Sure Electronics New Tripath Board tc2000+tp2050

[matherp]

Signal Generator safe to use

I received my 2 * 100 today together with the Meanwell S-350-36 which I'm running at 32V to start with. I want to do some testing of my whole system by using it to amplify the signal from a signal generator which can do a variety of waveforms from 20hz to 40khz with controllable amplitude. The problem is that it creates the waveform centred around +8V relative to its ground. Is it safe to input this into the Sure amp and what RMS amplitude would you recommend.

Thanks

Peter

 

[soldersmoker]

By "centered" are you saying 8 Vp-p (2x) or 8 Vrms (0.707x) or 8 Vavg (0.637x)? That's for a sine wave, of course. 8 volts sounds like a lot to me, and Sure warns against a too-high input voltage.

 

[matherp]

An example would be a sine wave with its most negative peak at 6V and its most positive at 10V i.e. the "zero" crossing is at +8V. I can then increase or decrease the amplitude so for example 1V to 15V but always centred at 8V

Thanks

 

[ElFishi]

An example would be a sine wave with its most negative peak at 6V and its most positive at 10V i.e. the "zero" crossing is at +8V. I can then increase or decrease the amplitude so for example 1V to 15V but always centred at 8V

Thanks

The input cap of the stock board can take 50v according to the schematic.
So your offset DC voltage should do no harm.
The AC amplitude should be lower than 2V not to fry the chip.

 

[MikeHunt79]

I know there was a fairly lengthy discussion on inductor replacement in this thread, but I was wondering if anyone has tried any of the 41Hz.com inductor kits?

41Hz Audio:Two toroids T106-2 with Litz wire (AMP2)

I've got some of arjenhelders inductors, and while it's really nice having a ready made inductor, I have a hunch the wire is a little on the thin side. I have get to check how much current the Tripath datasheet specifies tho...

 

[sendler]

Same core at 41Hz

I know there was a fairly lengthy discussion on inductor replacement in this thread, but I was wondering if anyone has tried any of the 41Hz.com inductor kits?

41Hz Audio:Two toroids T106-2 with Litz wire (AMP2)

I've got some of arjenhelders inductors, and while it's really nice having a ready made inductor, I have a hunch the wire is a little on the thin side. I have get to check how much current the Tripath datasheet specifies tho...

Arjen and 41Hz are using the same T106-2 core that I have verified sounds way better than the tiny shielded bobbins that come on the 2X100, the difference being the Litz wire versus solid. All of the reports on 41Hz like the Litz wire better. I haven't tried their wire but with the Litz wire that I have, I like the solid core Belden better. The Ferroxcube cores would be much better still, very close to the air core toroids, but are hard to get. I only have 4 samples of the smaller TN23 cores and am still waiting for some TN26 Ferroxcubes which should be the best according to the class D app note. Maybe Arjen will start winding on these. Then we would be all set.

 

[dr_vega]

Shielding Air Core Inductors

Has anyone tried shielding air core inductors with aluminum or copper foil? If so, what were the results?

Did the shielding effectively block the RFI? Did it affect the effectiveness of the inductors? Did it affect the sound?

Where did you ground the shielding? Can you ground it to the negative power connector? My amps are out in the open so I can't ground them to a case.

-dr_vega

 

[MikeHunt79]

Arjen and 41Hz are using the same T106-2 core that I have verified sounds way better than the tiny shielded bobbins that come on the 2X100, the difference being the Litz wire versus solid. All of the reports on 41Hz like the Litz wire better. I haven't tried their wire but with the Litz wire that I have, I like the solid core Belden better. The Ferroxcube cores would be much better still, very close to the air core toroids, but are hard to get. I only have 4 samples of the smaller TN23 cores and am still waiting for some TN26 Ferroxcubes which should be the best according to the class D app note. Maybe Arjen will start winding on these. Then we would be all set.

Excellent, after looking at the spec for the T106-2, it's rated for 100W so I guess this size is fine for this amp. I think I'll stick with Arjen's inductors, I also noticed a notable improvement from the stock inductors.

Shielded aircore inductors sound like a nice idea however, maybe you could use some of those little mu metal boxes that are designed to shield line-level trafo's, I guess grounding to the -ve would work.

 

[ChuckT]

Arjen and 41Hz are using the same T106-2 core that I have verified sounds way better than the tiny shielded bobbins that come on the 2X100, the difference being the Litz wire versus solid. All of the reports on 41Hz like the Litz wire better. I haven't tried their wire but with the Litz wire that I have, I like the solid core Belden better. The Ferroxcube cores would be much better still, very close to the air core toroids, but are hard to get. I only have 4 samples of the smaller TN23 cores and am still waiting for some TN26 Ferroxcubes which should be the best according to the class D app note. Maybe Arjen will start winding on these. Then we would be all set.

