rajeev luthra said:
Rajeev,
My estimate of 8 pairs was with a 700va transformer that sags the rails to about 63-65vdc where 8 pairs will survive. However once you said that you want 850 watts into 4 ohms... the equation changed.
Jens is quoting 75vdc unloaded rails for 5 pairs which will not give you more than 500 watts 4 ohms unless you use a huge transformer, by huge I am talking over 2kVA.
As Stuart says and as I have said, for each of the following variables, you need to increase the number of devices:
1) Power needed
2) Current demanded by load
3) DC rails
4) Operating temperature of the device PLUS ambient. You should shoot of 55c tops for reliable operation. For typical class-AB you want to stay below 45C.
5) Reactance of the load.
6) How stable your DC rails are under load.
All of these factors will affect the number of devices you need. I still mantain that if you want 850 watts into 4 ohms, you are looking at rails of 96vdc or above and 24 pairs to have a safe and reliable amplifier. You are better off getting MJ21193/94's or settle for lower power.
If you have 11 pairs, with rails of +/-85 vdc (75 fully loaded) I think you can get safely get (in hot Indian weather) 350 watts 8 ohms, 600 watts 4 ohms and momentarily 1000 watts 2 ohms. However as the amp warms up, 2 ohm operation will be very flakey.
Yes like K-amps says:
For continuous use at 2 ohms , heat will be the killer.
There is however a big difference between the heat generated at the heat sink by musical signals and continuous sinewave test signals.
But you have to be careful also about transient heat at the chip itself . That rises much faster than the heat sink temperature.
Never use the devices at or close to their maximum current rating . Long term relability is inversely proportion to operating temperature for most equipment.
Cheers.
For continuous use at 2 ohms , heat will be the killer.
There is however a big difference between the heat generated at the heat sink by musical signals and continuous sinewave test signals.
But you have to be careful also about transient heat at the chip itself . That rises much faster than the heat sink temperature.
Never use the devices at or close to their maximum current rating . Long term relability is inversely proportion to operating temperature for most equipment.
Cheers.
Another option. Like Jens says use 75 volt rails but bridge the amp. You will get your 850 watts nto 4 ohms bu tuse at least 10 pairs.
I chose the Toshiba because now they are availible here at reasonable
rates nowadays , but those who are in the US and for whom price and
availibility is not a problem can go for THE BEST that is
MJL4281A/MJL4302A has a very high SOA 1.0A/100v@1sec,
2.0A/75v@1sec , and 350v Vces . A much better device than the
Toshiba .
Also if I am not wrong the thumb rule is that the dissipation of the devices is 40%of the max output , if we add other losses for safety and make it 50% the max power per device at 1500w will be 750/22=34watt per device , and at 35w the SOR should be within limits . Comment .
rates nowadays , but those who are in the US and for whom price and
availibility is not a problem can go for THE BEST that is
MJL4281A/MJL4302A has a very high SOA 1.0A/100v@1sec,
2.0A/75v@1sec , and 350v Vces . A much better device than the
Toshiba .
Also if I am not wrong the thumb rule is that the dissipation of the devices is 40%of the max output , if we add other losses for safety and make it 50% the max power per device at 1500w will be 750/22=34watt per device , and at 35w the SOR should be within limits . Comment .
The current at 34watts will be around 0.4A max , and the SOA for the
Toshiba are 110v continious , 130v /100ms , & 180v /10ms , ofcourse
the Tc is = 25digrees c .
Even if we derate due to increase in temp we should be within SOA .
Also the amp will be amplifying music and NOT sine wave and this is the
absolute max power .
With experiance I have realised that the best way of mounting the
devices on the heatsink for max heat transfer is to mount it directly
without any insulator , I have designed the layout such that the pcb is
above the heatsinks , devices can be mounted on both sides on different
heatsinks and the heatsinks which are the +ve and -ve rails can be
insolated from the main chasis , further my two heatsinks for one
channel are weighing 2.7kg and have a 4" cooling fan at full speed ,
the extreme day temp here is never above 42digrees that too only for a
few hours in a day , and mostly the programmes are at night when temp
falls below 35 digrees , also this extreme is only for one month in a year
, rest of the year the day temp never goes above 30 digrees .
Toshiba are 110v continious , 130v /100ms , & 180v /10ms , ofcourse
the Tc is = 25digrees c .
