TPA3110 and TPS3116 amps

Markvdv · « **Reply #1200 on:** 17 Oct 2014, 07:39 am »

I had harsh high and limited bass with cheap transformer. I believe I had pot befor transformer too, can't remember with 100% certainty. For LDR removing R3 audiobah will improve sound, and a lower outputimpedance from source probably too. With transformer you will need to know outputimpedance LDR.

Markvdv · « **Reply #1201 on:** 17 Oct 2014, 07:55 am »

Quote from: Salis Audio on 16 Oct 2014, 09:57 pm

20k is simply high. These amps would do well with a buffer. As is 20db gain with a 10k volume is about as good as you'll get for analog/attenuator control.

It is actually your source that needs the buffer. Sometime ago cdplayers were tested/measured for being able to drive a passive pot. Only 1 in this test had perfect result, which was not the most expensive in test, but more expensive than most:) Picture shows excessive attention to powersupply, but that probably was not reason it didn't have trouble.

fleawattaudio · « **Reply #1202 on:** 17 Oct 2014, 03:00 pm »

I would also be curious how a tube audio legend like Roger likes the measurements of the 3116 with his ears.

Quote from: rhing on 5 Oct 2014, 03:28 pm

Nice work Randy. You must be a neurosurgeon to be able to modify the Audiobah board like that. Indeed, having high quality transformers like the Lundahl's on your DAC output and feeding the signal directly to the differential inputs of the TPA3116 amp IC is an ideal setup.

I just received my DUG TPA3116 PBTL amp PCB's with the TPA3116 IC's mounted on them. I purchased these PBTL boards through the diyAudio.com Group Buy thread:

http://www.diyaudio.com/forums/group-buys/257996-gb-tpa3116-8-pbtl-bare-pcb.html

I can work with the other SMD's, but I didn't want to risk wasting my money or boards trying to mount the TPA3116 IC's myself. DUG will test them too for an additional charge.

In the meantime, I've been assisting Roger Modjeski of Music Reference and RAM Tubeworks in setting up his audio engineering school and laboratory near Berkeley, CA. The other day I was at his lab, he lent me a prototype of a single 12AT7 dual triode line stage preamp. I've been listening to this preamp in my system, and I am really amazed with how well this works with my modified Yuan Jing TPA3116 blue amp. It is giving my Audio Research LS7 some serious competition. I am trying to talk Roger into making this preamp a DIY project to offer through the school.

Roger is willing to test my modified TPA3116 amp in his lab and get measurements. Should be interesting what we measure.

rhing · « **Reply #1203 on:** 18 Oct 2014, 02:42 am »

Quote from: fleawattaudio on 17 Oct 2014, 03:00 pm

I would also be curious how a tube audio legend like Roger likes the measurements of the 3116 with his ears.

I'll be at his lab tomorrow with the modified YJ TPA3116 amp. It marries very well with his single tube 12AT7 line stage. His Berkeley Hi Fi School has successfully launched in the Bay Area.

http://berkeleyhifischool.com

Folsom · « **Reply #1204 on:** 18 Oct 2014, 03:22 am »

Rhing, if you haven't built the DUG boards, consider doing 2x per decoupling cap (for same value) as opposed to the single they're setup for; cost is almost nothing for lower impedance towards noise. They can literalyy sit on each other. It'd be better with opposing poles but that requires a new board. One could get carried away with the concept, but the benefits at two are substantial.

Mark, source vs. Volume vs. Amp is more complicated than that.

rhing · « **Reply #1205 on:** 19 Oct 2014, 03:32 pm »

I went to Roger Modjeski's lab yesterday to help get it organized to hold lab classes for the Berkeley Hi Fi School. I brought my modified Yuan Jing blue amp and my slightly modified Sure Electronics TPA3116 amp for measurements using his Sound Technologies 1700A Series Distortion Analyzer and Hitachi oscilloscope. We used my upgraded Astron RS-12A regulated linear power supply with both amplifiers.

Modified Yuan Jing TPA3116 blue amp

Power..............1kHz..........2kHz.. ........6kHz..........10kHz..........20 kHz
1 watt............ .03%......... .06%........ .35%......... .60%.......... .80%
10 watts......... .03%......... .20%........ .25%......... .30%.......... .25%

Modified Sure Electronics TPA3116 amp

Power..............1kHz..........2kHz.. ........6kHz..........10kHz..........20 kHz
1 watt............ .14%......... .14%........ .35%......... .55%.......... .65% (first 2 are in the noise)
10 watts...... .10%......... .13%........ .33%......... .30%.......... .40%

As we went through the measurements, Roger explained the interesting waveforms we viewed on his scope. These amps both exhibited crossover distortion. What's interesting to me is that the THD significantly increases above 6kHz at 1 watt. However, we measured lower distortion at 10 watts. Looking at the TI datasheet, this may explain why they do not show distortion measurements above 6kHz. My hearing rolls off around 16kHz.

