Post your Solid State pics here

Here's the progress on my new preamp. This'll be sitting on my bedside table, so it's important that it looks reasonably nice.

I've used a Hammond 1455 125 x 220 x 30 black anodised case, and machined a front panel from red perspex sheet. This sits over the normal aluminium front panel, hiding the mounting screws for the display board and allowing the nut for the rotary encoder to be recessed.

There's just enough room for my discrete pre board, a little power supply board using a pair of 15W 15V Traco SMPS modules, and the display board.

Once I've got the PIC programmed to drive the volume control stuff there'll be a pair of red seven segment LED displays showing the volume level to the left of the volume adjustment knob I've increased the brightness of the picture showing the front, so the displays can just be made out behind the acrylic front panel. Pressing the knob will toggle power.

Very nice ! :)
 
Administrator
Joined 2004
Paid Member
Hi xrk971,
The computer industry has made this space age cooling technology so accessible and affordable that it’s archaic to use natural convection - the least efficient of all means of cooling. Forced air convection can be 100x more efficient and much more compact.
I disagree with that statement. While forced air cooling is far more effective, you are introducing two major points of failure that I can see right off the bat. First you need filters and a maintenance plan for the cleaning of same. Normal people are terrible with "normal maintenance" and can't be trusted to follow directions or common sense. Sorry, but that is true.

Secondly, fan or controller failure rapidly overheats the protected circuit elements and will result in eventual failure that is unrepairable. Overheated PCBs tend to delaminate and the copper lifts off easily. So even if all the parts are replaced there is a good chance it will never be reliable.

Normal convection cooling is far more reliable long term (or even short term sometimes). Generally heat damage only occurs when there is catastrophic failure or poor circuit design. I see this in both small systems (consumer electronics) and very large installations (> 2,500 watt plant paging systems) where fan cooled amplifiers typically do not receive any maintenance. I make sure the products I use are reliable convection cooled products for both controllers and amplifiers for example. Those amplifier rooms typically do require good ventilation on top of individual amplifiers, but room ventilation does not typically cause system failures and are subject to normal maintenance. I guess if a system is large enough it can't escape the attention of maintenance staff.

-Chris
 
Founder of XSA-Labs
Joined 2012
Paid Member
The final design has a micro controller for auto shutoff in case of fan failure. The world’s Internet is powered by miles of rack mount micro controller PWM fan cooled CPUs. They run 24x7 years at a time. I agree you need fan filters if you run all the time. No one leaves a 400 w dissipation Class A amp on all the time - only when listening. It’s not a case of thermal delamination of the board - as the MOSFET is thermally isolated from the PCB, but will melt in about 3-5 minutes without the fan. I also plan to have a backup mechanical auto shutoff bimetallic switch like they have on coffee makers. It goes open circuit at 75C mounted on clamp bar above MOSFET.
 
Here's the progress on my new preamp. This'll be sitting on my bedside table, so it's important that it looks reasonably nice.

I've used a Hammond 1455 125 x 220 x 30 black anodised case, and machined a front panel from red perspex sheet. This sits over the normal aluminium front panel, hiding the mounting screws for the display board and allowing the nut for the rotary encoder to be recessed.

There's just enough room for my discrete pre board, a little power supply board using a pair of 15W 15V Traco SMPS modules, and the display board.

Once I've got the PIC programmed to drive the volume control stuff there'll be a pair of red seven segment LED displays showing the volume level to the left of the volume adjustment knob I've increased the brightness of the picture showing the front, so the displays can just be made out behind the acrylic front panel. Pressing the knob will toggle power.

Looks like a lot of stages in the prea. Looks super pro though.
 
Hi Suzy,
That looks really well packaged, nice project. How about ventilation ?

-Chris

As this one is being used as a pre rather than driving speakers, I set the bias to 5mA per output stage, which keeps quiescent dissipation down to 80mW per output transistor. I’m moderately confident it’ll be fine stuffed in a small box.

For the power amps, I’m mounting the PCBs upside down with a spreader plate glued into the case with thermally conductive epoxy to transfer heat from the transistor heatsinks to the case, and gap pads on top of the L-channel to transfer heat to the spreader plate. That works really well - the whole lot stays lovely and cool. Their cases are a tad bigger (220 x 125 x 50) to allow for a little 50VA toroidal.