Even if we derate due to increase in temp we should be within SOA .
Also the amp will be amplifying music and NOT sine wave and this is the
absolute max power .
With experiance I have realised that the best way of mounting the
devices on the heatsink for max heat transfer is to mount it directly
without any insulator , I have designed the layout such that the pcb is
above the heatsinks , devices can be mounted on both sides on different
heatsinks and the heatsinks which are the +ve and -ve rails can be
insolated from the main chasis , further my two heatsinks for one
channel are weighing 2.7kg and have a 4" cooling fan at full speed ,
the extreme day temp here is never above 42digrees that too only for a
few hours in a day , and mostly the programmes are at night when temp
falls below 35 digrees , also this extreme is only for one month in a year
, rest of the year the day temp never goes above 30 digrees .
Attachments
like the comment about it being a power supply part,
It would be difficult NOT to find a better transistor. The original RCA 2n3055 had real specs, but over the years everyone put leftover die on t03 packages and called them 3055.
It would be difficult NOT to find a better transistor. The original RCA 2n3055 had real specs, but over the years everyone put leftover die on t03 packages and called them 3055.
rajeev luthra said:
That is part 1 !
Part 2 is to calculate the average die temperature.
Taking in account:
- the thermal resistance of the heatsinks you are using.
- design air temperature.
- the effectiveness of the fans at given air temperature.
- the number of devices.
Part 3 is to check if there is a bottleneck with the derated SOA.
The thing to keep in mind is that with BJT's leaving SOA is not allowed, not even for 1 mS !
In contrast to Mosfets, that can survive 1st thermal breakdown.
Even if the average power load you estimated is acceptable, at the resulting die temperature there may not be enough margin on the SOA for the amplifier to deliver the desired numbers.
Mathematically:
[V *I] / t should be within the SOA but also [dV * dI] / dt
Suppose you are driving on a highway, not allowed to do more than 60 mph, or you will receive a ticket.
If at any point you are doing less than 60 mph on the speedometer, you will not reach an average speed of 60.
If at any point you do less than 60, and you desire to have driven 60 miles after 1 hour, you will need at least once to go faster than 60.
But then you will be fined !
imo, it is more accurate to use the impedance graph of the loudspeaker system you intend to hook up to the amplifiers and calculate average load from that.
I usually draw derating lines on my BJT SOAR charts, -10% in the solid second breakdown region and to 75degrees for the thermal.
i.e. 90% and 70% for a 200deg max one. Then I compare and contrast!
i.e. 90% and 70% for a 200deg max one. Then I compare and contrast!
Hi
Hi Rajeev,
I absolutely doesnot agree with you.
Here in Punjab[province of India] the average temp. at mid-day summer is around 50deg C.
The DJ parties are mostly at nights but the famous long duration festivals [babeya da mela , pira da mela , maa de navratre ]
for others like carnivals, are all likely to be non stop programs running continuously for days and that too, day & night.
maybe your indore doesnt have these type of carnivals.
In this type of situation the amplifier must have an OVERBUILT ouput stage to stand still in these harsh and hardcore conditions.
Therefore its mandatory to use some extra protectiveness in terms of famous SOA contours.
regards,
Kanwar😉
rajeev luthra said:
Hi Rajeev,
I absolutely doesnot agree with you.
Here in Punjab[province of India] the average temp. at mid-day summer is around 50deg C.
The DJ parties are mostly at nights but the famous long duration festivals [babeya da mela , pira da mela , maa de navratre ]
for others like carnivals, are all likely to be non stop programs running continuously for days and that too, day & night.
maybe your indore doesnt have these type of carnivals.
In this type of situation the amplifier must have an OVERBUILT ouput stage to stand still in these harsh and hardcore conditions.
Therefore its mandatory to use some extra protectiveness in terms of famous SOA contours.
regards,
Kanwar😉
Kanwar,
50c - that's cooking before you turn the power on!
How's the new project? 0.002% THD at 20KHz and below yet, (while it's driving nails) - that's when they sing!
50c - that's cooking before you turn the power on!
How's the new project? 0.002% THD at 20KHz and below yet, (while it's driving nails) - that's when they sing!
