Hello all
I am looking for recommendations on a small power class D that can be reasonably easily soldered with discrete through components other than the chip, and hopefully there is a chip that isn't too difficult to solder. There are plenty of prebuilt boards with various chips out there, but I want to try my hand at incorporating a power supply on the same PCB and make it easily conform to a stereo or bridged mono depending on use at hand
For me, it would be nice to have some sort of switching power supply all on one board
Around 50wrms into 4R x 2 channels is plenty for what I would do with this. If anyone takes an interest in a project like this, then I also welcome any type of group effort on the PCB and parts. I am also happy to have a small batch of PCBs made and send to contributors for cost plus postage
My reason for wanting such a unit is to use it in a bass amp that I am building for sale. It contains two biamped drivers, and one is a high-passed full range that needs this smaller amplifier. For the bass driver, I am sourcing the amplifier board from a forum member, Eric. This board has a built-in PSU. This is a small amp using 6.5" drivers and my jfet pre sections
Thanks and regards
Randy
I am looking for recommendations on a small power class D that can be reasonably easily soldered with discrete through components other than the chip, and hopefully there is a chip that isn't too difficult to solder. There are plenty of prebuilt boards with various chips out there, but I want to try my hand at incorporating a power supply on the same PCB and make it easily conform to a stereo or bridged mono depending on use at hand
For me, it would be nice to have some sort of switching power supply all on one board
Around 50wrms into 4R x 2 channels is plenty for what I would do with this. If anyone takes an interest in a project like this, then I also welcome any type of group effort on the PCB and parts. I am also happy to have a small batch of PCBs made and send to contributors for cost plus postage
My reason for wanting such a unit is to use it in a bass amp that I am building for sale. It contains two biamped drivers, and one is a high-passed full range that needs this smaller amplifier. For the bass driver, I am sourcing the amplifier board from a forum member, Eric. This board has a built-in PSU. This is a small amp using 6.5" drivers and my jfet pre sections
Thanks and regards
Randy
TPA3116D2 reference schematic looks almost DIY doable 🙂
Wonder if there is a 26v dc PSU reference schematic of similar difficulty level
I'll have a go at a one PCB diy solution
Wonder if there is a 26v dc PSU reference schematic of similar difficulty level
I'll have a go at a one PCB diy solution
Looking for some DIY-friendly chip: To my knowledge TPA3125 is the last THT-class-d amp from TI.
Hi Randy, for Class D diy builds it looks like you will have to learn smd soldering. Here is a possible project:
https://www.diyaudio.com/community/threads/transformer-input-tda8932-mono-amp.357678/post-6283660
You can order kits here and read about member's listening impressions:
https://www.diyaudio.com/community/...8932-25w-8r-mono-amp-kits.359193/post-6318454
Interesting project, thank you for bringing it to my attention. SMD actually seems easier in concept, as long as visual aids are used and have some appropriate equipment. I have plans to add workshop equipment for this and then this will be a viable project for me
The TPA3116D2 reference schematic looks easy enough to implement with thru-hole parts. And I may have found a complementary reference schematic for a PSU
PSU reference design
I have also found a source for 22v-0-22v 200w ring cow (coolest thing a toroid ever got called 🙂 )
About AUD $40 delivered, which tempts me to try this for a simpler approach. Trying to make a TPA3116D PCB containing the capacitors and rectifier seems viable
Here is a Yummy thot!
