Do you mean qnty 8 of each or 4 N-channel and 4 P-channel? At quanty 10 they are 3.74GBP.
The ECW20 are 7.75GBP ea. or 10.70GBP ea. At current exchange rate os $1.23USD to 1GBP we are talking $106US not including shipping and duties/taxes. A premium price for a premium amp.
ECX10N20 Plastic Lateral MOSFET | Profusion
ECW20N20 Plastic Lateral MOSFET | Profusion
That's per channel, so maybe it makes more sense to get 10 and have 2 spares for a stereo set.
@Astx, trust me, I am not the only one who finds higher order odd harmonics (when greater than even order) fatiguing, and maybe makes me uneasy after longer extended sessions of listening.
Although 150mA x 70v x 4 sets of MOSFETs is 42w = same as 25w Class A amp!
Last edited:
Please note Keantoken indicates the S types, which are more expensive (GBP4,51 and 8,57 resp.).
Group buys??
4 pairs of output fets would dissipate 84w at idle.
Last edited:
Not to derail, but Keantoken suggests that a 300x210x4cm heatsink will do (0.18C/W).
It's an A/B amp after all.
You can use OITPC compensation by Dadod. I used it for same topology of your amplifier, but I used VMOS.
https://www.diyaudio.com/forums/solid-state/317335-oitpc-output-inclusive-tpc-tmc-11.html#post5605697
https://www.diyaudio.com/forums/solid-state/317335-oitpc-output-inclusive-tpc-tmc-11.html#post5605697
Well, the reason I added another pair was mainly because I saw that 20Hz worst case phase angle exceeded the thermal time constant of the FETs. But again, that assumes someone playing a full power sine into the worst speakers at 20Hz. Who does that? When I move the worst case frequency up to 40Hz, 3 pairs is fine.
For 3 pairs, we would be looking at this:
70V rails, 0.18C/W heatsink:
3 pairs ECX10x20-S 300W 8ohm 417W 6ohm
3 pairs ECW20x20-S 600W 4ohm
If we have 5 builders, we can order 60 S grade FETs to get the best price. That is $59 for 12 10x20 per amp, or $111 for 12 20x20 per amp.
I guess the question is, is there anyone here who would actually build the amp when it is finished?
70V rails, 0.18C/W heatsink:
3 pairs ECX10x20-S 300W 8ohm 417W 6ohm
3 pairs ECW20x20-S 600W 4ohm
If we have 5 builders, we can order 60 S grade FETs to get the best price. That is $59 for 12 10x20 per amp, or $111 for 12 20x20 per amp.
I guess the question is, is there anyone here who would actually build the amp when it is finished?
As will I of course. I can probably guess that another member from Long Island will also go in on this big amp. 
That’s 3 builders.
In better ask Jhofland to design a SSR DC protect that can handle 140Vpp outputs. I think we may have just the MOSFETs to do this as I found some that are used in my SFP soft start.
That’s 3 builders.
In better ask Jhofland to design a SSR DC protect that can handle 140Vpp outputs. I think we may have just the MOSFETs to do this as I found some that are used in my SFP soft start.
Last edited:
I am working with Jhofland and we have a design for a thermal cutoff combined with a SSR for soft start circuit. It will require a manual reset if it cuts off. We also have a SSR DC protect and turn on delay circuit that can be added directly to the main amp PCB. You would never want to run a DC coupled 300w amp without speaker protection. We have this all covered based on previous designs. The thermal cutoff aspect is new but fairly simple to implement.
I plan on using a mechanical bimetallic temperature switch set at say, 85C bolted to the heatsink. When the temp switch activates (N/O type, it closes a low level logic circuit) an optotriac disconnects mains AC power from the amp.
I plan on using a mechanical bimetallic temperature switch set at say, 85C bolted to the heatsink. When the temp switch activates (N/O type, it closes a low level logic circuit) an optotriac disconnects mains AC power from the amp.
Last edited:
Because the combined thermal resistance of the output transistors is so close to the thermal resistance of the heatsink, the thermal cutoff needs to be set lower, something like 60C. That is low enough to prevent exceeding the thermal time constant at 20Hz. If we are not worried about that then 70C or 80C would be fine.
6x 10x20 have a combined thermal resistance of 1.1/6=0.183C/W. Our heatsink is 0.18C/W. Therefore at a 30C ambient, when the output transistors reach 150C the heatsink temp will be 89C.
6x 10x20 have a combined thermal resistance of 1.1/6=0.183C/W. Our heatsink is 0.18C/W. Therefore at a 30C ambient, when the output transistors reach 150C the heatsink temp will be 89C.
Last edited:
What does that mean in practice: that the heatsink should be mounted inside a chassis, and cannot be touched?
IMHO it would be better to turn off the amplifier if the heatsink or something inside reaches 60 degree celsius.
BR, Toni
What happen if the energy does not enough to supply current to output transistors. I means that transformers maybe limited the power.
The newest ''Aurum'' are using 2 x 300VA/55VAC x 2 transformer for 2 channel (the transformers are connected as parallel), each channel using 4 output MOSFETs. So it obviously the output MOSFETs will be limited before reach maximum due worse case of loads conditions. Because the transformers can't run 200% in long term.
The newest ''Aurum'' are using 2 x 300VA/55VAC x 2 transformer for 2 channel (the transformers are connected as parallel), each channel using 4 output MOSFETs. So it obviously the output MOSFETs will be limited before reach maximum due worse case of loads conditions. Because the transformers can't run 200% in long term.
Hi Keantoken,
Are we getting closer to locking down the design for going to layout for Rev 1 of this amp? Keep in mind that we will have a flexible layout that has a separate mezzanine board for the input stage and the MOSFETs will be on Molex quick connects so that even the main amp board can be replaced quite quickly.
Thanks,
X
Are we getting closer to locking down the design for going to layout for Rev 1 of this amp? Keep in mind that we will have a flexible layout that has a separate mezzanine board for the input stage and the MOSFETs will be on Molex quick connects so that even the main amp board can be replaced quite quickly.
Thanks,
X