That appears to be a reasonable amount of heat sinking. It can be made more effective if there is a gap, say two inches, between the bottom of the heat sink and the metal base. Alternatively, provide ventilation holes or slots in the metal base at the heat sink locations and be sure that air can have unimpeded access to the cooling holes/slots. If the metal base in situated on a flat surface, have feet under the base to raise the it for the air to get to the cooling holes/slots.
There's a small gap of about 3mm between the bottom of the sink and the chassis, but this "bottom half" of the chassis that I'm using was manufactured with these heatsinks, so it's got a bunch of openings underneath the heatsinks that almost perfectly match with the fins, so it's pretty much open on the bottom. I have a fan down there now, but I'm going to be adding fans to each side. I have a couple of slow/silent computer fans that should be just the ticket to knock a few degrees off.
Edit: Ill tell ya what though, these thermistors on the edges of the ps board are trying to give me a hot brand on my wrist every time I move the clippies around trying to peek at the bias lol
I just had a closer look at your pictures and I do see the slots. The slots being located right under the fins and with only a 3mm gap between the fins and the slots, there isn't much room for air to get from the bottom of the amp base to the fins. The heat sink would be better situated for air flow if the slots were lined up with the space between the fins. For best heat transfer from heat sink to air, you want to establish an upward air flow between the cooling fins.
Class A amplifiers will be hot. A heat sink temperature of 50 degrees C is quite acceptable, and 55 degrees C is still safe.
Class A amplifiers will be hot. A heat sink temperature of 50 degrees C is quite acceptable, and 55 degrees C is still safe.
Yep, it's good. I've been checking it with the infrared thermometer. I'm seeing a little over 60c if I aim it just right at the top of the fet, but on the sinks it's mid 50s.
Infrared thermometers on silver surfaces may give inaccurate readings.
How long can you keep your hand on the heat sink?
Or do you have a contact thermometer that you can try? I use a cooking thermometer placed against the heat sink when I check the temperature.
How long can you keep your hand on the heat sink?
Or do you have a contact thermometer that you can try? I use a cooking thermometer placed against the heat sink when I check the temperature.
I can keep my hands on the sinks as long as I want, but I worked at jiffy lube for 10 years, so it's not a fair comparison, lol.
I actually have a patch of the heatsink colored black with marker so I can hit it. 😉
I don't remember where the post was but Pass mentioned having had amps running with the FET case temp of up to 80C or 85C heatsinks around 65C or so. In other words, these parts are pretty rugged.
The F5m probably has somewhere between 14-18db gain which is what you would expect from these smaller 2 stage designs. Less stuff in the way but less gain.
I just picked up the kit with the single power supply. I am excited to play with it and get some parts cookin'. I am upgrading my dad's Aleph Mini with it.
The F5m probably has somewhere between 14-18db gain which is what you would expect from these smaller 2 stage designs. Less stuff in the way but less gain.
I just picked up the kit with the single power supply. I am excited to play with it and get some parts cookin'. I am upgrading my dad's Aleph Mini with it.
I have some heat test observations which I think are also relevant here. I'm using fan cooled vertical aluminium extrusions as heat sinks on F5T's running +-27.5V rails with IRF240/9240 devices which have plastic case tops on metal tab bases.
(See photos in post on the F5T thread https://www.diyaudio.com/community/threads/f5-turbo-builders-thread.207103/post-7745945)
At an ambient of 25C, I use a thermocouple contact probe to measure at three different spots, ;
1. The case temp at the metal tab backing, which can be accessed with the temp probe from above through the half moon cutouts either side of the mounting hole.
2. The plastic case temp where the thermistor on top touches output device.
3. Metal heat sink temp at mid point between the two output devices.
@0.8A bias & 21W per device - metal tab case temp = 85C, plastic case temps = 70C, heat sink temp = 47C. Projected junction temp would presumably be 17C higher than metal tab temp which would be 102C.
@0.9A bias & 25W per device - metal tab case temp = 106C, plastic case temp = 84C, heat sink temp = 54C. Projected junction temp would be 21C higher than metal tab temp which would be 127C.
