continuing on with the power supply section build...
I forgot to say that putting the voltage regulators in at the end isn't a testing problem anymore due to the new JP18 and JP19 disconnects on the board. The rest of the board isn't hooked into the power supply section until those two are jumpered across. The arrrangement makes it simple test the power supply section (output voltage on one half of JP18 and JP19), and only then connect it to the rest of the board once everything checks out OK.
* There are no polarity to the MLCC ceramic capacitors, of course. The next photo shows the 0.1uf C22 and C23 in the power supply section installed, along with the 220pF MLCCs for the RF filter around the rear RCA jack (C1 & C2).
* The next two photos shows two of the "tombstoned" end-mount diodes in the power supply section. Note the "band up" or "band down" marking, just as in the CRC section. Double check to make sure the 4 diodes go in the right way. Here D3 has the band up, away from the hole, while D5 has the band down. Try to keep the "flying lead" on these parts fairly straight up and down, over the hole, so that it won't interfere with inserting the nearby regulator chips later on.
* The next photo shows the two Schottky diodes, D12 and D13, soldered in. The polarity of these two is opposite so make sure they go in the right way, with the bands oriented to match the PCB marking. Then finally the two 274R resistors near the RCA jack and the 12 0.5R resistors in the output section. The leads on all the 1/8W-sized resistors have to be bend fairly close to the resistor body. If needed pull the lead through from the bottom to get that last 1/8" or so down through the hole and the resistor to lie flat on the board.
* The next photo is a shot of all 12 output section 0.5R resistors in place. These are the resistors that I pack with the PC board, the ones Mouser and Digikey don't stock. Try to solder the resistors in as straight as possible, so they sit right over their symbol on the board. This helps insure the resistors won't interfere with inserting the bypass caps and output chips later on either side. After one resistor lead is soldered just grab the other lead, reheat the one joint, and you can move the resistor back and forth sideways as needed to get it straightened.
* The final two photos are the net result on the underside of the board with all the parts in this section soldered in. In the last photo you can see how JP18 and JP19 with their 4 holes work. One side of each, two holes, is attached to a power supply rail. The other half, the other two holes, goes out to the rest of the board. By jumpering across those two sections (either hole in one half to either hole in the other half) on JP18, for example, you connect the V+ positive power supply rail to the rest of the board. Then same with JP19 to connect the V- negative rail. Those jumpers will be the last things installed in the entire build, once the power supply voltage regulators are installed and all is tested.
All of these photos are also posted out at the project Google Drive link now.
I forgot to say that putting the voltage regulators in at the end isn't a testing problem anymore due to the new JP18 and JP19 disconnects on the board. The rest of the board isn't hooked into the power supply section until those two are jumpered across. The arrrangement makes it simple test the power supply section (output voltage on one half of JP18 and JP19), and only then connect it to the rest of the board once everything checks out OK.
* There are no polarity to the MLCC ceramic capacitors, of course. The next photo shows the 0.1uf C22 and C23 in the power supply section installed, along with the 220pF MLCCs for the RF filter around the rear RCA jack (C1 & C2).
* The next two photos shows two of the "tombstoned" end-mount diodes in the power supply section. Note the "band up" or "band down" marking, just as in the CRC section. Double check to make sure the 4 diodes go in the right way. Here D3 has the band up, away from the hole, while D5 has the band down. Try to keep the "flying lead" on these parts fairly straight up and down, over the hole, so that it won't interfere with inserting the nearby regulator chips later on.
* The next photo shows the two Schottky diodes, D12 and D13, soldered in. The polarity of these two is opposite so make sure they go in the right way, with the bands oriented to match the PCB marking. Then finally the two 274R resistors near the RCA jack and the 12 0.5R resistors in the output section. The leads on all the 1/8W-sized resistors have to be bend fairly close to the resistor body. If needed pull the lead through from the bottom to get that last 1/8" or so down through the hole and the resistor to lie flat on the board.
* The next photo is a shot of all 12 output section 0.5R resistors in place. These are the resistors that I pack with the PC board, the ones Mouser and Digikey don't stock. Try to solder the resistors in as straight as possible, so they sit right over their symbol on the board. This helps insure the resistors won't interfere with inserting the bypass caps and output chips later on either side. After one resistor lead is soldered just grab the other lead, reheat the one joint, and you can move the resistor back and forth sideways as needed to get it straightened.
* The final two photos are the net result on the underside of the board with all the parts in this section soldered in. In the last photo you can see how JP18 and JP19 with their 4 holes work. One side of each, two holes, is attached to a power supply rail. The other half, the other two holes, goes out to the rest of the board. By jumpering across those two sections (either hole in one half to either hole in the other half) on JP18, for example, you connect the V+ positive power supply rail to the rest of the board. Then same with JP19 to connect the V- negative rail. Those jumpers will be the last things installed in the entire build, once the power supply voltage regulators are installed and all is tested.
All of these photos are also posted out at the project Google Drive link now.
Attachments
Last edited:
Good things come to those who wait. 
With any luck I should be able to ship boards out the week after next. The LT3015's and LT1963A's arrive from Arrow this coming Thursday. I'll post the details in the vendor thread when the time comes.
I've ordered one rear panel for testing from front panel express, just to check my measurements. Still looking into lower cost panel options. Nothing has changed on the ODA front panel from the last ODA version so the last CAD file dimensions for the previous version still apply. I have those moved over to an FPE CAD file but need to go back and double check it.
With any luck I should be able to ship boards out the week after next. The LT3015's and LT1963A's arrive from Arrow this coming Thursday. I'll post the details in the vendor thread when the time comes.I've ordered one rear panel for testing from front panel express, just to check my measurements. Still looking into lower cost panel options. Nothing has changed on the ODA front panel from the last ODA version so the last CAD file dimensions for the previous version still apply. I have those moved over to an FPE CAD file but need to go back and double check it.
