As per the SOA chart of Toshiba 5200/1943's can safely handle 0.8A 80v Continuous and 95v 100ms pulsed
Think the datasheet has T=25 on it, Rajeev.
Now place your amplifier in India around lunchtime, and you have been watching a Bollywood for half an hour !
Now place your amplifier in India around lunchtime, and you have been watching a Bollywood for half an hour !
Hi
Hi Jacco,
That's It ,
What I also said earlier in this forum that you can't beat the Indian summer with average temp. of about 55degree celcius for continuously high SPL playing with Bipolars amps.
regards,
Kanwar
jacco vermeulen said:
Hi Jacco,
That's It ,
What I also said earlier in this forum that you can't beat the Indian summer with average temp. of about 55degree celcius for continuously high SPL playing with Bipolars amps.
regards,
Kanwar
Kanwar,
If you have a nice cold frigg for me to sit in i'll come over to visit.
I'd say that is amplifier killer territory.
If you have a nice cold frigg for me to sit in i'll come over to visit.
I'd say that is amplifier killer territory.
Hi buddy
Hi Jacco,
You are right about the amp killer territory, but wait man ,
the amps which are hunted by these Indian summers are QSC, Studiomaster, Peavey .
You cannot run these amps continuously for 24hrs nonstop,7 days a week.
But our stallions are the few lucky ones which are indian summer friendly , pumping continuous RMS irrespective of 55 deg C. temp , long durations.
Regards,
Kanwar 😀 😉
jacco vermeulen said:
Hi Jacco,
You are right about the amp killer territory, but wait man ,
the amps which are hunted by these Indian summers are QSC, Studiomaster, Peavey .
You cannot run these amps continuously for 24hrs nonstop,7 days a week.
But our stallions are the few lucky ones which are indian summer friendly , pumping continuous RMS irrespective of 55 deg C. temp , long durations.
Regards,
Kanwar 😀 😉
derating...
Jens, a couple of points, sorry, I wasn't reading closely enough, I didnt see that the SOA and loadline you showed was for 150c...but then again I'm not sure it can be...150c is the drop dead point for the transistor...it has no loadline at all at that temp...ie none, no power rating whatsoever...
I think your charts are essentially resistive, not worst case reactive. Depending on who you ask, worst case reactive has the entire rail voltage across the output transistors of one pole, probably at the maximum current the power supply can deliver, ie 90degree lead or lag...
Rajeev, from what I can see you are reading the 25c chart, do you really think you can keep your transistors at 25c? My chart has an insert that says "Curves must be derated linearly with increase in temperature'...you have to know the expected temp of the parts, then recalc the curves based on that temp...at 100c, you will get half the rating, at best...so you can assume maybe 0.4A at 80v...if you want a flaky amp...
If you re-plot the curves at 100c, you'll conclude a very different number of devices is needed for bulletproof operation into 2 ohms, if 5 pairs is 'enough' at 8ohms, you will need at least 20 for 2 ohms, and 4 pairs of drivers...and so on...
AFAIK the datasheet pulse ratings are plotted assuming assuming a 25c ambient, giving very optimistic results. It seems to me the SOA charts should never be presented at 25c ambient, power transistors of this sort are rarely used at 25c.
Good luck, you are gonna need it...
Stuart
Jens, a couple of points, sorry, I wasn't reading closely enough, I didnt see that the SOA and loadline you showed was for 150c...but then again I'm not sure it can be...150c is the drop dead point for the transistor...it has no loadline at all at that temp...ie none, no power rating whatsoever...
I think your charts are essentially resistive, not worst case reactive. Depending on who you ask, worst case reactive has the entire rail voltage across the output transistors of one pole, probably at the maximum current the power supply can deliver, ie 90degree lead or lag...
Rajeev, from what I can see you are reading the 25c chart, do you really think you can keep your transistors at 25c? My chart has an insert that says "Curves must be derated linearly with increase in temperature'...you have to know the expected temp of the parts, then recalc the curves based on that temp...at 100c, you will get half the rating, at best...so you can assume maybe 0.4A at 80v...if you want a flaky amp...