Hi, Sendler,
thanks for posting the listening result of the various cores you have tried. Very useful.
So do u mean the Ferroxcube TN23 3C65 sounds better than the T106-2 (assuming same construction). Does the TN23 gets warm in the circuit? It is normally a sign of saturation.
Also, have you tried the MPP cores too.
Thanks again.

 

[sendler]

Ferroxcube sounds better than type 2 cores

Hi, Sendler,
thanks for posting the listening result of the various cores you have tried. Very useful.
So do u mean the Ferroxcube TN23 3C65 sounds better than the T106-2 (assuming same construction). Does the TN23 gets warm in the circuit? It is normally a sign of saturation.
Also, have you tried the MPP cores too.
Thanks again.

The Ferroxcube cores sound better than than the type 2 cores that became the de-facto core when Tripath recommended them. Unfortunately, type 2 was chosen for it's gentle saturation curve which allows the on board over current protection to work better in case of a shorted speaker wire. This same feature means that it not linear anywhere on the inductance versus current chart. Type 2 was chosen for safety not sound.
.
The TN23 cores show a saturation point of 6 amps and do get a bit warmer than ambient. I checked again last night but there is still no availability for the TN26-A113 cores that are recommended in the Ferroxcube class D application note and saturate at 10 amps. Frustrating! I think they will sound even better if I can ever get any.
.
Where can I buy MPP cores?

 

[dr_vega]

Hi, Sendler,
thanks for posting the listening result of the various cores you have tried. Very useful.

I'd also like to thank sendler for the great information. It's so good to see his observations from actually building and listening to the mods. So much opinion floats around the forums based on theory (often dubious) instead of actual empirical experience.

For those interested in some more discussion on air core inductors in class D amps, check out this thread. Sendler is a valuable contributor and, again, he's the guy who brings real world testing to the discussion:

http://www.diyaudio.com/forums/class-d/153262-ever-tried-type-0-toroids.html

I'm winding my air cores now. For formers, I'm using 1/2" pvc pipe couplers. They're 1 1/8" OD and 1 1/2" long. 22 wraps of 18 gauge magnet wire is right about at 10 uH and is almost exactly 1" long, leaving 14" of the former exposed at each end. They're 19 cents each at Ace Hardware. Figure about 9feet of wire for each inductor giving yourself long leads so you have some options in positioning the monsters on your amp.

I'll see (or hear) how they sound this weekend and then address the issue of controlling EMI. Any further ideas on shielding?

-dr_vega

 

[sendler]

Shields tonight

Any further ideas on shielding?

-dr_vega

I can try an aluminum foil shield tonight but the consensus from the engineers in my other threads is that they don't expect it to attenuate that amount of EMI power from a cylindrical coil nearly enough. I was thinking of placing the coils in a capped piece of larger copper pipe to try but then decided it was easier to just form the coils into a toroid. Toroids use more wire though.

 

[dr_vega]

I can try an aluminum foil shield tonight but the consensus from the engineers in my other threads is that they don't expect it to attenuate that amount of EMI power from a cylindrical coil nearly enough. I was thinking of placing the coils in a capped piece of larger copper pipe to try but then decided it was easier to just form the coils into a toroid. Toroids use more wire though.

I was going to use copper tubing, then I thought, why not copper foil, it would be easier. Once I got to foil, of course, aluminum foil came to mind. I have pie plates made of heavy aluminum foil that I should be able to solder to.

I believe that anything that is conductive will work, it doesn't have to be magnetic shielding, does it?

-dr_vega

 

[sendler]

Aluminum foil shield seems to work

I threw some aluminum foil over my amp for a quick test of the possible shielding effect and it seems to cut the remaining emissions of the Ferroxcube cores to zero even before terminating the foil to ground. I don't have my cylindrical air cores, which blast out much more EMI, mounted to an amp right now so I will have to try that another day but even the Ferroxcube cores are audible in an AM radio, although at a much lower level, and the foil covering has eliminated the noise. There is hope for the air cores yet.

 

[dr_vega]

I threw some aluminum foil over my amp for a quick test of the possible shielding effect and it seems to cut the remaining emissions of the Ferroxcube cores to zero even before terminating the foil to ground. I don't have my cylindrical air cores, which blast out much more EMI, mounted to an amp right now so I will have to try that another day but even the Ferroxcube cores are audible in an AM radio, although at a much lower level, and the foil covering has eliminated the noise. There is hope for the air cores yet.

