Thanks Lars ,
I think you will have to incorporate the timer protection in the Zeta and we will follow ,
I will postpone making the pcb until then ,
Regards
Rajeev
I think you will have to incorporate the timer protection in the Zeta and we will follow ,
I will postpone making the pcb until then ,
Regards
Rajeev
The TLC555 protection circuit (thermal overload, and short circuit protection) will be incorporated in the N-channel amplifier very soon. Stay tuned .... 🙂
Also i had a few other ideas for improvements to try out. Lower noise, better immunity to power supply, lower rail loss, balanced input a.o.
Anybody have suggestions for a heat sink for this mother? I think it should be able to accommodate up to 10 pairs of IRFP's (25-30 cm width or length), for 1000 Watts RMS output power. 0.06 K/W or better.
Also i had a few other ideas for improvements to try out. Lower noise, better immunity to power supply, lower rail loss, balanced input a.o.
Anybody have suggestions for a heat sink for this mother? I think it should be able to accommodate up to 10 pairs of IRFP's (25-30 cm width or length), for 1000 Watts RMS output power. 0.06 K/W or better.
Using the best available silicone pads (st. Gobain type 1877), the absolute limit of dissipation for one IRFP250N is 44 watts (100 C heat sink temp).
This corresponds to 147 Watts of audio power. But remember this is THE limit, so put in a safety margin.
This corresponds to 147 Watts of audio power. But remember this is THE limit, so put in a safety margin.
Lars Clausen said:
Farnellinone have some coolingfins for fans. They are not that expensive. Have you seen them?
Sonnya: yes they are very very nice.
But unfortunately also very very expensive 😀 120 USD per channel.
(PLUS the fans.....).
Now i realize once again why pulse amplifiers always end up being cheaper at the end of the day 😉
Any other ideas?
🙂
But unfortunately also very very expensive 😀 120 USD per channel.
(PLUS the fans.....).Now i realize once again why pulse amplifiers always end up being cheaper at the end of the day 😉
Any other ideas?
🙂
Hey Lars,
0.06°C/W means an heat seenk long 600mm high 300mm with 35 fins 70mm high. Isn't it a bit too big?
0.06°C/W means an heat seenk long 600mm high 300mm with 35 fins 70mm high. Isn't it a bit too big?
It's just counting backwards: used power supply voltage, output power at a known load, duty cycle known => Power must be dissipated somehow. Cool away 500-1000 watts needs a huge heatsink.pro said:
You have the power and how many degrees warmer (or hotter) you want the heatsink to be. This gives you the deg/Watt ot the heatsink.
Lars Clausen said:Sonnya: yes they are very very nice.
But unfortunately also very very expensive 😀 120 USD per channel.
Now i realize once again why pulse amplifiers always end up being cheaper at the end of the day 😉
Any other ideas?
🙂
Maybe You should double the number of fets to be able to degrade the W/degress C ratio... ??
0.06 °C/W means that with 500W the rise of temperature is 30 °C
over the ambient temperature. 60 °C at 1000W and it doesn't depends from how many fete Do you use.
Ciao.
over the ambient temperature. 60 °C at 1000W and it doesn't depends from how many fete Do you use.
Ciao.
Lars ,
A wonderful idea , we will be waiting for your next version of the ZETA .
Regarding heatsinks I have a friend in California who makes custom made heatsinks , if suitable to you to call from there I will send you his e-mail id , and also talk to him , please let me know if you are interested .
with best wishes
Rajeev
A wonderful idea , we will be waiting for your next version of the ZETA .
Regarding heatsinks I have a friend in California who makes custom made heatsinks , if suitable to you to call from there I will send you his e-mail id , and also talk to him , please let me know if you are interested .
with best wishes
Rajeev
pro said:
The heat is the same, on this point we agree.
Your base is your ambient temp.
Then add the heatsink temperature.
Every fet has a temp. coefficient of 0.7degress/watt
A standard isolation pad has a temp. coefficient of 0.3degress/watt
The result is then 1degress/watt per fet.