If you re-plot the curves at 100c, you'll conclude a very different number of devices is needed for bulletproof operation into 2 ohms, if 5 pairs is 'enough' at 8ohms, you will need at least 20 for 2 ohms, and 4 pairs of drivers...and so on...
AFAIK the datasheet pulse ratings are plotted assuming assuming a 25c ambient, giving very optimistic results. It seems to me the SOA charts should never be presented at 25c ambient, power transistors of this sort are rarely used at 25c.
Good luck, you are gonna need it...
Stuart
From the datasheet of the MJL3281/1302:
"The data of Figure 13 is based on TJ(pk) = 150°C; TC is variable
depending on conditions."
How does this information help? If the junction temp is 150 what does it matter what the ambient temp is, as long as it is lower?
\Jens
"The data of Figure 13 is based on TJ(pk) = 150°C; TC is variable
depending on conditions."
How does this information help? If the junction temp is 150 what does it matter what the ambient temp is, as long as it is lower?
\Jens
Jens,
that means the soa curve ends where current =0, at 150 C, the straight vertical line from 0.1 amps to 0 is for T=25C to T=150C
The rest of the curve is for 25C.
that means the soa curve ends where current =0, at 150 C, the straight vertical line from 0.1 amps to 0 is for T=25C to T=150C
The rest of the curve is for 25C.
weird...
Jens,
My motorola datasheets from 1995 are different from yours in a very strange way...
My chart is listed as allowing TjPK of 200c, ie the max temp the transistor junction experiences is 200c, kind of strange that the max. operating temp listed on the same datasheet is 150c...it seems they fixed the error in your datasheets...
All they are telling you is that if you start with the junction temp at 25c with the part mounted on an infinite heatsink (think block of copper in bucket of water), your transistor will hit 150c in 1 sec when subjected to current and voltage params shown in the chart...what they don't mention is that it blows up in about another millisecond...or less...
Stuart
Jens,
My motorola datasheets from 1995 are different from yours in a very strange way...
My chart is listed as allowing TjPK of 200c, ie the max temp the transistor junction experiences is 200c, kind of strange that the max. operating temp listed on the same datasheet is 150c...it seems they fixed the error in your datasheets...
All they are telling you is that if you start with the junction temp at 25c with the part mounted on an infinite heatsink (think block of copper in bucket of water), your transistor will hit 150c in 1 sec when subjected to current and voltage params shown in the chart...what they don't mention is that it blows up in about another millisecond...or less...
Stuart
Hi all,
we had this discussion recently on another thread.
Tjmax =150degC for plastic and =200degC for To3 (usually).
SOA is based on Tc=25degC and Tjmax <150degC or <200degC.
The temp difference from junction to case allows the heat to escape but keeps the smoke in.
All the power figures on the SOA charts whether dc or short term must be derated for a higher Tc i.e. @ Tc(plastic)=87.5degC orTc(To3)=112.5degC all figures are 50% of SOA curves, etc. These 50% figures might well be the starting point for an Indian summer!!
Yes, Tc is variable and the design process allows for the derating to take account of elevated temeratures.
we had this discussion recently on another thread.
Tjmax =150degC for plastic and =200degC for To3 (usually).
SOA is based on Tc=25degC and Tjmax <150degC or <200degC.
The temp difference from junction to case allows the heat to escape but keeps the smoke in.
All the power figures on the SOA charts whether dc or short term must be derated for a higher Tc i.e. @ Tc(plastic)=87.5degC orTc(To3)=112.5degC all figures are 50% of SOA curves, etc. These 50% figures might well be the starting point for an Indian summer!!
Yes, Tc is variable and the design process allows for the derating to take account of elevated temeratures.
an order of magnitude?
Seems to me the chart is at least a factor of 10 too high for current, even for the 25c ratings...
...in principal, yes, reducing the power handling of the transistors by 50% for a operating temp of 55c is exactly what I was talking about...
But if ambient is 50c, the transistors will be much, much hotter than 55c...so probably should derate them another 50%...