IRS2092 - scaling
Same shop has a 24v-0-24v 600w and 1600w transformer, I could possibly land with a decent small amp as well as use boost and get the higher voltage to turn this into a larger power amp
I found a schematic for a SMPS that seems easy to make and could remove the need for a power transformer. The circuit looks easy enough to add to the amp PCB
These two choices appeal most at the moment. I can be a good parrot and follow the reference circuits and try something with thru parts and see if it goes up in smoke before succumbing to SMD
By placing the two chips, transformers and a boost circuit in the part's bin, we can have three versions
DIY70D (TPA3116D2)
DIY120D (IRS2092)
DIY200D(IRS2092)
The numbers represent a clean mono output
I also have a CNC router and 3D printer to make custom parts and fittings following this design and make it available. Things like mounting plates and IO plates. Front panel, knobbery holes. I think I have picked two good chips for something we all can turn into something very DIY community accessible
Think about having an open source type community design based around a reference chassis in modular format that anyone can cut, print or buy and a parts list that anyone with skill level can again make or buy. I am shifting into working retirement so can support such a venture full time with the craft and execution skills that I have. I shut down a very successful small manufacturing business of many years as my consciousness would no longer allow any involvement in a blood sport. The idea of doing things like these audio project appeals greatly as its pure innocent joy to the soul
Would be good to hear some thots from folks experienced in this regarding the viability of following this path. I would prefer to get some PCBs made and use highly regarded components
For anyone interested
I propose starting with a baseplate that can accommodate either chip design and additional width option for a stereo. This baseplate can then either be fitted to active speakers or have sides, top, front and back panels attached for a self-contained power amp. The baseplate will have accommodations for a transformer, a chosen community based preamp/tone board. I'll design one and present it here later. I'll pick some chassis options like fans too. Base material would be aluminium sheet with iron foil option. Both can be worked on CNC and laser. Other panels will be combination of wood, acrylic and aluminium with iron foil sandwich. I am skilled with lamination and led edge lighting and fibre optic lighting effects
As time permits, I'll try to marry the reference schematics to a PSU circuit for the proposed transformer but would prefer someone who actually knows what they are doing to take charge of the circuit details. I'll stick to the general arrangement, chassis work, graphics design, file sharing type things as I am well versed in that
What do you guys think? Thanks and regards
Randy
Parrot
Slight change in naming convention. This amp is all about parroting reference design elements in DIY, so the name Parrot appeals to me and the idea of drawing up a badge for it, 🙂 . I'll be cooking up a bunch of bespoke chassis parts and will help catalogue parts better
Aims - Should Accommodate:
Versions
Parrot 70
Parrot 120
Parrot 200
x2 suffix will designate stereo
I have limited ability with it, but I should be ok to at least draw up the factory reference circuit in LTspice. If a spice guru wants to take charge of this then you would be more than welcome. I'll take main initiative in general arrangement and chassis and I think anyone interested in this would find my design ideas appealing and will welcome open source chassis files for it
Looking forward to hearing some thots on this. Thanks and regards
Randy
Slight change in naming convention. This amp is all about parroting reference design elements in DIY, so the name Parrot appeals to me and the idea of drawing up a badge for it, 🙂 . I'll be cooking up a bunch of bespoke chassis parts and will help catalogue parts better
Aims - Should Accommodate:
- Factory reference evaluations boards for TPA3116D2 and IRS2092 and their clones
- The DIY thru hole reference board that results from this thread query
- DIY reference SMPS or
- Transformer based supply DIY reference schematic
- Community selected reference pre stage
- H2
- Open source design files uploaded for chassis elements
Versions
Parrot 70
Parrot 120
Parrot 200
x2 suffix will designate stereo
I have limited ability with it, but I should be ok to at least draw up the factory reference circuit in LTspice. If a spice guru wants to take charge of this then you would be more than welcome. I'll take main initiative in general arrangement and chassis and I think anyone interested in this would find my design ideas appealing and will welcome open source chassis files for it
Looking forward to hearing some thots on this. Thanks and regards
Randy
Hi by now the TPA3116D2 is yesterday's news. Please have a look at the recent Infineon IC's. Class D is still developing.
SMD is not a novelty anymore since at least 30 years and it is a necessity with class D. While many parts can be through hole some parts really need to be SMD.
SMD is not a novelty anymore since at least 30 years and it is a necessity with class D. While many parts can be through hole some parts really need to be SMD.
The basic idea is good but the devil is in the details 🙁
Say you stick to TPA311x simply because of wide use and availability, you will still need to design your own PCB for it.
Not easy in high frequency switching stuff.
A conventional "iron" PSU is easy and you can get transformers for various suppliers, only needed data is voltages and currents needed to get one from a shelf, ready to use.
