I thought it was 100W mono?
This would be cool to split the signal or maybe have an internal charge pump to run it at 18-24V
TPA3118 60W Power Amp Module
- Thread starter falzhobel
- Start date
This would be cool to split the signal or maybe have an internal charge pump to run it at 18-24V
Paul.Ruby
Well-known member
It is mono. It's advertised at 100W but not into 8-ohm. The board is behaving better than the original real 3116 boards (which apparently no longer exist except in fake form). This is going to make a VERY loud guitar amp with a ton of headroom.
This looks like it might be the same board for $10. Gonna buy one...
If you trust AliExpress....
Paul.Ruby
Well-known member
I've never blown one up, of about a dozen or so, using 24V power supply. But, I have a 1N508 diode inside all of my 3118 boxes to drop a bit of that voltage. And, I use power supplies that I've personally tested to have accurate output level and no overshoot at power-up or under transient load. I don't understand "34." The chips are rated at 26V max and the 3118 boards use 25V caps. So, 24V is the true max these should ever be allowed to see.
Paul.Ruby
Well-known member
Some info about this alternative TPA3116 board I have.... It's a real chip. I bought it from this amazon link but it is available cheaper elsewhere if willing to wait for shipping from china. https://www.amazon.com/dp/B07H339BFP
Just got this board. Older revision of the board but same circuit and real chip: https://www.amazon.com/dp/B0BWVD518D
Still waiting on the same thing from China...
TLDR: I'm going to remove the entire preamp and directly drive the TPA3116 with my Marshall TMB board. Also removing the volume pot, which has a mute switch, and removing the power and input jacks. Not enough room in a 125B for a soft-touch footswitch for mute. The heat sink of the amp will hit it. Using a 5s power-on delay only. The amp will be bolted to the bottom of the case with the heat sink coupled directly to the metal case.
The TMB preamp is documented here: https://forum.pedalpcb.com/threads/tpa3118-60w-power-amp-module.2015/page-9#post-110531
The board barely fits in a 125B enclosure. Here's a picture. It squeezes tight to the sides before bottoming out on the top surface of the case.
It obviously needs some re-working to put in a pedal: remove the volume pot, power connector, 1/8" jack. I'm not sure if I will keep the existing preamp in place or not. (EDIT: It's removed.) It's not adding value for a guitar amp because some alternative preamp is still needed to get to line level and 10K impedance to drive it.
Picture after removing the preamp and connectors and belt sanding the edges to fit into the bottom of a 125B case. There was plenty of room on the sides to sand down. The only small electrolytic remaining is coupling the negative input of the TPA3116 to ground. All others gone. (EDIT: I need to put the 2.2uf cap next to it back in. That's also part of the power amp circuit. Oops.) It's just a power amp now. The 1uf cap that originally coupled the preamp to power amp is removed to make a good spot to add a wire from the new preamp. That's the 1uf at the lower left of this picture.
Here's the original preamp schematic. This same board can be used as a subwoofer driver, which is the reason for the unpopulated capacitors. The 33nf and 68nf caps would create a 100hz low pass filter. It takes a stereo input, which is mixed together after the ganged volume pot. But, I will likely not use any of this. I'll likely use my Marshall TMB board to drive it.
Here's the power amplifier schematic. I started with the schematic right from the TPA3116 datasheet because there isn't much different except values.
The choice of gain resistors does not appear in the datasheet so I don't know what the gain setting is when powering up. The chosen values are in between the 32db and 36db setting but closer to 32db. So, I suspect it's using 32db, which is the same as the previously used boards, both 3118 and 3116.
The mute pin is wired to the on/off switch on the volume pot. So, this amp doesn't shut off when off. It is muted. There is also an LED wired to the switch to turn on when unmuted and that acts as the pull-up on the mute pin to activate it when the switch is off.
Modified for ~5s power on delay after removing the volume pot, 4.7K and LED. The holes for the switch on the volume pot now has 100K. 4.7K resistor is replaced with a 4.7uF cap and the LED is replaced with 220K. A bit of a pain since the LED and resistor were SMT but not too bad.
