If this is such a 'great way to start', why then did you have to make the half-dozens or so mods to it. If I read this thread correctly, it will barely suffice and possibly blow up out of the box. What kind of amp solution is that!? That's the problem. Everyone has a different take on what makes a good amplifier operate with certain built-in precautions, etc. etc. I want my amp to operate safely and within spec, regardless of the speakers I have or how the long the cables may be. That's not too much to ask is it? I know this is DIY and I take to the level I feel comfortable at. I like a steak on the grill too, but I don't redesign the grill!
High current Outputs
I'm considering the mx50 se kit and have a couple of transformer pwr. supply options. One is 28-0-28 AC (38VDC =-) and another transformer at 45-0-45 AC. I also have a large heatsink with complimentary pairs of MJ15003 / MJ15004 T-03 outputs. I'm thinking of using these outputs driven by the MX50 se's due to their high output capabilities ( IC 20A and 250W PD). Plus I have them and they're already mounted. My concern is that these drive boards do not have adjustable bias. Will these outputs be an match adequate. I suppose all I can due is try them and check for crossover notch on the scope.
I'm considering the mx50 se kit and have a couple of transformer pwr. supply options. One is 28-0-28 AC (38VDC =-) and another transformer at 45-0-45 AC. I also have a large heatsink with complimentary pairs of MJ15003 / MJ15004 T-03 outputs. I'm thinking of using these outputs driven by the MX50 se's due to their high output capabilities ( IC 20A and 250W PD). Plus I have them and they're already mounted. My concern is that these drive boards do not have adjustable bias. Will these outputs be an match adequate. I suppose all I can due is try them and check for crossover notch on the scope.
redjr
You MUST use the amp within its limits. You must use ANY amp within its limits. It has no short protection, no I/V limiting, no input clamps, etc. But it will sound better than if it did. If you know you have a well behaved load above 6 Ohms at all times, you are good for 40W or so. Cleaner at 20. Outside those limits, the distortion will skyrocket. Just a fact of a single output amplifier. Notice my harsh criticism is for the OP who wants to use 4 Ohm speakers, or the 2.5 Ohm car speakers. This is not a fault of the amp, it is a bad system design choice. This amp has a single output pair, that for me, means it is a 40W max ( 8 ohm) amp. Blowing up or not is not my only limit criteria. It may be yours. Yea, I can tweak it to make it better. I bet LJM could make it better than I can next time around. It might cost more though.
It is a "classic" example of how to build a small amp. Nothing fancy. No bad mistakes. ( looking at many amplifier designs, I can list many many flat out mistakes.) No output mute/DC protection? External. ( ESP has a nice one. Costs more than the MX50). No output RL network? That should be mounted right at the terminals not on the board. Your problem. He sold us AMPLIFIER BOARDS. Not an amplifier. We have to do the rest of the engineering ourselves.
a109
Put in a POT to adjust the bias of course! If you fear pots long term ( many do), use them to set it up and then put in a fixed resistor. A CFP output is not too terribly sensitive as long as it is not too lean ( and pots are expensive, like as much as the rest of the parts combined) so they were left out. If changing the outputs, you WILL need to adjust the bias. At least with a CFP the thermal tracking is on the drivers, so that should not hurt you. I do not know the needs of the bigger outputs if you want to increase the VAS current a little. If you don't know how to do this, then please do a little reading on amplifier design basics so we can lead you through it.
One more time. There is no perfect amp. There is no universal amp. There are only amps that will satisfy the constraints they were designed to live within. When you think you know better than the laws of physics, you will find out the hard way. Mother nature always wins.
You MUST use the amp within its limits. You must use ANY amp within its limits. It has no short protection, no I/V limiting, no input clamps, etc. But it will sound better than if it did. If you know you have a well behaved load above 6 Ohms at all times, you are good for 40W or so. Cleaner at 20. Outside those limits, the distortion will skyrocket. Just a fact of a single output amplifier. Notice my harsh criticism is for the OP who wants to use 4 Ohm speakers, or the 2.5 Ohm car speakers. This is not a fault of the amp, it is a bad system design choice. This amp has a single output pair, that for me, means it is a 40W max ( 8 ohm) amp. Blowing up or not is not my only limit criteria. It may be yours. Yea, I can tweak it to make it better. I bet LJM could make it better than I can next time around. It might cost more though.