But an SMPS needs first to be designed, (and developed, it´s not as straightforward as a classic one), its PCB will also need to be designed (and tweaked) and to make matters worse, the ferrite core transformer will need designing, tweaking, and then blueprints sent to some OEM supplier ... I guess significant orders involved (at least hundreds).
Not easy for prototyping or small scale production.
Say you stick to TPA311x simply because of wide use and availability, you will still need to design your own PCB for it.
Not easy in high frequency switching stuff.
A conventional "iron" PSU is easy and you can get transformers for various suppliers, only needed data is voltages and currents needed to get one from a shelf, ready to use.
But an SMPS needs first to be designed, (and developed, it´s not as straightforward as a classic one), its PCB will also need to be designed (and tweaked) and to make matters worse, the ferrite core transformer will need designing, tweaking, and then blueprints sent to some OEM supplier ... I guess significant orders involved (at least hundreds).
Not easy for prototyping or small scale production.
I just can't see an easy DIY implementation with those. I did read all the datasheets. If you have a strategy for an easy implementation than I'll be all for it, once I am set up for SMD
Hey man, you might have missed the previous posts. I have found a supplier for transformers that ships to Australia for around $4 and transformers are $36, $80 and $90 for the 200w, 600w and 1600. So myself and any interested Australians, this would be a bargain. One of the aims is for the chassis to fit a transformer power supply in these three ranges
I have come across a reference schematic for a SMPS that will suit the TPA3116D2, as well as one for the IRS2092
I don't have any problems blowing up a circuit or two to see if it can work with thru hole and would welcome any help
The beauty of the TPA3116D2 is the number of discretes needed to hang off the chip, and their roles. Which one of these sub chip circuits would cause switching issues that you mentioned?
As I proposed, a transformer based design first, with the continued efforts to implement SMPS with thru holes as an alternative to transformers. Has to be a worthy cause
Self answering, isn´t it?
You´ll need to design and make/order your PCB, unless you can find unpopulated ones somewhere.
I didn´t.
I said "iron" PTs are easy.
You found generic ferrite ones?
That fit your voltage requirements?
If so, please post direct links.
Also schematics for them.
Don they also offer unpopulated PCBs for you to assemble?
Post links to all, I am not aware of any, yet never too old to learn.
Your post is still confusing, are you talking iron or ferrite transformers?
Please post links.
Can you please post it here?
Does it include a PCB design? (Gerbers and such)
Does it have transformer data? (core - bobbin - windings - insulation) so you can wind or order your own?
Fi9ne, but to blow them you first need to build them.
Do you have what it takes? (on various levels).
Have you ever built a Class D amplifier or an SMPS at component level?
Hey man, it's a two-way street. Previously, I have asked many questions about many projects that I had been starting without constructive assistance and have not returned to show what I ended up with, as I don't see the point of only feeding lurkers with schematics and links. Very few ever decide to get involved, and it is usually with something very pessimistic or dismissive. For example, instead of being dismissive, if someone offered to examine to see if the power input of the chip can handle holeys then it's an encouragement to share found resources to the project regarding cheap transformers and SMPS schematics
It would be great for a knowing person to step in and examine the TPA3116D2 and show which parts of the circuit is very suitable for DIY and which parts can only work with SMD. There are options to get partially fitted boards too. One of the vendors here could possibly offer a board with the switching circuitry fitted, and we can mess with our choice of input and output discretes?
