Near match resistor values: If the LED calculator says 9.4928371K, you don't need a variable resistor to dial that in exactly because an inexpensive 10K will do fine. Just use its "nearest" resistor estimator.
When in doubt, use less current, less milliamps.
You can drain a capacitor by putting a resistor in parallel with the capacitor. See attached photo.
Attachments
For the power supply building, here's a few examples made for dual secondaries transformer.
Basic power supply
First example is the basic power supply with good performance.
Ordinary CRC power supply
Second example is the CRC power supply with noise filtering added at the cost of increased rail sag. This ordinary CRC power supply works for monobloc chip amplifier.
High Power CRC power supply
Third example is a CRC power supply with added diode that limit the filter sag. The diodes turn on and "brace up" the rails when the total amplifier power passes 50w. This high power CRC power supply is nice for Stereo chip amplifier.
Basic power supply
First example is the basic power supply with good performance.
Ordinary CRC power supply
Second example is the CRC power supply with noise filtering added at the cost of increased rail sag. This ordinary CRC power supply works for monobloc chip amplifier.
High Power CRC power supply
Third example is a CRC power supply with added diode that limit the filter sag. The diodes turn on and "brace up" the rails when the total amplifier power passes 50w. This high power CRC power supply is nice for Stereo chip amplifier.
Attachments
Last edited:
Hi,
does anybody actually realize that ilimzn has shared the most complete and incredibly competent discussion about the TDA7293/94, mentioning all its issues and the remedies for it? His two posts really deserve to be split from the thread and become a sticky topic "getting the max out of TDA7294".
It's the best summary on the chip and how to use it that you will ever find, period. There is effectively nothing I could add to it, other than saying I can back up any sentence he wrote, notably the supply issues. If you precisely follow the design hints he is giving then you'll be rewarded with an amp that sounds as good as it measures...
...that is, you also need to take great care about the layout which mostly means indentify and seperate all the different current loops and make their enclosed area and hence mutual electromagnetic coupling as small as possible, as well as making any I*R drops irrelevant by referencing "GND" to the correct points. Then take care about higher impedance nodes and make their exposed surface area small to avoid electrostatic pickup.
does anybody actually realize that ilimzn has shared the most complete and incredibly competent discussion about the TDA7293/94, mentioning all its issues and the remedies for it? His two posts really deserve to be split from the thread and become a sticky topic "getting the max out of TDA7294".
It's the best summary on the chip and how to use it that you will ever find, period. There is effectively nothing I could add to it, other than saying I can back up any sentence he wrote, notably the supply issues. If you precisely follow the design hints he is giving then you'll be rewarded with an amp that sounds as good as it measures...
...that is, you also need to take great care about the layout which mostly means indentify and seperate all the different current loops and make their enclosed area and hence mutual electromagnetic coupling as small as possible, as well as making any I*R drops irrelevant by referencing "GND" to the correct points. Then take care about higher impedance nodes and make their exposed surface area small to avoid electrostatic pickup.
KSTR,
That would bring up the obvious question - are there members and resources available and/or willing to do what is necessary to develop a board that incorporates those ideas? Perhaps if a base PCB could be posted here, or on something like Google Docs, that development process could begin.
Anyone with good chops on a layout program willing to get the ball rolling ??? There may be the plausibility of sectioning off segments if it is unrealistic for just one person to do the complete base design.
That would bring up the obvious question - are there members and resources available and/or willing to do what is necessary to develop a board that incorporates those ideas? Perhaps if a base PCB could be posted here, or on something like Google Docs, that development process could begin.
Anyone with good chops on a layout program willing to get the ball rolling ??? There may be the plausibility of sectioning off segments if it is unrealistic for just one person to do the complete base design.
Last edited:
Jay also posted an interesting feedback network at post#53
And like Bob, I have a similar question. Where is the double-layer board for it? Due to the mad pinout of the chip in combination with the added complexity, it probably takes an expert PCB layout.
Do you see an easily doable solution for working with the boards that we already have? Adapter? Daughtercards?
Yes, we saw it earlier. That thread is here: http://www.diyaudio.com/forums/chip-amps/3857-tda7293v.html#post890087
And like Bob, I have a similar question. Where is the double-layer board for it? Due to the mad pinout of the chip in combination with the added complexity, it probably takes an expert PCB layout.
Do you see an easily doable solution for working with the boards that we already have? Adapter? Daughtercards?
Last edited:
I am not in a position to judge the validity of any suggestions related to this chip; I think KSTR meant that it might be better to change the path being taken at the moment (by following some advices made by some members he suggested) in order to truly get the best out of this chip. As Bob noted already, this requires a PCB to be designed. I believe this is one thing Bob has been trying to avoid since the beginning of this thread.
