I saw one DIY'er , he set the EI gap to .2mm and measured the inductance of the trafo. For the least loss LC must be tuned to the
SMPS frequency. Or if you have a prebuilt trafo , use the controller to tune the freq. to the LC. These are "tuned" circuits. Sort of
like the crystal oscillators on the Arduino's. Oh , thank you for those nano/uno's ... I'm addicted to these little CPU's now. I got a ESP32 , as well.
OS
It's not the frequency , it's the lack of "features" (current sensing , wave shaping, and more). Safety , efficiency , reducing losses from
all the passives . That's what's sets (primitive/advanced) apart.
PS- "Advanced" will change freq. as load/ inductance changes (even at saturation).
Last edited:
ESP32 is really cool! If I ever get some time that will be the front end for my preamp remote control.
Due to the high frequencies these run at the transformers don't actually need very many windings. They look easy to wind, but I've never actually done one yet.Jeff,never i haven't done this ..
Let's see what Pete can do and i will think again.
Do you know if there is already an implementation of a balanced input precision op-amp buffer IPS? Otherwise it would really be great to enable the Wolverine to accept balanced input
Stuart is a person of the highest class of quality and exceedingly generous, based on personal experience, as well as following the entire two threads. I have no doubt Harry, Daniel, and the rest are first class people as well. I feel privileged to have access to this labor of love. I’ve met a few such designers - Darryl Wilson does not rush, and is extraordinarily meticulous and generous. There are a handful of others. There are plenty of professionals who think too highly of themselves to produce anything of value.
Is there any interest in another design ? I will use Stuart's IPS standard but ALSO design a single pair power device stereo OPS board.
The advantages of the hellraiser.
-PSRR is quite superior. -120db easy. Single supply will match a monoblock wolverine.
-NO caps (just input cap) in audio (or DC) FB . servo can make for uV offset - LED CCS's are controlled by servo.
-300V/us slew - I don't know whether this would be audible , but - what the hell ... FAST !!
-A small 300W SMPS would make for a quite small and accurate 60w/ch amp with plenty of headroom (100w+ peak).
-this design seems to have the same 8-10ppm regardless of freq./ load - could this be the current FB ?
-not as many silver mica caps.
disadvantages -
-warm 1W feedback resistors , you will waste a whole watt .. ha ha !
Seems a quite ideal endeavor as far as a design. Even if there is no interest - I shall have it.
below I have a FINAL schema to layout what ( I ) want. Off to the Sprint layout.
PS - this is a "branch" of the "Wolverine build thread" , I will make any design adaptable to the "wolverine EF3" , but
have no great interest in that thread. I want to at least match that design's performance with different topologies , but also want to
include our poorer hobbyist's in this project.
The "Wolverine EF3" is actually a modified HK680 (Slewmaster) OPS. I really don't have any rights to it ,but neither does anybody else.
I'm glad it was tested to the very limits of testing. COOL ! So , I will use it to to it's limit - I still can make it more robust ...
OS
The advantages of the hellraiser.
-PSRR is quite superior. -120db easy. Single supply will match a monoblock wolverine.
-NO caps (just input cap) in audio (or DC) FB . servo can make for uV offset - LED CCS's are controlled by servo.
-300V/us slew - I don't know whether this would be audible , but - what the hell ... FAST !!
-A small 300W SMPS would make for a quite small and accurate 60w/ch amp with plenty of headroom (100w+ peak).
-this design seems to have the same 8-10ppm regardless of freq./ load - could this be the current FB ?
-not as many silver mica caps.
disadvantages -
-warm 1W feedback resistors , you will waste a whole watt .. ha ha !
- It MIGHT approach Wolverine THD - plugged into "fireanimal's" 4 pair Sanken EF3.
- Typically , complimentary ips's cancel even order harmonics. But , the odd's are so low it don't matter.
Seems a quite ideal endeavor as far as a design. Even if there is no interest - I shall have it.
below I have a FINAL schema to layout what ( I ) want. Off to the Sprint layout.
PS - this is a "branch" of the "Wolverine build thread" , I will make any design adaptable to the "wolverine EF3" , but
have no great interest in that thread. I want to at least match that design's performance with different topologies , but also want to
include our poorer hobbyist's in this project.
The "Wolverine EF3" is actually a modified HK680 (Slewmaster) OPS. I really don't have any rights to it ,but neither does anybody else.
I'm glad it was tested to the very limits of testing. COOL ! So , I will use it to to it's limit - I still can make it more robust ...
OS
Attachments
The spook was done. I just never updated to Stuart's final IPS layout (he kept changing it - excessively ).
The Hellraiser and Spooky behave similar , have the same LED cascode. CFA = 250v++/uS Spooky = 120+
Both at 1-2ppm 20Khz/8R.
All these IPS's will run from dual 35-80V supplies. 70+ might need a few component tweaks (larger R on zener supplies , close attention to
Vce for devices).
The Hellraiser and Spooky behave similar , have the same LED cascode. CFA = 250v++/uS Spooky = 120+
Both at 1-2ppm 20Khz/8R.
All these IPS's will run from dual 35-80V supplies. 70+ might need a few component tweaks (larger R on zener supplies , close attention to
Vce for devices).