always), Rahul
Thanks for the kind words! It makes the effort worthwhile! The original development was with an F4, so you are spot on. It must be incredible with the HGF.
Member "birdbox" has the essentials (PCB's and FET's) available if someone wants to give the HGF a try. Dan is one of the good guys for sure.
@Rahul - Did you mention that you had something in the works with a follower OS using laterals?
@Rahul - Did you mention that you had something in the works with a follower OS using laterals?
Thanks @william2001! I am working on a speaker crossover and helping a friend with some projects. HGF is in the future some time for sure.
Yes, I am currently running a lateral mosfet follower, similar to the MoFo. I'm calling it Elmo, as in L-Mo 🙂. It sounds pretty fantastic. Low distortion at one Watt and clears about 14 Vrms at about 1% THD at no load (2nd harmonic dominant). The problem is it needs a beefy driver. Right now, I'm using a beefed up follower on the SCG, but my goal is to share it with others so I am trying to make it a standalone design. A driver needs to swing symmetrically and the challenge is to do it without a bipolar supply to avoid extra complexity. My solution right now is to put a choke load on the buffer. driver follower. But haven't tested it yet.
If you want to try the sound though, it is the Exicon 10N20 in place of the mosfet in the MoFo. Bias at 1.5A and supply anywhere from 20-30V. Needs good heatsinking. If you have the HGF, it will probably sound too sweet in comparison. But if you mostly operate at 1-10W levels, it can sound amazing. It has this openness and organic or live quality that is very alluring.
Yes, I am currently running a lateral mosfet follower, similar to the MoFo. I'm calling it Elmo, as in L-Mo 🙂. It sounds pretty fantastic. Low distortion at one Watt and clears about 14 Vrms at about 1% THD at no load (2nd harmonic dominant). The problem is it needs a beefy driver. Right now, I'm using a beefed up follower on the SCG, but my goal is to share it with others so I am trying to make it a standalone design. A driver needs to swing symmetrically and the challenge is to do it without a bipolar supply to avoid extra complexity. My solution right now is to put a choke load on the buffer. driver follower. But haven't tested it yet.
If you want to try the sound though, it is the Exicon 10N20 in place of the mosfet in the MoFo. Bias at 1.5A and supply anywhere from 20-30V. Needs good heatsinking. If you have the HGF, it will probably sound too sweet in comparison. But if you mostly operate at 1-10W levels, it can sound amazing. It has this openness and organic or live quality that is very alluring.