I agree on Carl's point 2)
there is no need for the buffer.
I'll go further there is no need for the input gain stage for normal Line Level Sources.
Make the gain stage bypassable, or apply it to a dedicated input that is reserved for low level source equipment.
The 22uF on the NFB leg is far too small.
The four 2u2F could be a little on the low side for deep and extended bass reproduction, but should be OK with normal restricted bandwidth speakers.
there is no need for the buffer.
I'll go further there is no need for the input gain stage for normal Line Level Sources.
Make the gain stage bypassable, or apply it to a dedicated input that is reserved for low level source equipment.
The 22uF on the NFB leg is far too small.
The four 2u2F could be a little on the low side for deep and extended bass reproduction, but should be OK with normal restricted bandwidth speakers.
About the preamp stages, if you use a volume pot it might be better to implement either a line stage OR a buffer stage, to isolate the pot from the 3886. But just one stage.
Don't get me wrong: I have used my amps, including a 3886 Penasa version, with no buffer between pot and amp, but if the stage is there as on this case, why not try it? But just one stage.
Even if I agree with you on the 22uF cap being a bit small, it is what National puts on their datasheet.
Also agree on the 2.2uF caps being small, but that I think was improved with 3.3uF our friend is using, isn't it?
What I do not agree on the power supply design is about the high current stage feeding the 3886 also feeding the regulator.
If the schematic is right, which I think it might not be, some things could be modified:
1) Simple mod. Add an extra cap, 220uF or what could fit there, between 22k and 470 resistors. Another cap, between 100uF and 220uF (also what could fit on copper side), after the 470 resistor and before entering the regulator. That should filter any fluctuations that might happen on larger current pumps.
2) Complicated mod. This is what I would have done. Add another diode bridge and capacitors to feed the regulators. The CRC filtering could be kept. To do this you would need a small separate board.
Carlos
Don't get me wrong: I have used my amps, including a 3886 Penasa version, with no buffer between pot and amp, but if the stage is there as on this case, why not try it? But just one stage.
Even if I agree with you on the 22uF cap being a bit small, it is what National puts on their datasheet.
Also agree on the 2.2uF caps being small, but that I think was improved with 3.3uF our friend is using, isn't it?
What I do not agree on the power supply design is about the high current stage feeding the 3886 also feeding the regulator.
If the schematic is right, which I think it might not be, some things could be modified:
1) Simple mod. Add an extra cap, 220uF or what could fit there, between 22k and 470 resistors. Another cap, between 100uF and 220uF (also what could fit on copper side), after the 470 resistor and before entering the regulator. That should filter any fluctuations that might happen on larger current pumps.
2) Complicated mod. This is what I would have done. Add another diode bridge and capacitors to feed the regulators. The CRC filtering could be kept. To do this you would need a small separate board.
Carlos
Thanks for your reply's and suggestions they are much appreciated. I don't really want mod the board right now its really nice as it is, even though they maybe some flaws in the design.
What do you think of these boards? It seems to be exactly the same as the LM3886 example circuit on the data sheet.
eBay - The UK's Online Marketplace
For the power supply I was think of this.
eBay - The UK's Online Marketplace
What do you think of these boards? It seems to be exactly the same as the LM3886 example circuit on the data sheet.
eBay - The UK's Online Marketplace
For the power supply I was think of this.
eBay - The UK's Online Marketplace
The boards seem fine. They are designed for discrete and better diodes on the supply.
Whether the circuit is the same as on the datasheet, it's quite likely, but not for sure. It's the one most people are using, recommended by the manufacturer, and this is a commercial product.
It's up to us, DIYers, to try to improve on things and see what happens.
Whether the circuit is the same as on the datasheet, it's quite likely, but not for sure. It's the one most people are using, recommended by the manufacturer, and this is a commercial product.
It's up to us, DIYers, to try to improve on things and see what happens.
I need some help with these LM3886 boards I was talking about in my last post, I know its going off topic a little as the thread is for the Yuanjing Gainclone.
I'm unsure which value of resistor to use for R1 (see the pic) On the data sheet the seller emailed it was expressed as 0.15. After seeing ready made boards on ebay they were using 0.1R 5w resistors.
