DoZ cooling

[tschrama]

The schematic seems be errorous... better remove all 0.1 Ohm resistors... all the output transistors should have at a individual 0.22 Ohm emitor resistor at their emmitor...

[starbase218]

Tschrama: Why should I use 0R22 resistors instead of 0R1? (note: I already know the schematic is errorous, I attached it because the GIF image I attached earlier didn't display correctly).

Mattyo5: 50wpc in Class-A and you want more power? What massive heatsinks do you have to keep things cool? (pics? )

[Mattyo5]

Heatsinking is no problem....house wiring to handle those toroids i'm using...thats an issue. hehe FIRE

I really want to have a fairish comparo between Aleph 2 and DOZ... just for the heck of it. Of course, its a wee bit expensive experiment, but in the end,

here's a pic of the heatsinks, this was my first DOZ at about 50wpc, each sink dissipates 150 watts without fans

this became a stereo amp, my aleph 2's will look similar but be monoblock 75lb monsters

-Matthew K. Olson

[DocLorren]

See attached my DoZ project.
It runs hot since it dissipates 68Watts per channel. The heatsinks are relatively small but actually the whole construction (40 volts times 1.7A quiescent current) radiates heat (the base plate is made from 4mm aluminum) even the toroid top plate warms up nicely. Soon, I will put it in a full metal enclosure that will add additional heatsink 'mass'. No problems until now after approx. 600 hours of playing. What do you think of this approach?

DocLorren

[Mattyo5]

Its better if the heatsinks were turned upright, so the fins could have airflow up through them, as you have it, its close, but the air cannot flow up through them from the bottom. Otherwise, should be fine Heatsink orientation isn't easy to get right sometimes b/c its convenient to have it more like you have it, or oriented in a different direction (See Nelson Pass Volksamp 30)

-Matthew K. Olson

[megajocke]

For better temperature stability the resistors should be connected like this. The original didn't use any resistors so I think 0R1:s will be enough for balancing.

[starbase218]

DocLorren: Looks nice! Are those Fischer SK34 heatsinks? I use the same, only 4 of them (2 transistors per heatsink, 2 heatsinks per channel), they're 10 cm high. Right now they're not mounted to the case (which is made of steel with an aluminium front plate). If relocating the 0R1 transistors helps enough to counteract thermal runaway, I'll just mount them to the case upright. If not, I'll mount the sinks with fins horizontally and leave room for some fans (like I mentioned earlier) in between the heatsinks.

megajocke: I'm going to relocate them that way if I have the time (I'm a little busy right now).

[DocLorren]

Starbase, thx for your comments.

The heatsinks are from Conrad.
http://www1.nl.conrad.com/scripts/wg...c=X&~cookies=1

I tested for thermal runaway in a situation very close to what the final construction would be like (same amount of aluminum to dissipate heat) and let it warm up for 3-4 hours. The quiescent current stayed within 20 mA of the aimed 1.7A. I would therefore say it is stable. In practice (much more important) it does not seem to experience any problem by running hot. By the way, Rod E. himself acknowledges it becomes hot. Well, what do you expect? We are running Class A here!
Below find the CAD pictures of the complete DoZ with DoZ pre-Amp. I will probably finish the casings next month. A true and worthy D-o-Zen minimalismistic design! Stay tuned...

DocLorren

[starbase218]

Well, I got around to it and re-wired the amp so that the resistors are in the emitter leads. And.. it does stabilize at around 43C with 1.35A quiscent current (I wrote down the current and temperature every 5 minutes for 1.5 hours after it stabilized). But after I closed the case and put the amp back in my rack, it heated up further. If closing the case allready makes such a difference, I figure the margin I have is very, very small.

So I'll probably buy some fans after all, and use a resistor and maybe a thermistor to keep the noise down.

[tja]

Hi,

If you check out the manufacturers K/W graphs vs. what Conrad claims in their catalog or website there are quite a few discrepancies. SK34 10cm is more like 0.9K/W instead of Conrads claim of 0.75K/W if you check Fischer's graph on page A54 of this pdf. Same thing for the other heatsinks in the Conrad catalog.

Take