Here's what this mod looks like.
Here's the fully altered amp mounted to the bottom of the case. 1/4" aluminum spacers on the bolts to the heat sink and nylon spacers under the other two mounting holes because there are PCB traces there. I added some lock washers as spacers and a thermal pad to raise the heat sink slightly. The heat sink might touch the SMT caps under there as it was. Made me nervous. The heat sink has M3-0.5 threads.
The heat sink is coupled by the bolts to the case for better heat transfer to the outside.
Here it is showing final soldering the two halves together and ready to be folded up and bolted together.
No room for a footswitch with the heatsink on there. You'll need a minimum 3A power supply for this. Use 24V to drive 8 ohm and 19V for 4 or 2 ohm. This thing is LOUD. Holy cow. Tons of low end. Had to roll the bass down quite a bit. Not a trivial build but not the hardest either.
Just got this board. Older revision of the board but same circuit and real chip: https://www.amazon.com/dp/B0BWVD518D
Still waiting on the same thing from China...
TLDR: I'm going to remove the entire preamp and directly drive the TPA3116 with my Marshall TMB board. Also removing the volume pot, which has a mute switch, and removing the power and input jacks. Not enough room in a 125B for a soft-touch footswitch for mute. The heat sink of the amp will hit it. Using a 5s power-on delay only. The amp will be bolted to the bottom of the case with the heat sink coupled directly to the metal case.
The TMB preamp is documented here: https://forum.pedalpcb.com/threads/tpa3118-60w-power-amp-module.2015/page-9#post-110531
The board barely fits in a 125B enclosure. Here's a picture. It squeezes tight to the sides before bottoming out on the top surface of the case.
It obviously needs some re-working to put in a pedal: remove the volume pot, power connector, 1/8" jack. I'm not sure if I will keep the existing preamp in place or not. (EDIT: It's removed.) It's not adding value for a guitar amp because some alternative preamp is still needed to get to line level and 10K impedance to drive it.
Picture after removing the preamp and connectors and belt sanding the edges to fit into the bottom of a 125B case. There was plenty of room on the sides to sand down. The only small electrolytic remaining is coupling the negative input of the TPA3116 to ground. All others gone. (EDIT: I need to put the 2.2uf cap next to it back in. That's also part of the power amp circuit. Oops.) It's just a power amp now. The 1uf cap that originally coupled the preamp to power amp is removed to make a good spot to add a wire from the new preamp. That's the 1uf at the lower left of this picture.
Here's the original preamp schematic. This same board can be used as a subwoofer driver, which is the reason for the unpopulated capacitors. The 33nf and 68nf caps would create a 100hz low pass filter. It takes a stereo input, which is mixed together after the ganged volume pot. But, I will likely not use any of this. I'll likely use my Marshall TMB board to drive it.
Here's the power amplifier schematic. I started with the schematic right from the TPA3116 datasheet because there isn't much different except values.
The choice of gain resistors does not appear in the datasheet so I don't know what the gain setting is when powering up. The chosen values are in between the 32db and 36db setting but closer to 32db. So, I suspect it's using 32db, which is the same as the previously used boards, both 3118 and 3116.
The mute pin is wired to the on/off switch on the volume pot. So, this amp doesn't shut off when off. It is muted. There is also an LED wired to the switch to turn on when unmuted and that acts as the pull-up on the mute pin to activate it when the switch is off.
Modified for ~5s power on delay after removing the volume pot, 4.7K and LED. The holes for the switch on the volume pot now has 100K. 4.7K resistor is replaced with a 4.7uF cap and the LED is replaced with 220K. A bit of a pain since the LED and resistor were SMT but not too bad.
Here's what this mod looks like.
Here's the fully altered amp mounted to the bottom of the case. 1/4" aluminum spacers on the bolts to the heat sink and nylon spacers under the other two mounting holes because there are PCB traces there. I added some lock washers as spacers and a thermal pad to raise the heat sink slightly. The heat sink might touch the SMT caps under there as it was. Made me nervous. The heat sink has M3-0.5 threads.