K-Amps
This is what I had written in my post no 103 page 11,
""I am planning to make a class AB amp ,
Transformer 66v-o-66v ie DC will be +/- 93v
Power Output approx 450w 8E, 800w 4E , and 1500w 2E
The amp will rarely be used at 2E but I want to over kill ,
Will 8 pairs of 2SA1943/2SC5200 be ok ,
Drivers MJE15034/35 or 2SA1302/2SC3281
Filter caps 30,000MFD per rail
here the Toshiba drivers 2SA1302/2SC3281 are costing 50% more
than MJE15034/35 which ones should I choose.
please advise and give your recomendations ""
Your reply on the same page was ;
"" Post #105
I don't know the exact model but a Mark Levinson 4 series monoblock
(rated at 300 8E, but actually does 500 8E, 9004E and 1.5kW 2E) uses
8 pairs of a Toshiba class TO-3PL devices.
It has 40mF per rail or 80,000uF in total with a 2372 VA toroid.
Your design seems similar to that in terms of power, I'd say as a rough
estimate, you are on target. ""
Now you say that you did not know I was looking for 800w at 4ohms ??
How ever nearly every one on the forum had passed 8pairs of
2SA1943/2SC5200 , only AndreasW had said that they were less ,
hence I increased from 8pairs to 11pairs , Andrews had okeyed 10pairs
.
This is what I had written in my post no 103 page 11,
""I am planning to make a class AB amp ,
Transformer 66v-o-66v ie DC will be +/- 93v
Power Output approx 450w 8E, 800w 4E , and 1500w 2E
The amp will rarely be used at 2E but I want to over kill ,
Will 8 pairs of 2SA1943/2SC5200 be ok ,
Drivers MJE15034/35 or 2SA1302/2SC3281
Filter caps 30,000MFD per rail
here the Toshiba drivers 2SA1302/2SC3281 are costing 50% more
than MJE15034/35 which ones should I choose.
please advise and give your recomendations ""
Your reply on the same page was ;
"" Post #105
I don't know the exact model but a Mark Levinson 4 series monoblock
(rated at 300 8E, but actually does 500 8E, 9004E and 1.5kW 2E) uses
8 pairs of a Toshiba class TO-3PL devices.
It has 40mF per rail or 80,000uF in total with a 2372 VA toroid.
Your design seems similar to that in terms of power, I'd say as a rough
estimate, you are on target. ""
Now you say that you did not know I was looking for 800w at 4ohms ??
How ever nearly every one on the forum had passed 8pairs of
2SA1943/2SC5200 , only AndreasW had said that they were less ,
hence I increased from 8pairs to 11pairs , Andrews had okeyed 10pairs
.
Jacco,
In part 1 , I had added 10% tolerance extra ,
Next re part 2&3 ,
If I compare with QSE MX3000 ,
heatsinks are eight times larger ,
I have two 4inch cooling fans at full speed while that has one 3" cooling
fan ,
I have 11 pairs of output devices, while that has 6 pairs ,
I am mounting devices directly on heat sinks , while that has insolating
washers ,
My rated power output is lesser than MX3000,
The main differance is that my amp will be class AB and most proberbly
that is class H .
I wonder how the branded poweramp manufacturers getaway with very
limited material , and QSE MX3000 is a very good amp it works very
well in our weather .
Ampman,
I am making the amp for my own use , and the weather conditions mentioned were for Indore , that is my city , I do not know of temp in your area ,
However the temp you mention must be outside in full sun , do sound operaters in your area keep their gear in the open without shade or fan , I also wonder how the operators themselves can bear such extreme tempratures ??
In part 1 , I had added 10% tolerance extra ,
Next re part 2&3 ,
If I compare with QSE MX3000 ,
heatsinks are eight times larger ,
I have two 4inch cooling fans at full speed while that has one 3" cooling
fan ,
I have 11 pairs of output devices, while that has 6 pairs ,
I am mounting devices directly on heat sinks , while that has insolating
washers ,
My rated power output is lesser than MX3000,
The main differance is that my amp will be class AB and most proberbly
that is class H .
I wonder how the branded poweramp manufacturers getaway with very
limited material , and QSE MX3000 is a very good amp it works very
well in our weather .
Ampman,
I am making the amp for my own use , and the weather conditions mentioned were for Indore , that is my city , I do not know of temp in your area ,
However the temp you mention must be outside in full sun , do sound operaters in your area keep their gear in the open without shade or fan , I also wonder how the operators themselves can bear such extreme tempratures ??
rajeev luthra said:
Rajeev,
ML's and Krells etc use lower rails and larger and better regulated power supplies. Lower rails means you can have less devices for a given power output.
