Those of you that use the Hypex UCD400's , Do you heatsink the 2 side BJT's (the ones that get real hot !) ???
Look (below) , I made a "L" shaped aluminum plate (below 1) to pull the heat out of them.
My Hypex is bolted on a 100W class AB sub amp heatsink , that never gets even warm. I was scared
to have them 2 TO-220 devices hot to where my finger hurt.
The amp runs with the hot BJT's , but I like real cool amps.
BTW , does anyone know what these devices do ? , and why do they run so hot ?- circled (below 2).
I don't like anything to run 60-70C on any of my amps , even all my AB ones !
Thanks...
OS
Look (below) , I made a "L" shaped aluminum plate (below 1) to pull the heat out of them.
My Hypex is bolted on a 100W class AB sub amp heatsink , that never gets even warm. I was scared
to have them 2 TO-220 devices hot to where my finger hurt.
The amp runs with the hot BJT's , but I like real cool amps.
BTW , does anyone know what these devices do ? , and why do they run so hot ?- circled (below 2).
I don't like anything to run 60-70C on any of my amps , even all my AB ones !
Thanks...
OS
Attachments
Last edited:
Full amp - Tests OK !!
My FIRST Class D amp ! Yah ....
It has a +/- 12V supply and low pass filter (left PCB's). I reused the Sonance "virtuoso" original circuitry.
In the middle is a DC/delay to protect my lovely sub and the "star" (Hypex uCD400 w/hxd). Nice cool (L) bracket , only 40C now.
Main heatsink barely gets warm with normal use - It should not replacing a 100W AB amp !
The right is a 250VA toroid and 40Kuf cap bank to give me +/- 55V. This voltage seem to be mid spec for the uCD400.
The build looks more OEM than some DIY "toy". Matches the other Sonance 60W AB amp (below 2). A nice matching 2.1
system to play my new Topping E-30 II with !
I haven't Really let "it rip" into the Ultra 500W subwoofer yet (below 2). But I will....
PS - I researched and all I could find on the transistors is "active load" ??? Bruno Putzney's paper is out there , but is just
a basic overview of the simpler uCD180 without Hxd or the load.
OS
My FIRST Class D amp ! Yah ....
It has a +/- 12V supply and low pass filter (left PCB's). I reused the Sonance "virtuoso" original circuitry.
In the middle is a DC/delay to protect my lovely sub and the "star" (Hypex uCD400 w/hxd). Nice cool (L) bracket , only 40C now.
Main heatsink barely gets warm with normal use - It should not replacing a 100W AB amp !
The right is a 250VA toroid and 40Kuf cap bank to give me +/- 55V. This voltage seem to be mid spec for the uCD400.
The build looks more OEM than some DIY "toy". Matches the other Sonance 60W AB amp (below 2). A nice matching 2.1
system to play my new Topping E-30 II with !
I haven't Really let "it rip" into the Ultra 500W subwoofer yet (below 2). But I will....
PS - I researched and all I could find on the transistors is "active load" ??? Bruno Putzney's paper is out there , but is just
a basic overview of the simpler uCD180 without Hxd or the load.
OS
Attachments
They seem to be voltage regulators for the control/drive/switch circuit. They are isolated from the rest of the switching circuitry - moved to the side of the PCB... which makes me believe that they are indeed NOT part of the drive/switch circuit. I'd measure voltages at the collector and emitter and see what I get, or just check the TO220 metal cases and see what you read in ref to common. Being located at the side of the PCB should make it easy to attach them to a copper rod or similar and keep them cool. 60-70degC is nothing (me with my class-A hat on)... I wouldn't worry about it at all.
Thank you - someone replied.
The collectors read +/- 55V (rails) , emitters read +/- 15.7V exactly.
I did put that aluminum "L" in , I'm just 40'ish C now.
PS - the patented/hyped HXr regulators are also warm , no documentation other than their (regulating) virtues.
Are they Shunts ? , does having the HXR option stress the TO-220's ?