I think I mentioned before that heatsinking for these things is rather fiddly.
 

Attachments

  • IMG_0433.jpg
    IMG_0433.jpg
    747.2 KB · Views: 1,054
Last edited:
Administrator
Joined 2004
Paid Member
Hi xrk971,
Also well thought out. You might want to blow that out after a few years in (intermittent) operation.

Most devices with forced air cooling don't even have a piece of foam for a filter. I have to blow out my computers every year for example. I'd much rather clean a foam filter out every few months. Too bad the case is run at negative pressure! Power supply fan pressurizes, case fan sucks the air out, drawing dust in from every opening. You just can't win.

-Chris
 
The final design has a micro controller for auto shutoff in case of fan failure. The world’s Internet is powered by miles of rack mount micro controller PWM fan cooled CPUs. They run 24x7 years at a time. I agree you need fan filters if you run all the time. No one leaves a 400 w dissipation Class A amp on all the time - only when listening. It’s not a case of thermal delamination of the board - as the MOSFET is thermally isolated from the PCB, but will melt in about 3-5 minutes without the fan. I also plan to have a backup mechanical auto shutoff bimetallic switch like they have on coffee makers. It goes open circuit at 75C mounted on clamp bar above MOSFET.

Hi X
The micro controller is most probably less reliable than the mechanical temperature switch you plan to use. This thermal sw will cover all failure modes due to high temperature while it cutoff the main line voltage (normally open).

Fab
 
Here's the progress on my new preamp. This'll be sitting on my bedside table, so it's important that it looks reasonably nice.

I've used a Hammond 1455 125 x 220 x 30 black anodised case, and machined a front panel from red perspex sheet. This sits over the normal aluminium front panel, hiding the mounting screws for the display board and allowing the nut for the rotary encoder to be recessed.

There's just enough room for my discrete pre board, a little power supply board using a pair of 15W 15V Traco SMPS modules, and the display board.

Once I've got the PIC programmed to drive the volume control stuff there'll be a pair of red seven segment LED displays showing the volume level to the left of the volume adjustment knob I've increased the brightness of the picture showing the front, so the displays can just be made out behind the acrylic front panel. Pressing the knob will toggle power.

Oh my...

That looks really well done.
I've not heard a discrete component preamp before.
 
Here's the progress on my new preamp. This'll be sitting on my bedside table, so it's important that it looks reasonably nice.

I've used a Hammond 1455 125 x 220 x 30 black anodised case, and machined a front panel from red perspex sheet. This sits over the normal aluminium front panel, hiding the mounting screws for the display board and allowing the nut for the rotary encoder to be recessed.

There's just enough room for my discrete pre board, a little power supply board using a pair of 15W 15V Traco SMPS modules, and the display board.

Once I've got the PIC programmed to drive the volume control stuff there'll be a pair of red seven segment LED displays showing the volume level to the left of the volume adjustment knob I've increased the brightness of the picture showing the front, so the displays can just be made out behind the acrylic front panel. Pressing the knob will toggle power.

Hi Suzy, wow that looks great, fantastic. Can I ask please where did you order the Hammond case from?
 
Element14 drives me nuts. Make one order, get five or more deliveries (and five charges on my card) over the next three weeks. Even for stuff that's in stock in Sydney it's still a week to get to my place.

They do have pretty good stock of things like MELF resistors though. That said, some years ago I ordered a pile of MELFs. One of the packets was a single resistor short (like 499 out of 500). Sure enough, a few weeks later a single resistor shows up on my desk at work, in it's own little plastic bag, that gets it's own padded envelope, via courier, with an 8c charge to my credit card.

By contrast, Mouser are incredibly good. Free delivery for orders over $60 - everything shows up in a single box, and it takes three or four days to get to my place in regional WA.
 
Administrator
Joined 2004
Paid Member
Newark / Element14 drives me crazy for the same reason. Many things have to come from England. I figure that's their problem and not mine. So they don't get he sale as soon as I see stock is across the pond. I have them hold the entire order until it's filled for the simple reason that I need to have the other parts as well.

Mouser and Digikey have this down pat. I haven't any idea why Newark thinks the way they do.

-Chris