Following on from my Paralleled Class A with Class D experiment, I decided to dig out an old TA0104 evaluation board and compare the performance. Something about the output from the parallel amp reminded me of the TRIPATH... the "spread spectrum" and relatively high switching frequencies... could it even be the trick behind so-called "Class T"?
I shelved the evaluation board almost 10 years to the day, but instantly remembered the bodge-up they did for the 12V lower gate drive supply when I got it on the bench again. This bit was a switching reg. originally fronted by a 51V Zener drop - to allow it to run between the main 0V and -V supply. A transistor follower was used to help the dissipation, but they still hung a label on the board warning against higher supplies than 70V. This time I got out the desolder pump and cleared all nonsense away. I've built a separate transformer supply for the +5 and +12 along with startup sequencing and muting. While I'm at it I've also got it a rack-case and its own toroid plus capacitor bank. Now it's running comfortably at +/-80V
There is an uncanny similarity in the output waveforms - definately some kind of self-oscillating hysteretic control is involved. But I get a more pleasurable sound from my experimental parallel amp - probably due to the minimalistic (naive?) single-ended front end consisting of nothing more than a couple of transistors and two hot resistors.
Peeking under the TRIPATH module there's obviously a lot more going on than I've got on the parallel amp but I'm still rather pleased with my DIY effort.
I shelved the evaluation board almost 10 years to the day, but instantly remembered the bodge-up they did for the 12V lower gate drive supply when I got it on the bench again. This bit was a switching reg. originally fronted by a 51V Zener drop - to allow it to run between the main 0V and -V supply. A transistor follower was used to help the dissipation, but they still hung a label on the board warning against higher supplies than 70V. This time I got out the desolder pump and cleared all nonsense away. I've built a separate transformer supply for the +5 and +12 along with startup sequencing and muting. While I'm at it I've also got it a rack-case and its own toroid plus capacitor bank. Now it's running comfortably at +/-80V
There is an uncanny similarity in the output waveforms - definately some kind of self-oscillating hysteretic control is involved. But I get a more pleasurable sound from my experimental parallel amp - probably due to the minimalistic (naive?) single-ended front end consisting of nothing more than a couple of transistors and two hot resistors.
Peeking under the TRIPATH module there's obviously a lot more going on than I've got on the parallel amp but I'm still rather pleased with my DIY effort.
Attachments
Still need to fit a 2U front panel and lid but all the wiring is complete. Having a logic level mute is handy - so I've added a 3.5mm jack to the back panel - into which I can plug a USB/jack lead. This allows the amp to be automatically muted if I'm using it with my PC. This is almost essential as the M-audio 2496 sound-card I use creates a *ridiculous* amount of hash when the PC shuts down.
> One fright I gave myself remains unexplained though - with the amp running and connected to speakers and sound-card I pulled out one of the phono inputs and a very loud screech came out of that channel instantly accompanied by a plume of smoke from the TRIPATH board 😱
The amp shutdown pretty fast as the -ve 80V rail fuse blew. The smoke came from a 33R snubber resistor on the channel I was disconnecting. Nothing seems to have taken damage though - after replacing the fuse and plugging everything back in after a careful inspection it all works OK again. I have to confess that I'm too chicken to try pulling the phono plug again - not while I don't understand what happened.... anyone got any ideas?
> One fright I gave myself remains unexplained though - with the amp running and connected to speakers and sound-card I pulled out one of the phono inputs and a very loud screech came out of that channel instantly accompanied by a plume of smoke from the TRIPATH board 😱
The amp shutdown pretty fast as the -ve 80V rail fuse blew. The smoke came from a 33R snubber resistor on the channel I was disconnecting. Nothing seems to have taken damage though - after replacing the fuse and plugging everything back in after a careful inspection it all works OK again. I have to confess that I'm too chicken to try pulling the phono plug again - not while I don't understand what happened.... anyone got any ideas?
Attachments
I had a TC2001/TP2350 board do this to me once. It was connected to a car audio SMPS. There was a very loud "whine" when I powered it up and the output zobels started to smoke.
It turned out it was from the two sections not being grounded properly.
BTW, I have a deep hatred for these damn Tripath modules. I think I've blown something like 6 TA0104A modules and one TDA2500 from various experiments. They are not very robust and seem to die if you so much as glance at them the wrong way.
I have one TA0104A module left if anyone is interested. I don't ever plan on putting it in a powered board, I know what will happen to it if I do.
It turned out it was from the two sections not being grounded properly.
BTW, I have a deep hatred for these damn Tripath modules. I think I've blown something like 6 TA0104A modules and one TDA2500 from various experiments. They are not very robust and seem to die if you so much as glance at them the wrong way.
I have one TA0104A module left if anyone is interested. I don't ever plan on putting it in a powered board, I know what will happen to it if I do.
I'm thinking that the reason mine smoked was feedback - switching noise getting into the freshly opened input. The event seemed to coincide with the phono plug just disconnecting the shroud (inner still making contact). Just to be sure I've added a 200K resistor and a 100pF bypass between the input and ground. Still to chicken to deliberately pull a plug though.
TBH, I'll probably shelve this amp again but it's been interesting digging it out for a review. I just noticed serial no. 008 in silver ink on the board. I guess this makes it a fairly early model!?!
TBH, I'll probably shelve this amp again but it's been interesting digging it out for a review. I just noticed serial no. 008 in silver ink on the board. I guess this makes it a fairly early model!?!
Board that I have now, has no S/N, and what you are saying you are putting +/-80v to it? is that ok with board and ta104?
- Status
- Not open for further replies.