Very nice build there. Thanks for sharing. Also thank you for elaborating a bit on your build experience.
If you haven't already, double-check that the ground terminal of the IEC inlet is connected to the chassis. The grounding lug generally does a good job of digging through the anodization even if you don't remove it, but it's always good to check.
I'm glad to hear the amp came together nicely for you. Enjoy!
Tom
If you haven't already, double-check that the ground terminal of the IEC inlet is connected to the chassis. The grounding lug generally does a good job of digging through the anodization even if you don't remove it, but it's always good to check.
I'm glad to hear the amp came together nicely for you. Enjoy!
Tom
This works well, 4 users over here have had good results with it and myself with Mod 86.
https://www.diyaudio.com/forums/gro...ter-toroidal-transformers-14.html#post5728917
New group buy just started.
Yes, a good connection is present - checked and double-checked.
And thanks misterdog, Tom and BrianL for the thoughts on the DC trap. The humming amp is destined for another location where the listening position is much further from the amplifier, so I'm not sure if the hum will remain audible. I'll certainly look at your suggestions though if it continues to irritate in the future.
Nope. No more THAT Driver. I may have a replacement for it at some point. If I do, it'll be a fully assembled module as all the components will be SMD. I need to look at the math, ask folks what they'd be willing to pay, and see what makes sense.
BTW: My website shows the inventory. If a product is marked "in stock" there's at least five left. Once the inventory drops to five, my website will show the actual number in stock.
Tom
BTW: My website shows the inventory. If a product is marked "in stock" there's at least five left. Once the inventory drops to five, my website will show the actual number in stock.
Tom
Last edited:
Tom,
I hope you consider replacing the THAT Receiver at some point as well.
After my successful build of the Modulus-86 (post #2580 for the prospective amp builders out there), I was hooked on amp building. My next project was the First Watt Sony VFET, but I made sure to purchase a Neurochrome THAT Receiver to have a balanced connection from my preamp/DAC to the amp. The VFET project stayed at the back of my workbench for two years, because every time I listened to my audio system with the Modulus-86, I realized that I was in no hurry to add another amp.
Nevertheless, a differential receiver is a good addition to any amp project. While it may be unnecessary for your Modulus-x86 offerings, it would complement your power supply, soft start, and speaker protection boards for those builders using these circuits with other amp boards.
Curtis
I hope you consider replacing the THAT Receiver at some point as well.
After my successful build of the Modulus-86 (post #2580 for the prospective amp builders out there), I was hooked on amp building. My next project was the First Watt Sony VFET, but I made sure to purchase a Neurochrome THAT Receiver to have a balanced connection from my preamp/DAC to the amp. The VFET project stayed at the back of my workbench for two years, because every time I listened to my audio system with the Modulus-86, I realized that I was in no hurry to add another amp.
Nevertheless, a differential receiver is a good addition to any amp project. While it may be unnecessary for your Modulus-x86 offerings, it would complement your power supply, soft start, and speaker protection boards for those builders using these circuits with other amp boards.
Curtis
What I have in mind would replace both the THAT Driver and the THAT Receiver.
I agree. Using a differential link is a good way to push all the ground-related crap down by some 90 dB.
I agree that some "preamp stuff" would be a nice addition to my current portfolio. The real question is whether people will be willing to pay for any additions to my portfolio. Given that any product that involves SMD parts will be machine-assembled, I'm likely looking at a $3-4k investment (plus my time) to get started.
Making an investment like that is not something I do without at least thinking about it and doing the math. It also brings up questions about whether I could get a better return on investment by investing elsewhere, for example in a fully built headphone amp retail product.
My top priority right now is to wrap up my psychology degree. My last exam is exactly one month from today. I hope to be able to take some time off after I graduate. Maybe pitch my tent in the mountains for a few days. After that I expect to be focused on audio stuff full time.
Tom
Good luck with that. I fear you will have your new degree before I power up my amplifiers...
I sure hope so!
Nothing, except it used the LME49710NA, which has been obsolete and out of production for a few years now.
All the modern opamps are in SMD, so if you want to maintain performance, you'll have to accept SMD. At that point I might as well take full advantage of the modern parts and improve performance. This means going all-SMD. As I've noted elsewhere, that means automated assembly. I can deliver a product that delivers much more consistent and better performance at a lower price point by going with automated assembly. It also drastically lowers my support cost. Some will be annoyed that they don't get to solder itty-bitty parts (and lose half of them to the Carpet Gods). Such is life.
Neither the THAT Driver nor the THAT Receiver were strong sellers. That's another reason I need to do the math before getting sucked too far into that project. I do have an idea for a sort of "Puzzle Preamp". Buy some combination of boards and you get a fully customizable preamp. That'd be cool ... after graduation.
Tom
Last edited:
automated assembly kills DIY imo,
people just have to learn how to solder smd, so-8 is the most popular package, 1206/0805 are easy, no need for smaller for DIY unless you have a special parts to do. DFN is a PITA, so avoid that one if you can. BGA out of the ? for DIY.
funny how THT film caps are in great supply with new ones coming out all the time, so it is usually a mix tech world we live in, has been like that since the late 80's when smt was first coming on. I worked at one of the first smt lines in Canada back then. Motorola had a FR4 and a hybrid ceramic substrate line going beside each other. I still have parts that gathered off the floor when fixing the Excellon pick n' place machines. Codes on the sot-23 parts are a PITA to decipher. I have never figured out why they do not label MLCC, I remember Motorola forced the mfg's to print codes on them for inspection.
