Brother of Quasi

Status
Not open for further replies.
a newer one

i have so far builted 3 of the brother of quasi in power 180-300w depending on the number of outs and rail voltage all of them are used for small pa application and so far they manage realy zouper

only last night friend and costumer asked to use one with his own speakers for a party .... i give him a big one the
300+300W with
1500W trafo
120 ma bias /
70+ rails /
75.000+75.000 mfd/
10x MJW21194 per ch/
45 cm of heatsink per chanel plus
tunel ventilation with 2 x 7.5cm double ventilators per chanel

to my surprize ( poped in to the party for a drink ) my friend was driving the most stupid speaker available in the market
Bought from MAkro super markets
including 2x15" woofers without any xover
including 5 8x8 cm piezo tweeters also i think without any xover
including 1.5 " piezo driver + horn also i think without any xover ....

i said to my shelf that by the end of the night either tweeters or amp is going to be history after driving this so stupid load for full power all night ....

well nothing happend .... full power for many hours always very next to cliping no serious temperature rising speakers presented a load i think lower than 4 ohms ( but i can never tell ) and by the end of the night all the piezos where still working ....

any comments ????
 
Until sourcing some parts to building process, i continue to work on board. I want to try 2 pair Sanken BJT - 2SC2922- version with about +/- 30Vac (38-40Vdc rails) 660Va. In same order i want to swap MJE15030/15031 to 2SC4790/2SA1837, but i don't if Toshiba Ic capability is enough compared to MJE one : 1A compared to 8A...The amp must be able to drive 3.5 ~4ohms speaker (worst situation and not along all frenquencies). I'am on building this MT bookshelf .

nbip_210.jpg


nbip_211.jpg


Marc
 
1837 won't drive a highly reactive 4ohm speaker.

The maximum Ic @ 40Vce through the 1837, when Tc=25degC, is 300mA.

assume a gain on a high current transient for the 2pair output stage is ~hFE=40
The maximum output is 12Apk shared between the two devices.

40Vce and 12Apk is OK for the output devices driving a 3r1 resistive load.
You need ~25Apk to 30Apk to drive a 4ohm speaker.

The cold 10ms SOA shows only 630mA @ 40Vce. Even this cannot drive a transient through a 4ohm speaker. The gain of the outputs will drop too low to maintain sensible loading on the driver.

1837 will drive an 8ohm speaker from +-50Vdc.
 
AndrewT said:
1837 won't drive a highly reactive 4ohm speaker.

The maximum Ic @ 40Vce through the 1837, when Tc=25degC, is 300mA.

assume a gain on a high current transient for the 2pair output stage is ~hFE=40
The maximum output is 12Apk shared between the two devices.

40Vce and 12Apk is OK for the output devices driving a 3r1 resistive load.
You need ~25Apk to 30Apk to drive a 4ohm speaker.

The cold 10ms SOA shows only 630mA @ 40Vce. Even this cannot drive a transient through a 4ohm speaker. The gain of the outputs will drop too low to maintain sensible loading on the driver.

1837 will drive an 8ohm speaker from +-50Vdc.

Thanks Andrew for explanation, I will keep the on semi MJE15030/15031 i have in stock.

Marc
 
a 3pair output stage with big sinking on the drivers may just drive 4ohm from 1837 if the speakers are not too reactive. Moderate would be OK, max phase angle 45degrees and min impedance ~3ohms.
But, for domestic use only, NOT PA, NOT disco.
MJE15034/5 might be better still. 4A but much better specs for all else.

Which ground should R11 be connected to? GND1 (signal ground) or GND (power ground)?
 
MJE15030/31 drivers look funny with Sanken's 2SC2922, sorta like throwing out the baby with the bathwater.

Not so long ago people built 200W power amps with 3 pairs of the same Sankens and Toshiba 2SA968/2SC2238 drivers, as in >56Vdc rails.
For 40Vdc rails, Toshiba's A1837/C4793 are not much worse than the A968/C2238 oldies.
If you had picked the A1930/C5171, maybe one might argue that OnSemi drivers would be a preferable choice.
Quel dommage.
 
its been

way too many times that in the forum is the question
"""" how many pairs ber board """
"""""how much rail voltage""""

no mater what the forum people say i have to admit that i see almost every day that japanese people might have another opinion

just come in to my shop for repair/remake a NIKKO ALPHA II wich means only 2 transistors from NEC 2sb600 2SD555 per board and the astronomical rail voltage of
63 VOLTS

any comments ????

( amplifier belongs to a costumer and it come in for recaping and upgrading up to a possible margin ....though i will try to borrow it and try to push it to the really hard limits to see how its going to be .... )
 
AndrewT said:
a 3pair output stage with big sinking on the drivers may just drive 4ohm from 1837 if the speakers are not too reactive. Moderate would be OK, max phase angle 45degrees and min impedance ~3ohms.
But, for domestic use only, NOT PA, NOT disco.
MJE15034/5 might be better still. 4A but much better specs for all else.

Which ground should R11 be connected to? GND1 (signal ground) or GND (power ground)?

Hi Andrew, according to quasi layout R11 is connected to power ground.

My goal is to build a amp for domestical use. My room mesure under 40m². I juste want to be able to drive the speaker i shown you above or may be an other combo Scanspeak 15W4531G00 4Ohm/ Fountek JP3. I was able to drive this combo with difficulty with a lateral mosfet amp (2 pair of SK1058/SJ162) and comparable PSU i will use with BOQ (660Va 2x30Vac for both chanel, 4x68.000µf/63v). So no PA appication for me but only normal domestical use with all requierement in dynamics.....

Marc
 
jacco vermeulen said:
MJE15030/31 drivers look funny with Sanken's 2SC2922, sorta like throwing out the baby with the bathwater.

Not so long ago people built 200W power amps with 3 pairs of the same Sankens and Toshiba 2SA968/2SC2238 drivers, as in >56Vdc rails.
For 40Vdc rails, Toshiba's A1837/C4793 are not much worse than the A968/C2238 oldies.
If you had picked the A1930/C5171, maybe one might argue that OnSemi drivers would be a preferable choice.
Quel dommage.


Hi, i look foward on SA1930 (that i can pick) DS but SOA indicated that at 40V Vce Ic will only be 180mA. With The two pair Output BJT HFE=40 it provide 7.2 Apk according Andrew method. It's wide 25~30Apk needed to drive 4 ohms speakers, No?

regards Marc.
 
Re: its been

sakis said:
way too many times that in the forum is the question
"""" how many pairs ber board """
"""""how much rail voltage""""

no mater what the forum people say i have to admit that i see almost every day that japanese people might have another opinion

just come in to my shop for repair/remake a NIKKO ALPHA II wich means only 2 transistors from NEC 2sb600 2SD555 per board and the astronomical rail voltage of
63 VOLTS

any comments ????

( amplifier belongs to a costumer and it come in for recaping and upgrading up to a possible margin ....though i will try to borrow it and try to push it to the really hard limits to see how its going to be .... )

Good adjust all parts in a amp is not as easy as it seems for people with the whole kwnoledge. That why i like so Andew method explanation. I can yet evaluate how good is a couple of transistor for my use. Yet the goal is to find the best candidate regarding SOA/use and how fast it is...

Marc
 
Idefixes said:
It's wide 25~30Apk needed to drive 4 ohms speakers, No?

For 30A peak, 2 parallel output devices will be the major problem, not the drivers.

Arithmatic example :
Your Troelstra loudspeakers do a minimum of 4 Ohms at 600 Hz.
600 Hz means each phase lasts something like 8.33ms, which is a bit shorter than 10ms.
According to the Sanken datasheet, the output devices handle some 17A for a 10ms square signal burst.
Derated by 50% for a raised die temperature : 8.5A max for each output device.
2 output devices will do 17A, for the magical 30A number you'll need 4 power devices in parallel.