Clipping detect & gain stage sections build up
Next up is the clipping detect section since the parts (IC5 and the 2 blue LEDs) sit right in the middle of the board and are harder to reach later on. Then some of the gain stage stuff, the pot, rotary gain switch, 3.5mm input jack and input select switch, and several associated resistors and capacitors.
I'm saving the two gain stage LME49990 chips and surrounding parts for later. Less chance of static damage to the chips if other attached parts are installed first, like the gain switch and 3.5mm jack.
* The first photo shows the setup. Time to break out the anti-static mat and wrist strap. I've also connected that blue wire going to the ground holes on the board to ground, since the jaws on the vise I'm using are rubber and would insulate the board from the grounded metal vise. The whole goal of the antistatic stuff is just to get the hands, tools, parts, and board traces all at the same potential (voltage) so there is no potential difference across anything that would exceed part parameters.
* The next two photos show the sections involved on the top and bottom of the board. The red circles are the clipping indicator parts while the purple is gain-stage related. Not all that much on the bottom of the board this time, just the gain setting resistors under the gain switch, the current limit resistor for the clipping LED, and a couple of RF filter resistors under the 3.5mm jack.
The spacing around the gain setting resistors under the gain switch is fairly tight. In this particular case it is best to solder the rotary gain switch in first before those surface mount parts. That way you won't get any solder splashed into the small holes for the gain switch. Once the switch is in then solder in the gain resistors and the other board-bottom SMD parts in the circles.
You will see a few surface mount parts missing in the photos. I ran out of parts! Sending an order off to Mouser today. Just as with the power supply section the same value of resistor is used in two spots for all the gain resistors. So while you have that particular bag from Mouser open it is easiest just to solder both in, such as the 301 ohm resistors R7 and R11 .
* The next photo shows some special attention needed for R1 and R2, the resistors that connect ground to the input 3.5mm jack's switch. These are equivalent to the traces you cut on the O2 headamp if you want to use the holes next to the input jack for an offboard jack. With R1 and R2 in place the switch grounds the inputs when nothing is plugged in, just as with the O2 amp. With them removed the inputs are no longer grounded so you can use the JP2 & JP3 holes (ground, left channel, right channel) next to the 3.5mm jack for an off-board input jack, similar to P1 on the O2 headamp. The rear panel RCA jack is not grounded when that is selected by the input source selection switch on the ODA So there is something to keep in mind. If you have the input source switch set for the 3.5mm jack and R1 and R2 in place you will have grounded inputs and should have a perfectly silent output from the ODA. But if you switch to the rear RCA jacks they won't be grounded unless you insert grounding plugs or run RCA cables to a source component. So you may hear some background noise in that position until you connect your rear RCA jack to a source.
R1 and R2 are tiny 0603-sized SMD 0 ohm "jumper" resistors. You can solder those in, as I've done here, or you can just use a piece of wire soldered across the pads. Just tin the end of some small wire, solder to both pads in R1, and cut the excess off. Probably simpler and easier than actually soldering in a 0603 resistor.
* Next up is soldering the clipping detect chip, IC5, and the parts around it in the red circle, including the two blue LEDs that set the reference voltage for the window comparator. The first trick here is figuring out which end of IC5 is up! That gold bar marked on the top of the chip in the next photo is the "dot" or "top" end that goes toward the pot and gain switch side of the board.
* The next 2 photos show soldering the IC5 chip in. First the chip is placed over the pads in the first photo by holding it from the sides with the tweezers. Then the top right pin is soldered. If the alignment of the 14 pads still looks good after that joint then solder the bottom left corner as the second photo shows. If the alignment doesn't look good after the top right pin just reheat and move it around as needed before doing the bottom left corner. The finally solder the rest of the pins. I just dragged a small ball of solder on the iron tip around.
* The next photo shows all the pins soldered and some solder wick. Something interesting in this photo. Note that I solder bridged pins 8-9 and 12-13 on the right side. At the time I took this photo I didn't know that! Couldn't see it on the small camera screen or looking at the chip on the top through a magnifying glass. It wasn't until I soldered the LEDs in the next step and looked sideways across the chip that I discovered the bridges. Easy enough to remove the excess solder between the pins with that solder braid and touch the pins up again. But it goes to show the importance of looking at the pins from a few angles after soldering.
* The next two photos show soldering in the two LEDs, LED1 and LED2, and the current limit resistor betwen them R28. Do R28 first to get it out of the way. Notice those 3 little polarity dots near the bottom pad of each LED on the board. On the LED the polarity marking is extremely hard to see. It is a tiny amount of green color on the bottom solder tab, just before the LED lens. To see it I had to use a 6x magnifying glass held only about an inch away from the LED. There is an easier way. Look though the LED lens and you will see a small metal square right in the center where the LED mounts. Trace the lead inside the LED lens from that square and that is the LED tab that matches up with the dots on the board.
I'm going to list two alternernate zener diodes in the BOM to replace the LEDs in case anyone would rather not have the interior of their ODA lit blue! The zener diodes would mount with the opposite polarity to the LEDs, with their banded end away from the three dots on the board.
* Finally go ahead and solder in the diode logic for the comparator, D7 - D10. I have a picture of that in the next post.
* JP14 and JP15 in the last photo, near the LEDs, are good to keep in mind. Those are test points on the output of the LME49990 gain stage chips (left and right channel). If you want to look at the output of the gain stage with a scope, here is where to get it. These two points are also where you can connect the front panel pre-amp RCA jack, as one of the options, if you want that to hook into the output of the gain stage. If you want that pre-amp jack to run the rear panel RCA input straight through you can use JP1, 4, and 5 near the rear panel jack. If you want the pre-amp out to pass the 3.5mm jack through then use JP2 and JP3. If you want the pre-amp jack to pass through whatever the input select switch has chosen, use the JP7 holes on the board. And then with any of this you can run the wires right to the pre-amp out RCA jack, or you can run it through the pre-amp chip set up either for unity gain current buffer, or with some voltage gain. All sorts of options on using that pre-amp out RCA jack. The build instructions goes through this.