Somebody on YouTube told me its 10R resistor with 10 turns of coil wire in parallel.
So I'm a little confused, I will also post the schematic here too but I'm 95% sure its the wrong one.
http://i317.photobucket.com/albums/mm398/haz_1000/LM38863150WAmplifierAmpboardPCBDIY.jpg
I'm unsure which value of resistor to use for R1 (see the pic) On the data sheet the seller emailed it was expressed as 0.15. After seeing ready made boards on ebay they were using 0.1R 5w resistors.
Somebody on YouTube told me its 10R resistor with 10 turns of coil wire in parallel.
So I'm a little confused, I will also post the schematic here too but I'm 95% sure its the wrong one.
http://i317.photobucket.com/albums/mm398/haz_1000/LM38863150WAmplifierAmpboardPCBDIY.jpg
Then it is a resistive buffer between amp output (after feedback) and the feed to the speaker cables. This is there to help prevent the reactive load affecting the amplifier feedback and thus changing stability margins with changes in frequency and reactance.
The resistor value is usually quite low, typically 0r1 to 0r2.
The resistor value is usually quite low, typically 0r1 to 0r2.
That R//L assembly is there to protect the amplifier, when highly capacitive loads are connected. If you look at the impedance curve of a capacitor, you see that the impedance becomes very low at high frequencies. An amplifier that reproduces such a high frequency, would effectively work into a short-circuit. To avoid that it gets damaged, you put a current limiting resistor in series, usually 10 Ohms.
When you leave it at that, you loose more than half of the amplifier power in that resistor with a typical 4 or 8 speaker connected. Now the impedance curve of the coil comes into play, which falls to very low values for low frequencies. Connected in parallel to the resistor, the coil shorts it out for all audio frequencies. Typical values are in the range of single-digit µH, and the absolute value is not critical, because either way its value is so small that the audible frequency range will be unaffected.
The resistor has no influence on the sound, because it is effectively shorted out for our hearing range.
The coil can have effects. People use air-core coils to avoid distortion from core saturation. Due to the small size of the coil, the copper resistance is negligible. That leaves us with the undesired effect of electro-magnetic fields. Such a coil can induce a signal into other parts of the circuit, so it is a good idea to place it in a way that keeps the effect small. E.g. place it far away from line level wires, traces and components. Put it so that the coil runs in parallel to the signal wires or traces, to keep the inductive coupling low. If you put it at a right angle, the coupling is high.
You may now wonder when you need those components. That is the case each time you cannot be sure, which load is connected, and also when you know that the load can be highly capacitive. E.g. in an active speaker with nothing but a short piece of wire in between the amplifier and an inductive speaker, there will be no need for the R//L circuit. For everything else you should provide it however, because there is always the possibility that an electrostatic speaker is connected to the amplifier or that very long runs of cables go from the amp to the speaker or that the speaker's cross-over poses a highly capacitive load or that your children play around with the components you have lying around, and they check what happens, when they stick the pins of those big capacitors into the speaker terminals.
When the designer put 0,1 Ohm in that place, he may have thought about providing a place for a load sharing resistor in case you want to connect amps in parallel.
When you leave it at that, you loose more than half of the amplifier power in that resistor with a typical 4 or 8 speaker connected. Now the impedance curve of the coil comes into play, which falls to very low values for low frequencies. Connected in parallel to the resistor, the coil shorts it out for all audio frequencies. Typical values are in the range of single-digit µH, and the absolute value is not critical, because either way its value is so small that the audible frequency range will be unaffected.
The resistor has no influence on the sound, because it is effectively shorted out for our hearing range.
The coil can have effects. People use air-core coils to avoid distortion from core saturation. Due to the small size of the coil, the copper resistance is negligible. That leaves us with the undesired effect of electro-magnetic fields. Such a coil can induce a signal into other parts of the circuit, so it is a good idea to place it in a way that keeps the effect small. E.g. place it far away from line level wires, traces and components. Put it so that the coil runs in parallel to the signal wires or traces, to keep the inductive coupling low. If you put it at a right angle, the coupling is high.
