"I'm not complaining, but I can't read the numbers on the Hadley schematic.. Can I get 150watts using those cheap transistors? WOW "
Its just a bridge amp, although I like it better than most.
If you need some values, go look at 'el-cheapo' at ESP. The Hadley is even less expensive to build as it has no VAS.
http://164.195.100.11/netacgi/nph-P...,376,899.WKU.&OS=PN/5,376,899&RS=PN/5,376,899
You need a TIFF viewer, click on 'images', go to figure 9 on page 8.
Its just a bridge amp, although I like it better than most.
If you need some values, go look at 'el-cheapo' at ESP. The Hadley is even less expensive to build as it has no VAS.
http://164.195.100.11/netacgi/nph-P...,376,899.WKU.&OS=PN/5,376,899&RS=PN/5,376,899
You need a TIFF viewer, click on 'images', go to figure 9 on page 8.
SkinnyBoy said:Basically I wan to make a 150watt amplifier using any of the following parts...
MOSFETS:
IRFZ44 MOSFET: 60V / 50A (N channel, TO-220 package), 0.028ohm on resistance: (IRFZ44) $3 each
IRF820 MOSFET: 500V / 2.5A (N channel, TO-220 package), 3.0ohm on resistance: (IRF820) $3 each
SGSP239 MOSFET: (TO-126 package): (SGSP239) $1 each
MTP3055E MOSFET: 60V / 12A (N channel, TO-220 package), 0.15ohm on resistance: (MTP3055) $2.50 each
MTP2955E MOSFET: 60V / 12A (P channel, TO-220 package), 0.3ohm on resistance: (MTP2955) $3 each
--------------------------------------------------------------------------------
TRANSISTORS:
(NEW) POWER TRANSISTORS 2N3055: New TO3 package power transistor with metal case. Large but limited quantity: (2N3055) $1.20 each or 10 for $8.00
C8050 NPN Transistors: (2W 1.5A Hfe 120 - 200): (C8050) 20 for $5
C8550 PNP Transistors: (2W 1.5A Hfe 120 - 200): (C8550) 20 for $5
AD161 / AD162 TRANSISTORS: Obsolete germanium power transistors. These were once commonly used in the output of class B audio amplifiers. Max current 1A, Vce is 20V. AD161 is NPN, AD162 is PNP: (AD1612) $3 per pair
Some are failry low powered, but I have no money, and using more parts helps with heat disapation, so, basically, whats a safe number of what parts to use for a 150watt (into 8ohm) amplifier?? I am happy to use 10 or more devices, it makes the amplifier look more powerfull anyway..lol.. I have searched for a few of these parts, but only found low power amps..
ALSO ALSO ALSO ALSO!!! I don't want to brigde incase one day I want to use 4ohm speakers....
You may want to look into what type of MOSFet devices these are. I know that IRFZ44 is a hexfet device which is very non-linier and would make a poor audio amplifier. These are designed for switching applications, power supplies and such. I'm not sure of the other types you listed.
You want to use lateral mosfets for audio amps.
You could realistically use 2n3055 & mj2955 (PNP) and make two seperate channels, 75W rms 180 degrees phase shift(use a small input audio transformer with a grounded CT.) design each circuit for 4 Ohms and bridge them accross 8Ohms.
I am afriad that the Hadley amplifier weblink doesn't work any more... and this:
http://qt.tn.tudelft.nl/publi/2002/RSI73Pieter.pdf
was the only Hadley amplifier I could find on the net....
http://qt.tn.tudelft.nl/publi/2002/RSI73Pieter.pdf
was the only Hadley amplifier I could find on the net....
"I am afriad that the Hadley amplifier weblink doesn't work any more... "
It looks like not many can link to it direct. Try it this way:
http://home.kimo.com.tw/skychutw/
Then click on 'circuits', then 'Hadley 622 pdf'
It looks like not many can link to it direct. Try it this way:
http://home.kimo.com.tw/skychutw/
Then click on 'circuits', then 'Hadley 622 pdf'
Yes, you can, i made it once.
And sounds reasonable..... you can hear..... but not too long.
I also heard germanium too... my second amplifier was made with AC187/188.... 2.5 Watts...hummm good bass.
And them i enter hy class... 10 Watts with AD161/162
and today, will not like anymore this sound.... we have better sound now..better designs....but i could hear without big problems... but not too long, and i had not the experience i have today, maybe, today, no possible to hear again.
But is great to see those ideas... wonderfull....
Carlos
And sounds reasonable..... you can hear..... but not too long.
I also heard germanium too... my second amplifier was made with AC187/188.... 2.5 Watts...hummm good bass.
And them i enter hy class... 10 Watts with AD161/162
and today, will not like anymore this sound.... we have better sound now..better designs....but i could hear without big problems... but not too long, and i had not the experience i have today, maybe, today, no possible to hear again.
But is great to see those ideas... wonderfull....
