The Zenductor 2's got finally moved up to the 'HiFi table' in the living room, looped in between a stand-alone bluetooth receiver (DS220) and a speaker/amplifier switch box (more amps than speakers 😎).
They were still running a bit hot at 1.2 to 1.25V across the inductor (the hottest of the four transistors was close to 90°C on the case), so I dialed the bias down to 1.0V.
Plenty loud with the KEF R3's!
They were still running a bit hot at 1.2 to 1.25V across the inductor (the hottest of the four transistors was close to 90°C on the case), so I dialed the bias down to 1.0V.
Plenty loud with the KEF R3's!
Simplified the setup a bit, before I lose living room privileges for my amps...
The Zenductor2's fit nicely on top of the ET30 VU meter/switch box. Has anybody built an integrated enclosure with VU meters?
And how does everybody handle setups with multiple amplifiers? Multiple input and output (speaker) switch boxes, or set up only a single amp at a time, and swap amps by recabeling?
The Zenductor2's fit nicely on top of the ET30 VU meter/switch box. Has anybody built an integrated enclosure with VU meters?
And how does everybody handle setups with multiple amplifiers? Multiple input and output (speaker) switch boxes, or set up only a single amp at a time, and swap amps by recabeling?
Just wanted to point out that the amps are just sitting on top of the VU meter/speaker/amp switch box, so nothing was built there (yet).
I'm thinking of maybe getting a Modushop Galaxy Maggiorato 4U GX287 230 x 170 mm chassis, and attaching the Mosfets (and the Zenductor 2 boards) to the side walls. I believe this is the smallest Modushop chassis that can hold a pair of Zenductor 2 with the Mosfets 'folded down' into the same plane as the PCB. The side walls of this case, which Modushop points out are not 'real' heat sinks, should about triple the cooling surface by my estimates (from ~190 cm^2 per Mosfet to ~1100 cm^2 for a pair of Mosfets).
Alternatively, a Modushop 'Mini Dissipante 4U 300mm' would provide almost 5,000 cm^2 heat sink area per side (0.31 K/W, corresponding to ~6.5 W/(K*m^2), which sounds about right for still air), but that would be cooling overkill, and this chassis is almost twice as deep and almost one and a half times wider than the GX287.
I'm thinking of maybe getting a Modushop Galaxy Maggiorato 4U GX287 230 x 170 mm chassis, and attaching the Mosfets (and the Zenductor 2 boards) to the side walls. I believe this is the smallest Modushop chassis that can hold a pair of Zenductor 2 with the Mosfets 'folded down' into the same plane as the PCB. The side walls of this case, which Modushop points out are not 'real' heat sinks, should about triple the cooling surface by my estimates (from ~190 cm^2 per Mosfet to ~1100 cm^2 for a pair of Mosfets).
Alternatively, a Modushop 'Mini Dissipante 4U 300mm' would provide almost 5,000 cm^2 heat sink area per side (0.31 K/W, corresponding to ~6.5 W/(K*m^2), which sounds about right for still air), but that would be cooling overkill, and this chassis is almost twice as deep and almost one and a half times wider than the GX287.
Hi all
A question for you guys who have a kit to play with.....Is equipment needed to eliminate "Poof sound" during startup and shutdown?
Another question..... Will the new PCB include options to choose "boost transformer" between Triad, Cinemag CMOQ-4 or Jensen JT-123-FLPCH?
// Mats
A question for you guys who have a kit to play with.....Is equipment needed to eliminate "Poof sound" during startup and shutdown?
Another question..... Will the new PCB include options to choose "boost transformer" between Triad, Cinemag CMOQ-4 or Jensen JT-123-FLPCH?
// Mats
The 1GX388, available from the DIYaudiostore, could also work: the 310mm internal width allows to place the Zenductor2 PCBs back to back on the baseplate, with the Mosfets bolted to the side walls and connected to the PCB with about 3 inches of wire. Cooling area is about 900 cm^2 per side wall (28*8=224 cm^2 per side, and about doubled by the channels), so that should give about 1.7K/W, or a Delta T of about 30 K to ambient, more than a factor 2 better than the default heat sinks.