I'm inclined to think that the temps measured at 0.9A bias may be on the high side for long term reliability of output devices, even though they're a bit less than absolute max of 150C?
And I'm also not sure what to make of the large temp differential between metal and plastic parts of the case, and whether to interpret Mr Pass' mention of 85C as being OK temp measured on the metal or plastic part of the case?
Any comments appreciated.
(See photos in post on the F5T thread https://www.diyaudio.com/community/threads/f5-turbo-builders-thread.207103/post-7745945)
At an ambient of 25C, I use a thermocouple contact probe to measure at three different spots, ;
1. The case temp at the metal tab backing, which can be accessed with the temp probe from above through the half moon cutouts either side of the mounting hole.
2. The plastic case temp where the thermistor on top touches output device.
3. Metal heat sink temp at mid point between the two output devices.
@0.8A bias & 21W per device - metal tab case temp = 85C, plastic case temps = 70C, heat sink temp = 47C. Projected junction temp would presumably be 17C higher than metal tab temp which would be 102C.
@0.9A bias & 25W per device - metal tab case temp = 106C, plastic case temp = 84C, heat sink temp = 54C. Projected junction temp would be 21C higher than metal tab temp which would be 127C.
I'm inclined to think that the temps measured at 0.9A bias may be on the high side for long term reliability of output devices, even though they're a bit less than absolute max of 150C?
And I'm also not sure what to make of the large temp differential between metal and plastic parts of the case, and whether to interpret Mr Pass' mention of 85C as being OK temp measured on the metal or plastic part of the case?
Any comments appreciated.
Last edited:
Heat sink temperature close to either device measure only around 1 to 3 degrees C different from mid point i.e. essentially the same. The higher device generally seems a few degrees hotter than the device 8cm below, irrespective of whether it is N or P type. These small differences in measurement seems to vary with how exactly the metal probe touches the metal heat sink, and also for how long, so may not be really relevant all the time.
Right now, only 1 of the 6 amplifiers I have on hand requires me to almost bury the volume knob. Granted, the pre I'm using doesn't offer a ton of gain, if I can reduce the input resistors to bring this amp more in line with my others, that feels like a cure.
Lowering the sensitivity of another amplifier sounds like an increase in problems, not a decrease.
Just so I'm clear, this pic is why I'm wondering these things outloud.
A completely maxed volume control that's still "not enough" is certainly a problem, and an adjustment or alteration would need to be made somewhere to get some more gain. But keep in mind with any resistive control like in that picture for example attenuates by wasting energy, by literally turning the excess signal into heat. I would rather "preserve" my signal, rather than throwing it away as waste heat personally. You will be rewarded with better resolution, detail, and clarity if you can play your cards right. The higher you can run a resistive control, the more it "gets out of the way", subjectively. There is some good information about system gain structure on the diyAudio forums that is easy to find with a search.
You mention "input resistors". If we are talking about R1 and R2 on the F5M, messing with those values won't be the correct way to alter the gain.
You mention "input resistors". If we are talking about R1 and R2 on the F5M, messing with those values won't be the correct way to alter the gain.
The control in this pre is a relays and resistors affair, I would imagine sonically it doesn't really matter much where it is, it's just that it exacerbates an already existing issue I have, that it's a linear tapered control, and it's always felt like the first half of the control is a waste, and now, the first 3/4 are a waste.
It's playing in the background right now and it's well over halfway up. It's a compounded problem. Luckily my speakers are about 94db sensitive or I'd be really boned, lol
Honestly what would be great is if these passworks crossover boards I'm using had a little gain 😉
😆 Kind of the opposite from the old Kustom amps that I remember. You couldn't hardly touch the volume knob before they got really loud, maybe near full volume for them.
I do not know where pass measured the temp of the case. I would assume the plastic or middle pin. However, the difference between your case and heatsink temps may be able to be reduced. Using the below insulator and thermal paste could help to get the difference between the two down:
Avid 4180G:
https://www.mouser.com/ProductDetail/532-4180
I use arctic silver MX-4 thermal paste which seems to be a lifetime supply.
The other solution is simply to have larger heatsinks. Although maybe the glow of your heatsinks should match the glow of your tubes...
Yeah that's a whole new meaning of cherry plating, lol