I've finally had enough of your baseless arrogance and rudeness. Most people I know in our age group have picked up enough life experience to not be cursed with those anymore. I was unfamiliar with you, and wanted to judge for myself what some of the others were saying about you. Unfortunately, you have fully confirmed them. I'm done with you.
Those look great George! Πολύ ωραία !
I have an AMP 10 kit that is still not built. I'm a bit disappointed to hear the HF is not as good as on the AMP 6. Is the HF not as extended, or perhaps not as detailed as on the AMP6? Is it very different, or just a little?
By the way, I lived and worked in Athens during 1995-1996. I was the designer and Engineering Director for a large military radio communications system commissioned at the Army base in Menidi. I love Greece so much, and that was the best year of my life!
I met some DIY audiophile friends there, and we used to frequent a beer bar named "TAR" where the owner was also a DIY audiophile. He had a home-made KT88 push pull amplifier driving home-made speakers at either end of the bar. There was a DJ there at night who spun excellent music on a good turntable. It was great fun to sit and drink beer while listening to good quality audio at the same time. Just before I left Greece, the owner of TAR showed me some VT-4C (211) triodes that he was going to build into a single-ended amp for his bar.
My best greetings to you in the wonderful city of Athina!
Hi George
Your Truepath looks like the one that Connexelectronic is selling, is it ?
Thanks
Eric
Attachments
Hopefully the following links work.
A trog about the Patent system throws up this one...
Espacenet -
Bibliographic data
WO2006124491A2
I wouldn't care to claim to know much about noise shaping and the other words. I kind of get 1 bit delta-sigma A/D but at my knuckle scraping hand waving level it is a simple analog feedback comparison of the output switching waveform to the input demand switching waveform which arrives at the final hysteresis comparator block in order to remove relative errors. I wouldn't be surprised if some sort of resampling occurs such that the output switching frequency ends up being some sub multiple of the input switching frequency but again I'm just using words.
In respect of the soft clipping that might result from this one...
Espacenet -
results view
US5974089A
Again this is me waving hands but the description references narrow pulse widths which in a fixed frequency Class D amplifier is exactly the behaviour you get at the extremes of modulation, AKA clipping. As a result I can imagine that the circuits as described would only be expected to be active and providing the suggested compensation at such extremes. This may be the source of the clipping behaviour and the resulting sound signature.
One thing that kind of drops out of that is that as things approach or enter clipping the switching frequency will drop which is something very much akin to what happens in UCD self oscillation circuits. Noise shaping and similar words are used to describe the operation of those circuits. It may be that Tripath is/was forcing a condition that comes as a natural result of UCD. There are lots of things that appear as if by magic in UCD that are of natural benefit for switching amplifiers.
Hopefully the analogies will not impact on your confirmation bias should you consider the possibility of NCore or Purifi being a route to rediscovering a sound similar to Tripath.
Oh... Click on link to patent then top left click on link to 'original document' then click on download. Solve captcha...
Last edited:
Diogenio, thanks for the kind words
On the AMP 10 I 've built, the highs are not as detailed as on the AMP6. But maybe it’s because of the implementation, the TA2022 may not to be blamed.
If I would revisit it's build, I would certainly:
Wind the output coils not using the solid core magnet wire provided with the kit. Instead I would use Litz wire or make one myself twisting together many thin magnet wires .
As a second mod , I would use two 4700uF one at the top the other at the bottom of pcb for each 10000uF at the psu section.
Re “TAR”
I’ve enjoyed a lot of long nights there sitting on a bar stool by the deck, talking to others, many of them gray-hair fellow diyers, drinking beer and looking at tubes some of them took out from their pockets (“what did we buy today”). I still visit one-two times a year with my wife. Great place, great food, splendid beer. Nothing has changed since 1988 except a mediocre mixing console inserted into the pure-glass home-made audio chain.
