high power
For design of high power aplifiers is good to understand SOA (safe operation area) of transitors - serial connection looks better than simply make more pair of end transistors... look to www.aldax.cz - module SPD800 with +/-95V gives 800W/4ohm.
It is nice module cost about 100 dollars.
For better understanding what i mean with serial connection see this shematic: http://margo.student.utwente.nl/el/ampz/pics/dpa880b.gif
The origin of this schematic was made because in 80' there where no transistors with Uce bigger than 80V in east europe countries driven by Soviet union. But still this have some advantages about SOA when you need to use more then +/-100V source...
For design of high power aplifiers is good to understand SOA (safe operation area) of transitors - serial connection looks better than simply make more pair of end transistors... look to www.aldax.cz - module SPD800 with +/-95V gives 800W/4ohm.
It is nice module cost about 100 dollars.
For better understanding what i mean with serial connection see this shematic: http://margo.student.utwente.nl/el/ampz/pics/dpa880b.gif
The origin of this schematic was made because in 80' there where no transistors with Uce bigger than 80V in east europe countries driven by Soviet union. But still this have some advantages about SOA when you need to use more then +/-100V source...
Ahead of time
Seems like SAE and GAS were ahead of time with series output connections. (Over 30 years ago!)
http://www.audio-circuit.dk/Schematics/GAS_Ampzilla.pdf
http://www.audio-circuit.dk/Schematics/Dynaco_ST400.pdf
Seems like SAE and GAS were ahead of time with series output connections. (Over 30 years ago!)
http://www.audio-circuit.dk/Schematics/GAS_Ampzilla.pdf
http://www.audio-circuit.dk/Schematics/Dynaco_ST400.pdf
.
AndrewT: I do now my own project - partly based on this schematic http://www.stavebnice.net/dps/a338sch.pdf with 16 pcs of 2SA1943/SC5200 in serial connection in B class.
AndrewT: I do now my own project - partly based on this schematic http://www.stavebnice.net/dps/a338sch.pdf with 16 pcs of 2SA1943/SC5200 in serial connection in B class.
Swagboy,
I would go for full bridge, it's not complicated and you can do it with 63 V filter electrolytes, using series connection means you must use higher voltage for capacitors and cost goes up considerably for 100 V lytes compared to 63 V lytes, moreover I think it's more difficult to implement it and maybe not for beginners (but that's perhaps not the case for you..) not to forget to mention the risk working with higher voltages!
Using a bridge topology is also better in that respect the output tranies SOA is better for lower voltages, at higher voltages the current durability start to fall steeply which may again ad up to the cost (this is not the case with series connection of course where OPT's share the voltage fall over each other).
Just my 2 c's.
Cheers Michael
I would go for full bridge, it's not complicated and you can do it with 63 V filter electrolytes, using series connection means you must use higher voltage for capacitors and cost goes up considerably for 100 V lytes compared to 63 V lytes, moreover I think it's more difficult to implement it and maybe not for beginners (but that's perhaps not the case for you..) not to forget to mention the risk working with higher voltages!
Using a bridge topology is also better in that respect the output tranies SOA is better for lower voltages, at higher voltages the current durability start to fall steeply which may again ad up to the cost (this is not the case with series connection of course where OPT's share the voltage fall over each other).
Just my 2 c's.
Cheers Michael
All these series connected output devices in Class-AB amp, do nothing in terms of increasing efficieny but they only facilitate the low VCE rated devices to be used with high voltage rails....
Whereas as series connected devices in Class-G , really provide a leap in efficiency and increase the output device robustness when driving reactive loads....
Class-H provides one step further along with reduced costs and increased efficiency...
K a n w a r
Whereas as series connected devices in Class-G , really provide a leap in efficiency and increase the output device robustness when driving reactive loads....
Class-H provides one step further along with reduced costs and increased efficiency...
K a n w a r
Hey guys,
Thanks for all the help thus far!
So I have another question for my 600Watt stereo amp. . .
If I went with a two tier class-G amp (I think I could build one of those) I have
no idea whatsoever on how to calculate the necessary rail voltages, the current
per rail needs, or the VA of the transformer/s.
Thanks to Randy Slone, I can figure out a class B supply, but do not know how
to design a class G supply.
I searched the forum, and have not found an answer to this question.
I am thinking about building a test amp first with basic incorporation of class-G,
that way if I blow it up I won't be AS ticked. . .
Thanks!
Thanks for all the help thus far!
So I have another question for my 600Watt stereo amp. . .
If I went with a two tier class-G amp (I think I could build one of those) I have
no idea whatsoever on how to calculate the necessary rail voltages, the current
per rail needs, or the VA of the transformer/s.
Thanks to Randy Slone, I can figure out a class B supply, but do not know how
to design a class G supply.
I searched the forum, and have not found an answer to this question.
I am thinking about building a test amp first with basic incorporation of class-G,
that way if I blow it up I won't be AS ticked. . .
Thanks!
A simple output stage (TIFF viewer required).
http://patft.uspto.gov/netacgi/nph-...,387,876.WKU.&OS=PN/5,387,876&RS=PN/5,387,876
If you like, I have some 80-40-0-40-80 toroids rated at 2KVA 25% duty cycle that will give you about 600W/8R with this output stage. I also have some heatsink tunnels drilled for the TO-3 outputs.
You can buy an empty Borberly fully complementary diff input driver board to run this with from AudioXpress.com for about $12
http://www.audioxpress.com/bksprods/pcbs/solidstate.htm
ITEM # PCBP-3 Borbely 60W Mosfet Amplifier
From Audio Amateur Issue 2/82. Borbely EB-60 60W single-channel Mosfet amplifier. 3-3/8" x 6-1/8". 1 lb. $11.75
It's very similar to the Leach, and has provisions for a Vbe multiplier for bias.
If you want something simpler, the $6 Pass board will also work:
http://www.audioxpress.com/bksprods/pcbs/nelsonpass.htm
ITEM # PCBJ-5 Pass A40 Power Amplifier
From Audio Amateur Issue 4/78. Pass A-40 single-channel power amplifier. 3" x 3". 1 lb. $6.00
The Crest 7000 runs on this type of circuit (Peavey now owns Crest), and uses the same kind of rail voltage (±114V, ±57V) with a total of 80,000µF, or 4 pcs 10,000µF/63V per channel. The 6000 uses 120,000µF total.
Surplus 9000 toroids are available for $150 each, but the 9000 puts out 1200W/8R.
http://patft.uspto.gov/netacgi/nph-...,387,876.WKU.&OS=PN/5,387,876&RS=PN/5,387,876
If you like, I have some 80-40-0-40-80 toroids rated at 2KVA 25% duty cycle that will give you about 600W/8R with this output stage. I also have some heatsink tunnels drilled for the TO-3 outputs.
You can buy an empty Borberly fully complementary diff input driver board to run this with from AudioXpress.com for about $12
http://www.audioxpress.com/bksprods/pcbs/solidstate.htm
ITEM # PCBP-3 Borbely 60W Mosfet Amplifier
From Audio Amateur Issue 2/82. Borbely EB-60 60W single-channel Mosfet amplifier. 3-3/8" x 6-1/8". 1 lb. $11.75
It's very similar to the Leach, and has provisions for a Vbe multiplier for bias.
If you want something simpler, the $6 Pass board will also work:
http://www.audioxpress.com/bksprods/pcbs/nelsonpass.htm
ITEM # PCBJ-5 Pass A40 Power Amplifier
From Audio Amateur Issue 4/78. Pass A-40 single-channel power amplifier. 3" x 3". 1 lb. $6.00
The Crest 7000 runs on this type of circuit (Peavey now owns Crest), and uses the same kind of rail voltage (±114V, ±57V) with a total of 80,000µF, or 4 pcs 10,000µF/63V per channel. The 6000 uses 120,000µF total.
Surplus 9000 toroids are available for $150 each, but the 9000 puts out 1200W/8R.
Hi Swagboy,
The Class-G is not difficult, but its easy than Class-H
600W is very easy to achieve with just 4 pairs of MJL21196/95 complementary devices if you use them in Class-G topology....
As DJK said he has the necessary elements for your project, you could build the amp in very simple manner.....
I have build both Class-G [upto 1KW] & Class-H[2KW] amps....very highly efficient than class-AB when playing real music into real world speaker loads......
regards,
K a n w a r
The Class-G is not difficult, but its easy than Class-H
600W is very easy to achieve with just 4 pairs of MJL21196/95 complementary devices if you use them in Class-G topology....
As DJK said he has the necessary elements for your project, you could build the amp in very simple manner.....
I have build both Class-G [upto 1KW] & Class-H[2KW] amps....very highly efficient than class-AB when playing real music into real world speaker loads......
regards,
K a n w a r
Hi DJK,
The New Crest Pro9200 is much better and simple with 3250W@2ohms/CHannel...and its 2 Tier Class-H with rails +- 75V & +-150V and uses IGBT as Switchers along with SMPS to keep its weight low.....
With just 8 pairs of A1302/C3281 Transistors it achieves tremendous power rating due to Class-H....
K a n w a r
The New Crest Pro9200 is much better and simple with 3250W@2ohms/CHannel...and its 2 Tier Class-H with rails +- 75V & +-150V and uses IGBT as Switchers along with SMPS to keep its weight low.....
With just 8 pairs of A1302/C3281 Transistors it achieves tremendous power rating due to Class-H....
K a n w a r
"SMPS to keep its weight low....."
That's why the transformers are available surplus.
Actually, I'm note sure if there is a 0V rail switch in the 9000. AB International uses them. The max CE voltages will be a lot lower with a 0V rail, and the efficiency goes up of course. It's real easy to make a 0V rail too, no extra transformer windings, rectifiers, or filter caps.
That's why the transformers are available surplus.
Actually, I'm note sure if there is a 0V rail switch in the 9000. AB International uses them. The max CE voltages will be a lot lower with a 0V rail, and the efficiency goes up of course. It's real easy to make a 0V rail too, no extra transformer windings, rectifiers, or filter caps.
Hi DJK,
Yeah Zero voltage Rail.....Imagine the robustness when driving reactive loads.....Then how much current is going to pass through the non-conducting device with its VCE at Zero......hahahahaa which is forced to conduct....and then all these SOA VI limiters were simply become very easy with rail switchers although......
Zero Voltage rail cost Zero USD also!
Cheers
K a n w a r
Yeah Zero voltage Rail.....Imagine the robustness when driving reactive loads.....Then how much current is going to pass through the non-conducting device with its VCE at Zero......hahahahaa which is forced to conduct....and then all these SOA VI limiters were simply become very easy with rail switchers although......
Zero Voltage rail cost Zero USD also!
Cheers
K a n w a r
OK. . .
Andrew, Thanks for the vote of confidence. Building my own sound gear is my dream.
If we don't follow our dreams, we never will be truly happy.
Contra: "It does not do well to dwell on dreams and forget to live," (Professor Dumbledore).
So Workhorse:
1) can I run 600 watts with MJL21193/MJL21194's (4pairs) in a two tier class-G?
2) What would you recommend for pre-drivers, MJE340/MJE350's?
3) would you put them in series as Douglas Self, or parallel?
djk:
1) the crown Macrotech looks like it uses 1302/3281's for predrivers, is this your
assesment as well?
2) how much would you want for a transformer, and I don't suppose you know which
wires are which (for idiots like me. . .)?
Assessment for all (correct me if I am stupid. . . ):
Douglas Self's book demonstrates Class-G amplifiers with +/-15V and +/-50V rails.
That would make the lower voltage rail 30% smaller than the higher rails.
Taking this into account for my amplifier design . . .
V2 should be my desired top tier RMS voltage: 600 watts into 8 ohms, 69.28 volts RMS
secondaries (P=Vsquared/R)
V1 should be 20.784 volts RMS secondaries, 54 watts into 8 ohms. . .
Does this sound about right?
Thanks!
Andrew, Thanks for the vote of confidence. Building my own sound gear is my dream.
If we don't follow our dreams, we never will be truly happy.
Contra: "It does not do well to dwell on dreams and forget to live," (Professor Dumbledore).
So Workhorse:
1) can I run 600 watts with MJL21193/MJL21194's (4pairs) in a two tier class-G?
2) What would you recommend for pre-drivers, MJE340/MJE350's?
3) would you put them in series as Douglas Self, or parallel?
djk:
1) the crown Macrotech looks like it uses 1302/3281's for predrivers, is this your
assesment as well?
2) how much would you want for a transformer, and I don't suppose you know which
wires are which (for idiots like me. . .)?
Assessment for all (correct me if I am stupid. . . ):
Douglas Self's book demonstrates Class-G amplifiers with +/-15V and +/-50V rails.
That would make the lower voltage rail 30% smaller than the higher rails.
Taking this into account for my amplifier design . . .
V2 should be my desired top tier RMS voltage: 600 watts into 8 ohms, 69.28 volts RMS
secondaries (P=Vsquared/R)
V1 should be 20.784 volts RMS secondaries, 54 watts into 8 ohms. . .
Does this sound about right?
Thanks!
swagboy said:
Hi Boy,
1...Yes you can run 600W with 4 pairs of 21193/94 but only in Class-G/H
2... MJE15032 & MJE15033 as pre drivers
3...Series only....not in parallel
You could choose +-60VDC as First Tier and +-120VDC as Second Tier...for Proper 600W@8 Ohms output power requirement....Use 10000MFD/75V Caps for each tier as Minimum.....in Series....
Check this out for series supplies and output configuration....for Class-G
http://www.yorkville.com/downloads/servman/smap6020.pdf
regards,
K a n w a r
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.