Has anybody else tried any compact chassis? (Besides Mega_amp's Mega-chassis in post #193...)
I believe that the options will be Triad and either Jensen or Cinemag.
The IRFP048 that came with the Zenductor 2 is in a TO-247AC package, with an insulated screw hole (which is why we could use all metal hardware (bolt, washers, nut) to attach it). At Burning Amp we mounted them with a thermal pad that I believe is also insulating.
Question 1: Does one need the insulating pad, i.e. is the back side of the case electrically connected to one of the electrodes (the drain, most likely), or is the back side insulated as well, and one could just use thermal grease? The Vishay package drawing shows a ~1mm layer at the back of the device labelled 'E1', but it doesn't say what that layer is made from.
Question 2: I assume I'm not supposed to measure case-drain resistance with a standard DVM, as that could let the smoke out? Is there a safe procedure to check this resistance?
Question 1: Does one need the insulating pad, i.e. is the back side of the case electrically connected to one of the electrodes (the drain, most likely), or is the back side insulated as well, and one could just use thermal grease? The Vishay package drawing shows a ~1mm layer at the back of the device labelled 'E1', but it doesn't say what that layer is made from.
Question 2: I assume I'm not supposed to measure case-drain resistance with a standard DVM, as that could let the smoke out? Is there a safe procedure to check this resistance?
There is metal at the back of the TO-247 package and it is connected to the drain. So you need an electrical insulating pad between the mosfet and the heat sink if the heat sink is connected electrically to ground. The metal at the back of the package helps thermal transfer to the heat sink as metal is a better thermal conductor than plastic.
I do not believe the DVM will damage the mosfet. However static electricity may.
I do not believe the DVM will damage the mosfet. However static electricity may.
Thanks for the quick reply, Ben Mah! Is this (back metal electrically connected to drain) also true for the TO-247AC (with the insulated hole)? The Vishay web page for the IRFP048 has three drawing pages, the middle one you posted with 'version 2 : facility code = Y', which does show a drain connection on the back side, and a first drawing 'version 1 : facility code = 9' and a third drawing with 'version 3 : facility code = N', which both do not have lead assignment labels. The devices I have are labelled 'IRFP048N'.
Last edited:
I think all TO-247 have an insulated hole.
You can easily check the resistance between the back metal and the drain pin with a meter to confirm the electrical connection between the back metal and the drain pin.
I have drilled TO-247 packages to fit M4 bolts with no issues. There is enough plastic surrounding the hole to keep the hole insulated, at least in the mosfets that I drilled. So drill at your own risk. 🤓
You can easily check the resistance between the back metal and the drain pin with a meter to confirm the electrical connection between the back metal and the drain pin.
I have drilled TO-247 packages to fit M4 bolts with no issues. There is enough plastic surrounding the hole to keep the hole insulated, at least in the mosfets that I drilled. So drill at your own risk. 🤓
I ordered the 1GX388 from the diyaudiostore (https://diyaudiostore.com/products/galaxy?variant=12174834308). I'm going to drill (& saw & file) the panels myself, but I thought I try some preliminary layouts in the free front panel designer software available at frontpanelexpress.
I'd appreciate comments and criticism, as these are my first amplifier front & back panels.
The front panel has two rectangular 23mm x 10mm cutouts for small bias voltage displays (voltage over the inductor), potentially a small 1/4" round hole for a volume potentiometer (part of a chip preamp that I'm not sure yet I'll need & want), and a 30mm hole for the biggest on/off button I could find.
The back panel has the speaker connections near the top, the RCA inputs in the middle, and the DC input jacks near the bottom. The chassis is big enough to hold the switching power supplies for the two Zenductor 2 amps, but I think I'd rather have the smaller DC cables running into the chassis. These should all be 10.7mm holes, maybe 11mm.