Vangelis Diakakis, the owner and soul of the place is a teacher of music, a beloved kind of low profile guy.
tar πατησια - Αναζήτηση Google
Hi Eric
It has to be build somehow along the same lines. What I don’t see there are the current sensing resistors (x4). Attached is how the board from 41Hz looks like.
If you decide to build it and you’ll find out that you need some good documentation , PM me your e-mail address, so I can send you all the documentation I have on this. Take it easy and don’t run, it’s not an easy project.
George
Attachments
Diogenio, thanks for the kind words
On the AMP 10 I 've built, the highs are not as detailed as on the AMP6. But maybe it’s because of the implementation, the TA2022 may not to be blamed.
If I would revisit it's build, I would certainly:
Wind the output coils not using the solid core magnet wire provided with the kit. Instead I would use Litz wire or make one myself twisting together many thin magnet wires .
As a second mod , I would use two 4700uF one at the top the other at the bottom of pcb for each 10000uF at the psu section.
Re “TAR”
I’ve enjoyed a lot of long nights there sitting on a bar stool by the deck, talking to others, many of them gray-hair fellow diyers, drinking beer and looking at tubes some of them took out from their pockets (“what did we buy today”). I still visit one-two times a year with my wife. Great place, great food, splendid beer. Nothing has changed since 1988 except a mediocre mixing console inserted into the pure-glass home-made audio chain.
Vangelis Diakakis, the owner and soul of the place is a teacher of music, a beloved kind of low profile guy.
tar πατησια - Αναζήτηση Google
Hi Eric
It has to be build somehow along the same lines. What I don’t see there are the current sensing resistors (x4). Attached is how the board from 41Hz looks like.
If you decide to build it and you’ll find out that you need some good documentation , PM me your e-mail address, so I can send you all the documentation I have on this. Take it easy and don’t run, it’s not an easy project.
George
Hi George
When you talk about the current sensing resistors (x4), can you see them on p.2 of the attached schematic?
If you believe it’s a well implemented TA3020 design I’ll certainly buy one.
Thanks
Eric
Eric, the current-sensing resistors detect an over-current condition in the output MOSFETS and then shut the amp down. They are shown on the Connexelectronic schematic as R4,R26 and R5, R25. The Connex circuit description also calls out and describes the function of these resistors, so it is highly doubtful they would not be included on the PCB.
George, thanks for the link to "TAR" photos. Warms my heart to see the place again, brings back many fine memories of chatting with the local DIY "tube guys" the same as you do!
Back to the Tripath:
1) Did you use Litz on your 2020 amps? Did you ever do a listening comparison test between solid wire inductors and Litz - wound in a Tripath amp?
2) Can you please share your Truepath power supply implementation with us - voltage selected, size of transformer, type of caps and rectifiers used? You could PM me with that if you prefer to open a discussion.
3) Worth nothing that a number of responders on this post said they prefer the Truepath amp sound to the totally-integrated Tripath chips (2020, 2022, etc). There is a bit of apples vs. oranges going on, as the Tripath chips are very low power compared to the Truepath. My 2020 is just 10W into 8 ohms, while the Truepath is 10x that amount, so at normal listening levels, the Truepath is operating way down low in its distortion curve. The little 2020 is already seeing rising distortion at the same listening level. Perhaps this could be the reason for Truepath sounding better?
4) Last question (for now!). What output MOSFETs are you using on your Truepath amp?
Τις καλύτερες ευχές,
Diogenio, my answer to your questions:
(1)I can’t remember on which amplifier I did the listening comparison but the Litz style winding was better at the treble.
(2) I had a good 28-0-28V/3.5A transformer. Consider 30000uF per rail as a minimum for good bass.
Remember, max Vdc is +/- 65V. You should allow some 5V minimum headroom, safe to stay below +/-50Vdc
(3) Most probably true.
(4) The kit came with IRFP4231 MOSFETs. When I blew one channel, I used IRFP150N on that. My notes says that heatsink of the 150 is 4 d. Celsius higher than the heatsink of the 4321.