Playing Inagodavida on the radio almost the same time frame.
people just have to learn how to solder smd, so-8 is the most popular package, 1206/0805 are easy, no need for smaller for DIY unless you have a special parts to do. DFN is a PITA, so avoid that one if you can. BGA out of the ? for DIY.
funny how THT film caps are in great supply with new ones coming out all the time, so it is usually a mix tech world we live in, has been like that since the late 80's when smt was first coming on. I worked at one of the first smt lines in Canada back then. Motorola had a FR4 and a hybrid ceramic substrate line going beside each other. I still have parts that gathered off the floor when fixing the Excellon pick n' place machines. Codes on the sot-23 parts are a PITA to decipher. I have never figured out why they do not label MLCC, I remember Motorola forced the mfg's to print codes on them for inspection.
Playing Inagodavida on the radio almost the same time frame.
Last edited:
I agree with rsavas. To be honest I even prefer to solder SMD parts as there is no hassle with cutting leads and "flat" PCB is easier to handle. I'm perfectly fine even with 0603 but I guess that even one part in big SMD package is dealbreaker for lots of DIYers. THT is much easier to handle with shaky hands or bad eyesight.
In that case, SMD kills for-profit DIY businesses, such as mine. This will leave DIY for those with eyesight and patience good enough to take on an SMD build, and hobbyist circuit designers who can sell their goods at prices low enough that people won't mind throwing a non-working build in the trash if they can't fix it.
Fact is that if you hand-solder a stack of SMD boards, you will get considerable board-to-board variation in performance. I know because I've measured it. That's not consistent with my philosophy of providing state of the art circuits. You also get weird support calls, such as, "it sounds distorted in one channel" from your DIY builders. Try debugging that via email! In that specific build, the issue was a ceramic cap that had cracked during hand-soldering. The crack wasn't visible to the naked eye, but the cap fell apart when de-soldered using hot tweezers.
I cringe at the DIY Audio posts from hobbyist DIY vendors with the terms "frying pan solder reflow" and "high-end" or "high performance" in the same sentence. The folks who make those claims have obviously never measured their circuits or taken a circuit to mass production.
Some may argue that I should have an "entrance exam" or somehow vet buyers so only experienced SMD builders get to buy the bare boards. That might work in theory, but I don't enjoy being a cop and many don't enjoy dealing with cops, so it seems like the wrong direction to take a business. In addition, I've had beginner builders who took on the Modulus-286 Rev. 1.1 as their first SMD build, and who were successful, while more experienced builders ran into trouble with their builds.
Splitting the project into fully built and bare board options also increases the cost for the fully built option, which increases the project cost overall. Having two product options also does nothing to address the support cost of the SMD builds.
It is very, very, very difficult to debug SMD circuits via email. I've been successful so far, but it has been insanely stressful and very time-consuming every time. If I wanted that stress level, I would have stayed at TI. I do not want to spend my life that way.
By letting SMD builds be fully assembled, I can deliver better and more consistent performance at a lower cost. It also allows me to dedicate more of my time to circuit design rather than support.
Basically my options are:
- Through-hole only -> lower performance, issues with parts availability, no access to modern parts.
- SMD, bare PCBs -> Support cost too high and I will hate life.
- SMD, pre-assembled boards -> High-end audio DIY is possible (if going fully assembled does not kill DIY).
- Use pre-assembled SMD -> Get good performance. Pre-assembly kills DIY, so I drop DIY and make retail products instead.
All true. All my SMD circuits are 0805 or larger except for maybe a few SOD-123 diodes.
I agree. I seem to be able to complete SMD builds much faster. It's a bit annoying when a part goes 'tink' off the tweezers, but that's easily solved by ordering a few extra of the "trouble parts" (like those SOD-123 diodes I mentioned earlier).
I can solder 0603 or 0402 without aids, though, my work is a lot prettier if I use a microscope. I prefer 0805 or larger as all the parts are marked with their values - except ceramic caps. Some 0603 parts are marked with their actual value. Others just have some code on them.
That's certainly the case for some.
To be clear - and to approach the actual topic of this thread - the Modulus-86 is here to say. I'll keep making boards for as long as I can get parts. Currently, all parts are in production and available.
Tom
I hear you Tom, you have a good understanding of what is involved. Support is a challenge, people do not want to admit to there screw ups and it makes it that more difficult with remote troubleshooting.
I am not sure how much sub-contract mfg experience that you have done. In my past life I have worked in production test/support for a few of them and as a pcb designer/HW EE. We mfg'd at a number of them, Celestica, Samina/SCI etc. It is not a cake walk, it adds a whole new level to a companies operations. One small mistake can get multiplied real fast. Some quick thoughts to ponder about sub-contract mfg for you.
1) testing an smt/mixed tech assembly is a big one, how? ict, functional? Partially assembled are challenging. need fixturing?
2) minimum production runs, need production forecasting. component availability
3) components need to be in waffle packs, tubes/rails, reels.
4) need big runs to reduce cost and eat setup/NRE charges.
I could go on, but you should get my drift of what the challenges are.
I am not sure how much sub-contract mfg experience that you have done. In my past life I have worked in production test/support for a few of them and as a pcb designer/HW EE. We mfg'd at a number of them, Celestica, Samina/SCI etc. It is not a cake walk, it adds a whole new level to a companies operations. One small mistake can get multiplied real fast. Some quick thoughts to ponder about sub-contract mfg for you.
1) testing an smt/mixed tech assembly is a big one, how? ict, functional? Partially assembled are challenging. need fixturing?
2) minimum production runs, need production forecasting. component availability
3) components need to be in waffle packs, tubes/rails, reels.
4) need big runs to reduce cost and eat setup/NRE charges.
I could go on, but you should get my drift of what the challenges are.
I hope so too.
Last edited: