It is interesting that you mentioned that. The 3116 chip has pins out specifically for snubber placement but 7492/7298 do not. Also, the 7492/7498 use AD modulation LC filter whereas 3116 employes BD modulation LC filter. This might explain the different set up. May be someone more knowledgeable could shed some lights on the details.
Regards,
Ok, doing some digging on this. On the bottom of page 10 of datasheet the switching frequency is specified with this equation:
fSW = 106 / ((16 * ROSC + 182) * 4) kHz
so I need to look up what Rosc is on my board. I have a feeling it is 39k as that seems to be default value they keep referring to. On TPA3116, the frequency is internal to chip and set with logic on 3 pins. Here it is with R and has max of 400kHz and minimum of 250kHz.
I did say it was on page 10 and only mentions 400k Max. Typical around 310-330k. It also has a fixed 60k input resistance due to the internal R.
Anyhow. Last night conseeded to using 2nd order for my pllxo woofer low pass, and along with tweeter got crossover perfectly between at 1octave above woofer optimum top frequency and 4.5octave below tweeter optimum top frequency. Why do I mention as its for 3116 board?
Because I played around with the saved file by adding the 3116 lc and zobel filter circuits after it, AND THEN...
I changed the 3116 outputs to resemble the 7492 schematic, with the cap/resistor zobel BEFORE the LC.
Very interesting results.
With the 3116 circuit in place the speaker + and - traces were a bit wobbly here and there compared to each other and didn't remain at a constant 180 opposite phase, even worse around crossover point and tweeter high pass in bottom end. These aligned with spikes in Voltage traces. (Not there when do xo circuit on its own).
When I switched the output circuit zobel to the left (IC) side of the LC, as per 7492 schematic, the spikes reduced somewhat by dB, but the dramatic change was the phase.
Not only did the phase traces keep a constant 180 apart across the whole 20-20k spectrum. But, the switching in phase at the spike points became perfectly vertical!!!! For both + and - traces. Even when I added the 3116 zobel back into the circuit as well as the 7292 zobel the vertical, 180degree +/- phase lines remained.
The other interesting part was the zobel before LC shifted the traces up in dB on tlspice, so they ended up +/-40db, with to crossover voltage spike virtually at the 0db. Whereas, with just the 3116 output filter circuit both phase traces well below 0db, and the voltage spike at xo also well below 0db.
This is a circuit variation I am keen to explore on for real on 3116 board now. Try it in LTspice.
Anyhow. Last night conseeded to using 2nd order for my pllxo woofer low pass, and along with tweeter got crossover perfectly between at 1octave above woofer optimum top frequency and 4.5octave below tweeter optimum top frequency. Why do I mention as its for 3116 board?
Because I played around with the saved file by adding the 3116 lc and zobel filter circuits after it, AND THEN...
I changed the 3116 outputs to resemble the 7492 schematic, with the cap/resistor zobel BEFORE the LC.
Very interesting results.
With the 3116 circuit in place the speaker + and - traces were a bit wobbly here and there compared to each other and didn't remain at a constant 180 opposite phase, even worse around crossover point and tweeter high pass in bottom end. These aligned with spikes in Voltage traces. (Not there when do xo circuit on its own).
When I switched the output circuit zobel to the left (IC) side of the LC, as per 7492 schematic, the spikes reduced somewhat by dB, but the dramatic change was the phase.
Not only did the phase traces keep a constant 180 apart across the whole 20-20k spectrum. But, the switching in phase at the spike points became perfectly vertical!!!! For both + and - traces. Even when I added the 3116 zobel back into the circuit as well as the 7292 zobel the vertical, 180degree +/- phase lines remained.
The other interesting part was the zobel before LC shifted the traces up in dB on tlspice, so they ended up +/-40db, with to crossover voltage spike virtually at the 0db. Whereas, with just the 3116 output filter circuit both phase traces well below 0db, and the voltage spike at xo also well below 0db.
This is a circuit variation I am keen to explore on for real on 3116 board now. Try it in LTspice.
Last edited:
I am just taking it from the data sheet available online from them. It clearly states 60k in spec boxes, and even mentions it to confirm the need for input decoupling caps - suggesting1uF.
If they have changed mind I look forward to finding the revision sheet. lol.
Just to note. I tried this with the 15uh / 470nf LC as per suggestion to me by gmarsh in 3116 forum, and used the 3116 data sheet output filter values.
I inserted the pre-LC zobel according to 7492 data sheet values. Even tried with the stock 10uH/680nf. But, I neglected to compare the difference between the 3116 LC and 7492 LC with and without zobel changes.
I might play around with that tonight as I think I can perfect the pre-LC zobel for the 3116, which might allow me to go back to the pllxo 1st order LP I wanted on the woofer.
datasheet says 60k for differential input, I believe that is 30k for SE input.
Found this info regarding the pre-LC zobel filter. As it happens its by an accomplished ST engineer.
http://www.embedded.com/print/4015876
Liking the fact the pre-lc zobel incorporates any impedance fluctuation effect added by the LC, whereas the 3116 output filters only account for the speakers impedance/resistance.
I am guessing those big 33uh inductors, combined with the AD modulation, must fluctuate resistance seen by IC output legs by a significant amount. I am wondering if control of this through the pre-lc zobel is an element of the better sounding highs!
As I said, I am gonna play with LT spice some more on this. Maybe a ferrite bead output with pre-LC zobel on 3116 board. But that's for the 3116 forum, I'll keep findings for the 7492 filters here, just using the 3116 findings as comparisons.
http://www.embedded.com/print/4015876
Liking the fact the pre-lc zobel incorporates any impedance fluctuation effect added by the LC, whereas the 3116 output filters only account for the speakers impedance/resistance.
I am guessing those big 33uh inductors, combined with the AD modulation, must fluctuate resistance seen by IC output legs by a significant amount. I am wondering if control of this through the pre-lc zobel is an element of the better sounding highs!
As I said, I am gonna play with LT spice some more on this. Maybe a ferrite bead output with pre-LC zobel on 3116 board. But that's for the 3116 forum, I'll keep findings for the 7492 filters here, just using the 3116 findings as comparisons.
Last edited:
xrk added a second pre L filter, it compensated in part the peaking of the outputfilter, with minidsp filter on input xrk tried to compensate peaking outputfilter further after adding the second pre L filter. Output filter still peaks and rings, but at least frequency respons looks more flat now. xrk also mentions he liked the effect, as do many with cheap tubepreamps. It adds false detail, probably a little more than 3116 outputfilter that also already adds a little with SS10F load, recorded playback (Reed) seemed to have more or clearer detail than original cd, its nice, coincidence maybe, but not accurate.
Irrebeo,
How do you know the stock filter had peaking? I am not sure if I detect any additional ringing from the speaker frequency and impulse response measurement. The treble was a bit brighter with the TDA7492 and I had to EQ it down a bit - is this brightness the same as the peaking? What about the bass, why does it seem to have more impact with the TD7492 vs the 3116?
How do you know the stock filter had peaking? I am not sure if I detect any additional ringing from the speaker frequency and impulse response measurement. The treble was a bit brighter with the TDA7492 and I had to EQ it down a bit - is this brightness the same as the peaking? What about the bass, why does it seem to have more impact with the TD7492 vs the 3116?
I read somewhere concerning the RC zobel that it has effect across whole frequency range, so maybe what is actually being heard is more true to original than you may think. Maybe our ears have settled for "the norm" and deceived into thinking new sounds don't belong! lol.
The pre-LC zobel, or Boucherot cell, may have something to do with it. Reading through some electric circuits stuff I found out it both prevents energy storing, much like a battery, in the magnetic flux of inductors, and it reverses the effect of whether current follows voltage slope or vice versa. Read it an hour ago, think it said, that in this instance it would make the current follow voltage slope. It also shows zero potential across it.
Would preventing energy storage in either direction flowing through coil make sound more open, less muddy? From what I remember that's what the idea of using it across speaker terminals is.
I know this is not the 3116 thread, but I am talking about the Pre-LC zobel.
Been messing around with the 3116 board tonight by adding the pre-LC zobel as with the 7492. Got a good hour of comparing, albeit low volume because kids in bed.
Here's what I had set up....
3116 board running off 20V 5A laptop brick, signal fed from a headphone socket of Samsung galaxy 3. The amp end the headphone wires feeds through 2+ and 2- wires each having a 470nf nichicon passing to the pins of the input caps underneath, basically bypassing the stock input caps.
The board is still a pretty stock YJ black/blue. Only mods so far are heatsink removed, DC decoupling caps are now Lelon 470uF RZW, and the 1nF smd/t replaced with a leaded Philips version.
Both channels wired to 8inch full range woofers direct without xo. These are paper cone woofers from a 1978 Advance VII speaker, which I recently refoamed. Just to note the speaker wires were about 8ft cheapo diy store stuff.
What I did was to add the zobel to 1 channel only, enabling a side by side comparison without swapping and changing.
As I said, this wasn't to talk about the 3116, but more what I found with the mod.
All round the zobel improved sound. I am not sure it made the bass deeper, or louder, but it definitely became more precise. I could clearly here the strike of the bass without it sounding punching, whereas the other channel lost the strike a little to the deep bass. The bass in voices even became clearer, which to some would sound as if the bass was lost to highs, but the bass was still there as deep as ever, just less muddled together.
With the highs, I didn't notice excessive highs. Only managed to listen to voices on WiFi internet radio during chat segments, and yes I noticed a tiny bit of sibilence coming through, but this was a radio station talk show, I could also clearly hear intake of breaths as well.
Music wise, I listened to jazz, classical, dance music and heavy metal. All sounded cleaner, clearer, deeper, higher, and had better separation of the different types of sounds/instruments all round. One of my downloads I like to test with is the instrumental of the Clean Bandit tune, think its the one with Jess Stone in vocal version, those in UK would recognise it as the Sainsburys TV advert music. The difference was very subtle, but when I placed ear about 12inches from cone I could clearly hear these experiences....remember I wasn't listening at a particularly high volume at all.
Unfortunately, I wasn't able to crank up the volume to get a feel as to whether the soundstage would have shifted slightly towards the pre-LC zobel channel. Hopefully I'll get the chance with speakers outside on a sunny day this week.
I don't believe I was hearing an inaccurate representation of so many different types of music, nor of the human voices discussing the bands musical inspirations. Everything became more open, less muddled, sounds more separated.
All in all for my final system design the mod fits in nicely as the 1inch tweeter which will accompany the woofer on an OB is going to be 24-35db down by the time
Its xo slope hits vocal ranges.
Just my two penneth on the effect of the pre-LC zobel as on the 7492.
Last edited:
Can you draw a diagram of your pre-LC Zobel on the 3116? Is it simply a 330pF cap in series with 22R and ends connecting the two inputs of the two inductors together?
Very interesting though. If this is true you have developed another mod that the 3116 thread should try out and put on the Wiki.
Could this be the reason the bass with the TDA7492 sounds more precise and clear compared to the TPA3116? I guess one way to find out is to add the Pre-LC Zobel to the 3116.
You should post this on the 3116 thread and people will be all over it. It wil really help to post photos of how you did it on the YJ blue black. I just got a brand new second gen YJ blue black in the mail. Maybe I can try this.
Last edited:
I've added a copy of post to 3116 thread, didn't think of photo. I'll add a photo there tomorrow.
To answer you questions. No its not the same values, I believe the values need to be calculated according to the L after it, that's what the web page below says....
http://www.embedded.com/print/4015876
As I knew I am only testing for short period I used what caps/resistor I had in spares box, and fitted what the above page mentions most manufacturers do as they are lazy or incompetant.
The resistor was around 8ohm on multimeter, it was one of a bundle I stripped off an old 1990's stacking stereo PCB before throwing out, to fill a spares tin for testing new circuit ideas. The 100nf is a new PP no brand name cap I got as a lot of 5 from eBay. Its one advertised as being good for guitar amps especially. Bought for the smd 100nf DC bypass I want to do, but were bigger than advertised so didn't fit in there.
I basically did what you mention. I soldered the resistor leg to cap leg, the soldered other resistor leg to input of the positive "inner" inductor i.e. the solder exposed just at the edge closest to bootstraps.
Then soldered the other cap leg to the same spot on the negative "outer" inductor.
Just as the 7492 schematic.
And note..... Easier to solder cap to resistor first, then resistor to inductor while holding cap. Started the other way around forgetting how bleeding hot the resistors get when soldering!!!! lol.
Try mocking it up on you 3116. I would be interested to know if it makes same difference with the bootstrap snubber mod in place already. You also have those lovely sounding reference speakers and measuring equipment.
To answer you questions. No its not the same values, I believe the values need to be calculated according to the L after it, that's what the web page below says....
http://www.embedded.com/print/4015876
As I knew I am only testing for short period I used what caps/resistor I had in spares box, and fitted what the above page mentions most manufacturers do as they are lazy or incompetant.
The resistor was around 8ohm on multimeter, it was one of a bundle I stripped off an old 1990's stacking stereo PCB before throwing out, to fill a spares tin for testing new circuit ideas. The 100nf is a new PP no brand name cap I got as a lot of 5 from eBay. Its one advertised as being good for guitar amps especially. Bought for the smd 100nf DC bypass I want to do, but were bigger than advertised so didn't fit in there.
I basically did what you mention. I soldered the resistor leg to cap leg, the soldered other resistor leg to input of the positive "inner" inductor i.e. the solder exposed just at the edge closest to bootstraps.
Then soldered the other cap leg to the same spot on the negative "outer" inductor.
Just as the 7492 schematic.
And note..... Easier to solder cap to resistor first, then resistor to inductor while holding cap. Started the other way around forgetting how bleeding hot the resistors get when soldering!!!! lol.
Try mocking it up on you 3116. I would be interested to know if it makes same difference with the bootstrap snubber mod in place already. You also have those lovely sounding reference speakers and measuring equipment.
I read the same information as Stanley - input impedance is set internally at 60 K - as listed on the data sheet.
Regards,
Ordered this version from EBAY. Someone reported in another thread that this one sounded fine too. A little more expensive and it uses different components (not sure whether it is good or bad). It does have the mute and standby options built-in, and the 3.3 V output (as depicted in the ST schematic) also.
TDA7498 100W 100W High Power Digital Amplifier Board AMP Board With Radiator | eBay
TDA7498 100W 100W High Power Digital Amplifier Board AMP Board With Radiator | eBay
Yep. I double checked, and my copy is the 5th edition. If anyone has a later edition please post link.