Time for a small update of what I did.
Because I wanted to hear how different front ends will sound I adopted a modular approach. The output stage and it`s control circuitry is on the main pcb and the front ends on slave pcbs that connect between them with pin strips.
To change the modules you will need to turn off the amp.
The schematic for the OS pcb:
Here is the schematic for the P core. This is a small jfet buffer that has no voltage amplification. When using this front end you will need a preamp. This is used by ZM in his Warbler amp.
Some thd measurements using the LU1014 as output device
Here is the schematic for the C core. This is a 2 stage current feedback front end with voltage gain so no preamp needed. I found that in some situations C5 was needed so I added it. There is also a cascode feedback connection (for mixed global feedback) but didn`t test this yet. This is used by same ZM in his Scryer.
The thd results that I shared for the initial Gaincore that had the C core on the main pcb are similar to what I get with the modular stuff so won`t share them again.
Now the schematic for the V core. This core is a voltage feedback front end that has voltage gain and also xlr inputs. This is the update on the Stasis fe that Pa did.
I took some thd measurement for this core which is close in performance to the C core but unfortunately lost them. Will take them again at a later time.
Because I wanted to hear how different front ends will sound I adopted a modular approach. The output stage and it`s control circuitry is on the main pcb and the front ends on slave pcbs that connect between them with pin strips.
To change the modules you will need to turn off the amp.
The schematic for the OS pcb:
Here is the schematic for the P core. This is a small jfet buffer that has no voltage amplification. When using this front end you will need a preamp. This is used by ZM in his Warbler amp.
Some thd measurements using the LU1014 as output device
Here is the schematic for the C core. This is a 2 stage current feedback front end with voltage gain so no preamp needed. I found that in some situations C5 was needed so I added it. There is also a cascode feedback connection (for mixed global feedback) but didn`t test this yet. This is used by same ZM in his Scryer.
The thd results that I shared for the initial Gaincore that had the C core on the main pcb are similar to what I get with the modular stuff so won`t share them again.
Now the schematic for the V core. This core is a voltage feedback front end that has voltage gain and also xlr inputs. This is the update on the Stasis fe that Pa did.
I took some thd measurement for this core which is close in performance to the C core but unfortunately lost them. Will take them again at a later time.
Now there is on more thing.
The last week I bought two meanwell sp-320-24 power supplies. One for each channel. Intially I was using notebook power supplies.
When I first connected the new power supplies there was an unpleasant buzz coming from my speakers which wasn`t there before.
It looked like this on the fft analyzer
Having the A2Cmx filter connected between the power supply and the amp, it was less probable that the buzz was coming from the power supply so I supposed it is a gnd loop. After trying all possible combinations of connecting the gnd and earth and still having the buzz there last thing which I could think of was that I could have a gnd loop on the pcb.
I was using a twisted pair of unshielded wires to carry the signal from the preamp to the amp.
As a last attempt to cure the problem I said to test with a lousy pair of rca shielded cables (that i had around) because the amp was quiet with the input shorted to gnd.
Now the story changed a lot. What is displayed at 50hz is coupled from the wiring and I can`t hear a thing in my 96db eff speakers. So the lousy shielded rca cables did their job.
So at this point one thing I can think of is that the notebook power supplies are screened better than the meanwell and if you don`t need the higher voltage and power from a meanwell it`s probably better to stay with notebook ps. Mine are 90w/20v lenovo and got them used for very cheap. These are good up to 2.4a bias.
The second thing that I am thinking of is that I need some coaxial wires to build a pair of rca cables 🙂
The last week I bought two meanwell sp-320-24 power supplies. One for each channel. Intially I was using notebook power supplies.
When I first connected the new power supplies there was an unpleasant buzz coming from my speakers which wasn`t there before.
It looked like this on the fft analyzer
Having the A2Cmx filter connected between the power supply and the amp, it was less probable that the buzz was coming from the power supply so I supposed it is a gnd loop. After trying all possible combinations of connecting the gnd and earth and still having the buzz there last thing which I could think of was that I could have a gnd loop on the pcb.
I was using a twisted pair of unshielded wires to carry the signal from the preamp to the amp.