Carlos
Interesting to look at this discussion after Self's circuit for a power amplifier massively paralleling opamps! Just a couple of points to add here:
1. using many lower power transistors (in parallel, bridge-tied, etc... all with problems to solve like power sharing with thermal stability and so on) does have an advantage that can make it worthwhile... look at the hFE vs current curves for most power transistors - when you try to use most transistors at even half their maximum current the hFE is falling significantly; in some situations this might not be a big problem (but very often it is), and high-power transistor tend to have much lower hFEs at the best of times. Compared with that the AD161/162 combination has a high gain that is releatively constant with current and using many in parallel gives a more linear device than a single much higher powered device. As I say, it depends on the circumstances as to whether this a big problem or not, but most driver stages are high impedance (even in the Silicon Chip 200W MOSFET circuit I'd say the VAS/driver needs a better approach), so hFE is likely to be darned important.
2. There are many side issues with the examples given that need to be looked at in their own right, to avoid clouding the issue of paralleling many devices. Some of the options use old transistors that will have poor high frequency performance (anything like 2N3055 vintage, and unfortunately the germanium ones) that could rule them out (yet I know of beautiful-sounding 2N3055 guitar amplifiers... so it is going to depend on how they are used). And some MOSFET designs make great projects because they will take all manner of ill treatment, and within MOSFET options the cheaper devices like hexfets are not going to sound so great as laterals - where the complementary devices match better (hence lower THD) and the temperature coefficients are more suited to square-law designs like Ian Hegglun's... but they have limited current capability so really would benefit form using several in parallel.. yet doing that could spoil the square-law goal is source resistors have to be used. That said, I did come up with a (still untested) design that allows many transistors to be used in parallel without being super-matched (i.e. have some series resistors to ensure good current sharing), yet retains (and possibly enhances) square-law efficiency and distortion cancellation. I might publish it here if the co-designer of the circuit doesn't mind.
1. using many lower power transistors (in parallel, bridge-tied, etc... all with problems to solve like power sharing with thermal stability and so on) does have an advantage that can make it worthwhile... look at the hFE vs current curves for most power transistors - when you try to use most transistors at even half their maximum current the hFE is falling significantly; in some situations this might not be a big problem (but very often it is), and high-power transistor tend to have much lower hFEs at the best of times. Compared with that the AD161/162 combination has a high gain that is releatively constant with current and using many in parallel gives a more linear device than a single much higher powered device. As I say, it depends on the circumstances as to whether this a big problem or not, but most driver stages are high impedance (even in the Silicon Chip 200W MOSFET circuit I'd say the VAS/driver needs a better approach), so hFE is likely to be darned important.
2. There are many side issues with the examples given that need to be looked at in their own right, to avoid clouding the issue of paralleling many devices. Some of the options use old transistors that will have poor high frequency performance (anything like 2N3055 vintage, and unfortunately the germanium ones) that could rule them out (yet I know of beautiful-sounding 2N3055 guitar amplifiers... so it is going to depend on how they are used). And some MOSFET designs make great projects because they will take all manner of ill treatment, and within MOSFET options the cheaper devices like hexfets are not going to sound so great as laterals - where the complementary devices match better (hence lower THD) and the temperature coefficients are more suited to square-law designs like Ian Hegglun's... but they have limited current capability so really would benefit form using several in parallel.. yet doing that could spoil the square-law goal is source resistors have to be used. That said, I did come up with a (still untested) design that allows many transistors to be used in parallel without being super-matched (i.e. have some series resistors to ensure good current sharing), yet retains (and possibly enhances) square-law efficiency and distortion cancellation. I might publish it here if the co-designer of the circuit doesn't mind.
a 10pair mje15034/5 each biased to optimum ClassAB would give a high total bias for good ClassA output.
10*2*50 = 1000W of total output stage wattage.
Divide by 5 to give ~200W of maximum power capability.
Reduce this by 30% to 50% for the higher operating temperature of the high biasing current and you end up with a low power copy of a very famous Power amplifier. Is it Halcro that does similar?
10*2*50 = 1000W of total output stage wattage.
Divide by 5 to give ~200W of maximum power capability.
Reduce this by 30% to 50% for the higher operating temperature of the high biasing current and you end up with a low power copy of a very famous Power amplifier. Is it Halcro that does similar?
History repeats...
Yes... but notice how the idea of paralleling multiple low-power devices came to life in the relatively modern Elektor design using opamps. It is a good example of how "old" threads don't really become outdated. I bet there will be more designs for very good amplifiers using cheap-but-good devices (and have some of my own designs under way), and not just when somebody has a heap of devices sitting around without a use - look at the cost of some modern low-power high-gain good-linearity transistors when bought in quantity, compared with the price and quality of one big transistor they might replace. Heatsinking can be easier (since the heat is distributed over a large area). Matching devices becomes very easy too.
Mark
Yes... but notice how the idea of paralleling multiple low-power devices came to life in the relatively modern Elektor design using opamps. It is a good example of how "old" threads don't really become outdated. I bet there will be more designs for very good amplifiers using cheap-but-good devices (and have some of my own designs under way), and not just when somebody has a heap of devices sitting around without a use - look at the cost of some modern low-power high-gain good-linearity transistors when bought in quantity, compared with the price and quality of one big transistor they might replace. Heatsinking can be easier (since the heat is distributed over a large area). Matching devices becomes very easy too.
Mark
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.