I do hope that future enlightening suggestions will come from members who can offer something that offers a single-sided PCB design, schematic simplicity, and components availability, yet flush the best out of this chip, why not? It is not always necessary that a complicated design will yield better results than a simple one.
I do hope that future enlightening suggestions will come from members who can offer something that offers a single-sided PCB design, schematic simplicity, and components availability, yet flush the best out of this chip, why not? It is not always necessary that a complicated design will yield better results than a simple one.
Last edited:
Let me try to clarify. I thought, and still do, that the sound from a simple $30 kit was pretty amazing. What was lacking in the quality of the top end sounded like it could be improved upon by primarily a few component changes. Though I'm still lagging a bit in implementing some of those suggestions, I plan to go through that process soon. I thought this thread might attract other builders of the same or similar minimalistic design (several eBay vendors offer it) and some comparative results and suggestions would be posted. Still think that's a good idea.
It quickly became apparent many advanced DIYers had recognized the potential of the TDA7294 and proposed a much more sophisticated implementation. That was far beyond my expectations and quite encouraging. Someone earlier stated the development of a new board would require a lot of time, commitment and money to pull it off.
What I think I see now is a dual track, one being the collection of work done by many into a concentrated effort to actually create a new high-grade PCB, the other being the original tweaking. I'm just one member with no design experience (though I am trying to learn some layout programs) but am fully aware a new board would require a lot from more than one source/person. I may have missed it, but I didn't see an actual/published board on the thread Daniel linked to this morning.
I experienced the collective effort that took place in the multi-year development on the MyRef amps, feeling it was enjoyable and productive for all involved. I looks like a lot of the theory work has already been done for this chip. Whether that is a track that proves useful here will be determined by the continuing discussions.
It quickly became apparent many advanced DIYers had recognized the potential of the TDA7294 and proposed a much more sophisticated implementation. That was far beyond my expectations and quite encouraging. Someone earlier stated the development of a new board would require a lot of time, commitment and money to pull it off.
What I think I see now is a dual track, one being the collection of work done by many into a concentrated effort to actually create a new high-grade PCB, the other being the original tweaking. I'm just one member with no design experience (though I am trying to learn some layout programs) but am fully aware a new board would require a lot from more than one source/person. I may have missed it, but I didn't see an actual/published board on the thread Daniel linked to this morning.
I experienced the collective effort that took place in the multi-year development on the MyRef amps, feeling it was enjoyable and productive for all involved. I looks like a lot of the theory work has already been done for this chip. Whether that is a track that proves useful here will be determined by the continuing discussions.
I'll happily volunteer for circuit design and PCB layout of some varieties of TDA7294 amplifiers, trying to cover all the important aspects.
Of course a clean implementation of the typical "datasheet page 1 circuit" already gives, say, 80% of the possible perfomance. If we stick to that basic**) implementation a good circuit and layout is not that hard a task but will take time and thought (give me 1..2 weeks or so). A simple trick for easy chip amp layout is either to use an aluminum L-bracket or mount the chip to the heatsink edge, in order to extend the PCB around the chip's pins to any length/area needed.
**) noninverting config. Personally I'd always go for inverting unless one wants to use a passive pre (volume pot >10k) ahead of it or anything exotic which cannot handle a load in the 1k...10k region. Also I'd omit any protection circuitry and I'd DC-couple the whole thing. Except for "gross user fault" this tends to be safe IME.
Eventually I'm heading to at least a 3 month's sabbatical (starting in March) so I'll finally find the time to delve more deeply into DIY projects where I can give full and sustained support (note that I'm planning to publish a composite chip amp if the thingy works out as expected).
Of course a clean implementation of the typical "datasheet page 1 circuit" already gives, say, 80% of the possible perfomance. If we stick to that basic**) implementation a good circuit and layout is not that hard a task but will take time and thought (give me 1..2 weeks or so). A simple trick for easy chip amp layout is either to use an aluminum L-bracket or mount the chip to the heatsink edge, in order to extend the PCB around the chip's pins to any length/area needed.
**) noninverting config. Personally I'd always go for inverting unless one wants to use a passive pre (volume pot >10k) ahead of it or anything exotic which cannot handle a load in the 1k...10k region. Also I'd omit any protection circuitry and I'd DC-couple the whole thing. Except for "gross user fault" this tends to be safe IME.
Eventually I'm heading to at least a 3 month's sabbatical (starting in March) so I'll finally find the time to delve more deeply into DIY projects where I can give full and sustained support (note that I'm planning to publish a composite chip amp if the thingy works out as expected).