You have to appreciate that some of us prefer to make our things if we can. I have never built a class d or smps, does that mean I should withdraw from the attempt or not even inquire to see if anyone else is like minded? I find it difficult to read the actual meaning and tone of your post. More so because I regard you as one of the real knowing types who does appear to enjoy getting into finding means and making things happen
It's easy enough for me to stick to good production boards, like the factory eval boards. Whether for personal or commercial projects. I have no real need to inquire or share there. But to me, it's like buying a boat hull and an outboard and bolting them together when I know that I can try to make a boat that looks like a lambo on water. Took me a whole year to learn and make, but now I have a boat that looks like a lambo on water, exceeds design objectives, easily replicable. In learning outboard performance predictions, I also got to learn about electric and air props, air foils and such, so now that has evolved to another concept of developing airlift and water separation differently to traditional WIG. Soon a manned vehicle will eventuate if my model trials go well
All the threads about these chips seem to be about commercial boards or SMD projects. Seems no one is even confident enough in their ability to want to make a real attempt, or such beings do not have need to DIY something like this
I think it's enough energy spent on the why's of can't for this project without even exploring a single means to achieve. This is not a thread about why we can't, it's a thread about how we can
The transformer I mentioned is a toroidal with taps for main CT secondary as well as 15v and 12v aux windings. My knowledge is only that a toroidal is great for power amps. I have never stopped to check the core material and didn't know I had to
That shop's price structure makes for interesting shopping. Quantity of one has a shipping of around $4 but adding a second makes the shipping over $100. So each ordered separately makes for a cheaper per-unit cost. These types of shops can buy some amazing shipping packages that work great with weight limits
Someone has published a SMPS schematic to the public that has been built over and over and the guy even amazingly keeps helping anyone inquire any details or troubleshooting and modifying
Feed the lurkers with zero project constructive input?
"Maybe it´s rigged but it´s the only poker game in town" 😉
Meaning: the Forum is what it is 🙂
Most people are average, quite a few happen to have experience and knowledge and offer useful help, maybe a few others are, as you suspect, lurkers or time wasters ... who knows?
In your case I think you really want to build it and have shown it before, aren´t you the one who built a beautiful Bass for his young Daughter? Unless I am confusing you with someone else.
As of this project, I am NOT telling you NOT to do it, just (in good faith) mentioning a couple practical problems you will meet:
* only by sheer chance (although not impossible), some site or magazine could have published such a project, obviously offering PCB artwork, suggested suppliers, etc.
Thinking Elektor Magazine, the famous Australian one etc.
* maybe somebody offers an unpopulated PCB so you finish it at home?
Again, schematics are fine, but not enough, and high frequency switching PCBs make seasoned designers hair turn grey 🙁
Good luck and keep us updated.
PS: as of NON SMD components modern projects: not too sure about that.
Meaning: the Forum is what it is 🙂
Most people are average, quite a few happen to have experience and knowledge and offer useful help, maybe a few others are, as you suspect, lurkers or time wasters ... who knows?
In your case I think you really want to build it and have shown it before, aren´t you the one who built a beautiful Bass for his young Daughter? Unless I am confusing you with someone else.
As of this project, I am NOT telling you NOT to do it, just (in good faith) mentioning a couple practical problems you will meet:
* only by sheer chance (although not impossible), some site or magazine could have published such a project, obviously offering PCB artwork, suggested suppliers, etc.
Thinking Elektor Magazine, the famous Australian one etc.
* maybe somebody offers an unpopulated PCB so you finish it at home?
Again, schematics are fine, but not enough, and high frequency switching PCBs make seasoned designers hair turn grey 🙁
Good luck and keep us updated.
PS: as of NON SMD components modern projects: not too sure about that.