The RCA and DC inputs are positioned so they line up with the back edge of amplifier PCB, making the connections a straight shot. The speaker posts are moved a bit outwards, in order to be out of the way.
I'd appreciate comments and criticism, as these are my first amplifier front & back panels.
The front panel has two rectangular 23mm x 10mm cutouts for small bias voltage displays (voltage over the inductor), potentially a small 1/4" round hole for a volume potentiometer (part of a chip preamp that I'm not sure yet I'll need & want), and a 30mm hole for the biggest on/off button I could find.
The back panel has the speaker connections near the top, the RCA inputs in the middle, and the DC input jacks near the bottom. The chassis is big enough to hold the switching power supplies for the two Zenductor 2 amps, but I think I'd rather have the smaller DC cables running into the chassis. These should all be 10.7mm holes, maybe 11mm.
The RCA and DC inputs are positioned so they line up with the back edge of amplifier PCB, making the connections a straight shot. The speaker posts are moved a bit outwards, in order to be out of the way.
Since you mention heat dissipation: I redid my guesstimate for the Galaxy 'non-heatsink' side panels. Counting just the cross-connect segments, i.e. the horizontal surfaces that add to the surface area of the inside and outside walls, that are open to the sides (and not the ones open to the top and bottom), and assuming they are 8mm deep each, I arrive at a total area of (2*8cm+14*0.8cm)*28cm=760 cm^2. Assuming a heat transfer coefficient 6.5 W/(m^2*K), extracted from the quoted thermal resistance and the (estimated) dimensions of the 'real' Modushop heatsinks, I get a thermal resistance of 2 K/W. Earlier I had guessed 1.7 K/W, by assuming that the channels double the side wall area. I believe the new estimate is more realistic, since the top and bottom channels are cut off from air circulation by the top and bottom plates.
This is a lot worse than the real heatsinks, which range around 0.3 to 0.5 K/W, but still a factor two better than the (double) 8 K/W heatsinks mounted on the Zenductor 2. So hopefully the Mosfets will run a factor two (in difference to ambient) cooler.
Reading up on the IRFP048, according to the data sheet it is good up to 175 °C. So running it at 90 or 100 °C, as seems to be the case with the default heatsinks, isn't really a problem for the Mosfet.
This is a lot worse than the real heatsinks, which range around 0.3 to 0.5 K/W, but still a factor two better than the (double) 8 K/W heatsinks mounted on the Zenductor 2. So hopefully the Mosfets will run a factor two (in difference to ambient) cooler.
Reading up on the IRFP048, according to the data sheet it is good up to 175 °C. So running it at 90 or 100 °C, as seems to be the case with the default heatsinks, isn't really a problem for the Mosfet.
The Vishay datasheet mentions a typical case-to-sink thermal resistance of 0.24 K/W, presumably the number they want you to strive for, and a maximum (i.e. worst case) junction-to-case resistance of 0.80 K/W. That would add about 6 K (or °C) to the case temperature, for 8 W per device.
If one only counts the surface area exposed to the outside (i.e. no case ventilation), then the relevant area is (8cm+10*0.8cm)*28cm= 460cm^2, with a resulting thermal resistance of 3.3 K/W. This is not significantly better than the existing heatsinks (at 4 K/W combined). We'll see.
I measured the thermal resistance of the Modushop Galaxy 1GX388 2U (80 mm) by 280 mm sidewall profile to be 1.4 K/W or °C/W. See the thread at https://www.diyaudio.com/community/...for-modushop-galaxy-1gx388-2u-chassis.426972/ .
More pics of my Zenductor 2s in the 2U Modushop Galaxy chassis are in a separate thread.
The final setup, with an Elekit TU-8200R stacked on top:
Sidewall temperatures are 53°C (127 °F): touchable.
The final setup, with an Elekit TU-8200R stacked on top:
Sidewall temperatures are 53°C (127 °F): touchable.