I am trying to keep my outputs at 55c and below, and I'm starting from 25c, I don't envy you guys in india, it will need to be a truly huge set of heatsinks and fans that keep the smoke in...
Stuart
Seems to me the chart is at least a factor of 10 too high for current, even for the 25c ratings...
...in principal, yes, reducing the power handling of the transistors by 50% for a operating temp of 55c is exactly what I was talking about...
But if ambient is 50c, the transistors will be much, much hotter than 55c...so probably should derate them another 50%...
I am trying to keep my outputs at 55c and below, and I'm starting from 25c, I don't envy you guys in india, it will need to be a truly huge set of heatsinks and fans that keep the smoke in...
Stuart
Hi Jens,
nice graphs but they don't make sense.
the new blue line seems to be reduced by 50% to account for higher Tc, can you confirm?
The new red line (75v) does not match with the currents shown on the second from top red line on the previous graph, what's gone wrong?
nice graphs but they don't make sense.
the new blue line seems to be reduced by 50% to account for higher Tc, can you confirm?
The new red line (75v) does not match with the currents shown on the second from top red line on the previous graph, what's gone wrong?
The fun part: the soa graphs have logarithmic scale.
You can redraw the curve and slide it to the left if you wish to keep max current, slide it down for keeping max voltage.
You only need 1 proper operation point for the selected temperature.
Simple redrawn on a piece of transparent, clever Dickies use the pc.
You can redraw the curve and slide it to the left if you wish to keep max current, slide it down for keeping max voltage.
You only need 1 proper operation point for the selected temperature.
Simple redrawn on a piece of transparent, clever Dickies use the pc.
Hi Jens,
I saw that and multiplied out the extracted numbers but they do not tally.
eg. @40v the current graph shows 3.7A or so. the power graph says 70w divide by 40 =1.75A. Your blue line has come down from 6A to 120w =3A so looks right.
BUT your red line is wonky or am I misreading these?
I saw that and multiplied out the extracted numbers but they do not tally.
eg. @40v the current graph shows 3.7A or so. the power graph says 70w divide by 40 =1.75A. Your blue line has come down from 6A to 120w =3A so looks right.
BUT your red line is wonky or am I misreading these?
Hi Jacco,
what happens when you have to reduce the current and the power at the same time? Does sliding it down still apply?
what happens when you have to reduce the current and the power at the same time? Does sliding it down still apply?
Andrew,
does not one imply the other ?
Every datasheet specifies to derate linearly with temperature.
A device is able to deliver 150 watts at 25C, 0 watts at 150C.
Inbetween power is derated with the factor:150/125W/C.
So either lower current, same voltage.
Or same current, lower voltage.
Hence the logarithmic scale.
Think capacitor:
If you put half of the operation voltage on the minus side of the capacitor the remaining potential is half too.
As the capacity of the capacitor is linear with the voltage, the maximum current the capacitor is able to deliver is also half too.
Or think heatsink:
place the same heatsink in an environment with twice the temperature difference, the heatsink will be able to dissipate twice the heat.
does not one imply the other ?
Every datasheet specifies to derate linearly with temperature.
A device is able to deliver 150 watts at 25C, 0 watts at 150C.
Inbetween power is derated with the factor:150/125W/C.
So either lower current, same voltage.
Or same current, lower voltage.
Hence the logarithmic scale.
Think capacitor:
If you put half of the operation voltage on the minus side of the capacitor the remaining potential is half too.
As the capacity of the capacitor is linear with the voltage, the maximum current the capacitor is able to deliver is also half too.
Or think heatsink:
place the same heatsink in an environment with twice the temperature difference, the heatsink will be able to dissipate twice the heat.
The red line is based on the power in one transistor.
P = Vce*I = Vce * Vout/(RL*N) = Vce * (VCC-Vce)/(RL*N)
N = number of parallel devises
Is that not ok?
\Jens
P = Vce*I = Vce * Vout/(RL*N) = Vce * (VCC-Vce)/(RL*N)
N = number of parallel devises
Is that not ok?
\Jens
- Status
- Not open for further replies.