Next up is the clipping detect section since the parts (IC5 and the 2 blue LEDs) sit right in the middle of the board and are harder to reach later on. Then some of the gain stage stuff, the pot, rotary gain switch, 3.5mm input jack and input select switch, and several associated resistors and capacitors.
I'm saving the two gain stage LME49990 chips and surrounding parts for later. Less chance of static damage to the chips if other attached parts are installed first, like the gain switch and 3.5mm jack.
* The first photo shows the setup. Time to break out the anti-static mat and wrist strap. I've also connected that blue wire going to the ground holes on the board to ground, since the jaws on the vise I'm using are rubber and would insulate the board from the grounded metal vise. The whole goal of the antistatic stuff is just to get the hands, tools, parts, and board traces all at the same potential (voltage) so there is no potential difference across anything that would exceed part parameters.
* The next two photos show the sections involved on the top and bottom of the board. The red circles are the clipping indicator parts while the purple is gain-stage related. Not all that much on the bottom of the board this time, just the gain setting resistors under the gain switch, the current limit resistor for the clipping LED, and a couple of RF filter resistors under the 3.5mm jack.
The spacing around the gain setting resistors under the gain switch is fairly tight. In this particular case it is best to solder the rotary gain switch in first before those surface mount parts. That way you won't get any solder splashed into the small holes for the gain switch. Once the switch is in then solder in the gain resistors and the other board-bottom SMD parts in the circles.
You will see a few surface mount parts missing in the photos. I ran out of parts!
Sending an order off to Mouser today. Just as with the power supply section the same value of resistor is used in two spots for all the gain resistors. So while you have that particular bag from Mouser open it is easiest just to solder both in, such as the 301 ohm resistors R7 and R11 .* The next photo shows some special attention needed for R1 and R2, the resistors that connect ground to the input 3.5mm jack's switch. These are equivalent to the traces you cut on the O2 headamp if you want to use the holes next to the input jack for an offboard jack. With R1 and R2 in place the switch grounds the inputs when nothing is plugged in, just as with the O2 amp. With them removed the inputs are no longer grounded so you can use the JP2 & JP3 holes (ground, left channel, right channel) next to the 3.5mm jack for an off-board input jack, similar to P1 on the O2 headamp. The rear panel RCA jack is not grounded when that is selected by the input source selection switch on the ODA So there is something to keep in mind. If you have the input source switch set for the 3.5mm jack and R1 and R2 in place you will have grounded inputs and should have a perfectly silent output from the ODA. But if you switch to the rear RCA jacks they won't be grounded unless you insert grounding plugs or run RCA cables to a source component. So you may hear some background noise in that position until you connect your rear RCA jack to a source.
R1 and R2 are tiny 0603-sized SMD 0 ohm "jumper" resistors. You can solder those in, as I've done here, or you can just use a piece of wire soldered across the pads. Just tin the end of some small wire, solder to both pads in R1, and cut the excess off. Probably simpler and easier than actually soldering in a 0603 resistor.
* Next up is soldering the clipping detect chip, IC5, and the parts around it in the red circle, including the two blue LEDs that set the reference voltage for the window comparator. The first trick here is figuring out which end of IC5 is up! That gold bar marked on the top of the chip in the next photo is the "dot" or "top" end that goes toward the pot and gain switch side of the board.
* The next 2 photos show soldering the IC5 chip in. First the chip is placed over the pads in the first photo by holding it from the sides with the tweezers. Then the top right pin is soldered. If the alignment of the 14 pads still looks good after that joint then solder the bottom left corner as the second photo shows. If the alignment doesn't look good after the top right pin just reheat and move it around as needed before doing the bottom left corner. The finally solder the rest of the pins. I just dragged a small ball of solder on the iron tip around.
* The next photo shows all the pins soldered and some solder wick. Something interesting in this photo. Note that I solder bridged pins 8-9 and 12-13 on the right side. At the time I took this photo I didn't know that! Couldn't see it on the small camera screen or looking at the chip on the top through a magnifying glass. It wasn't until I soldered the LEDs in the next step and looked sideways across the chip that I discovered the bridges. Easy enough to remove the excess solder between the pins with that solder braid and touch the pins up again. But it goes to show the importance of looking at the pins from a few angles after soldering.
* The next two photos show soldering in the two LEDs, LED1 and LED2, and the current limit resistor betwen them R28. Do R28 first to get it out of the way. Notice those 3 little polarity dots near the bottom pad of each LED on the board. On the LED the polarity marking is extremely hard to see. It is a tiny amount of green color on the bottom solder tab, just before the LED lens. To see it I had to use a 6x magnifying glass held only about an inch away from the LED. There is an easier way. Look though the LED lens and you will see a small metal square right in the center where the LED mounts. Trace the lead inside the LED lens from that square and that is the LED tab that matches up with the dots on the board.
I'm going to list two alternernate zener diodes in the BOM to replace the LEDs in case anyone would rather not have the interior of their ODA lit blue! The zener diodes would mount with the opposite polarity to the LEDs, with their banded end away from the three dots on the board.
* Finally go ahead and solder in the diode logic for the comparator, D7 - D10. I have a picture of that in the next post.
* JP14 and JP15 in the last photo, near the LEDs, are good to keep in mind. Those are test points on the output of the LME49990 gain stage chips (left and right channel). If you want to look at the output of the gain stage with a scope, here is where to get it. These two points are also where you can connect the front panel pre-amp RCA jack, as one of the options, if you want that to hook into the output of the gain stage. If you want that pre-amp jack to run the rear panel RCA input straight through you can use JP1, 4, and 5 near the rear panel jack. If you want the pre-amp out to pass the 3.5mm jack through then use JP2 and JP3. If you want the pre-amp jack to pass through whatever the input select switch has chosen, use the JP7 holes on the board. And then with any of this you can run the wires right to the pre-amp out RCA jack, or you can run it through the pre-amp chip set up either for unity gain current buffer, or with some voltage gain. All sorts of options on using that pre-amp out RCA jack.
