I decided to make a seperate PSU that I can connect to the amplifier enclosure with an umbilical cord using Neutrik
Speakon STX 8-pole connectors.
The reason to go for a separate PSU enclosure where for me:
1 - cost reduction: PSU can be used in other future amplifiers (cost of this PSU is sofar €606)
2 - size an weight: with the PSU I had in mind, it was not possible to put everything inside one enclosure
Features and build up of this completely dual mono PSU
--------------------------------------------------------------------
1: EMI-filter
2: DC-blocker à la Rod elliot (10000µF/63V VBC-components caps and RHRP1560 15A/600V diodes)
4: two safety loop breakers with R10/5W resistors and 15A/600V diodes
5: Soft start circuit with:
- standby/ON led identication (Led's can be mounted on front panel togheter with Standby/ON switch)
- reduced primary voltage of 21V/50Hz for all R-cores during standby => +-6Vdc on all capacitors
to increase life time and reduce inrush currents (darTZEEL does something similar if I'm right))
- 230V/50Hz auxilliary output when PSU is switched on (For XFMR's inside amplifier enlcosure to fed for example the output protection circuittry)
- all output's are fused (= primary of XFMR's)
6: 2x 500VA/2x36V R-cores for +rail ann -rail for output stage
7: 2x 30VA/2x9V R-cores for ++rail and --rail for input, VAS and Driver stage.
8: HFA25PB60 (2x8pcs) and HFA08TB60 (2x8pcs) diodes used for rectification
9: 22000µF/100V (output stage) and 4700µf/100V (input-, vas- and driver stage) as capacitance.
The rectified voltage of the 30VA R-cores is in series with the rectified voltage of the 500Va R-cores, but all capacitors are referenced to ground. All voltages will thus vary the same, dependant of the mains voltage.
For the input stage, VAS and Driver section, adjustable (Vout-Vin) capacitance multipliers have to be used.
I would like to have the possibility to use capacitance multipliers for the driver section. While I'm having a preference of using Fet's as drivers, I've choosen for a higher rail voltage for the driver section.
For the driver section I would like to have a PSU that can deliver easily 1A, with the series solution of the XFMR's, a 30VA R-core was more than sufficient to increase rail voltage and allow the minimum current of 1A.
Everything is build up on 8mm HPL (high pressure laminate) board. Very easy to drill holes and cut threath.
Enjoy the pictures!
Speakon STX 8-pole connectors.
The reason to go for a separate PSU enclosure where for me:
1 - cost reduction: PSU can be used in other future amplifiers (cost of this PSU is sofar €606)
2 - size an weight: with the PSU I had in mind, it was not possible to put everything inside one enclosure
Features and build up of this completely dual mono PSU
--------------------------------------------------------------------
1: EMI-filter
2: DC-blocker à la Rod elliot (10000µF/63V VBC-components caps and RHRP1560 15A/600V diodes)
4: two safety loop breakers with R10/5W resistors and 15A/600V diodes
5: Soft start circuit with:
- standby/ON led identication (Led's can be mounted on front panel togheter with Standby/ON switch)
- reduced primary voltage of 21V/50Hz for all R-cores during standby => +-6Vdc on all capacitors
to increase life time and reduce inrush currents (darTZEEL does something similar if I'm right))
- 230V/50Hz auxilliary output when PSU is switched on (For XFMR's inside amplifier enlcosure to fed for example the output protection circuittry)
- all output's are fused (= primary of XFMR's)
6: 2x 500VA/2x36V R-cores for +rail ann -rail for output stage
7: 2x 30VA/2x9V R-cores for ++rail and --rail for input, VAS and Driver stage.
8: HFA25PB60 (2x8pcs) and HFA08TB60 (2x8pcs) diodes used for rectification
9: 22000µF/100V (output stage) and 4700µf/100V (input-, vas- and driver stage) as capacitance.
The rectified voltage of the 30VA R-cores is in series with the rectified voltage of the 500Va R-cores, but all capacitors are referenced to ground. All voltages will thus vary the same, dependant of the mains voltage.
For the input stage, VAS and Driver section, adjustable (Vout-Vin) capacitance multipliers have to be used.
I would like to have the possibility to use capacitance multipliers for the driver section. While I'm having a preference of using Fet's as drivers, I've choosen for a higher rail voltage for the driver section.
For the driver section I would like to have a PSU that can deliver easily 1A, with the series solution of the XFMR's, a 30VA R-core was more than sufficient to increase rail voltage and allow the minimum current of 1A.