It is a "classic" example of how to build a small amp. Nothing fancy. No bad mistakes. ( looking at many amplifier designs, I can list many many flat out mistakes.) No output mute/DC protection? External. ( ESP has a nice one. Costs more than the MX50). No output RL network? That should be mounted right at the terminals not on the board. Your problem. He sold us AMPLIFIER BOARDS. Not an amplifier. We have to do the rest of the engineering ourselves.
a109
Put in a POT to adjust the bias of course! If you fear pots long term ( many do), use them to set it up and then put in a fixed resistor. A CFP output is not too terribly sensitive as long as it is not too lean ( and pots are expensive, like as much as the rest of the parts combined) so they were left out. If changing the outputs, you WILL need to adjust the bias. At least with a CFP the thermal tracking is on the drivers, so that should not hurt you. I do not know the needs of the bigger outputs if you want to increase the VAS current a little. If you don't know how to do this, then please do a little reading on amplifier design basics so we can lead you through it.
One more time. There is no perfect amp. There is no universal amp. There are only amps that will satisfy the constraints they were designed to live within. When you think you know better than the laws of physics, you will find out the hard way. Mother nature always wins.
I'm not bemoaning this little amp board per se, just the general frustration of which amp design to try next. It just seems to me that a design could be made that incorporates most, if not all, of the standard 'features' of a reasonable amp for a reasonable cost. I know there are limits at both ends, but that's not what concerns me. I just would like to know which of all the amp modules available (from the TI TPA3116D2 class D chip amp at $14.95, to the LM3886, to the LM4702 (which I built awhile back), to the powerful Honey Badger featured on DIY, or even the more expensive Hypex NC400 modules @ $400/channel offer the 'best' or a 'sound' design to start with. No pun intended. I have no interest (well maybe a little
) in tweaking whatever I buy to death, because I'm reasonably certain I won't be able to tell the difference! And, for me, cost is not that big a factor(within limits of course), if I know from the get-go I'm getting a solid design, that will serve my general listening habits. I want to be reasonably confident it won't go up in smoke if I accidentally trip and pull out my speakers wires! I love the thrill of planning, buying, assembling and executing and then a little gloating at pulling it all together in a nice sounding, and looking box.Hi everyone
Yesterday , had a day off work , instead of drinking + BBQ I spent my time in hot weather to assemble the board , I bought the one with complete PSU + Speaker delay ( Protected ..not test ) + sanken output devices fused, filtered caps .. complete in one single board , the assembly process went smooth as the kit is well prepare down to screws fuses ..etc , took me around 3 hours to complete , it use 2SC5171 + 2SA1930 instead of 669/649, power it up no problem the output transistor run so cool that I think it is not working , connect to a pair of speakers 4 ohms and play all night , there is no hum or noise, and the sound is pretty good for the price paid , I think the price is hard to beat if someone want a budget amp to play with , things I have to change is I/P cap that use a tiny 10uF/16V samyong to a better quality one , the bias is so low that it barely get warm , DC offset is 6mV for both channels relay delay is around 3 seconds , I match HFE of the BC556B and got two pairs out of 4 pairs supply , I think I can put it in a box without heatsink , I can't comment much as I am not in electronic ( mechanical ) .
Thanks
Yesterday , had a day off work , instead of drinking + BBQ I spent my time in hot weather to assemble the board , I bought the one with complete PSU + Speaker delay ( Protected ..not test ) + sanken output devices fused, filtered caps .. complete in one single board , the assembly process went smooth as the kit is well prepare down to screws fuses ..etc , took me around 3 hours to complete , it use 2SC5171 + 2SA1930 instead of 669/649, power it up no problem the output transistor run so cool that I think it is not working , connect to a pair of speakers 4 ohms and play all night , there is no hum or noise, and the sound is pretty good for the price paid
, I think the price is hard to beat if someone want a budget amp to play with , things I have to change is I/P cap that use a tiny 10uF/16V samyong to a better quality one , the bias is so low that it barely get warm , DC offset is 6mV for both channels relay delay is around 3 seconds , I match HFE of the BC556B and got two pairs out of 4 pairs supply , I think I can put it in a box without heatsink , I can't comment much as I am not in electronic ( mechanical ) .Thanks
@ +/2 20Vdc: about 16 watts in 8 ohms.. that is about 46 watts in 2.7 ohms, provided rails-voltage does not drop.20VDC
The MX50 is very much a reasonable amp. So is a gain clone. Either should serve one fine as long as what you ask them to do is reasonable. ( Ironic, I call my first scratch design "The Reasonable")
I am not too impressed with any class D amp yet. OK for subs and cars, but they have a way to go.