Please see the list with recommended MOSFETS in 1st attachment
Remember, when testing, proper sequencing of supply voltages is important (ask me how do I know)
Note: I had made a search at the time (9 years ago) for Tripath patents: US5777512, US6549069, US6785392, US7746935, US20020089376, US20050122168
George
Attachments
George, thanks for all the info and attachments. I think I'll get a Connex module and build up a Trupath, instead of building my remaining 41 Hz TA2022 kit. Why not go big? I'll give that 2022 to a young engineering student here in town as a Class D learning example. She has yet to explore Class D - and as a future audio engineer...she will certainly have to understand the basic principles.
Let's lift a beer at TAR next time I'm in Athens. Was just there 2 years ago. I'm usually found hanging out in Kolonaki, sipping frappe and watching all of the celebrities and beautiful people walk by! I once had Terzis sit at our table!
I have already suggested to keep the 2020 kit
By all means support the young engineer to learn about class D amplification. There are so many class D modules on ebay from $5 to $50plus.
Do try the TA3020 set.
Certainly I would like to enjoy some beers together whenever you'll make it to Athens. PM me.
Terzis eh? Now you are talking! He's a fine person.
George
By all means support the young engineer to learn about class D amplification. There are so many class D modules on ebay from $5 to $50plus.
Do try the TA3020 set.
Certainly I would like to enjoy some beers together whenever you'll make it to Athens. PM me.
Terzis eh? Now you are talking! He's a fine person.
George
Tripath 3020 pcb with chip
This may be the wrong place for this post, but for those interested in the Tripath 3020, I have some Tripath 3020 pcbs and chips. 12 pcbs and 9 chips.
Make me an offer for for 1pcb and chip.
Please send me a PM regarding these as I don't want to clutter up this thread.
Thanks,
George
This may be the wrong place for this post, but for those interested in the Tripath 3020, I have some Tripath 3020 pcbs and chips. 12 pcbs and 9 chips.
Make me an offer for for 1pcb and chip.
Please send me a PM regarding these as I don't want to clutter up this thread.
Thanks,
George
Attachments
Last edited:
I have already suggested to keep the 2020 kit
By all means support the young engineer to learn about class D amplification. There are so many class D modules on ebay from $5 to $50plus.
Do try the TA3020 set.
Certainly I would like to enjoy some beers together whenever you'll make it to Athens. PM me.
Terzis eh? Now you are talking! He's a fine person.
George
OH, I'm keeping my 2020 amplifier...I've loved it for 15 years now!
Terzis - yes, and I chatted with Vangelis on the platia over on Hydra...it was 5 AM when we pulled our boat in after sailing all night from Kalamaki under a full moon. Nobody was awake, just this one table sitting all by itself with a disheveled Vangelis and another guy sitting there with coffee all hung-over! I thought that was great!
I've seen all of the Laika and Skiladiko greats in concert, Bill Karras, Mitropanos, Terzis, Angie D, Remos, Samiou, Garbi...etc!
Last edited:
I am interested in 2-3 pcbs. I have a few chips. Can you point me to some history on the pcbs? Ideally I would like a Trupath to match my other boards, but I think if the implementation is good, it might work well.
Andy
Andy
I move often and I move far! Some of the stuff follows me around, some gets sold off or donated. Some disappears. The nice thing about the Tripath amps is that they are generally small enough to keep. Alas speakers rarely are.
Diogenio: If I find a spare AMP6, I'll be happy to donate it.
I know anyone who has been on this site know already, but the Motorola DCP-501 (Motorolas ill fated attempt to get into the Home Theater market - at least they got the amps right) has five of the Tripath 3020 amps in it...all with power supply, for a laughable $50 on ebay and CL...
Just rehouse and change the input caps from electrolytic to film and your good to go...not bad...
Just rehouse and change the input caps from electrolytic to film and your good to go...not bad...