Bob - What you describe is what is known as scope creep in the project world. What started out as a few simple mods to a $30 PCB based amp, now has the potential to involve a complete redesign of not only the amplifier PCB, but the PSU PCB too! If we go down that path, then our minimalist $30 amp, now becomes, in all likely, several times over the original cost. As I mentioned in my earlier post, while this may be seen as a challenge for some, the additional investment in time, effort and money, may not justify the very subjective incremental improvements to the sound characteristics of the original amp for some. 
For others, it's like waving a red flag - implying - bring it on!
At least in the DIY community, people can decide for themselves the degree (and extent) to which they implement mods. I know the proper design of a PCB is no trivial matter. It takes a certain degree of knowledge and understanding of electrical characteristics to be done properly. Way outside my level of expertise. Perhaps there are such members that would be up to the challenge. As has been stated, both Daniel and ilimzn have provided a wealth of information relative to improving the 7294 design. And now, KSTR has volunteered his time to redesign the 7294 PCB. So, maybe there will be new 7294 amp in the near future.
Rick
For others, it's like waving a red flag - implying - bring it on!At least in the DIY community, people can decide for themselves the degree (and extent) to which they implement mods. I know the proper design of a PCB is no trivial matter. It takes a certain degree of knowledge and understanding of electrical characteristics to be done properly. Way outside my level of expertise. Perhaps there are such members that would be up to the challenge. As has been stated, both Daniel and ilimzn have provided a wealth of information relative to improving the 7294 design. And now, KSTR has volunteered his time to redesign the 7294 PCB. So, maybe there will be new 7294 amp in the near future.
Rick
redjr,
The "scope creep" you mention should definitely be explored and stated openly and honestly. My first LM3886 project was the BrianGT where the full build cost for two momoblocks was~ $125. My last MyRef Fremen Edition came in around $650 with all the top choice components - and no chassis yet. Such differences in investment dollars might indeed be involved here. The larger outlay proved justified for that project.
Again, I think your question, paraphrasing, "Just what do we get for all that effort" is surely valid and hopefully members will share their views on that here.
The "scope creep" you mention should definitely be explored and stated openly and honestly. My first LM3886 project was the BrianGT where the full build cost for two momoblocks was~ $125. My last MyRef Fremen Edition came in around $650 with all the top choice components - and no chassis yet. Such differences in investment dollars might indeed be involved here. The larger outlay proved justified for that project.
Again, I think your question, paraphrasing, "Just what do we get for all that effort" is surely valid and hopefully members will share their views on that here.
I know from personal experience that scope creep can, and does often derail the original project resulting in unanticipated and unforeseen costs overruns - especially in the business world. My DIY projects - as well as I try to plan them - often involve costs I didn't foresee.
So, it can be a very real component/issue of an overall DIY project too.Now the perceived 'value' issue for mods relative to the investment is surely a subjective factor and will vary with the user. I for one, am more skeptical than most, in searching for the ultimate sonic holy grail. But I don't fault those who try. More power to them! I'm more of an integrator of existing modules to see how I can pull them all together in a well executed, finished project. To me that is very satisfying. I just happen to know my limits when I tackle an amp and the degree to which I start to delve in at the component (or arrangement) level.
Maybe we could start a 'kickstarter' project to defray the cost that any single member might bear in helping make this happen.
Rick
On my way back home in the subway I was thinking about amplifier concepts that make sense doing a new PCB. For the most basic circuit it doesn't make too much sense to me, just hook it up point-to-point (see Daniel's example) or use any of the boards available and see how it goes with any mods (both on PCB tracks and components).
I'm more inclined to do a topology that isn't bog standard and which takes care of the sonically most important issue of having the negative front-end supply (of the chip) at a clean AC ground -- virtually free of supply and load current ripple -- which is not too difficult with an inverting configuration, also having the benefit that there is no common-mode signal on the chip's input pins.
Further, I'm not a fan of fully independent mono-blocks (even when in the same chassis), so a complete stereo PCB with most of the supply components on board seems more adequate for most builders I would think. For DIY we also have the option to make use of unorthodox layout/component placement as we do not need to care about feasibilty for mass production (which means sticking to all SMT parts located on the same side).
Finally I always strive for low complexity and very moderate BOM cost using standard parts (like I have to do in the day job). Design and layout is much more important than using expensive boutique parts IMHO. Using those is the final step of fine tuning if all else is done right already.
Does this sound like a good starting point to everbody?
For evaluation and comparison I'd first need to finish the MyRef FE boards that are still waiting for being stuffed, though... atm I only have one channel of a pre-RC version of it at hand.