As a last attempt to cure the problem I said to test with a lousy pair of rca shielded cables (that i had around) because the amp was quiet with the input shorted to gnd.
Now the story changed a lot. What is displayed at 50hz is coupled from the wiring and I can`t hear a thing in my 96db eff speakers. So the lousy shielded rca cables did their job.
So at this point one thing I can think of is that the notebook power supplies are screened better than the meanwell and if you don`t need the higher voltage and power from a meanwell it`s probably better to stay with notebook ps. Mine are 90w/20v lenovo and got them used for very cheap. These are good up to 2.4a bias.
The second thing that I am thinking of is that I need some coaxial wires to build a pair of rca cables 🙂
Last edited:
I always wanted to try the lu1014 at a higher bias but was limited by the to247 cascode. Well not anymore..
I plan to run these up to 3.5a, above this my 4a choke starts to saturate.
Will give them a try the next days.
For today I put together a pair of FC cores.
FC from folded cascode.
Here is the schematic.
With the other gain stages(except P core) that I have the amp is a 3 stage amp. With the FC core it becomes a 2 stage amp, yes something new to explore.
The FC core inverts the phase so the speaker terminals need to be swapped.
All good till here but now it became even better.
The last days I found Pathos Acoustics
because their heatsinks caught my attention. They look nice.
At a closer look not only the heatsinks look nice but also the INPOL amplifier concept.
Here is a part of the description from their site: https://pathosacoustics.it/en/about-us/
From the above description seems the os uses same inductive coupling as Gaincore.
At this point I think that for now I will try to change some of the cores(that I have) with voltage gain and global feedback to local feedback.
A core with valves would be interesting too 🙂
I plan to run these up to 3.5a, above this my 4a choke starts to saturate.
Will give them a try the next days.
For today I put together a pair of FC cores.
FC from folded cascode.
Here is the schematic.
With the other gain stages(except P core) that I have the amp is a 3 stage amp. With the FC core it becomes a 2 stage amp, yes something new to explore.
The FC core inverts the phase so the speaker terminals need to be swapped.
All good till here but now it became even better.
The last days I found Pathos Acoustics
because their heatsinks caught my attention. They look nice.
At a closer look not only the heatsinks look nice but also the INPOL amplifier concept.
Here is a part of the description from their site: https://pathosacoustics.it/en/about-us/
From the above description seems the os uses same inductive coupling as Gaincore.
At this point I think that for now I will try to change some of the cores(that I have) with voltage gain and global feedback to local feedback.
A core with valves would be interesting too 🙂
First I started with the 2SK182es SIT and the FC core.
Some measurements at 4r
With a bit more bias the thd at 4r drops considerably.
8r measurements
Here a bit more bias helps too even if the measurement shows otherwise.
I managed to take take a screenshot with the shape of the h2 at 1w/8r.
Then things got more serious 🙂
Here a few measurements.
First at 4r
And here some measurements for 8r
Here a bit more bias helps too with the thd at 30W and can get you 35W
One more thing that I noticed all this time with the GC.
For the small 3.4r bookshelf speakers(that I have and I use to evaluate this amp) I like more the sound with the cascoded lu1014 in the os, on the 6r Tannoys the SIT is king.
Some measurements at 4r
With a bit more bias the thd at 4r drops considerably.
8r measurements
Here a bit more bias helps too even if the measurement shows otherwise.
I managed to take take a screenshot with the shape of the h2 at 1w/8r.
Then things got more serious 🙂
Here a few measurements.
First at 4r
And here some measurements for 8r
Here a bit more bias helps too with the thd at 30W and can get you 35W
One more thing that I noticed all this time with the GC.
For the small 3.4r bookshelf speakers(that I have and I use to evaluate this amp) I like more the sound with the cascoded lu1014 in the os, on the 6r Tannoys the SIT is king.
Nice Russian nest doll you got going on there, Lucian! Why do you like the Lu on the 3.4r better?
When pushed harder the lu1014 has a bit more control, gives more output power but sounds a bit drier.
The sit on the other side is a bit more detailed with fatter lows but when you push it, it falls quickly.
This with the 3.4r small woofie bookshelfs.
With the bigger(impedance, spl, size and everything) Tannoys the SIT wins easily.