Then divide by 20 for 10 pairs => .05degress/watt
Or divide by 40 for 20 pairs => .025degress/watt
This way you can degrade the heatsink peformance from .06 up to .085 => This will make the heatsink a lot cheaper.
And i think with the price of the IRFP250N the increased number of Fets will be cheaper than a bigger heatsink.
It should also be mentioned that a given size heatsink will have better performance with a distributed load, ie:
5 pcs X 200W vs 1 pc 900W, the 5 pcs heatsink can be about 30% smaller, or looked at the other way, the 1 pc heatsink would need to be about 50% bigger.
5 pcs X 200W vs 1 pc 900W, the 5 pcs heatsink can be about 30% smaller, or looked at the other way, the 1 pc heatsink would need to be about 50% bigger.
I think cooling 500W of heat passively wont be good, if fan is used much smaller heatsink can be used.
I think of using 32 cm long and 12x12cm high and wide heatsink, then ill mount 12x12c vent on each side of it. Ill make it of 4 heatsinks which are 16cm long 12cm wide and 6cm high. So at the end ill have a tunnel with high air flow which will take away much more heat than just passive use of those 4 heatsinks in standard way..
I think of using 32 cm long and 12x12cm high and wide heatsink, then ill mount 12x12c vent on each side of it. Ill make it of 4 heatsinks which are 16cm long 12cm wide and 6cm high. So at the end ill have a tunnel with high air flow which will take away much more heat than just passive use of those 4 heatsinks in standard way..
I understand that increasing the no of output devices will reduce the temp /watt but I do not understand howcome the same fan cooled heatsink ie ;- 32cm long and 12cm high x12cm wide------ if made of four sections 16cm long and 12cm wide and 6cm high will be cooler ??? please explain , or did I understand your text wrongly ??
Rajeev
Rajeev
Start with a new one
Hello !
today i'm starting with a new "The ZETA!" in my Version it's a "The small ZETA!".
Please look on my website for this project: http://bauteile-fuer-die-elektronik.de/projekte/ZETA/index.php
I go on with the next Parts tomorrow.
Any Comments ?
Hello !
today i'm starting with a new "The ZETA!" in my Version it's a "The small ZETA!".
Please look on my website for this project: http://bauteile-fuer-die-elektronik.de/projekte/ZETA/index.php
I go on with the next Parts tomorrow.
Any Comments ?
The two heatsinks will be turned to each other and then followed by another two, creating something like a tunnel (like its done in some Pa amps..), then 12x12cm fan will be on each side.
my question was why four , ( The two heatsinks will be turned to each other ) I understand , will breaking the length help in better cooling ???
skaara,
With your plan to use 4 heatsinks, that 2 face the other 2, How exactly do you plan to distribute the FETs?
If anybody wants to build the 1000W version, and benefit the price of buying higher quantity, I'll be glad to order the parts for us from the vendors, as long as you can make a part list; something that I still don't have any experience with.
What's up with the board?
With your plan to use 4 heatsinks, that 2 face the other 2, How exactly do you plan to distribute the FETs?
If anybody wants to build the 1000W version, and benefit the price of buying higher quantity, I'll be glad to order the parts for us from the vendors, as long as you can make a part list; something that I still don't have any experience with.
What's up with the board?
I think after some thinking i have a good idea for this, even if it is a bit unconventional.
It uses two comb profiles, pressed together to form a windtunnel.
I think 250 mm long, and 120 mm by 120 mm will cool the 1000 Watt ZETA stage nicely, and it holds 10 pairs of IRFP's.
The comb profiles need to be something like 120 by 50 mm, and 250 mm long, does anyone have a good source?
It uses two comb profiles, pressed together to form a windtunnel.
I think 250 mm long, and 120 mm by 120 mm will cool the 1000 Watt ZETA stage nicely, and it holds 10 pairs of IRFP's.
The comb profiles need to be something like 120 by 50 mm, and 250 mm long, does anyone have a good source?