The heat sink is coupled by the bolts to the case for better heat transfer to the outside.
Here it is showing final soldering the two halves together and ready to be folded up and bolted together.
No room for a footswitch with the heatsink on there. You'll need a minimum 3A power supply for this. Use 24V to drive 8 ohm and 19V for 4 or 2 ohm. This thing is LOUD. Holy cow. Tons of low end. Had to roll the bass down quite a bit. Not a trivial build but not the hardest either.
Last edited:
Paul.Ruby
Well-known member
That's the newer version that replaces the 3116 from TI. Might have less hiss. But, this amp was SILENT in testing it.
Sorry, I misread 34w as 34v
This is with the Amazon 3116 board?
Paul.Ruby
Well-known member
Yep. If I do it again, I'll leave in the 1uF cap that couples the NE5532 to the TPA3116 and use pin 6 or 7 of the empty NE5532 socket for input to the board. Must not hook up a signal with any DC reference to the TPA3116 input. I removed the 1M to ground resistor at the output of my TMB board so it is only capacitively coupled to the TPA3116, direct to pin 4, with no DC bias.
I just ordered one to try out.
Paul.Ruby
Well-known member
This one is also real. It's an older revision of the PCB but same circuit and layout.
Here it is modified and ready for use...
Paul.Ruby
Well-known member
Decided to make another version of the volume-only preamp PCB that includes the soft-touch mute circuit and posts to directly mount the TPA3118 board to the back. The goal was to see if this would all fit in a 1590B case. Just finished the build and it worked out really well.
Volume only preamp details here: https://forum.pedalpcb.com/threads/tpa3118-60w-power-amp-module.2015/post-107643
Soft-touch toggle mute details here: https://forum.pedalpcb.com/threads/tpa3118-60w-power-amp-module.2015/post-108125
Here's the top of the board, which will face the inside top of the case. On the left are the posts for input and mute. On the right, two brass pins for power. The pairs of empty holes on the far left and far right are for the stand-by and power LEDs (one hole accidentally filled with super glue below white posts). Top half of the board is the preamp, bottom is mute toggle. Power supplies to each are isolated. Note the 220uf cap is a surface mount with it's base removed to make it through-hole. I did that for component height. Everything must be shorter than the volume pot to fit in the case and the volume pot is already low profile, which is needed to get the PCB sandwich to fit.
This shows the stacking of the two boards, back to back. The speaker output does not have posts. That will directly wire to the output jack. On final mounting, I added a thermal pad between the two boards, mainly to keep anything from shorting between the two. This also shows all the components shorter than the volume pot. The IC and socket have about 1mm clearance. Component leads are all trimmed very low to allow pushing the sandwich tight together to fit the case.
Unfortunately, I made a mistake after soldering the two together and had to take them apart. This damaged the power post PCB traces on my board, so had to directly wire the power to the TPA3118 after correcting the mistake. Always check everything before going beyond points of no return... Dang it... Here's the installed boards. There's 3 or 4mm clearance to the bottom of the case. I have to modify the speaker jack to squeeze into this tiny case and use the ring as the sleeve terminal. Sleeve terminals are completely removed from the jack. Very tight fit. And shows the added power lines to the TPA3118 after the damage done. No posts on the left side anymore.
Finished product...
ExpressPCB design shared here https://drive.google.com/drive/folders/1EUGnukuxIq0znhbYBT9JnLQSlYZEh74F?usp=drive_link
There's also a PDF of the drilling points for 1590B and 125B for the PCB.
Volume only preamp details here: https://forum.pedalpcb.com/threads/tpa3118-60w-power-amp-module.2015/post-107643
Soft-touch toggle mute details here: https://forum.pedalpcb.com/threads/tpa3118-60w-power-amp-module.2015/post-108125
Here's the top of the board, which will face the inside top of the case. On the left are the posts for input and mute. On the right, two brass pins for power. The pairs of empty holes on the far left and far right are for the stand-by and power LEDs (one hole accidentally filled with super glue below white posts). Top half of the board is the preamp, bottom is mute toggle. Power supplies to each are isolated. Note the 220uf cap is a surface mount with it's base removed to make it through-hole. I did that for component height. Everything must be shorter than the volume pot to fit in the case and the volume pot is already low profile, which is needed to get the PCB sandwich to fit.
This shows the stacking of the two boards, back to back. The speaker output does not have posts. That will directly wire to the output jack. On final mounting, I added a thermal pad between the two boards, mainly to keep anything from shorting between the two. This also shows all the components shorter than the volume pot. The IC and socket have about 1mm clearance. Component leads are all trimmed very low to allow pushing the sandwich tight together to fit the case.
Unfortunately, I made a mistake after soldering the two together and had to take them apart. This damaged the power post PCB traces on my board, so had to directly wire the power to the TPA3118 after correcting the mistake. Always check everything before going beyond points of no return... Dang it... Here's the installed boards. There's 3 or 4mm clearance to the bottom of the case. I have to modify the speaker jack to squeeze into this tiny case and use the ring as the sleeve terminal. Sleeve terminals are completely removed from the jack. Very tight fit. And shows the added power lines to the TPA3118 after the damage done. No posts on the left side anymore.
Finished product...
ExpressPCB design shared here https://drive.google.com/drive/folders/1EUGnukuxIq0znhbYBT9JnLQSlYZEh74F?usp=drive_link
There's also a PDF of the drilling points for 1590B and 125B for the PCB.
Last edited:
I'd be interested.
I am working through some ideas I wanna do with mine when it comes in.
Paul.Ruby
Well-known member
Link to my ExpressPCB designs added to my post above and here.
PCB - Google Drive
drive.google.com
Last edited:
Paul.Ruby
Well-known member
Took a stab on ebay at TPA3118 but received fakes. They refunded me right away, so they know they are sending fakes. The pictures in the auction are of real ones and they are watermarked, which made me think they had a bit of pride in what they sell. Nope. I think real ones are almost extinct.
DC 12V-24V TPA3118 BTL 60W Mono Digital Audio Power Amplifier Board Amp Module | eBay
Find many great new & used options and get the best deals for DC 12V-24V TPA3118 BTL 60W Mono Digital Audio Power Amplifier Board Amp Module at the best online prices at eBay! Free shipping for many products!
www.ebay.com
Paul.Ruby
Well-known member
Here's another real TPA3116 board from ebay:
The mute pin is routed to a jumper to ground, so is easy to attach to a switch, power-on delay circuit or what ever.
Pictured here next to the original boards that are all fakes at this point.
This new board is slightly smaller and fits in the base plate of a 125B case.
But, it has a single large capacitor making the total height about 1".
It allows for differential input, which isn't needed. Just ground the - input and use the + input for normal single ended use.
Also found an alternative TPA3118 board that is real but the originals are still available (10/6/2023).
Here's the original and the alternative.
This new one is larger with bigger caps and it does NOT give easy access to the mute pin. Mute is hooked to ground through a 100k resistor, which is R5 in the picture. Not impossible to solder a wire onto but a pain. The originals are still king.
Link to new ones: https://www.ebay.com/itm/264156797486
Link to originals: https://www.amazon.com/dp/B07Q6RGVHQ?psc=1&ref=ppx_yo2ov_dt_b_product_details
DC 12V-24V TPA3116 Mono Channel Digital Power Audio Amplifier Board BTL 100W | eBay
Find many great new & used options and get the best deals for DC 12V-24V TPA3116 Mono Channel Digital Power Audio Amplifier Board BTL 100W at the best online prices at eBay! Free shipping for many products!
www.ebay.com
Pictured here next to the original boards that are all fakes at this point.
This new board is slightly smaller and fits in the base plate of a 125B case.
But, it has a single large capacitor making the total height about 1".
It allows for differential input, which isn't needed. Just ground the - input and use the + input for normal single ended use.
Also found an alternative TPA3118 board that is real but the originals are still available (10/6/2023).
Here's the original and the alternative.
This new one is larger with bigger caps and it does NOT give easy access to the mute pin. Mute is hooked to ground through a 100k resistor, which is R5 in the picture. Not impossible to solder a wire onto but a pain. The originals are still king.
Link to new ones: https://www.ebay.com/itm/264156797486
Link to originals: https://www.amazon.com/dp/B07Q6RGVHQ?psc=1&ref=ppx_yo2ov_dt_b_product_details
Paul.Ruby
Well-known member
This is a bit superfluous but I've always wanted to try out the ATtiny for soft touch control. I made another version of the Volume-only PCB with an ATtiny25/45/85 socket to handle power on delay and soft-touch foot switch. I started with the MAS effect code and implemented mine using the same pins for the controls. MAS effect info starts here:
shop.mas-effects.com
Of course, you could just use the MAS switch with the relay being used in place of the NPN transistor in my schematic below. Just reprogram the ATtiny to add power-on delay or configure it to power up bypassed all the time so it only comes on when told to do so. But, the point of this was to dig in deep even though duplicating what I already have with the discrete soft-touch toggle flip-flop.
Schematic is dirt simple. EDIT: After measuring current draw to be 26mA, the voltage dropping resistor is changed to 470ohm 1/2W. With ~23V input, the resistor drops it to 11V to the 78L05. Then power in the 78L05 is 160mW. EDIT AGAIN: Current varying between 22mA and 34mA. A resistor is futile. Changing to zeners... I'm using two 6.2V zeners on mine to drop from ~23V to ~10V. A single 12V 1W zener was getting a bit hot.
My Express PCB files and the Arduino sketch file for this are here:
drive.google.com
There is a layout putting all the components on the PCB and one for mounting a Digispark rev 3, which is really easy.
I use the Sparkfun Tiny AVR programmer.
www.sparkfun.com
Must install the drivers for the programmer, too.
And follow the hook-up guide.
Arduino IDE version 1.8.5.I could not get any other version of the IDE to work. Have not figured out why.
UPDATE: Arduino IDE 2.2.1 works now. The mistake is that 1.8.5 only needs to "upload" the program but newer versions need "upload with programmer" so the upload button works on 1.8.5 but not 2.2.1. Must instead us "Ctrl+Shift+U" instead of "Ctrl+U"... I'm a noob.
www.arduino.cc
Must add ATTiny support once the IDE is installed. File|Preferences and add these additional board manager links:
Then Tools | Board | Board manager...
And install "attiny by David A. Mellis"
Once installed, if using the sparkfun programmer, tools should look like this...
Board is "ATtiny25/45/85"
Processor is which ever of those three you have.
Clock speed must be internal 16Mhz or the timers won't be correct.
This seems to be a problem with the David Mellis board package.
Programmer is "USBtinyISP" assuming you got the sparkfun drivers installed.
I'm going to try the digispark things, too...
DIY relay true bypass: Soft touch foot switch upgrade for Guitar Pedals
DOCUMENTATION: https://mas-effects.com/relay.pdf This is a true-bypass relay bypass kit for your DIY pedals. This is specifically to engage/bypass an effect. For general-purpose switching, see our other digital relay module instead. This bypass board has loads of features: Soft touch SPST foot...
shop.mas-effects.com
Of course, you could just use the MAS switch with the relay being used in place of the NPN transistor in my schematic below. Just reprogram the ATtiny to add power-on delay or configure it to power up bypassed all the time so it only comes on when told to do so. But, the point of this was to dig in deep even though duplicating what I already have with the discrete soft-touch toggle flip-flop.
Schematic is dirt simple. EDIT: After measuring current draw to be 26mA, the voltage dropping resistor is changed to 470ohm 1/2W. With ~23V input, the resistor drops it to 11V to the 78L05. Then power in the 78L05 is 160mW. EDIT AGAIN: Current varying between 22mA and 34mA. A resistor is futile. Changing to zeners... I'm using two 6.2V zeners on mine to drop from ~23V to ~10V. A single 12V 1W zener was getting a bit hot.