Should have changed the speakers long ago but good things in life take time to happen.
If you have small speakers usually you will need big amplifiers, with big speakers is the contrary.
Now one thing I can tell for sure, don’t dream(like I did) that small speakers with big amps will sound the same as big speakers with small amps because they are nowhere near.
The sit on the other side is a bit more detailed with fatter lows but when you push it, it falls quickly.
This with the 3.4r small woofie bookshelfs.
With the bigger(impedance, spl, size and everything) Tannoys the SIT wins easily.
Should have changed the speakers long ago but good things in life take time to happen.
If you have small speakers usually you will need big amplifiers, with big speakers is the contrary.
Now one thing I can tell for sure, don’t dream(like I did) that small speakers with big amps will sound the same as big speakers with small amps because they are nowhere near.
Pathos
fancy words for good ole choke loaded follower
more efforts invested in pamphlets and heatsinks than in circuitry
besides, when I see ECC83 used in anything else than phono stage or radio stage, my curiosity is suddenly vanished
fancy words for good ole choke loaded follower
more efforts invested in pamphlets and heatsinks than in circuitry
besides, when I see ECC83 used in anything else than phono stage or radio stage, my curiosity is suddenly vanished
As I have been following this thread, I'd share my (limited) experience:
I was playing with inductor loaded amps mainly because I got some inductors for cheap, and had some 19V switchers lying around. So, why not?
What I tried:
1. Mofo
2. Zenductor
3. Zen V8 -inductor loaded
Mofo worked well, but in the long run, I found the sound to be "too polite". I don't know how to better describe it.
Zenductor. Interesting effect. Like Zen V1 on some steroids. On some lower impedance loudspeakers, I liked the sound. That is, until you want to crank it up beyond its capabilities... (Maybe gotta build something with a higher PSU volrage, more current- A Turbo Zenductor) 😀
Zen V8 inductor. Nice. Pretty much like "normal" Zen V8 or Zen V9. Plays well with most of my loudspeakers. That is what I am using now.
In the past, an F2 (yes, the one with the "worst" numbers) loaded with a 1500w 220v lightbulb sounded very nice with Visaton B200 on Open Baffle.
So...
I think that, despite all the shortcomings, there is some magic that pulls me towards the simplest topology amp + a fullrange speaker.
So I imagine that I'd like build something with lots of lightbulbs, higher power supply, lots of capacitance in the PSU and a single SiT...
And that would drive a Sonido 12" (not there yet, but maybe) fullrange in some sort of ML TL...
I was playing with inductor loaded amps mainly because I got some inductors for cheap, and had some 19V switchers lying around. So, why not?
What I tried:
1. Mofo
2. Zenductor
3. Zen V8 -inductor loaded
Mofo worked well, but in the long run, I found the sound to be "too polite". I don't know how to better describe it.
Zenductor. Interesting effect. Like Zen V1 on some steroids. On some lower impedance loudspeakers, I liked the sound. That is, until you want to crank it up beyond its capabilities... (Maybe gotta build something with a higher PSU volrage, more current- A Turbo Zenductor) 😀
Zen V8 inductor. Nice. Pretty much like "normal" Zen V8 or Zen V9. Plays well with most of my loudspeakers. That is what I am using now.
In the past, an F2 (yes, the one with the "worst" numbers) loaded with a 1500w 220v lightbulb sounded very nice with Visaton B200 on Open Baffle.
So...
I think that, despite all the shortcomings, there is some magic that pulls me towards the simplest topology amp + a fullrange speaker.
So I imagine that I'd like build something with lots of lightbulbs, higher power supply, lots of capacitance in the PSU and a single SiT...
And that would drive a Sonido 12" (not there yet, but maybe) fullrange in some sort of ML TL...
Today I was in a small road trip so I started early in the morning, I made new friends too 🙂)
Back to Pathos..
Well I was happy to see that someone uses inductive coupling in their amp.
Then researching more I discovered that they have a bridged os and that is not for me so I kept only the idea of no global feedback which I intend to apply to GC. Now all the front ends that I have(except one) have global feedback.
Then I was thinking to try a valve front end to make an idea about the sound but because valves wear over time not sure I will proceed. I like smd jfets more.
