Did you measure Tripath dc offset with no load connected? According to Tripath it should be 2.5V. So problem with 3116 might be left channel having no dc offset with no load connected. I don't know, somebody here might know.
Thanks for the clarification, I was going to ask. Configuration A is the four gang/dual stereo config. I didn't save the link, but when I was researching balanced volume control, I found a picture of someone connecting two pots with gears. He attached a gear to each stereo pot, and put a third gear between them. So turning one pot made the other turn equally. IIRC, he trimmed the shaft of the "slave" pot, and hid the gears in the chassis. So in the finished product, only the single knob was visible.
Based on my "academic" understanding of this, I agree, the Panasonic EVJ pots probably don't have the precision for this scheme to work well. I might just use that Emotiva Control Freak in the interim... still thinking.
Irribeo, this is TPA3116 YJ2.0 board. I replaced suspected 0.68uf film cap, now there is no abnormal heating resistors nor any dc offset at the outputs, both channels has 0,02-4V DC at the output.
Also TA2021 base SMSL amps has two trimpots for dc offset adjustments. It should be close to 0V DC.
Edit: I think we are talking about two different things. 2.5V is input stage bias voltage for TA2021. DC offset should be close to zero, for all kinds of amp classes AFAIK.
Last edited:
Actually, the Panasonic EVJ pots should be okay since they have good tracking between both channels. However, if you're going to add input transformers, a stepped attenuator is probably the way to go. Depending on the gain and input impedance of your TPA3116 amp, you might need a different value like a 10k or 20k attenuator along with some precision matched resistors to wire in series with the pots. Those resistors need to be close in value since one is wired on the "+" input and the other on the "-" input.
I'm thinking to try input transformer instead of input caps.
If I don't want to use pot (I will use tube preamp), where should I add resistor and what value ?
You need to consider the following:
- Does the preamp or source equipment have unbalanced (RCA) or balanced (XLR) connectors?
- Output impedance of preamp or source equipment connected to TPA3116 amp
- Input transformer model and impedance
- Gain setting (and input impedance) of TPA3116
The CineMag CMLI-15/15B input transformers can be used with source output impedances around 600 ohms. The transformer's input impedance is 15kohms and exhibits very good frequency response with a transformer secondary impedance of 15kohm.
My YJ TPA3116's gain is set at 26dB with an input impedance of 30kohm. Therefore, I need to use a resistor of 30kohms in parallel with the amp's input impedance of 30kohms to match the transformer's secondary impedance of 15kohm (i.e., 1/30kohms + 1/30kohms = 1/15kohms). The result is exceptionally low noise, a very balanced frequency response from 20Hz to 20kHz, and exceptional transparency. This translates into a solid image, large soundstage and excellent tonality when using this amp in my main sound system.
Post 4912 is incorrect too. It's not clear to me why I can't display the picture I want, and uploaded to replace the other.
In the case of the (more affordable) Edcor there is the option of many different input and output impedances. Assuming 30k input impedance for the 3116 what is the most appropriate transformer output impedance? Should you match with 30k? Is a 1:1 ratio preferred?
For the transformer input impedance is there a preferred impedance that would match whatever your input device is or will an input impedance of 30k on the transformer match OK with any reasonable source?
Preamp is 4P1L with capacitor in output with CCS in anode with unbalanced outputs (RCA)
Output impedance is around 1.5K , I think 🙂
Input transformer is vintage Sowter 4069 with 1.2K:1.2K P/S impedance.
Gain is standard on YJ blue board, 26dB.
Output impedance is around 1.5K , I think 🙂
Input transformer is vintage Sowter 4069 with 1.2K:1.2K P/S impedance.
Gain is standard on YJ blue board, 26dB.
That's a very high output impedance!
If that's the real figure, you might want to look into an input buffer between preamp and amp (very high input impedance, and very low output impedance).
Hi Rhing, could I ask whether there is a theoretical reason for matching the amp input impedance so closely to the transformer output impedance? I was under the impression that the transformer impedances were only nominal figures - not accurate and ultimately depending on the circuit. I am not challenging your wisdom, just keen to understand as much as I can about what you are doing here. Sorry if this is covered earlier in the thread.
Also, experimentally, I found I preferred the sound when I adjusted the gain on my YBDZ board to increase the amp input impedance away from the transformer output impedance. Using 10K:10K transformer I got noticeably better bass when I went from 30K input impedance (26db gain) to 60K input impedance (20db gain). Maybe this is an anomaly to do with how the amp sounds at different gain settings (?) but it was a clear difference.
Input impedance and capacitance together are a frequency filter.
The datasheet recommends different capacitor values for the different possible impedances.
i have the same experience with changing the input capacitors from 1uF to 3.3uF on my b/b while staying with the 26db gain.
but i only hear the difference when i use the board with my subwoofer.
btw:
which values uses the sure-board for gain and input caps?
i think about buying one, since they are only 16 € on ebay including the shipping.
Well is about switching amps in general but sometimes also very specific Tripath 🙂 And Tripath did go bankrupt very long ago, reason I don't know, might be they didn't understand importance of dc offset with no load connected or because they published "dc offset" when they meant to write about "input stage bias voltage"
Application Note 9 [FONT=Arial,Bold][FONT=Arial,Bold]
DESIGNING WITH SWITCHING AMPLIFIERS FOR PERFORMANCE
AND RELIABILITY[/FONT][/FONT]©Tripath Technology
[FONT=Arial,Bold][FONT=Arial,Bold]
Section 5: DC Offset in Mute Mode
[/FONT][/FONT] There is normally a small DC voltage at the output of Tripath amplifiers in mute mode with no load present.
The voltage is approximately 2.5V and should not be a cause for concern since its Thevenin impedance is approximately 10k
W.
This residual voltage is negligible once speakers are connected.
re resistor temperature YJblue
The differences I mentioned in temperature of the 4 outputresistors on YJblue board are with speakers connected, fully functional YJblue, no signal on input that is connected to a source that is off, so amp "idle" but power connected and on. In this state YJblue inductors warm up and also do the resistors. Not all resistors warm up equally, TI datasheet right side resistors get warmer then left side. The resistors in positive output get warmer then the resistors in negative output. Both inductors and resistors warming up in idle is something I haven't noticed so much on the other 3116ampboards.
The differences I mentioned in temperature of the 4 outputresistors on YJblue board are with speakers connected, fully functional YJblue, no signal on input that is connected to a source that is off, so amp "idle" but power connected and on. In this state YJblue inductors warm up and also do the resistors. Not all resistors warm up equally, TI datasheet right side resistors get warmer then left side. The resistors in positive output get warmer then the resistors in negative output. Both inductors and resistors warming up in idle is something I haven't noticed so much on the other 3116ampboards.
Ideally, maximum signal transfer occurs when the source output impedance and amplifier input impedance are matched. In my case, I wanted the 15k primary "to see" 15k secondary. For myself, I vacillated between no resistors, 15kohm resistors and the 30kohm resistors. After listening to both conditions for an extended time, I finally settled with the 30kohm resistors across the transformer secondaries. The differences are subtle, but audible.
I think with anything in audio, one will need to experiment to obtain the best solution. This is why I haven't even started trying these input transformers with my Sure TPA3116 yet. Also, I wanted to try my modified YJ amp in different systems as a sanity check. It always helps to enlist the help of other listeners who are not biased in terms of component changes in a circuit.
My Sure TPA3116 amp has a loud turn-off pop and it is also noisier than when I first tried it. I am trying to sort out the issues, and even contacted Sure Electronics. However, I have not received any response yet.
Sorry to hear that. My Sure3116 arrived 2 days ago. I listened last night. Sound is better than red YJ. Now I turned it on/off a couple of time and it produced on/off pop. Sh.....t. I bought it because red is POP @#&. Now, lucky me, there is an off pop.
Interesting when I turn vols up the pop is louder (borrowed Monacor ps-3030 variable PS 0-30V,3A), no connection to preamp is worse. Pop on and speakers go back, off and they go forward.
Interesting when I turn vols up the pop is louder (borrowed Monacor ps-3030 variable PS 0-30V,3A), no connection to preamp is worse. Pop on and speakers go back, off and they go forward.
Received my Sure 3116 yesterday,it too has the dreaded pop issue.
Also they have not corrected the right speaker out polarity issue.
Bad form Sure Electronics.
If I had bought this unit through Ebay,they would be receiving a negative rating.
Also they have not corrected the right speaker out polarity issue.
Bad form Sure Electronics.
If I had bought this unit through Ebay,they would be receiving a negative rating.
My board is bad labeled to.
Turn on pop is bigger if i turn volume, on preamp, up. I like sound but...
Is anybody have pop problems?
Turn on pop is bigger if i turn volume, on preamp, up. I like sound but...
Is anybody have pop problems?
So even with a couple of transistors on their board Sure can't get rid of the pop issue. I think I will stick with Ybdz Wiener for now. Going to order a couple more for a 4way miniDSP setup.