That would be me, I did that bass amp and guitar for my daughter. Thank you for the compliment
Links that you requested
Transformers - link suitable units for TPA3116D2 and IRS2092
SMPS schematic suitable for IRS2092
I think both the transformers and SMPS above can go into the parts bin and between them cover the three Parrot amps that I propose. I may have a SMPS design for the TPA3116D2, but too early to speak on that yet. I want to know first if I wouldn't be infringing in any way. If I find it is already in the public domain, then will present it here
I'll get cracking on ordering a TPA3116D2 reference BOM in mostly holeies as well as any suggested supply filter caps and diodes plus the 200w toroid. I mean, look at the thing below, 29 parts in some of the simplest layout I have ever seen. It looks quicker to hang a chip off a heat sink and direct solder the caps and resistors to each other without a PCB. Looking at the eval PCB, point to point with cut legs will be shorter than most traces
Would a breakout PCB for the TPA3116D2 be a genius idea? With matching rows of holes spaced out from each other to allow component attachment. For this approach, I can make the baseplate match part anchoring and heat sink. I can do this by designing terminal strip mounts into the baseplate that the passives connect to, while their other ends are attached to the breakout PCB. The chip would be mounted upside down with the thermal pad contacting the aluminium baseplate. I feel a PCB like this would be a good resource for the DIY community and something that can be ordered in numbers for value already stuffed with the chip
Maybe someone who also wants to try this and knows more can help with a holey PCB? If not, then I'll try to use some tools online and see what I can do
This below is the reference mono schematic that I find attractive for 70w cleanish power Parrot70 and x2 for the Parrot70x2
Now imagine an industry standard community developed front panel inlay plate containing the knobery holes, print, LED holes and effects that can be made from any fancy material. Files either downloaded and 3D/laser/CNC'ed at home or ordered from someone equipped. Those who want to customise can set their holes to suit on that inlay plate and that plate drops onto a matching chassis or another of choice
This bird is hatching and will fledge and fly in one form or another, I welcome any collaboration. Makes me smile, thinking of all the varied and colourful Parrots that can emerge 🙂
Last edited:
Thanks 🙂
I´ll browse those links slowly, there´s a lot of data in there.
EDIT: Still dazzled : NO WAY they can ship ,say, 5 to 10 lbs of iron and copper around half the World for:
Do not understand their trick, obviously there must be one.
Currently waiting for a small simple oscilloscope , price $22, shipping $9 ... for a pocket size 10 ounce plastic case.
How can it cost twice as much as shipping 10-20 times that weight?
Incredible.
I´ll browse those links slowly, there´s a lot of data in there.
EDIT: Still dazzled : NO WAY they can ship ,say, 5 to 10 lbs of iron and copper around half the World for:
and yet it´s there, black on white.
Do not understand their trick, obviously there must be one.
Currently waiting for a small simple oscilloscope , price $22, shipping $9 ... for a pocket size 10 ounce plastic case.
How can it cost twice as much as shipping 10-20 times that weight?
Incredible.
Last edited:
Yes, International shipping is crazy.
When I send a single speaker sample to neighbouring Brazil I pay around U$65 for a lonely 12" through EMS (cheapest option) , somewhat higher through the classic Couriers, but once product is accepted I send pallets by land, through an International truck company,, which charges me a whopping U$350 ... for 1 cubic meter which amounts to 63 speakers (7 layers of 9 ), or U$550 for two cubic meters=124 speakers.
They do not care about total pack weight or how many units it includes, only "how much floor space it takes in the truck"
Sea shipping from Buenos Aires to Santos (São Paulo) costs about the same but Customs processing is somewhat more complicated.
I guess your transformers must travel in bursting-at-the-seams containers, so shared individual freight cost is peanuts.
When I send a single speaker sample to neighbouring Brazil I pay around U$65 for a lonely 12" through EMS (cheapest option) , somewhat higher through the classic Couriers, but once product is accepted I send pallets by land, through an International truck company,, which charges me a whopping U$350 ... for 1 cubic meter which amounts to 63 speakers (7 layers of 9 ), or U$550 for two cubic meters=124 speakers.
They do not care about total pack weight or how many units it includes, only "how much floor space it takes in the truck"
Sea shipping from Buenos Aires to Santos (São Paulo) costs about the same but Customs processing is somewhat more complicated.
I guess your transformers must travel in bursting-at-the-seams containers, so shared individual freight cost is peanuts.
Ok, thru hole design with the TPA3116D2 is practical and already very well done and documented on this forum. I don't know why you guys are saying otherwise. There is a commercial board designed by a forum member here, which is for sale everywhere. There are also commercial thru hole designs for the IRS2092. Looks like the only thing standing in the way is bad advice
If I can somehow manage to make a PCB design, I am all for ordering some with the chip already fitted. Is anyone interested in something like this? A mono board for either chip?
If I can somehow manage to make a PCB design, I am all for ordering some with the chip already fitted. Is anyone interested in something like this? A mono board for either chip?