Playing with amps with and without global feedback I discovered that I like a bit more the thd from no global feedback.
If you have appropriate speakers then you can explore better choke loading, with better I mean at high spl levels.
Though initially I started to play with this stuff on 3.4r speakers and 87db(7” woofie 🙂) ) and I liked what I heard though at high output power you can hear a bit the amp screaming and for that I needed to use pucks and a higher bias.
So probably a mofo with only 2a bias is not enough for 3.4r so going to 3.5a like I do on gc makes a difference on low spl/low impedance but not miracles. If you want a miracle you need to kindly ask ZM to help you with a pair of proper speakers 🙂
With the Tanns I run only 1.8a bias. What a difference man..
Then Vix mentions about trafos in the front end..
My preamp hosts a Iron Pre(balanced) from which if I feed the output signal into the os of gc. I am getting 35w/8r like this and with the Tanns it doesn’t take long to have a neighbor at your door.
Like Vix says it sounds really nice. Was planning a front end with a trafo like Pa has in the Zenductor but other things got in the middle in the meantime and it got a bit delayed.
Only problem with this project are the funds but I will work my way there somehow.
I am curious about another thing..
If I get same inductor(inductance and dcr)
with 0.5mm laminations EI core and another one with 0.25mm laminations C core will there be a difference in performance? The price is close to 5x more.
Back to Pathos..
Well I was happy to see that someone uses inductive coupling in their amp.
Then researching more I discovered that they have a bridged os and that is not for me so I kept only the idea of no global feedback which I intend to apply to GC. Now all the front ends that I have(except one) have global feedback.
Then I was thinking to try a valve front end to make an idea about the sound but because valves wear over time not sure I will proceed. I like smd jfets more.
Playing with amps with and without global feedback I discovered that I like a bit more the thd from no global feedback.
If you have appropriate speakers then you can explore better choke loading, with better I mean at high spl levels.
Though initially I started to play with this stuff on 3.4r speakers and 87db(7” woofie 🙂) ) and I liked what I heard though at high output power you can hear a bit the amp screaming and for that I needed to use pucks and a higher bias.
So probably a mofo with only 2a bias is not enough for 3.4r so going to 3.5a like I do on gc makes a difference on low spl/low impedance but not miracles. If you want a miracle you need to kindly ask ZM to help you with a pair of proper speakers 🙂
With the Tanns I run only 1.8a bias. What a difference man..
Then Vix mentions about trafos in the front end..
My preamp hosts a Iron Pre(balanced) from which if I feed the output signal into the os of gc. I am getting 35w/8r like this and with the Tanns it doesn’t take long to have a neighbor at your door.
Like Vix says it sounds really nice. Was planning a front end with a trafo like Pa has in the Zenductor but other things got in the middle in the meantime and it got a bit delayed.
I am bugging ZM with same stuff since some time. 12” field coil woofie, a nice 8” field coil mid and a ribbon tweeter.
Only problem with this project are the funds but I will work my way there somehow.
I am curious about another thing..
If I get same inductor(inductance and dcr)
with 0.5mm laminations EI core and another one with 0.25mm laminations C core will there be a difference in performance? The price is close to 5x more.
can't say for all models, but I can for at least 3 of them - it's not coupling ...... it's inductive loading
coupling is capacitive
and, result of lack of interest ....... details I have are all for SE amps, not bridged
general reply - if metallurgy (material) is same for both lamm thickness, thinner is better
I ment that but wrote something else.
I said same thing when I seen that each half of amp in the Boo os seen 1/2 of my 3.4r elacs. Though it was easy to play with the phase of h2.
Would have been nice if same thing applied also for the wallet 🙂)
I had an impression that you meant Pathos, when speaking of bridged amps; again, I know only of SE Pathos amps
more thinner banknotes are probably better than less thicker ones
Yes, Pathos but mentioned about Boo because there I played with bridged os.
Initially I researched only about the flagship and that one is bridged. Searching again now I found this article where they present both se and bridged.
https://www.monoandstereo.com/pathos-inpol-legacy/
Earlier I ment that was not interested about the bridged version because I assumed that if the flagship is bridged then all of them are but seems not. They have also a inductive loaded cap coupled se version.
So in one word its better with big speakers and low power amps.