You may now wonder when you need those components. That is the case each time you cannot be sure, which load is connected, and also when you know that the load can be highly capacitive. E.g. in an active speaker with nothing but a short piece of wire in between the amplifier and an inductive speaker, there will be no need for the R//L circuit. For everything else you should provide it however, because there is always the possibility that an electrostatic speaker is connected to the amplifier or that very long runs of cables go from the amp to the speaker or that the speaker's cross-over poses a highly capacitive load or that your children play around with the components you have lying around, and they check what happens, when they stick the pins of those big capacitors into the speaker terminals.
When the designer put 0,1 Ohm in that place, he may have thought about providing a place for a load sharing resistor in case you want to connect amps in parallel.
Last edited:
Thanks for taking the time to post such detailed information, its much appreciated.
The amplifier will connect to the speakers with short cables, But I will add the coil in there anyway. Should the coil be made from enameled copper wire? I guess so otherwise the short coil would short out.
I have some enameled copper wire I could use form some old core inductors. Or even use some insulated solid core wire and make a coil from that?
Does the gauge of the wire used matter?, I herd you should use 1mm I have a capacitance and indutor meter coming soon so can experiment with different wire and the number of turns eg.
The amplifier will connect to the speakers with short cables, But I will add the coil in there anyway. Should the coil be made from enameled copper wire? I guess so otherwise the short coil would short out.
I have some enameled copper wire I could use form some old core inductors. Or even use some insulated solid core wire and make a coil from that?
Does the gauge of the wire used matter?, I herd you should use 1mm I have a capacitance and indutor meter coming soon so can experiment with different wire and the number of turns eg.
Enameled wire is the most practical, if you have it. You can also use insulated solid core wire, but the thicker insulation compared to enamel will increase the coil length and thus reduce the inductance. You will need more turns than with enameled wire.
The wire need not be very thick, because the effective copper resistance is small, and the resistor you wind the coil around will help with heatsinking. 1 mm is OK.
The wire need not be very thick, because the effective copper resistance is small, and the resistor you wind the coil around will help with heatsinking. 1 mm is OK.
yuanjing tda7293 amp 350W
Hi.
This seems to be the place to discuss / warn people about these cheap amps that you get on ebay. I wasn't using it as a hi fi amp, but I'm sure this will help people considering buying one of these.
My experience with the yuanjing 4 x tda7293 btl amp module
I was using it as a replacement power amp in a 70's Yamaha B100 bass amp. The B100 has a stunning discrete transistor preamp, great eq, beautiful. Unfortunately the triple bias diode in the power amp is very hard to obtain and very expensive if you can get it .. which is why I decided to give the tda7293 a go.
It's now used as the power stage in a peavey heritage.
The information with this product is very sketchy. One of the important things to note is that the you don't earth one side of the output. The output goes straight to the +ve and -ve termainals of the speakers.
Mine had a slight hum on it, acceptable for my purposes, but not hi-fi. It wasn't in a grounded enclosure, so that could have been improved upon, maybe.
The biggest problem with these is heat, I put mine on an ample 35 x 8 cm finned aluminium heatsink, I reckon @ 100W output, the sound starts cracking up in under 10 minutes - I can't see that you're ever going to get the rated 350W out of one, even with fans & a larger heatsink.
My impression of these amps is that they are a bit of an experiment - certainly not a tried and tested item from a reputable company, but for £25, what was I expecting? They do sound pretty good though and it works well enough at low volumes. They do have thermal shutdown and are forgiving to work with, so there is room for a bit of experimentation. If anyone else has had better luck with these, or if I can help them with their project, then please let me know.
Hi.
This seems to be the place to discuss / warn people about these cheap amps that you get on ebay. I wasn't using it as a hi fi amp, but I'm sure this will help people considering buying one of these.
My experience with the yuanjing 4 x tda7293 btl amp module
I was using it as a replacement power amp in a 70's Yamaha B100 bass amp. The B100 has a stunning discrete transistor preamp, great eq, beautiful. Unfortunately the triple bias diode in the power amp is very hard to obtain and very expensive if you can get it .. which is why I decided to give the tda7293 a go.
It's now used as the power stage in a peavey heritage.
The information with this product is very sketchy. One of the important things to note is that the you don't earth one side of the output. The output goes straight to the +ve and -ve termainals of the speakers.