Everything is build up on 8mm HPL (high pressure laminate) board. Very easy to drill holes and cut threath.
Enjoy the pictures!
Attachments
Last edited:
Congratulations on a very beautiful build, Bensen!
Since you've got a couple of diodes mounted back-to-back (with their case electrodes facing each other), you might want to possibly consider inserting an insulator between them. Just in case a service person (or gravity?) causes them to pinch together and touch cases some unhappy day in the future. Maybe a sheet of mylar, or mica, or even black electrical tape on the back of each TO-220, would bring additional peace of mind. Diagram below:
Since you've got a couple of diodes mounted back-to-back (with their case electrodes facing each other), you might want to possibly consider inserting an insulator between them. Just in case a service person (or gravity?) causes them to pinch together and touch cases some unhappy day in the future. Maybe a sheet of mylar, or mica, or even black electrical tape on the back of each TO-220, would bring additional peace of mind. Diagram below:
Attachments
Thanks Mark,
I had the same thought about these back to back rectifiers and have put some shrink tube (or what is the name of this stuff) over them yesterday.
Regards
I had the same thought about these back to back rectifiers and have put some shrink tube (or what is the name of this stuff) over them yesterday.
Regards
Looks nice,but why all those bridge rectifiers ?
By puting the two 36V in series and at every end a 9V winding you get two fases of 0-36-45V .For each output only two diodes.Those diodes are for the same current but double voltage.The gain is not only that half the number of diodes is cheaper but also,more interesting, half the heat to dissipate.
Mona
By puting the two 36V in series and at every end a 9V winding you get two fases of 0-36-45V .For each output only two diodes.Those diodes are for the same current but double voltage.The gain is not only that half the number of diodes is cheaper but also,more interesting, half the heat to dissipate.
Mona
Hello Ketje,
I definetely did not want to use the center tap of the xfmr's as ground level.
See as referentie post 38 from following thread.
For the rest, I didn't spend time on how to reduce the number of diodes.
http://www.diyaudio.com/forums/chip-amps/12162-2-rectifier-bridges-why-4.html#post143874
Greets
I definetely did not want to use the center tap of the xfmr's as ground level.
See as referentie post 38 from following thread.
For the rest, I didn't spend time on how to reduce the number of diodes.
http://www.diyaudio.com/forums/chip-amps/12162-2-rectifier-bridges-why-4.html#post143874
Greets
Yes,after looking at that post 38,he has a point there.But the netto current to the center is zero for a series PP.The momentary differences are absorbed by the (big) chemicals.You can even monitor the zero-line,if there is a current,DC is applyed to the speaker.
The diodes get rather hot,needs cooling.Cutting that in half doesn't interrest you?
Mona
The diodes get rather hot,needs cooling.Cutting that in half doesn't interrest you?
Mona
The dual bridge design for the high power section is a design choice from me after a lot of research. The same is applicable for the NOT in series connection of the secondaries before these are getting rectified.
In my previous amp, I'm using the same diodes, also not mounted on a Heatsink. They stay pretty cold even at continously high output power.
In my previous amp, I'm using the same diodes, also not mounted on a Heatsink. They stay pretty cold even at continously high output power.
If you prefer it that way,it's you choice, and your supply 🙂
I now of someone using a bridge rectifier for a Pass Aleph 5 (class A) that's getting rather hot.Probably your appication isn't drawing continously high current.
Mona
I now of someone using a bridge rectifier for a Pass Aleph 5 (class A) that's getting rather hot.Probably your appication isn't drawing continously high current.
Mona
It looks awesome! A sight to behold.
Any reason why the smoothing is handled by a single electrolytic per rail instead of several paralleled and/or bypassed with small value foil cap for lower ESR and high frequency performance??
Any reason why the smoothing is handled by a single electrolytic per rail instead of several paralleled and/or bypassed with small value foil cap for lower ESR and high frequency performance??
Thanks,
More big electrolytics (+-20000µF/rail) and bypass caps will be placed inside the amplifier enclosure.
The 22000µF/100V caps are Low ESR Types with a ripple current of 15A at 100Hz.
More big electrolytics (+-20000µF/rail) and bypass caps will be placed inside the amplifier enclosure.
The 22000µF/100V caps are Low ESR Types with a ripple current of 15A at 100Hz.
- Status
- Not open for further replies.