No, the LM3886 amplifier is 'single ended'; it uses a balanced power supply, and the output is referenced to ground.
As long as the motor you use is 'star wound', with a ground reference, then you need only feed each of the phase windings with the output of the amplifier, and connect the 'star' to the ground point.
I'm only using the motors the way I am because I already have the BTL amplifiers, and it's a shame to waste them 🙂
Correct.
To clarify a couple of points, the star wound motors from Anahiem Automation do not have a common point tied to ground. The 3 windings are tied to a common node, but there is no connection going to the amp or power supply.
The LM3886 3 channel amp I'm working on will have a spilt power supply so the outputs idle at ground, but it doesn't matter if it was single ended supply and they idled at Vcc/2. Since there is no ground connection on the coils, all that matters is the idle voltage is the same on all 3 legs.
I'm plan on using the preassembled LM3886 amps from eBay, as recommended in the 60 WPC amp for DIY Turntable Motor Drive post, which was recommended for use with the SG-4. I have two and was going to purchase another.
60W LM3886TF Sound Audio Amplifier 2-Channel Digital Power AMP | eBay.
Will they work? Maybe I should wait for a working schematic of the amp section?
60W LM3886TF Sound Audio Amplifier 2-Channel Digital Power AMP | eBay.
Will they work? Maybe I should wait for a working schematic of the amp section?
You should be fine with the motors Bill (Pyramid) is recommending, in fact those amps should be fine with pretty much any BLDC motor except those with very low impedance windings.
I think this would be a good idea. I hope to post the 3 phase amp project in the next day or two. Everything is working, I just need to clean up the documentation.
The 3 phase amp project will be more expensive than the e-Bay versions, but it will be better suited: Decoupled 12VDC output to power SG4, 10T pots on board for level setting, gain optimized for AA BLDC motors, terminal strips for AC input, SG4 I/O and motor outputs, summing test point for balancing the 3 motor phase amplitudes and compact size.
Will do. I can always get the new board and then rob the other ones for parts, etc. No big rush as I'm waiting for the AA BLWS and a pulley from Harry.
The 3 phase amp based on the LM3886 is working fine, but I'm concerned about a couple of things:
Cost-With a 50VA toroid to power it, it will come in at $91.
Size-The board is small, but the circuit requires a 50VA xfmr and needs a LOT of heatsinking for the 3 chip amps.
Heat-This will need a good sized case with lots of heat sink, and will still run warm (130°F).
I'm sketching out a design using TPA3125P class D amps that are DIP20 packages, so they would be relatively easy to solder. If the design works as well as the TPA3123 amps, it would solve all of the above problems. It would take a couple more weeks to prototype and test the new design. The TPA3125 is rated at 10W/ch, but I think it is because of the package which has no heat sink connection to the PCB. If a DIP type heatsink was used, I think thermal management and power dissipation would be OK.
Is doing a class AB amp design based on LM3886 worth pursuing at this point?
Cost-With a 50VA toroid to power it, it will come in at $91.
Size-The board is small, but the circuit requires a 50VA xfmr and needs a LOT of heatsinking for the 3 chip amps.
Heat-This will need a good sized case with lots of heat sink, and will still run warm (130°F).
I'm sketching out a design using TPA3125P class D amps that are DIP20 packages, so they would be relatively easy to solder. If the design works as well as the TPA3123 amps, it would solve all of the above problems. It would take a couple more weeks to prototype and test the new design. The TPA3125 is rated at 10W/ch, but I think it is because of the package which has no heat sink connection to the PCB. If a DIP type heatsink was used, I think thermal management and power dissipation would be OK.
Is doing a class AB amp design based on LM3886 worth pursuing at this point?
I'm not locked into the LM3886, just what I happened to have and what you mentioned. If a Class D makes more sense, than go for it.
At first Harry recommended a 600 RPM TNT pulley from the 90s, but I wanted the .250" shaft, taller than the TNT and three groves so I could run three belts (I'm running the HR-X pulley with three belts and like it a lot). So he is still pricing a one off for me. I have yet to mention that several people may want a similar pulley.
At first Harry recommended a 600 RPM TNT pulley from the 90s, but I wanted the .250" shaft, taller than the TNT and three groves so I could run three belts (I'm running the HR-X pulley with three belts and like it a lot). So he is still pricing a one off for me. I have yet to mention that several people may want a similar pulley.
I am also in favour of the class D amplifier; from a purely efficiency point of view it makes sense. The only downside I can easily see is the comparative lack of single ended chips, especially in medium power range.
I would be interested in seeing your design. Your designs are throughly thought out and optimized. I have some LM3886s I was thinking about using.
Class D does make a lot of sense from an efficiency, cost, and size perspective.
I already got on board with class D amps for my 2 phase controller when I saw your post in the amp thread of the 7492 driving the Lenco motor...it just seemed like a simpler, better solution.
I also picked up 3 of the eBay 3123 amps for $25 in a trigger happy binge while I was ordering components. If I don't use them it's no big deal, especially considering all of the features you are designing into your 3 channel amp to make it right for the SG4.
I'll be occupied building the 2 phase in the mean time 😉
I also picked up 3 of the eBay 3123 amps for $25 in a trigger happy binge while I was ordering components. If I don't use them it's no big deal, especially considering all of the features you are designing into your 3 channel amp to make it right for the SG4.
I'll be occupied building the 2 phase in the mean time 😉
I don't have all the documentation done, but here is the schematic, parts locator and BOM.
The link to OshPark is:www.oshpark.com/shared_projects/7l6vTLZK
You'll need a 50VA toroid to power it. 20-22AWG wire works best with the connectors.
Attachments
I'm going to try the LM3886, as I already have the amps. I will just change the gain and add the input pots and TP1. What is the AC voltage of the power transformer?
I used a dual secondary 12V xfmr with the secondaries in series. You might be able to get away with 9V secondaries; the only 9V xfmr I had was 10W and the voltage sagged too much. The 12V 25W xfmr produced ±18.5VDC which is probably why the amps (and xfmr) ran warm. The LM3886 needs at least 20VDC to operate.