My Express PCB files and the Arduino sketch file for this are here:
PCB - Google Drive
drive.google.com
There is a layout putting all the components on the PCB and one for mounting a Digispark rev 3, which is really easy.
I use the Sparkfun Tiny AVR programmer.
Tiny AVR Programmer - PGM-11801 - SparkFun Electronics
The ATtiny45 and 85 are a couple of really cool little MCUs but did you know you can program them in Arduino? That's right, now you can shrink your Arduino
www.sparkfun.com
Must install the drivers for the programmer, too.
And follow the hook-up guide.
Arduino IDE version 1.8.5.
UPDATE: Arduino IDE 2.2.1 works now. The mistake is that 1.8.5 only needs to "upload" the program but newer versions need "upload with programmer" so the upload button works on 1.8.5 but not 2.2.1. Must instead us "Ctrl+Shift+U" instead of "Ctrl+U"... I'm a noob.
Software
Open-source electronic prototyping platform enabling users to create interactive electronic objects.
www.arduino.cc
Must add ATTiny support once the IDE is installed. File|Preferences and add these additional board manager links:
https://raw.githubusercontent.com/damellis/attiny/ide-1.6.x-boards-manager/package_damellis_attiny_index.json
https://raw.githubusercontent.com/digistump/arduino-boards-index/master/package_digistump_index.json
Then Tools | Board | Board manager...
And install "attiny by David A. Mellis"
Once installed, if using the sparkfun programmer, tools should look like this...
Board is "ATtiny25/45/85"
Processor is which ever of those three you have.
Clock speed must be internal 16Mhz or the timers won't be correct.
This seems to be a problem with the David Mellis board package.
Programmer is "USBtinyISP" assuming you got the sparkfun drivers installed.
I'm going to try the digispark things, too...
Last edited:
Paul.Ruby
Well-known member
Update to previous post... I got the digispark CLONE boards today and they work fine.
These clones do NOT have the fuse set to use the reset pin as PB5. That's a pretty severe limitation.
The amazon link to these boards is here:
The instructions to use them with arduino is here:
iotprojectsideas.com
I might make another version of the PCB that allows these to plug right on the back, just like the amp board plugs on.
EDIT: Since these use the micronucleus bootloader to allow user programming over USB, they have an in-built power-on delay of a few seconds because they first try to connect to a usb port for programming before jumping to the user code. This is certainly no issue for use with a power amp since a power-on delay is a good thing. But, this might be a nuisance for some applications.
EDIT#2: I'm able to use Arduino IDE 2.2.1 now. It works fine with digispark and now with the sparkfun programmer. The mistake I was making is that 1.8.5 would "upload" the program but 2.2.1 requires "upload with programmer" so the upload button doesn't work.
These clones do NOT have the fuse set to use the reset pin as PB5. That's a pretty severe limitation.
The amazon link to these boards is here:
The instructions to use them with arduino is here:
Program Digispark Attiny85 with Arduino IDE
All about an inexpensive and Compact Alternative to Arduino called ATTINY85 so let's learn to Program Digispark Attiny85 with Arduino IDE.
iotprojectsideas.com
I might make another version of the PCB that allows these to plug right on the back, just like the amp board plugs on.
EDIT: Since these use the micronucleus bootloader to allow user programming over USB, they have an in-built power-on delay of a few seconds because they first try to connect to a usb port for programming before jumping to the user code. This is certainly no issue for use with a power amp since a power-on delay is a good thing. But, this might be a nuisance for some applications.
EDIT#2: I'm able to use Arduino IDE 2.2.1 now. It works fine with digispark and now with the sparkfun programmer. The mistake I was making is that 1.8.5 would "upload" the program but 2.2.1 requires "upload with programmer" so the upload button doesn't work.
Last edited:
