Front end board biased up to .780 Vdc now or so, slowly crept up from .750 once I put the lid on and let it cook a bit. The last BA-3 board I biased up stopped around low .780s. Once this happens I will power down, wire the out put boards to it and go at it again biasing them up and setting offset. Let the whole mess cook an hour, reset everything and assuming all is well, put it in the system for a trial run!
Attachments
![IMG_4078[1].jpg](/community/data/attachments/447/447897-44657929f028cb038bd113673ebeee50.jpg)
![IMG_4080[1].JPG](/community/data/attachments/447/447908-7094de68e426cf9778fd28e1267ae568.jpg)
![IMG_4083[1].jpg](/community/data/attachments/447/447917-d3aa9b32bedb0b3c014f4121f1162f0f.jpg)
I might be posting this in the wrong thread, but I am having trouble finding info on the BA-3SE. Lucky me just got 6 pairs of matched IRFP240s in the mail, so what would be more natural than building a dual mono BA-3SE with fat PSUs...?
The question in mind is what rail voltage I should use; am I stuck with 24V, or would 32V work just as well (or maybe even better)?
Thank You for listening!
Jorgen
The question in mind is what rail voltage I should use; am I stuck with 24V, or would 32V work just as well (or maybe even better)?
Thank You for listening!
Jorgen
So I'm scratching my head around heatsing dissipation calculations.
I'm planning on a 32v, 3 pairs per channel build.
I will try to bias around 0.5A per device, so around 3A total per side. At 32v this gives me 96W dissipation per channel.
The 5U store case has 4 sinks with 0.28W/C each, so each heatsink will have to dissipate 48W (3 devices per sink) which gives me this result:
Ta + (P * (Rjc + Rcs + Rsa))= Tj
30º + (48W* (0.83+0.24+0.28))=94.8º
This is way above the 65º recomended when reading the temperature in the mosfets.
I know I'm making a mistake somewhere as I'm not considering the dissipation being done by 3 devices.
So I thought about making the calculation with just one mosfet but dividing the heatsink Rsa by 3 since it shared by 3 mosfets, like this:
30º + (16W* (0.83+0.24+0.84))=60.56º
This is a much better result, it somewhat makes sence in my head but it's always good to double check.
Is my reasoning correct?
I'm planning on a 32v, 3 pairs per channel build.
I will try to bias around 0.5A per device, so around 3A total per side. At 32v this gives me 96W dissipation per channel.
The 5U store case has 4 sinks with 0.28W/C each, so each heatsink will have to dissipate 48W (3 devices per sink) which gives me this result:
Ta + (P * (Rjc + Rcs + Rsa))= Tj
30º + (48W* (0.83+0.24+0.28))=94.8º
This is way above the 65º recomended when reading the temperature in the mosfets.
I know I'm making a mistake somewhere as I'm not considering the dissipation being done by 3 devices.
So I thought about making the calculation with just one mosfet but dividing the heatsink Rsa by 3 since it shared by 3 mosfets, like this:
30º + (16W* (0.83+0.24+0.84))=60.56º
This is a much better result, it somewhat makes sence in my head but it's always good to double check.
Is my reasoning correct?
No !
3pair @ 0.5A each is 1.5A for the total
1.5A times 32Vdc = 48W dissipation for the upper half. Another 48W dissipation for the lower half. That then gives 96W total dissipation for the 3pair output stage biased at 1.5A using ±32Vdc. Right answer wrong calculation.
No.
You have omitted the temperature de-rating. This is usually given in the manufacturer's datasheet, but I'll bet you did not get that from the retailer.
For a cool heatsink the de-rating factor (DF) can be as high as 1.5
do the calculation in separate easy to follow stages.
Ta?
Ts?
apply DF
new Ts estimate
Tc?
apply device derating for Tc
Pmax? at this estimated Tc.
Last edited:
They will be mounted in a PCB and around and beneath it i will drill several holes to allow for maximum ventilation. I might also add a small low noise fan extrating heat from the case.
Assuming we are thinking the same thing, that is a good question.
These are rated at 0.18C/W Dissipatore 300 X 40 H 210 mm
But at what temperature above ambient? Heatsinks are more efficient the hotter they get.
Conrad suggest a derating figure for their heatsinks Conrad Heatsinks - Technical Details (bottom of page), but modushop offers no advice.
Last edited:
I can tell from experience the 5U 400mm case will easily dissipate 100W per side of the case or in other word easily dissipates 50W per heatsink (4 heatsinks on the case).
I don't trust the numbers modushop quotes
Last edited:
You have quoted it wrong, it should be 0.28C/W not 0.28W/C. Makes a huge difference in the calculation.
Assuming modushops specification is accurate then 48 x 0.28 = 13.4 degrees above ambient (in reality it maybe closer to 20 degrees above ambient)
Either way you are not going to have any issue with that case for BA3.
most heatsinks are rated for a deltaT of 70C to 80C degrees.
most heatsink manufacturers issue de-rating data/tables for their heatsinks when DeltaT is less.
For a deltaT of around 13C degrees expect the de-rating factor to exceed 1.5
i.e. 60W into 0.2C/W and using a DF of 1.6 gives DeltaT = 60 *0.2 *1.6 = 19.2Cdegrees
If Ta = 28°C, then Ts = 47°C
most heatsink manufacturers issue de-rating data/tables for their heatsinks when DeltaT is less.
For a deltaT of around 13C degrees expect the de-rating factor to exceed 1.5
i.e. 60W into 0.2C/W and using a DF of 1.6 gives DeltaT = 60 *0.2 *1.6 = 19.2Cdegrees
If Ta = 28°C, then Ts = 47°C