The build instructions goes through this.Attachments
Last edited:
remainder of clipping indicator, plus DC offset null circuit
Continuing on with the clipping indicator & partial gain stage photos. Plus I'm adding the build-up of the DC offset null section.
* The next photo shows the pot, gain switch, power LEDs (LED4, 5), and clipping LED (LED 3) installed. The LED current limit resistors R60 and R61 next to the LEDs should be soldered in first since they are harder to access when the other parts are installed. The 3.5mm jack RF filter caps C3 and C4 are also installed.
* The next photo shows the rest of the clipping indicator parts installed, along with the 3.5mm input jack.
* The next two photos shows the DC output offset null section of the board. There are several surface mount resistors under the board in this section, which should all be soldered in first again before the top through hole stuff (R42, R43, R45, R47, R48, R50) . Again those missing SMD resistors under the gain switch are just because I ran out of parts. Those would normally be installed by now.
* The next photo shows centering the 10-turn 10-ohm trimmer pots, R46 & R53, before installing them. Even though they are officially 10 ohn the actual reading for center was about 5.8 ohms on each end, measured center wiper to the ends. Again I ran out of parts and didn't have the second trimpot. But both should be installed at this point. Then install the rest of the through hole parts, R51,R54, R58, R59, D11, D14.
* The final two photos shows all the DC offset null parts soldered in. There are two "tombstoned" zener diodes again, D11 and D14. I forgot to add the "band down" wording on the board. Both diodes should have their cathode bands down toward the board.
Two of the 4.7uf 63V caps C40 & C42 are also part of the DC offset null circuit, but I'll install all twelve of those caps at once in a later step.
Continuing on with the clipping indicator & partial gain stage photos. Plus I'm adding the build-up of the DC offset null section.
* The next photo shows the pot, gain switch, power LEDs (LED4, 5), and clipping LED (LED 3) installed. The LED current limit resistors R60 and R61 next to the LEDs should be soldered in first since they are harder to access when the other parts are installed. The 3.5mm jack RF filter caps C3 and C4 are also installed.
* The next photo shows the rest of the clipping indicator parts installed, along with the 3.5mm input jack.
* The next two photos shows the DC output offset null section of the board. There are several surface mount resistors under the board in this section, which should all be soldered in first again before the top through hole stuff (R42, R43, R45, R47, R48, R50) . Again those missing SMD resistors under the gain switch are just because I ran out of parts. Those would normally be installed by now.
* The next photo shows centering the 10-turn 10-ohm trimmer pots, R46 & R53, before installing them. Even though they are officially 10 ohn the actual reading for center was about 5.8 ohms on each end, measured center wiper to the ends. Again I ran out of parts and didn't have the second trimpot. But both should be installed at this point. Then install the rest of the through hole parts, R51,R54, R58, R59, D11, D14.
* The final two photos shows all the DC offset null parts soldered in. There are two "tombstoned" zener diodes again, D11 and D14. I forgot to add the "band down" wording on the board. Both diodes should have their cathode bands down toward the board.
Two of the 4.7uf 63V caps C40 & C42 are also part of the DC offset null circuit, but I'll install all twelve of those caps at once in a later step.
Attachments
Last edited:
The remainder of the ODA minus voltage regulators
Here is the rest of it, minus the voltage regulators and RCA jacks for now, plus a few parts on order from Mouser that should get here tomorrow.
* Photo of the whole thing, front and back.
* Next photo is the gain chip section ready for the LME49990 chips, one on top and one under the board. Jumpers have been installed in the JPR1 and JPR2 positions using #22 solid insulated wire.
* At least the LME49990 chip has a visable dot on one end that matches up with the dot on the board, as the next photo shows. Both chips are oriented with their dots toward the edge of the board. And here I discovered another marking error, the dot for IC2 under the board wound up on top of the board. Just point IC2's dot toward the board edge. Same thing with the SMD soldering for IC1 and IC2 as with the other SMD chips. I soldered one corner, aligned it, soldered the other corner, then the rest of the pins.
* In the next photo the blue 0.22uF decoupling caps are added, along with the 560uF C38 and C39 caps, and C13 and C14 150pF compensation caps. I would be adding the R24 and R25 feedback resistors but that is another part on order from Mouser. I'm not adding bass boost to this build so the 4 bass boost parts (R22, R23, C15, C16) are not populated.
* The next 3 photos shows how the NJM4556AL output section SIP chips go in, with the gold bar toward the bottom of the board (toward the power supply parts). When you solder in the blue 0.22uF decoupling caps in this section try to get them as straight up and down as possible to prevent any interference with the output chips. With the SIP chips I solder one pin, flip the board to make sure the chip is straight up and down, then solder the rest of the pins in the row. These NJM4556AL chips are exactly the same chip as the NJM4556A output chips in the O2 headamp, just in the different inline SIP package that dissipates a bit more heat and allows for denser packing for more of them in parallel than the O2. The two 560uF capacitors in the middle of this section, C57 and C58, get soldered in along with the six output chips.
* The next photo shows all twelve of the red 4.7uF coupling and DC output null caps installed. There are no polarity to these caps so you may want to try turning some of them 180 degrees to line up with the markings on the board as closely as possible. It doesn't matter, I've left plenty of room around the caps, but just looks nice to have the spacings all the same.
* Finally there is the pre-amp section and output jacks. I'm leaving the pre-amp chip under the board off for now, so at this point I've just soldered in the bypass and decoupling caps C44 - C48. Then be sure to solder in jumpers in the R88 and R89 position if you are not inserting an optional damping resistor (usually not used). Then finally the 3.5mm and 1/4" output jacks.