I have seen some uCD builds without the suspected HXr shunt option ?
Edit here is the "HXR"....
OS
Attachments
Last edited:
So, they indeed seem to be series regulator BJTs... dropping a fair amount of voltage access them - hence the heat dissipation
You can trace the TO220 BJTs emitters towards the (presumably) shunts, to confirm conductivity. Then you'll know for sure if the TO220 BJTs are pre-regs for the shunts. The heat (stress as you refer to...) comes from the large voltage drop across them BJTs.
They do look like shunt regs to me. They probably drop very little voltage across IN and OUT pins... The HXR PCB seems to have two INs, two OUTs and one common pin... again - not really hard to check all of the assumptions with the multimeter.... They could be just a second set of series regs... The first set (TO220 BJTs) could be for the driver IC (buffer?); the second set (the HXRs) could be for the switch circuit (right in front of the output transistors). I suppose, if you don't have better things to do, you could buzz out everything stated above with a multimeter... and then you'll know for sure.
I'm scared of it !!! It does make sense that the 40V drop across the TO-220's is the cause of all that heat.
I am kind of amazed that a class D design is not ice cold , this is the selling point for class D. At idle , the Hypex dissipates
about the same as the other pictured amp (Sonance AB 60W).
I suppose it DOES "the class D thing" while running the sub at 300W , it does not get any warmer !
An AB amp running 300-400W would warm up the room or need a fan.
BTW - the HXR's power the input buffer op-amp and some of the SMD semi's (CCS , differential).
Since these HYPEX's are quite outdated , Putzney should open source the design. ALL of it....
Perhaps some of the epoxy potted daughter-boards contain part of the secret of the next gen ??
The HXR regulators are also "secret" potted on the back side....
He makes plenty of money off the next gen NC500's ($400).
Oh , well. It works. Hope it works for decades - I've given it the best "environment" I can (heatsink/good PSU).
PS - my SMD wolverine type AB actually runs much cooler than the Hypex (unless I crank it).
OS
I suppose the current draw through those TO220s is very little, keeping in mind that huge voltage drop. So, the class D efficiency statement is probably warranted.
I'd also like to see more details on that design... it will help others with implementation and better understanding in general.. the (original) UcD design is the one I'd get if I ever decide to go with class D amps... I did some research.. and concluded that UcD would be the one for me.
https://spectrum.ieee.org/the-consumer-electronics-hall-of-fame-philips-ucd-audio-amplifier
I suppose you could compare it (that class D amp) with your Wolverine class AB, driving full-range speakers, and see what happens...
I'd also like to see more details on that design... it will help others with implementation and better understanding in general.. the (original) UcD design is the one I'd get if I ever decide to go with class D amps... I did some research.. and concluded that UcD would be the one for me.
https://spectrum.ieee.org/the-consumer-electronics-hall-of-fame-philips-ucd-audio-amplifier
I suppose you could compare it (that class D amp) with your Wolverine class AB, driving full-range speakers, and see what happens...
Thoroughly tested out the smaller A04 case on the desktop. No fans and no vertical mounting needed for laminar air flow cooling....My All-Billet 6061-T6 amplifier case with integrated dial is complete and thoroughly tested. Will be building another soon...
Been running the XS-8TL's with this amp for better part of the last week with my phone DAC and FLAC files.
Barely gets warm running them half volume on the dial.
I took the dial back off and reset it to 6'oclock being zero.
WAF is at 100% too as opposed to the amp and CD player in the living room.
Here's a pic where I removed the fan so you can see direct airflow path to heatsink radiator.
I also added 3/8" height rubber feet
Attachments
Just modified my Hypex Nilai DIY500 Stereo amp by adding a gain switch (on-off-on 6 pin) to the rear panel. Seems to work and can be easily reversed if it has any negative effect on the sound.
detail posted in ASR.
https://www.audiosciencereview.com/...finished-amplifier-with-pffb-is-coming.54799/
https://www.audiosciencereview.com/...finished-amplifier-with-pffb-is-coming.54799/