20V rails would be excellent.
I am not too impressed with any class D amp yet. OK for subs and cars, but they have a way to go.
20V rails would be excellent.
Hi,
for those, who moan about too low power or too small SOA, have a look at LJM's L12-2 amp module and associated thread here at DIY.
You'll notice, that it doubles the output transistors, while Input stage and Vas are similar.
And the price tag is equally tempting
In general I'm fully with tvrgeek.
With those modules You get a bit more than what You've paid for.
but they are just amp modules, no complete amp.
If You wish for protection subassemblies and other stuff, then get Yourself some.
If You want to drive low impedance speakers at high levels, go get Yourself a way mor powerful amp.
Respect the design limits and all will be fine.
Disrespect them if You wish, but don't complain after.
jauu
Calvin
for those, who moan about too low power or too small SOA, have a look at LJM's L12-2 amp module and associated thread here at DIY.
You'll notice, that it doubles the output transistors, while Input stage and Vas are similar.
And the price tag is equally tempting
In general I'm fully with tvrgeek.
With those modules You get a bit more than what You've paid for.
but they are just amp modules, no complete amp.
If You wish for protection subassemblies and other stuff, then get Yourself some.
If You want to drive low impedance speakers at high levels, go get Yourself a way mor powerful amp.
Respect the design limits and all will be fine.
Disrespect them if You wish, but don't complain after.
jauu
Calvin
You need to do the thermal resistance calculations. Self, Cordel and other places will teach you this. Heat sink selection is not a pick and choose, it is an engineering problem that must be worked out. Dimensions of the heat sinks do not tell you their ability do do their job. Heat sinks have ratings.
Two ways to work it out. Correctly mathematically, or the "obviously big enough" way.
IMHO, if that heat sink is per side, it is probably sufficient using the second method.
Finaly I can have a go too!!
Some time ago I received the kit.
Looks good and complete. I must say the original LJM design is a bit different from the schematics on the internet... No problem! As long as TLSpice exists, I can solve anything
While (slowly) building the amps, I also do some simulating...
Using the topology on the board (and tweaking it) I got THD down to under 0.005% . I also noticed that @ 10 Hz THD rises to about 0.01%. Still trying to figure out what causes that phenomena... The FFT shows some garbage at 1Khz and @ higher harmonics..
Anyway: I started optimization not at 1Khz, but at 10Khz/20Khz, since THD in most amps rises at higher frequencies.
One of my tweaks: lowering the 100 Ohm resistors further to ca. 15 ohms (Tvrgeek uses 22.1). Leaving out the (original) 2k2 resistor, doesn't seem too optimal to me... about 600 Ohms appears to be a better value (some tenth-thousands % less THD).
I'll post the final schematic/values after I finished tweaking.
Some time ago I received the kit.
Looks good and complete. I must say the original LJM design is a bit different from the schematics on the internet... No problem! As long as TLSpice exists, I can solve anything
While (slowly) building the amps, I also do some simulating...
Using the topology on the board (and tweaking it) I got THD down to under 0.005%
. I also noticed that @ 10 Hz THD rises to about 0.01%. Still trying to figure out what causes that phenomena... The FFT shows some garbage at 1Khz and @ higher harmonics.. Anyway: I started optimization not at 1Khz, but at 10Khz/20Khz, since THD in most amps rises at higher frequencies.
One of my tweaks: lowering the 100 Ohm resistors further to ca. 15 ohms (Tvrgeek uses 22.1). Leaving out the (original) 2k2 resistor, doesn't seem too optimal to me... about 600 Ohms appears to be a better value (some tenth-thousands % less THD).
I'll post the final schematic/values after I finished tweaking.
Can somebody tell me if there is something to gain in heat-coupling the transistors in the input-stage?
Currents are low, so I guess there will not be too much difference in chip-temperatures of the transistors. In several other designs with comparable architectures coupled transistors are used, though...
Currents are low, so I guess there will not be too much difference in chip-temperatures of the transistors. In several other designs with comparable architectures coupled transistors are used, though...
I noticed, while simulating, another problem: frequency related DC component on the output. It's a couple of mV, but it messes-up the FFT diagram. I found a connection with the 470uF capacitor. Increasing that value gives better results. What causes this phenomena?