The Holy Grail Follower Output Stage

Hello Ihquam,
hello ZD25MINI-builders,
my ZD25MINI-project made some progress. I am building two monoblocks. Heavy aluminum frames
and big heatsinks.
The boards are stuffed. I am only waiting for another OPAmp (OPA445 instead of OPA551), because of my higher railvoltages (+-32 V DC unloaded). The transformers are 250W-types for each channel 230 V prim. /
2x 22V AC sec.. I wanted the higher rail voltages. Although Ihquam said - not higher than +-20 V AC on the trafos - or no more than +-30 V DC at the rails. Hmmm.
And I want to experiment with external frontends like the FE2022 and others. This is easily done how Ihquam
has designed the pcb. I like it.
So, work in progress - still a lot to do.
Cheers
Dirk 😉
 

Attachments

  • ZD25MINI_PCB_stuffing parts.jpg
    ZD25MINI_PCB_stuffing parts.jpg
    154.6 KB · Views: 216
  • ZD25MINI_Aluminum frame_transformer_PSU_top.jpg
    ZD25MINI_Aluminum frame_transformer_PSU_top.jpg
    105.3 KB · Views: 201
  • ZD25MINI_PCB_on_heatsink.jpg
    ZD25MINI_PCB_on_heatsink.jpg
    90.5 KB · Views: 196
  • ZD25MINI_Aluminum frame_transformer.jpg
    ZD25MINI_Aluminum frame_transformer.jpg
    152.1 KB · Views: 204
Hello out there,

today I got my ZD25_mini-monoblocks running.

I had some small trouble to get the offset into the ballpark. But the reason are my MosFets. I have matched a few ones
of the ONSEMI / Fairchild FQA28N15 and FQA36P15. I've got many real good N-matches and P-matches. But the Vgs
from the N- to the P-channel was wider. I had to make a little tweak to R26 (paralleling a 221kOhm resistor) and then
everything became super easy.
I have biased the blocks to 1.7A and offset adjustment became an easy task. Temperature at the outside (between fins, where
Mosfets are mounted inside) is at 40° C. So room for higher bias.
Listening to some music... :note::sax::eguitar:
I hope the floor will not break through! - there is a heavy sixpack of monoblocks... :nod:
Cheers
Dirk
 

Attachments

  • the_bad_sixpack_monoblocks.jpg
    the_bad_sixpack_monoblocks.jpg
    209.6 KB · Views: 178
  • ZD25MINI_sixpack-monoblocks2.jpg
    ZD25MINI_sixpack-monoblocks2.jpg
    243.4 KB · Views: 179
  • ZD25MINI_bias_1.7A._offset below1mV.jpg
    ZD25MINI_bias_1.7A._offset below1mV.jpg
    198 KB · Views: 157
  • ZD25MINI_bias_1.7A.jpg
    ZD25MINI_bias_1.7A.jpg
    213.2 KB · Views: 177
Good morning,

after some first listening to the ZD25-mini, I want to say:
'Thank you, Ihquam!'
You did some excellent work! 😉
This amp has some very good qualities. My first impression after listening to some well recorded music is:
it lets you hear very small details and the bass is very controlled and in no way 'fat'
I want to listen to it some days. Later I will try some external frontstages. And my casework has to be finished.
Greets
Dirk
 
If people want to experiment with an Amp-Camp-Amp type follower that runs off a 24v switching mono-rail supply, I have simulated this (see attached), and it seems to work fine, at least in simulation. It may not be a Holy Grail follower but more of a Monty Python—The Quest for the Holy Grail follower. It works without any trim pots and has a 10K input impedance. R127 sets the bias current, ~500mA at 2.1ohms and about 1A at 1.1ohms. My only concern is the optocoupler is running very lightly, with less than 1mA in the LED so I don't know if that will be consistent in real life, but optocouplers have a short life if you run them at 100% duty cycle over 5mA as the diode admittance drops. It has high capacitance but a headphone driver should be able to drive it fine. C25 stops the 1/2V from coming up too quickly to make the speaker pop just like the ACA mini. Protection diodes are included to prevent the MOSFETs from damage, such as driving the amp while off or ESD. ESP audio suggests putting a regular diode in series to reduce leakage current, this works well the simulation shows it. Please remember this schematic is not tested in real life! I have also not tested the power-up and down in simulation. A switch in series with R118 could be used to turn off the bias like the ACA mini.
 

Attachments

  • ACA_follower.png
    ACA_follower.png
    17.5 KB · Views: 231
  • Like
Reactions: triode_al
I have ZD25-mini mated to FE2022 with mixed feedback on the bench for a few days. Output FETs are Vishay IRFP240 / Harris IRFP9141. Initial results are promising with H2 at around -90 dB and H3 and higher at -100 dB and lower (at 2.83 Vrms into 8 Ohms). At some point I'll get around to investigating other output devices.
 
  • Like
Reactions: birdbox
L.H.Q.,

Your opto-coupler biased, no degeneration follower is brilliant. I own the the XA-25 and it’s the best amp I’ve ever heard, with a large period at the end. Of course it has the UGS front end, so it’s not just a follower.

Years ago I built the F4 and it remains one of my all time favorites. With the IXYS 600v HiperFET this could allow everyone on the forum the chance to have an XA-25 follower. Just brilliant.
This is interesting, HiperFet to be used. in F4.
however, how can the complementary be configured, as the HiperFET to my limited knowledge, has no P-device needed for the a symmetrical pair of source-coupled follower devices.?
Post #812 implies that the IXTQ36P15P P-fet might be used nevertheless. (as a combo ; I have several IXFX34N80 wanting some use beyond F6 I thought of first . . .)
 
Last edited:
@ triode_al:
but I didn't use them in the Holy Grail Follower Output Stage.
I used the complemetary IXYS Mosfets with the bias-circuit similar to Zen Mods XA252 bias -schematic and the driver stage
is a cascoded BA-3 frontend.
The IXYS - Mosfets ('hockeypucks') have often pretty high CISS.
Greets
Dirk
 
Thanks, Dirk. The square pucks are quite hefty.
What I'll do is buy some To-247 equivalents so they fit in my board (too bad the Ciss does not disappear in another form factor); I like the IXTQ36P15P p-fet , while I will look around for the Harris too.
 
I took some time today to put a schematic together for the front end I'm showing there at post #682. Sorry for the slight crude-ness, I'm much more of a soldering citizen than a CAD jockey.. I'm pretty sure it's correct, except I didn't show the LED indicator (top secret..). As I mentioned earlier, I am powering these front ends with a separate outboard power supply comprised of a pair of Balanced Zen Line Stage PS boards (+/-45V). Power transformer(s) are Antek AN-0222's.

Sound is quite good but we'll put some more miles on this stuff..
I like the Hiraga/LeMonstre input stage. You know of course the 2nd H can be tailored. (just reacting, interested).
Last week I made a front-end; I definitely preferred the BJT/Skilai output by the way.
 
@ triode_al:
but I didn't use them in the Holy Grail Follower Output Stage.
I used the complemetary IXYS Mosfets with the bias-circuit similar to Zen Mods XA252 bias -schematic and the driver stage
is a cascoded BA-3 frontend.
The IXYS - Mosfets ('hockeypucks') have often pretty high CISS.
Greets
Dirk
Sounds like a very nice amp. My BA3 is one of my favorites. Listening to Zenmod's PoP at moment. XA252 to be repaired, darn it.

I started this thread earlier, better go back and read over and all the new stuff. I am really taken by the square law amp OS, so here we go again...

Russell