Also I said uses Toshiba "class" devices not necessarily Toshiba's 5200/1943...because I did not know exactly what device numbers they used.
Hi everyone
That's the reason because Indian summer is always biased in class-A.......This is still very below from the average temp. at the SAHARA deserts of AFRICA..😀 😀 😉
In outdoor Melas a.k.a Carnivals you cant imagine an air conditioner in dusty and feeling hot hot summers to keep the amps cooooooool.😀
regards,
Kanwar
amplifierguru said:
That's the reason because Indian summer is always biased in class-A.......This is still very below from the average temp. at the SAHARA deserts of AFRICA..😀 😀 😉
rajeev luthra said:
In outdoor Melas a.k.a Carnivals you cant imagine an air conditioner in dusty and feeling hot hot summers to keep the amps cooooooool.😀
regards,
Kanwar
Hi,
K-Amps lets end it , its my baby and I am responsible for it ,
But I am wondering what is the use of a device which is 150w 230v
which is not safe at 34w80v/160v ??, I will mail and ask toshiba .
Next please tell me what is the MAX power I can get with 11 pairs of of
2SA1943/2SC5200 in the output , any supply voltage and any LS
impedance !!!!!
Max is that if you people advise I will not go below 4 ohms , here the dissipation per device will be 18 w only ....
K-Amps lets end it , its my baby and I am responsible for it ,
But I am wondering what is the use of a device which is 150w 230v
which is not safe at 34w80v/160v ??, I will mail and ask toshiba .
Next please tell me what is the MAX power I can get with 11 pairs of of
2SA1943/2SC5200 in the output , any supply voltage and any LS
impedance !!!!!
Max is that if you people advise I will not go below 4 ohms , here the dissipation per device will be 18 w only ....
Why don`t you make it class h !? you can fiind some QSC schematics showing a comparator and a fet to switch the voltage to higher rails 
B.I.G said:Why don`t you make it class h !? you can fiind some QSC schematics showing a comparator and a fet to switch the voltage to higher rails
He has already pre-fabbed the PCb's etc.
Rajeev,
11 pairs of MJL21193/94 will drive 1 ohm loads ;-) You only need new devices not change your whole design.
B.I.G,
Can you send me the schematic of QSE MX 3000 , I think it is a discontinued model now .
K- Amps,
Won,t MJL4281A/MJL4302A be better ?? What is the price of these in the US , they are not availible here .
However I have already purchased the Toshibas , first I will try to utilise them .
Can you send me the schematic of QSE MX 3000 , I think it is a discontinued model now .
K- Amps,
Won,t MJL4281A/MJL4302A be better ?? What is the price of these in the US , they are not availible here .
However I have already purchased the Toshibas , first I will try to utilise them .
does anyone expeienced an amplifier playing for a couple of hours
not knowing the fan use for cooling is dead and the thermal tracking
device not even placed on the heat sink?
I have experienced this before. after turning the volumes down ,I was wodering what was that tickling sound. I turn around looking where
that be and at the same time there was this smell, a smell of something that was very hot. Then I realised twas from my amp
the heat sink was so hot that could fry an egg, and you could hear an angry tickling sound as if angry enough to emmit fire.
I was very thankful that I was using some indestructible
outputs,
MJL21196 connected in a Quasi complementary topology
at plus/minus 80volts supply voltage, 4 ohm subwoofer load,
4 pairs per channel
so for a while after tunning off to cool down, I turned it on,
as if nothing happens, it was again kicking
hienrich
not knowing the fan use for cooling is dead and the thermal tracking
device not even placed on the heat sink?
I have experienced this before. after turning the volumes down ,I was wodering what was that tickling sound. I turn around looking where
that be and at the same time there was this smell, a smell of something that was very hot. Then I realised twas from my amp
the heat sink was so hot that could fry an egg, and you could hear an angry tickling sound as if angry enough to emmit fire.
I was very thankful that I was using some indestructible
outputs,
MJL21196 connected in a Quasi complementary topology
at plus/minus 80volts supply voltage, 4 ohm subwoofer load,
4 pairs per channel
so for a while after tunning off to cool down, I turned it on,
as if nothing happens, it was again kicking
hienrich
- Status
- Not open for further replies.