Not to mention how elegant it looks!

Good going, sendler. It looks like you didn't wrap around the bottom, right? And still it suppressed the EMI to a significant degree. Very promising.

And I've got the same coffee cup that you do!

-dr_vega

 

[wushuliu]

I threw some aluminum foil over my amp for a quick test of the possible shielding effect and it seems to cut the remaining emissions of the Ferroxcube cores to zero even before terminating the foil to ground. I don't have my cylindrical air cores, which blast out much more EMI, mounted to an amp right now so I will have to try that another day but even the Ferroxcube cores are audible in an AM radio, although at a much lower level, and the foil covering has eliminated the noise. There is hope for the air cores yet.

Looks like a tin foil hat

 

[sendler]

All around

Not to mention how elegant it looks!

Good going, sendler. It looks like you didn't wrap around the bottom, right? And still it suppressed the EMI to a significant degree. Very promising.

And I've got the same coffee cup that you do!

-dr_vega

Coffee good for DIY. The foil is all the way around. Now I am looking for cheap metal tins to put the amps in.

 

[dr_vega]

Coffee good for DIY. The foil is all the way around. Now I am looking for cheap metal tins to put the amps in.

We're a bit too large for the Altoid tin that was all the rage for T-Amps.

I've found very nice metal boxes for cheap at junk and antique shops. Some are quite lovely and only cost a couple of bucks. Very high WAF (wife acceptance factor).

Things like an old metal lunch box are easy to cut and drill and look great. They also give your project a nice "creative and imaginative" look (as if the mods aren't already creative and imaginative enough).

With something as large as a lunchbox, heat probably won't build up too badly and you might not even have to vent it.

I built a double T-Amp project into a beautiful Japanese laquer wooden box I bought at a junk shop for $3. For shielding I lined the inside with heavy aluminum foil I cut from a disposable roasting pan from Wal-Mart.

-dr_vega

 

[Neil Davis]

...and am still waiting for some TN26 Ferroxcubes which should be the best according to the class D app note. Maybe Arjen will start winding on these. Then we would be all set.

The gapped ferrite toroids are OK for applications where you need the toroid form-factor (such as an existing board layout), but there are better ferrite core solutions. The gapped toroid is going to have EMI issues from fringing around the gap.

It's worth noting that the older IR amp described in the class D app note used the 106-2 core, but all of the newer IR reference amps use the Sagami enclosed "low-distortion" gapped ferrite inductors (see this page: International Rectifier - Audio Reference Design Kits).

Toko, Sagami and Coilcraft all make "low-distortion" shielded gapped ferrites that should work as well as the Ferrocube core, or better. The picture isn't very good, but maybe you can see how this Coilcraft GA3416 is made. It's a dual inductor, so there are two identical chambers. Each chamber is an enclosed EP-type construction (see Magnetic core - Wikipedia, the free encyclopedia), with a 1mm gap on the center slug and flat "tape" wire. I haven't opened up a Toko yet, but I'm guessing it is similar.

 

[sendler]

Mostly too small

The gapped ferrite toroids are OK for applications where you need the toroid form-factor (such as an existing board layout), but there are better ferrite core solutions. The gapped toroid is going to have EMI issues from fringing around the gap.

It's worth noting that the older IR amp described in the class D app note used the 106-2 core, but all of the newer IR reference amps use the Sagami enclosed "low-distortion" gapped ferrite inductors (see this page: International Rectifier - Audio Reference Design Kits).

Toko, Sagami and Coilcraft all make "low-distortion" shielded gapped ferrites that should work as well as the Ferrocube core, or better. The picture isn't very good, but maybe you can see how this Coilcraft GA3416 is made. It's a dual inductor, so there are two identical chambers. Each chamber is an enclosed EP-type construction (see Magnetic core - Wikipedia, the free encyclopedia), with a 1mm gap on the center slug and flat "tape" wire. I haven't opened up a Toko yet, but I'm guessing it is similar.

Most of the Tokos, Sagamis, and the Coilcraft GA3416 ect, coils with 10uH values, are too small for a TK2050 amp. We need something closer to a 10 amps rating. I have the Coilcraft SER2915L-103 installed and they sound pretty good but have some blurring of the top octaves compared to the Ferroxcubes. Maybe ok for the woofers. I'm still hoping for something better off the shelf.