- Klaus
I'm more inclined to do a topology that isn't bog standard and which takes care of the sonically most important issue of having the negative front-end supply (of the chip) at a clean AC ground -- virtually free of supply and load current ripple -- which is not too difficult with an inverting configuration, also having the benefit that there is no common-mode signal on the chip's input pins.
Further, I'm not a fan of fully independent mono-blocks (even when in the same chassis), so a complete stereo PCB with most of the supply components on board seems more adequate for most builders I would think. For DIY we also have the option to make use of unorthodox layout/component placement as we do not need to care about feasibilty for mass production (which means sticking to all SMT parts located on the same side).
Finally I always strive for low complexity and very moderate BOM cost using standard parts (like I have to do in the day job). Design and layout is much more important than using expensive boutique parts IMHO. Using those is the final step of fine tuning if all else is done right already.
Does this sound like a good starting point to everbody?
For evaluation and comparison I'd first need to finish the MyRef FE boards that are still waiting for being stuffed, though... atm I only have one channel of a pre-RC version of it at hand.
- Klaus
Hey Klaus,
You really sound like what the doctor ordered. The first thing that comes to mind is whether you (and those who may just be watching) have actually heard what this chip produces. I looked back through the thread but couldn't tell for sure.
I've become a huge fan of SMD components where useful to reduce both costs and size - from a builder's perspective. Dario Inserra's combinations on the FE have certainly benefited that design.
Please do finish your MyRef project. In all honesty, I'm here because I am totally thrilled with that amp, but realized my praise of it was somewhat empty because I hadn't actually listened to anything other than LM3886 based amps since the beginning of this second (did Dynaco and Hefler kits years ago) venture into DIY - about three years.
So I set out to build something from the Pass community and a TDA????, as those were the categories about which I read the most raves. With the completion of Nelson's Amp Camp beginners project (a full blown F5 based build near completion) and this TDA kit it became clear that there wasn't going to be (for me) any need to determine the "best" power amp - more correctly a desire to take each approach as far as my skills and wallet would allow. The characters of the three categories are so unique that having several amps now is justified in my mind. It's similar to having a small car collection, a couple sets of golf clubs or even a nice selection of favorite wives
I don't know if it is jumping the gun, but it might be useful to establish a sign-up sheet just to determine the level of interest before the venture actually begins.
You really sound like what the doctor ordered. The first thing that comes to mind is whether you (and those who may just be watching) have actually heard what this chip produces. I looked back through the thread but couldn't tell for sure.
I've become a huge fan of SMD components where useful to reduce both costs and size - from a builder's perspective. Dario Inserra's combinations on the FE have certainly benefited that design.
Please do finish your MyRef project. In all honesty, I'm here because I am totally thrilled with that amp, but realized my praise of it was somewhat empty because I hadn't actually listened to anything other than LM3886 based amps since the beginning of this second (did Dynaco and Hefler kits years ago) venture into DIY - about three years.
So I set out to build something from the Pass community and a TDA????, as those were the categories about which I read the most raves. With the completion of Nelson's Amp Camp beginners project (a full blown F5 based build near completion) and this TDA kit it became clear that there wasn't going to be (for me) any need to determine the "best" power amp - more correctly a desire to take each approach as far as my skills and wallet would allow. The characters of the three categories are so unique that having several amps now is justified in my mind. It's similar to having a small car collection, a couple sets of golf clubs or even a nice selection of favorite wives
I don't know if it is jumping the gun, but it might be useful to establish a sign-up sheet just to determine the level of interest before the venture actually begins.
Last edited:
Bob,
Well, besides working with these chips for years on a private/hobbyist level, a reasonable amount of my day job in recent years was designing amplifiers around TDA729x series chips, as well as LM3886 (+LM4780). Most of them have been constrained by lowest BOM cost, smallest PCB real estate and such issues (in the context of the complete products they went into), therefore they do not always exploit the full capabilites of the chips.
So, that's where DIY shines, no real cost/feasibilty contraints and most importantly, no strict product deadlines that tend to block any creativity and experimenting. With the boss standing behind you reminding you to send out the PCB mfg data tomorrow as scheduled you really think twice before ripping up an almost finished layout in the quest for an even better routing... and then when the PCB arrives there seldom is time to fully debug and optimize it in all details (let alone the circuit, unless with real faults) before starting production....
Today, I also finally have all the professional lab equipment to dig down into what's really going on and then try correlate why it sounds the way it does...
Interesting
I would stick to SMDs on one side anyway... if some rework will become necessary it will be a pain desolder SMD parts among TH ones... strictly IMHO.
Yeah... I've some interests in reading your comments and measures...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.