Here is the rest of it, minus the voltage regulators and RCA jacks for now, plus a few parts on order from Mouser that should get here tomorrow.
* Photo of the whole thing, front and back.
* Next photo is the gain chip section ready for the LME49990 chips, one on top and one under the board. Jumpers have been installed in the JPR1 and JPR2 positions using #22 solid insulated wire.
* At least the LME49990 chip has a visable dot on one end that matches up with the dot on the board, as the next photo shows. Both chips are oriented with their dots toward the edge of the board. And here I discovered another marking error, the dot for IC2 under the board wound up on top of the board. Just point IC2's dot toward the board edge. Same thing with the SMD soldering for IC1 and IC2 as with the other SMD chips. I soldered one corner, aligned it, soldered the other corner, then the rest of the pins.
* In the next photo the blue 0.22uF decoupling caps are added, along with the 560uF C38 and C39 caps, and C13 and C14 150pF compensation caps. I would be adding the R24 and R25 feedback resistors but that is another part on order from Mouser. I'm not adding bass boost to this build so the 4 bass boost parts (R22, R23, C15, C16) are not populated.
* The next 3 photos shows how the NJM4556AL output section SIP chips go in, with the gold bar toward the bottom of the board (toward the power supply parts). When you solder in the blue 0.22uF decoupling caps in this section try to get them as straight up and down as possible to prevent any interference with the output chips. With the SIP chips I solder one pin, flip the board to make sure the chip is straight up and down, then solder the rest of the pins in the row. These NJM4556AL chips are exactly the same chip as the NJM4556A output chips in the O2 headamp, just in the different inline SIP package that dissipates a bit more heat and allows for denser packing for more of them in parallel than the O2. The two 560uF capacitors in the middle of this section, C57 and C58, get soldered in along with the six output chips.
* The next photo shows all twelve of the red 4.7uF coupling and DC output null caps installed. There are no polarity to these caps so you may want to try turning some of them 180 degrees to line up with the markings on the board as closely as possible. It doesn't matter, I've left plenty of room around the caps, but just looks nice to have the spacings all the same.
* Finally there is the pre-amp section and output jacks. I'm leaving the pre-amp chip under the board off for now, so at this point I've just soldered in the bypass and decoupling caps C44 - C48. Then be sure to solder in jumpers in the R88 and R89 position if you are not inserting an optional damping resistor (usually not used). Then finally the 3.5mm and 1/4" output jacks.
Attachments
Last edited:
Just wow! Looks amazing
Thanks! Maybe one of the O2 headamp builders in Europe will eventually sell ODA boards, parts, or assembled units.
There is one big problem I'm recently aware of though for Europe. Apparently an energy efficiency directive has just gone into effect and wall power adaptors with a transformer in them are banned. Everything apparently has to be a switching power supply now, and none of those will output AC (the ODA like the O2 headamp needs AC input). In another forum a few days ago a fellow was looking for a larger power adaptor for his O2 headhone amp. He snagged one of the last before the distributor (Maplin) returned them all to the manufacturer.
Here in the US AC-to-AC wall transformers are still available for 115Vac input, the WAU20-500, WAU20-2000, WAU24-750, WAU24-1000, and WAU-1800 all at Mouser. I'm not aware of any government directives here to phase them out yet - like the government here did with incandescent light bulbs.
Last edited:
Everything on but the negative LDO
The missing Mouser parts arrived today. Here are front and back photos of the ODA board with everything on but the positive LDO regulator, which should get here Thursday and I can finally do some testing.
I forgot to say in the writups above that the 1K attenuation resistors R31 & R32 should be populated at the same time the feedback resistors R24 & R25 are stuffed, after the IC1 and IC2 gain chips are soldered on.
Some other random notes:
* The JP12 and JP13 holes inthe gain stage section are used for an optional off-board bass boost switch, if wanted, which turns it on or off. If using bass boost those can just be jumpered to have it on all the time. More details in the build instructions. The gain stage has to have some gain (can't be set for unity voltage gain) for bass boost to work. However the standard attenuation resistors cut the gain stage output in half, so a stage gain of 2x does yield an end-to-end voltage gain of 1 going into the pot. Remember that the standard gaiin-switch voltage gains (the resistors in the BOM) for the ODA yeild gains of 1/2x, 1x, 2x and 4x (the first gain switch position is 1/2 attenuation, not gain). I set things up that way since two of my sources output over 2V and actually need that 1/2 attenuation to allow th epot to use full turning range. The gain stage is still useful, even with attentuation, for current-driving that 1K pot. Otherwise there would be a 1K pot across the source if at the front of the circuit. More info on it all in the build instructions, along with how to pick resistors for other gain settings on the gain switch.
* JP21 in the pre-amp section is hooked directly to the pre-amp output jack. It can be used to wire that jack to the RCA input jack, 3.5mm input jack, input select switch output, etc. if the pre-amp chip is left off.
* If the pre-amp chip is populated then JP21 isn't used and JP20 is used to feed the signal into the pre-amp chip. More info on pre-amp wiring and options in the build instructions.
* The R88 and R89 resistors are jumpered for nomal used, as mentioned. They are just in series with the output on each channel and can be used to add damping factor resistors. Wires can also be run from these hole to an external rotary switch to give several damping factor options. The part number for that optional switch is in the BOM. More info in the build instructions.
* R86, R87, C60 and C61 are there for an optional Zobel network. Zobels are added to headphone amps - and power amps - if certain speakers cause oscillation. But so far I've found the ODA output to be unconditionally stable with every load I've tried, similar to the O2 amp (not a surprise since it is using more of the same output chips in the same configuration). So these 4 parts are left off. If anyone building the amp does find a need for the Zobel network on a certain headphone, please let me know!
* In the power supply section JP16 and JP17 are just test points on the output of the pre-regulators going into the LDO regulators. They are not used for anything other than a voltage check during any testing. JP10 and JP11 are jumpered to JP8 and JP9 to connect the CRC filter output to the pre-regulators. I did it this way to keep that noisy sawtooth wave away from the "clean power" traces on the board, given the layout of having the regulators all spread out over the back panel for heatsinking.
The missing Mouser parts arrived today. Here are front and back photos of the ODA board with everything on but the positive LDO regulator, which should get here Thursday and I can finally do some testing.
I forgot to say in the writups above that the 1K attenuation resistors R31 & R32 should be populated at the same time the feedback resistors R24 & R25 are stuffed, after the IC1 and IC2 gain chips are soldered on.
Some other random notes:
* The JP12 and JP13 holes inthe gain stage section are used for an optional off-board bass boost switch, if wanted, which turns it on or off. If using bass boost those can just be jumpered to have it on all the time. More details in the build instructions. The gain stage has to have some gain (can't be set for unity voltage gain) for bass boost to work. However the standard attenuation resistors cut the gain stage output in half, so a stage gain of 2x does yield an end-to-end voltage gain of 1 going into the pot. Remember that the standard gaiin-switch voltage gains (the resistors in the BOM) for the ODA yeild gains of 1/2x, 1x, 2x and 4x (the first gain switch position is 1/2 attenuation, not gain). I set things up that way since two of my sources output over 2V and actually need that 1/2 attenuation to allow th epot to use full turning range. The gain stage is still useful, even with attentuation, for current-driving that 1K pot. Otherwise there would be a 1K pot across the source if at the front of the circuit. More info on it all in the build instructions, along with how to pick resistors for other gain settings on the gain switch.
* JP21 in the pre-amp section is hooked directly to the pre-amp output jack. It can be used to wire that jack to the RCA input jack, 3.5mm input jack, input select switch output, etc. if the pre-amp chip is left off.
* If the pre-amp chip is populated then JP21 isn't used and JP20 is used to feed the signal into the pre-amp chip. More info on pre-amp wiring and options in the build instructions.
* The R88 and R89 resistors are jumpered for nomal used, as mentioned. They are just in series with the output on each channel and can be used to add damping factor resistors. Wires can also be run from these hole to an external rotary switch to give several damping factor options. The part number for that optional switch is in the BOM. More info in the build instructions.
* R86, R87, C60 and C61 are there for an optional Zobel network. Zobels are added to headphone amps - and power amps - if certain speakers cause oscillation. But so far I've found the ODA output to be unconditionally stable with every load I've tried, similar to the O2 amp (not a surprise since it is using more of the same output chips in the same configuration). So these 4 parts are left off. If anyone building the amp does find a need for the Zobel network on a certain headphone, please let me know!
* In the power supply section JP16 and JP17 are just test points on the output of the pre-regulators going into the LDO regulators. They are not used for anything other than a voltage check during any testing. JP10 and JP11 are jumpered to JP8 and JP9 to connect the CRC filter output to the pre-regulators. I did it this way to keep that noisy sawtooth wave away from the "clean power" traces on the board, given the layout of having the regulators all spread out over the back panel for heatsinking.
Attachments
Last edited:
It works!
The LDO voltage regulators from Arrow arrived a day early. The ODA board works perfectly, including the new DC output offset null circuit!
I'll have a write-up posted in the vendor thread by Friday with the details of paying for the boards and parts (all at-cost, of course). I should be able to get the boards in the mail starting on Monday. The vendor thread is here:
http://www.diyaudio.com/forums/vendors-bazaar/237226-parallel-njm4556al-two-stage-amp.html
I'll also have a revised schematic and BOM up by Friday and the Google Drive link.
Below are some photos...
* First is the test setup. I have JP8 and JP9 jumpered to JP10 and JP11 using that black and red wire you can see under the board, that I tacked on during the CRC section build. Those wires connect the CRC filter to the pre-regulators. The real thing is soldered on to of the board into those JP holes. Between now and Monday I'm going to do some tests with the scope and QA400 analyzer to see if the real thing has to be mini-coax or if just a regular wire will do the job. I'll include the wire with the board and instructions in the build instructions.
* The next two photos shows the meter leads in the JP18 and JP19 output holes to test the power supply & the resulting meter readings. The voltage readings are right on the money. I purposely set the LDO resistors to aim for 15.7Vdc - 15.9Vdc to keep things under 16Vdc. Meter ground goes to the blue wire which goes to the ground holes, JP6. I specifically included R33 and R36 so those voltages could be micro-tweaked to get them just as close as you want to being equal. One of those resistors is 20R and the other 4R right now. The difference is the slightly different internal reference voltage between the LT3015 and LT1963A chips. Then there is a certain tolerance one way or the other on the chips themselves for that reference voltage. By swapping those resistors out with ones a few ohms more or less you should be able to hit +/-15.9Vdc or +/-15.8Vdc just right on the spot. It won't make any audio difference, but some folks may just want to do that for fun.
* The next photo shows the jumpers being installed in JP18 and JP19 to run the power to the rest of the board, now that the power supply output has been verified. Note that the jumper in JP18 goes up and down while the jumper in JP19 goes sideways. The jumpers are just pieces of bare 22 gauge wire. This photo also shows those blue LEDs working in the clipping indicator section.
* The next photo shows the DC output offset when the relay first closed, with the two trimpots R46 and R53 at their center setting from the build. About -525uV and -700uV (the two channels), which all by itself is an improvement. The previous ODA version without DC offset null had an offset of about 1.5mV (=1500uV), while the O2 headamp has an typical offset of 3mV (=3000uV). But by turning the pots about a few turns I was able to just completely null out the DC offset, as the second photo shows, 4uV and 2uV. That drifted a bit with temperature as I was expecting, after playing music for 30 minutes, but still just 30uV and 20uV, well within what I was expecting. All in all, the DC offset null circuit appears to be a big success.
* The final photo shows headphones plugged in and music playing. That is the audio output on the DMM main display, about 7mVac(rms), with the DC offset on the sub-display.
One note on testing. Since I don't have the rear panel yet, which is needed to heat sink the regulators, I cheated and used a 16Vac transformer (WAU16-1000) to keep the regulator dissipation down enough to be un-heatshinked for a while with the light load. 16Vac is low enough to cause the pre-regulators to barely start going into dropout. But it doesn't matter since those feed the final LDO regulators and the load here is very small with these high sensitivity headphones.
The LDO voltage regulators from Arrow arrived a day early. The ODA board works perfectly, including the new DC output offset null circuit!
I'll have a write-up posted in the vendor thread by Friday with the details of paying for the boards and parts (all at-cost, of course). I should be able to get the boards in the mail starting on Monday. The vendor thread is here:
http://www.diyaudio.com/forums/vendors-bazaar/237226-parallel-njm4556al-two-stage-amp.html
I'll also have a revised schematic and BOM up by Friday and the Google Drive link.
Below are some photos...
* First is the test setup. I have JP8 and JP9 jumpered to JP10 and JP11 using that black and red wire you can see under the board, that I tacked on during the CRC section build. Those wires connect the CRC filter to the pre-regulators. The real thing is soldered on to of the board into those JP holes. Between now and Monday I'm going to do some tests with the scope and QA400 analyzer to see if the real thing has to be mini-coax or if just a regular wire will do the job. I'll include the wire with the board and instructions in the build instructions.
* The next two photos shows the meter leads in the JP18 and JP19 output holes to test the power supply & the resulting meter readings. The voltage readings are right on the money.
I purposely set the LDO resistors to aim for 15.7Vdc - 15.9Vdc to keep things under 16Vdc. Meter ground goes to the blue wire which goes to the ground holes, JP6. I specifically included R33 and R36 so those voltages could be micro-tweaked to get them just as close as you want to being equal. One of those resistors is 20R and the other 4R right now. The difference is the slightly different internal reference voltage between the LT3015 and LT1963A chips. Then there is a certain tolerance one way or the other on the chips themselves for that reference voltage. By swapping those resistors out with ones a few ohms more or less you should be able to hit +/-15.9Vdc or +/-15.8Vdc just right on the spot. It won't make any audio difference, but some folks may just want to do that for fun.* The next photo shows the jumpers being installed in JP18 and JP19 to run the power to the rest of the board, now that the power supply output has been verified. Note that the jumper in JP18 goes up and down while the jumper in JP19 goes sideways. The jumpers are just pieces of bare 22 gauge wire. This photo also shows those blue LEDs working in the clipping indicator section.
* The next photo shows the DC output offset when the relay first closed, with the two trimpots R46 and R53 at their center setting from the build. About -525uV and -700uV (the two channels), which all by itself is an improvement. The previous ODA version without DC offset null had an offset of about 1.5mV (=1500uV), while the O2 headamp has an typical offset of 3mV (=3000uV). But by turning the pots about a few turns I was able to just completely null out the DC offset, as the second photo shows, 4uV and 2uV. That drifted a bit with temperature as I was expecting, after playing music for 30 minutes, but still just 30uV and 20uV, well within what I was expecting. All in all, the DC offset null circuit appears to be a big success.
* The final photo shows headphones plugged in and music playing. That is the audio output on the DMM main display, about 7mVac(rms), with the DC offset on the sub-display.
One note on testing. Since I don't have the rear panel yet, which is needed to heat sink the regulators, I cheated and used a 16Vac transformer (WAU16-1000) to keep the regulator dissipation down enough to be un-heatshinked for a while with the light load. 16Vac is low enough to cause the pre-regulators to barely start going into dropout. But it doesn't matter since those feed the final LDO regulators and the load here is very small with these high sensitivity headphones.
Attachments
Last edited:
Thanks!
I have the at-cost ordering price list posted over in the vendor thread now, along with all the ordering details and my paypal address:
http://www.diyaudio.com/forums/vend...llel-njm4556al-two-stage-amp.html#post3887668
Mouser just got a bunch of the input 3.5mm jacks in again yesterday, so now the only things they don't have are the pot and the LT regulators, I have all of those.
I'll have updated files posted at the Google Link tomorrow, along with posting the Gerbers if anyone wants to do a run of boards themselves. I will fix the few labelling errors I spotted in the photos above in the posted Gerber files.
I have also updated the top of the first post in this thread with the V2.0 information.
Last edited:
Good suggestion! I'll post some higher rez shots on the Google Drive link. I discovered a few months ago that I had the camera on the maximum rez setting and was generating 800KB images. I probably went to far in the other direction and set it for the minimum rez for the forum pictures.
Yes it is! Keep those cards and letters and paypals coming.
I've added a couple of columns to the "orders" spreadsheet out at the Google Drive link to show date payment received and date shipped.I have the Gerber files posted out at the Google Drive link now, if anyone wants to fab their own boards. I have all the minor text marking errors that I found in the photo build pictures fixed in the Gerbers. I have that errata list posted now too.
Some good news on the wiring, I did some tests today and the mini-coax won't be needed to connect the CRC filter jumpers to the pre-regulators. The CRC filter is rounding the corners on the sawtooth wave enough that EMI radiation of line harmonics isn't a problem. Just regular wire is fine and I'll include that (22 gauge silver-coated solid teflon-covered) with the boards. I will also do another posting showing where the two wires go.
Last edited:
I think agdr was looking for the screws for the B3 enclosure...found them on the box enclosure sits:
Box Enclosures - Product
FYI
Alex
Box Enclosures - Product
FYI
Alex
Information about the case screws
Thanks for letting me know the package made it! Now you have some soldering to do...