Driver :
40Vdc divided by 4 Ohm is 10A, suppose you desire twice as much, 20A peak.
For 2 parallel output devices, each one has to deliver 10A. (roughly half of the max number for 10ms)
The 2SC2922 has a Beta of 60 for 10A out, the driver has to deliver 166.7mA for each power device, times two is 1/3A.
At 40Vce and 10ms, the 2SA1837 does some 650mA Max, temperature derated by 50% again boils down to 325mA.
If the drivers are kept cool enough they'll handle the 1/3A without breaking down.

To be safe, a power amp with 40Vdc rails and 2 pairs of output devices will need some protection measures, even more so with a 660VA and 4x68kuF powersupply.
As in V/I-limiters, temperature and overcurrent protection, stuff like UPC1237ha or TA7317P and a solid output relay.
The reason why Nikko amps still keep going after +25 years.

Did Sakis just call my favorite wholesale store STUPID ?
It's an outrage. :clown:
 

Attachments

  • metro-makro.jpg
    metro-makro.jpg
    59.4 KB · Views: 1,447
  • Like
Reactions: NanoFarad
jacco vermeulen said:


For 30A peak, 2 parallel output devices will be the major problem, not the drivers.

Arithmatic example :
Your Troelstra loudspeakers do a minimum of 4 Ohms at 600 Hz.
600 Hz means each phase lasts something like 8.33ms, which is a bit shorter than 10ms.
According to the Sanken datasheet, the output devices handle some 17A for a 10ms square signal burst.
Derated by 50% for a raised die temperature : 8.5A max for each output device.
2 output devices will do 17A, for the magical 30A number you'll need 4 power devices in parallel.

Driver :
40Vdc divided by 4 Ohm is 10A, suppose you desire twice as much, 20A peak.
For 2 parallel output devices, each one has to deliver 10A. (roughly half of the max number for 10ms)
The 2SC2922 has a Beta of 60 for 10A out, the driver has to deliver 166.7mA for each power device, times two is 1/3A.
At 40Vce and 10ms, the 2SA1837 does some 650mA Max, temperature derated by 50% again boils down to 325mA.
If the drivers are kept cool enough they'll handle the 1/3A without breaking down.

To be safe, a power amp with 40Vdc rails and 2 pairs of output devices will need some protection measures, even more so with a 660VA and 4x68kuF powersupply.
As in V/I-limiters, temperature and overcurrent protection, stuff like UPC1237ha or TA7317P and a solid output relay.
The reason why Nikko amps still keep going after +25 years.

Did Sakis just call my favorite wholesale store STUPID ?
It's an outrage. :clown:

Thanks for explanation i will study that. Of course i planning to mount a speaker DC protection systeme (may be K4700 from velleman or a diy one)
 
well jacco !!!!!

As in V/I-limiters, temperature and overcurrent protection, stuff like UPC1237ha or TA7317P and a solid output relay.
The reason why Nikko amps still keep going after +25 years.
Did Sakis just call my favorite wholesale store STUPID ?

:clown: [/B][/QUOTE]


he he he actually not ..... on the other hand i may have to make them a big THANK YOU letter since i ve made so much money from the audio they sell in our country

total craaaaa******pppppp

but that is ok .... if we make money
 
All these discussions are quite helpfull to take decisions. My goal is really to build an amp that is able to drive 4ohms speakers. So i redirect my work. I will let 2SC2922 by side due the MT200 size. I will use the On semi MJL3281A i have or 2SC5200 in range of 8 transistor per board. I redrawn schema and board. Drive will be for moment MJE15030/15031 as i don't identify any other faster driver that will be suit to the job.

nbip_410.jpg


nbip_411.jpg


Marc
 
Status
Not open for further replies.