Mine had a slight hum on it, acceptable for my purposes, but not hi-fi. It wasn't in a grounded enclosure, so that could have been improved upon, maybe.
The biggest problem with these is heat, I put mine on an ample 35 x 8 cm finned aluminium heatsink, I reckon @ 100W output, the sound starts cracking up in under 10 minutes - I can't see that you're ever going to get the rated 350W out of one, even with fans & a larger heatsink.
My impression of these amps is that they are a bit of an experiment - certainly not a tried and tested item from a reputable company, but for £25, what was I expecting? They do sound pretty good though and it works well enough at low volumes. They do have thermal shutdown and are forgiving to work with, so there is room for a bit of experimentation. If anyone else has had better luck with these, or if I can help them with their project, then please let me know.
Soldering Question
I am just starting on my empty Yuanjing board from Jim's Audio. I've been reading this thread and have found a lot of great mods from linuxguru and others. I'm very exited to get this thing up and running.
I have two soldering questions related to this board:
1) I see that it's a double sided board but can anyone confirm if the Jim's Audio board has plated thru holes? It looks like it but I'm not certain.
2) If they are plated thru holes, does it make a difference if I solder the component on the top side of the board (to shorten the trace)? I feel kind of silly asking this question because the trace would only be shortened by the thickness of the board but I am a novice and who knows...
I am just starting on my empty Yuanjing board from Jim's Audio. I've been reading this thread and have found a lot of great mods from linuxguru and others. I'm very exited to get this thing up and running.
I have two soldering questions related to this board:
1) I see that it's a double sided board but can anyone confirm if the Jim's Audio board has plated thru holes? It looks like it but I'm not certain.
2) If they are plated thru holes, does it make a difference if I solder the component on the top side of the board (to shorten the trace)? I feel kind of silly asking this question because the trace would only be shortened by the thickness of the board but I am a novice and who knows...
Solder on the bottom side of the PCB.
Due to the extra heat flow through to the top side, you need a bit more temperature. Either hold the iron on for longer or increase the bit temperature.
Problem !!!
Don't overheat the components.
Try heating the pad only and then roll the tip across the pad to the component leg. Takes a bit of practice but helps with getting the PCB up to temp before overheating the component. This particularly suits the oval pads and oversize pads. Small pads don't allow for much rolling into contact room.
Due to the extra heat flow through to the top side, you need a bit more temperature. Either hold the iron on for longer or increase the bit temperature.
Problem !!!
Don't overheat the components.
Try heating the pad only and then roll the tip across the pad to the component leg. Takes a bit of practice but helps with getting the PCB up to temp before overheating the component. This particularly suits the oval pads and oversize pads. Small pads don't allow for much rolling into contact room.
Last edited:
Thanks AndrewT.
I have another question - this time about the voltage rating for the smoothing caps. The schematic from Jim's Audio uses 10,000 uF caps rated at 63V. I already have some decent smoothing caps but they are only rated for 35V. I am using a 300VA transformer with 24-0-24 taps (Antek AS-3224). Since 35V>24V, can I use the 35V soothing caps?
I have another question - this time about the voltage rating for the smoothing caps. The schematic from Jim's Audio uses 10,000 uF caps rated at 63V. I already have some decent smoothing caps but they are only rated for 35V. I am using a 300VA transformer with 24-0-24 taps (Antek AS-3224). Since 35V>24V, can I use the 35V soothing caps?
A 24Vac transformer only gives out 24Vac when the mains input voltage is at rated voltage and the output current is at rated current.
At all other combinations of mains voltage and output currents the output voltage varies, a lot.
The output voltage is AC, you have to convert that to DC for the amplifier.
When you use a capacitor input filter (smoothing caps and rectifier) the Vdc~1.4*Vac
Expect anywhere between 30Vdc and 45Vdc from a 24Vac transformer.
You should buy 50V caps for a 24V transformer.
At all other combinations of mains voltage and output currents the output voltage varies, a lot.
The output voltage is AC, you have to convert that to DC for the amplifier.
When you use a capacitor input filter (smoothing caps and rectifier) the Vdc~1.4*Vac
Expect anywhere between 30Vdc and 45Vdc from a 24Vac transformer.
You should buy 50V caps for a 24V transformer.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.