Hey good point about those Box Enclosure case screws. I've been meaning to post some more information there.
Box Enclosures ships the cases with two different sets of screws but provide no details. One set has a torx head and is "thread forming", which has finer pitched threads that are not as deep:
Box Enclosures - New Page
These apparentlly are meant to deform the aluminum rather than gouge it so that the screw can be taken out and put back in several times. Presumably this would leave fewer, if any, metal cuttings inside the case to short things out.
The other set is "self tapping" with a philips head:
Box Enclosures - New Page
They have coarser and deeper threads that are known to leave metal filings behind. NwAvGuy / RocketScientist posted about that shortly after the O2 amps came out when a couple of folks had shorted boards due to filings. The solution there is undo the screws and shake out any metal filings. I get the impression the "self tapping" screws are meant to be one-time or just a few times in and out, vs. many in/out cycles for the "thread forming".
I'm using the thread forming screws in my stuff here. I ordered one of the 50 packs from Box Enclosures at that link a while back when one of the boxes showed up short 2 screws. They have the 8-packs at that link too.
Thanks for letting me know the package made it! Now you have some soldering to do...
Hey good point about those Box Enclosure case screws. I've been meaning to post some more information there.
Box Enclosures ships the cases with two different sets of screws but provide no details. One set has a torx head and is "thread forming", which has finer pitched threads that are not as deep:
Box Enclosures - New Page
These apparentlly are meant to deform the aluminum rather than gouge it so that the screw can be taken out and put back in several times. Presumably this would leave fewer, if any, metal cuttings inside the case to short things out.
The other set is "self tapping" with a philips head:
Box Enclosures - New Page
They have coarser and deeper threads that are known to leave metal filings behind. NwAvGuy / RocketScientist posted about that shortly after the O2 amps came out when a couple of folks had shorted boards due to filings. The solution there is undo the screws and shake out any metal filings. I get the impression the "self tapping" screws are meant to be one-time or just a few times in and out, vs. many in/out cycles for the "thread forming".
I'm using the thread forming screws in my stuff here. I ordered one of the 50 packs from Box Enclosures at that link a while back when one of the boxes showed up short 2 screws. They have the 8-packs at that link too.
Last edited:
Panel update
I've posted the front panel CAD file today in Front Panel Express format (Front Panel Designer, ".fpd"). I've also posted the revised spreadsheets with the measurements for those wanting to drill their own panels that come with the B4-080 cases. Those are all at the Google Drive link under the 3_19_2014 V2.0 -> panels folder.
I do have the rear panel CAD file posted there too, and added that to the measurements, but I would advise waiting until Friday when I get the board in and test it to make sure those measurement are OK. The front panel file is good to go, those are the same measurements as the last ODA version.
If anyone would like a different font just download the free FPD program from FPE's website. Then open up the front or rear ODA .fpd file. Chose the arrow tool, then double left click on the text you want to change and popup will be launched. You can pick different fonts, font sizes, line spacings, color fills, italic, justification, etc. from that menu. You can also change the lettering itself.
So far Front Panel Express and CAM Expert are the two that I've been able to verify can do the job and use the .fpd file as it sits. FPE is more expensive, but I see some occasional 20% off coupon numbers at their twitter or facebook sites. With the 20% off the two would be about the same in cost.
I did get in contact with Pokono today, the excellent suggestion made a few posts above. They laser cut the parts and can't use anodized material. They won't work for this batch, but they would be good to keep in mind for a slick looking variation. The panel would be natural aluminum color and the lettering would be laser cut all the way through the panel. That would probably look pretty sharp! They use a different CAD system (can't use the .fpd files), so anyone interested could enter the measurements from the spreadsheet I've posted into their CAD editor.
Also, just a very important reminder, don't run the ODA with any significant load on the output until you have a rear panel in place to heatsink the voltage regulators. They definitely need the heatsink, especially if using a 24Vac transformer. With no load at all on the output (actually the ODA has a built-in 2.5K minimum load on each channel) you are OK for short periods with no heatsink, say 10 seconds or so, to test the power supply section voltages
I've posted the front panel CAD file today in Front Panel Express format (Front Panel Designer, ".fpd"). I've also posted the revised spreadsheets with the measurements for those wanting to drill their own panels that come with the B4-080 cases. Those are all at the Google Drive link under the 3_19_2014 V2.0 -> panels folder.
I do have the rear panel CAD file posted there too, and added that to the measurements, but I would advise waiting until Friday when I get the board in and test it to make sure those measurement are OK. The front panel file is good to go, those are the same measurements as the last ODA version.
If anyone would like a different font just download the free FPD program from FPE's website. Then open up the front or rear ODA .fpd file. Chose the arrow tool, then double left click on the text you want to change and popup will be launched. You can pick different fonts, font sizes, line spacings, color fills, italic, justification, etc. from that menu. You can also change the lettering itself.
So far Front Panel Express and CAM Expert are the two that I've been able to verify can do the job and use the .fpd file as it sits. FPE is more expensive, but I see some occasional 20% off coupon numbers at their twitter or facebook sites. With the 20% off the two would be about the same in cost.
I did get in contact with Pokono today, the excellent suggestion made a few posts above. They laser cut the parts and can't use anodized material. They won't work for this batch, but they would be good to keep in mind for a slick looking variation. The panel would be natural aluminum color and the lettering would be laser cut all the way through the panel. That would probably look pretty sharp! They use a different CAD system (can't use the .fpd files), so anyone interested could enter the measurements from the spreadsheet I've posted into their CAD editor.
Also, just a very important reminder, don't run the ODA with any significant load on the output until you have a rear panel in place to heatsink the voltage regulators. They definitely need the heatsink, especially if using a 24Vac transformer. With no load at all on the output (actually the ODA has a built-in 2.5K minimum load on each channel) you are OK for short periods with no heatsink, say 10 seconds or so, to test the power supply section voltages
Last edited: