Hi
Some pondering material.......Here is a little solid state amp I made up that might be of interest, works really well.😀 Detailed sound and very low noise floor. Using mostly scrap parts otherwise destined for the dump was the primary motivation. That and I haven't built anything lately and wanted to make sure I still have the skills.😛 I used an old busted car amp for the enclosure and heatsinking. Since it is only ~30W ave power per Ch the heatsink is plenty large enough. Figured it would be a good use for a couple of the many Radio Shack power transformers I have laying around.
Sorry for the bad drawings but I see no reason to take the time to draw the circuit out again on the computer. I just sketched out what I did.🙂
This amp has all the basic amenities such as clipping detection, DC detection/shut-off, over current protection/shut-off. If you short out the speaker wires while it is playing, the amp will go to shut-off/stand-by mode and not blow the outputs. You can short the output all you want.😀 The RGB LED indicates blue for stand-by, green for active, and will change from green to red for ~300mS for a clipping event. When the amp is in stand-by mode, the entire output stage circuits are de-energized. When the momentary switch is pressed, the LED turns green and the mosfet switches turn on the output stage. After about 1 second the relays will close if DC is not detected at the output.
There are two channels controlled by a fairly simple logic circuit. I know the circuit is not as typical as what is usually made around here, but how many of you go out of your way to design an amp to use TIP100/105 as outputs?
The output stage is a modified version of Dr Hawksford's amplified diode. This circuit was originally intended for use with Darlington power transistors and so is suited for these devices. The circuit does not seem to mind at all connecting rear 'surround' speakers to the outputs. This is done by simply connecting the plus of each surround speaker to the output and connecting the minus sides of the speaker together. IOW, they are connected between the two outputs in series, opposing phase.
In the schematic, all values are written so as the first two numbers is the value and the third is the multiplier. (101=100R; 102=1K; 103=10K; 104=100K; 105=1M and for capacitors, 101=100pf; 102=1nf; 103=10nf; 104=100nf; 105=1uf, ect)
Now I know some folks don't like J-fets, but I cannot see building a good amp without them, so useful they are.😎
Labeling is as.....
4117 J-fet is MMBF4117
4393 J-fet is PNBF4393
4091 J-fet is MMBF4091
595 J-fet is KSK595H
177 J-fet is MMBFJ177
5462 J-fet is MMBF5462
Then there are the typical BC850/860 and SOT-23 versions of 2N5087/5089 high hfe devices. SOT-23 versions of 2N1845/992 120Vceo devices. I used BS170 and VP2106 TO-92 mosfets for the driver stage.
In the logic circuit I pulled out a 556 timer and a 4747 flip-flop chip from the scrap pile of old PCBs and put them back to use.
So now it is on to designing my next power amplifier, not from scrap parts. It will be similar in circuit design, but it will be scaled up a bit and will use mosfet output stage.
Some pondering material.......Here is a little solid state amp I made up that might be of interest, works really well.😀 Detailed sound and very low noise floor. Using mostly scrap parts otherwise destined for the dump was the primary motivation. That and I haven't built anything lately and wanted to make sure I still have the skills.😛 I used an old busted car amp for the enclosure and heatsinking. Since it is only ~30W ave power per Ch the heatsink is plenty large enough. Figured it would be a good use for a couple of the many Radio Shack power transformers I have laying around.
Sorry for the bad drawings but I see no reason to take the time to draw the circuit out again on the computer. I just sketched out what I did.🙂
This amp has all the basic amenities such as clipping detection, DC detection/shut-off, over current protection/shut-off. If you short out the speaker wires while it is playing, the amp will go to shut-off/stand-by mode and not blow the outputs. You can short the output all you want.😀 The RGB LED indicates blue for stand-by, green for active, and will change from green to red for ~300mS for a clipping event. When the amp is in stand-by mode, the entire output stage circuits are de-energized. When the momentary switch is pressed, the LED turns green and the mosfet switches turn on the output stage. After about 1 second the relays will close if DC is not detected at the output.
There are two channels controlled by a fairly simple logic circuit. I know the circuit is not as typical as what is usually made around here, but how many of you go out of your way to design an amp to use TIP100/105 as outputs?
The output stage is a modified version of Dr Hawksford's amplified diode. This circuit was originally intended for use with Darlington power transistors and so is suited for these devices. The circuit does not seem to mind at all connecting rear 'surround' speakers to the outputs. This is done by simply connecting the plus of each surround speaker to the output and connecting the minus sides of the speaker together. IOW, they are connected between the two outputs in series, opposing phase. In the schematic, all values are written so as the first two numbers is the value and the third is the multiplier. (101=100R; 102=1K; 103=10K; 104=100K; 105=1M and for capacitors, 101=100pf; 102=1nf; 103=10nf; 104=100nf; 105=1uf, ect)
Now I know some folks don't like J-fets, but I cannot see building a good amp without them, so useful they are.😎
Labeling is as.....
4117 J-fet is MMBF4117
4393 J-fet is PNBF4393
4091 J-fet is MMBF4091
595 J-fet is KSK595H
177 J-fet is MMBFJ177
5462 J-fet is MMBF5462
Then there are the typical BC850/860 and SOT-23 versions of 2N5087/5089 high hfe devices. SOT-23 versions of 2N1845/992 120Vceo devices. I used BS170 and VP2106 TO-92 mosfets for the driver stage.
In the logic circuit I pulled out a 556 timer and a 4747 flip-flop chip from the scrap pile of old PCBs and put them back to use.
So now it is on to designing my next power amplifier, not from scrap parts. It will be similar in circuit design, but it will be scaled up a bit and will use mosfet output stage.
Attachments
Last edited:
I like your hand drawn schematics CBS, they lend a sense of painstaking care and an assuring knowledge of the design process that is missing with impersonal computer graphics and proof by simulation. I guess, since that's all I had to start with, it makes this all the more interesting, particularly the comprehensive protection scheme. It would be good to see that expanded to a separate project that could be fitted to other amplifiers.
Too many designs here use crude protection systems that were outdated and inadequate even when they were developed but the alternative of microcontrollers and programming is daunting for those who really only want to spend time on amplifiers. This looks very interesting and I hope your final amplifier development works out great too. 🙂
Too many designs here use crude protection systems that were outdated and inadequate even when they were developed but the alternative of microcontrollers and programming is daunting for those who really only want to spend time on amplifiers. This looks very interesting and I hope your final amplifier development works out great too. 🙂
Wow very impressive. It's rare you see design like this, I am impressed. It's just a shame it looks like that! Perhaps design a nice board for it?
That circuit is what it is, an experiment. However, I am working on another project, similar but larger.
although it takes a while to properly draw out a PCB with all the required details......
and then there must be proper review........
Still many hours to go......changes to make.😛
Attachments
Last edited:
So after all this time sitting not used, because I used the input pot for another project. I found another one and fired it up. I hate cheap ganged pots.😡 I opened it up and looked inside, and then hooked it up and ran a few tests.
You can see the discontinuity in the output stage transfer due to the cheap Darlington transistors. The slope of the error signal that is summed with the driver output into the base of the Darlington noticeably gets very steep around the crossover region of operation. Without the EC, the crossover distortion would be awful. At higher frequencies, these Darlingtons tend to move towards cross conduction and so the notch in the error signal where the slope increases is smaller and less pronounced than at 20Hz.
edit: I just noticed that the error signal is inverted in the photo, but it is in phase with the output. I also noticed that the error signal at 20Hz does not rise and fall equally. I suspect this could be due to the instantaneous heating effects on the Darlington device die, increasing the current gain.
You can see the discontinuity in the output stage transfer due to the cheap Darlington transistors. The slope of the error signal that is summed with the driver output into the base of the Darlington noticeably gets very steep around the crossover region of operation. Without the EC, the crossover distortion would be awful. At higher frequencies, these Darlingtons tend to move towards cross conduction and so the notch in the error signal where the slope increases is smaller and less pronounced than at 20Hz.
edit: I just noticed that the error signal is inverted in the photo, but it is in phase with the output. I also noticed that the error signal at 20Hz does not rise and fall equally. I suspect this could be due to the instantaneous heating effects on the Darlington device die, increasing the current gain.
Attachments
Last edited:
I forgot to add that the bias current in the TIP100/105 series Darlington output transistors is 10mA. In actuality, the the output device inside the Darlington package bias is only 5mA because the Vbe resistor of the output device is 120R so the driver inside the package bias is also 5mA. Probably a bit low for the output device, but it makes the idle power waste very small and certainly does demonstrate the effect of error correction drive.🙂
I do still have a bunch of TIP122/127 transistors but it is unfortunate that they just do not have enough Gm to drive a speaker.🙁
I do still have a bunch of TIP122/127 transistors but it is unfortunate that they just do not have enough Gm to drive a speaker.🙁
Peavey are fans of low output bias.
If the bias is sufficient to get rid or crossover distortion then why increase it ?
You are just wasting power in heat to the heat sinks.
If the bias is sufficient to get rid or crossover distortion then why increase it ?
You are just wasting power in heat to the heat sinks.
I agree. This is why I prefer to use nested feedback inside the global loop so that these distortions can be reduced to a minimum before returning to the input stage. The less "getting rid of" anything that the global loop has to do, the better, IMHO.
The output stage nested feedback essentially nullifies the ever changing effect of the speaker impedance upon the VAS. And in the case above, the VAS is a common base amplifier stage so it would like to see a constant impedance.
The output stage nested feedback essentially nullifies the ever changing effect of the speaker impedance upon the VAS. And in the case above, the VAS is a common base amplifier stage so it would like to see a constant impedance.
Last edited:
- Status
- Not open for further replies.