Another interesting observation is the amount of noise at 1kHz and 2kHz for the Sure amp. Roger pointed out that the distortion measurements were buried in the noise floor. He also noticed a 0.6dB rise at 20kHz, 1.5dB at 30kHz and 2dB rise at 40kHz. He attributed this to the output filter. We did not see this with the modified YJ blue amp. This may partly explain why I prefer my modified YJ amp (even before adding the CineMags) over the Sure amp.

Time didn't permit us to measure frequency response, but I'll probably bring both amps to the lab for more measurements and listening tests.

Markvdv · « **Reply #1206 on:** 19 Oct 2014, 04:50 pm »

You do not mention load. But you do mention frequency respons. For these amps these are connected.
These pictures might be interesting too:
http://www.diyaudio.com/forums/class-d/262964-class-d-output-spikes.html

Markvdv · « **Reply #1207 on:** 20 Oct 2014, 08:01 am »

OK silence, that brings noisefloor in region ear is most sensative to attention. 0.14% is audible hiss when music is playing, some might remember tape recorder hiss, about that level I think my Revox had? So is it there with your Sure or not? It might not be there, could also be a false read from distortionmeter. How? 6khz is considered worst case in measuring audioband, 2nd and 3rd harmonics within audioband. Modern devices can noise shape noise to places outside audioband, some devices just have noise above 20khz, inaudible but older analog distortion meters can not handle large amount of noise in places never expected when machine was constructed. That is why very steep inputfilters are added befor distortionmeter, so measurements in audioband are reliable, but also makes 6(.666)khz worst case scenario, in theory. See AES17 filtering. Now Sure amp was peaking visably on frequencies near audioband, more than your YJblue, this could cause false levels indicated by distortionmeter if not filtered steeply enough. Same problem might be there for for example 60khz noise with YJblue filter. So I don't know if measurements are reliable.

(2.83V 0.14%=4mV versus TI's 0.065mV EVM)

rhing · « **Reply #1208 on:** 20 Oct 2014, 09:00 am »

Load was 8 ohms. The Sure amp does have more audible hiss compared to the modified YJ amp.

Considering that Roger has measured many solid state and tube amplifiers including his own critically acclaimed designs, I trust his measurement results and his explanations. Certainly, the University of California in Berkeley feels the same way, and has him lecturing to their electrical engineering students and working with them in their labs.

Markvdv · « **Reply #1209 on:** 20 Oct 2014, 02:05 pm »

If you are used to measuring tube/ss amps, you could see funny things trying to measure digital age devices. Many do. Most of times just measuring artifacts that tell little about the device being measured.

Odal3 · « **Reply #1210 on:** 29 Oct 2014, 04:50 am »

Having a green Audiobah board with an Astron. Currently running it stock and it sounds great to me with the exception that it's not much bass at all. In fact, when connecting it to a sub via high-level there it's apparent that the low end is completely missing on some songs. I have connected it directly to a CD player, an ipad, and galaxy S4 via DAC. So what's the trick to get some more bass?

The only way I can get some bass on songs is to connect to the pre-out on an old Pioneer Elite receiver and use the loudness function but that makes the sound muddy and comletely removes the magic.

Poultrygeist · « **Reply #1211 on:** 29 Oct 2014, 10:20 am »

Quote from: Odal3 on 29 Oct 2014, 04:50 am

Having a green Audiobah board with an Astron. Currently running it stock and it sounds great to me with the exception that it's not much bass at all. In fact, when connecting it to a sub via high-level there it's apparent that the low end is completely missing on some songs. I have connected it directly to a CD player, an ipad, and galaxy S4 via DAC. So what's the trick to get some more bass?

The only way I can get some bass on songs is to connect to the pre-out on an old Pioneer Elite receiver and use the loudness function but that makes the sound muddy and comletely removes the magic.

I'd guess the weak link is somewhere other than the amp. I'm using the same combination with no lack of bass.

Odal3 · « **Reply #1212 on:** 29 Oct 2014, 12:22 pm »

Yes, wondering if the thin low end is due to connecting the source straight into the amp with no pre - amp. Hear the same with a Sure Tpa3110 2 x 8W using the same combo, but with a TA2020 Lepai I get punchy bass.

Poultrygeist - how is the sound in your system if you bypass your tube preamp and instead using a software control or only run the high levels into your Young plate amps? When I disconnect all other speakers and only run it out to the sub as a test, the driver is barely moving despite I feeding it a lot of gain. My work around has been to Y split the input signal and use RCA instead, but that doesn't work on my main speakers...
Thanks again for the tip of getting an Astron. I'm still amazed of how much better it sounds with it.

Poultrygeist · « **Reply #1213 on:** 29 Oct 2014, 03:07 pm »

Except for the integrated red board I've never tried any of the 3116's without a preamp.

At the moment I'm running a CDP source into the Yung RCA line in's and the Yung RCA line out's into the 3 tube preamp and then the preamp outs into the 3116.

It all works fine. The original signal is a pass through on the Yung line outs.

Markvdv · « **Reply #1214 on:** 29 Oct 2014, 10:18 pm »

Filterless Sure3110 has ~170nF input capacitors, audiobah's measure ~900nF. You stated that pre-out didn't have bass either unless you used loudness. So if you want to solder smd, increase value all 4 inputcapacitors audiobah. If you don't want to and have enough volumerange with sources you use, cd will be ok I guess, but ipad/iphone???, you could reduce gain by removing R3, that effectively is the same as doubling inputcapacitors value.

But audiobah standard should give more bass than little Sure, if not better check speakerpolarity, audiobah shouldn't be same as little Sure ampboard, 2 negatives are not in the middle!

Odal3 · « **Reply #1215 on:** 30 Oct 2014, 04:47 am »

Thanks Markvdv and Poultrygeist for the suggestions. To clarify, the bass is somewhat there with the pre-outs from the receiver (better than from Ipad/smart phone) but it sounds really flat on songs like From Gagarin's Point of View by Esbjorn Svensson Trio untill I boost it up with loudness.

Looks like post #720 (page 37) describes the same issue when using a DAC between a digital source with software controls and the TPA3116. I'm using the HifimeDIY android DAC via the USB output which gives some additional gain. Taking it straight into the sub plate amp via RCA gives very nice tight and full bass from sub but of course still miss it in the mains. Headphone outs = slightly more bass but still flat + less clarity at reduced volume.

Don't have the skills to solder smd, but removing R3 will increase bass?
Checked polarity, but it was OK. Perhaps the best solution is to get a pre-amp. Is the 6N3 + 6Z4 tube valve preamp the recommended way to go? I noticed the carry Yuan Jing pre-amps at PE as well but only with the 6N3 option.

Odal3 · « **Reply #1216 on:** 30 Oct 2014, 04:59 am »

Forgot to mention that when I moved cables around I managed to play the amp with no speaker load as well as only one speaker connected. To my surprise, I could hear soft music directly from the amp board itself!??? It sounded like when you stand next to someone playing music out of a pair of cheap headphones. Barely audible but I could distinguish the lyrics.

So where is this coming from? Is this anything I need to worry about? Did not hear this from the Sure board.

Markvdv · « **Reply #1217 on:** 30 Oct 2014, 08:48 am »

Quote from: Odal3 on 30 Oct 2014, 04:59 am

Forgot to mention that when I moved cables around I managed to play the amp with no speaker load as well as only one speaker connected. To my surprise, I could hear soft music directly from the amp board itself!??? It sounded like when you stand next to someone playing music out of a pair of cheap headphones. Barely audible but I could distinguish the lyrics.

So where is this coming from? Is this anything I need to worry about? Did not hear this from the Sure board.

Ceramic classII capacitors have some distortion, one of them is piezoelectric, they vibrate with voltage applied, in other words they act as a loudspeaker. Better check all 4 outputs, capacitance to ground, after playing ampboard without load.

Poultrygeist · « **Reply #1218 on:** 30 Oct 2014, 08:50 am »

"Taking it straight into the sub plate amp via RCA gives very nice tight and full bass from sub but of course still miss it in the mains. Headphone outs = slightly more bass but still flat + less clarity at reduced volume"

If you connect the plate amp's line outs to your main amp's inputs you shouldn't miss anything.

The signal coming out of the plate's line outs is unaltered and the same as the source.

Odal3 · « **Reply #1219 on:** 31 Oct 2014, 04:57 am »

Good idea, but my plate-amp doesn't have line-outs.