Xmas Amp - Dibya's TDA7293 by Jhofland

Jump to Latest
#261
xrk971 said:
So I should thank my local power company for the flat top sine wave power? It may be more efficient, but doesn’t that generate higher order harmonics?
Click to expand...

Actually it smooths out the rectifier pulses so they last longer, so despite the apparent distortion on the mains waveform I think it probably reduces high order harmonics. And I would assume the rectifier charging pulses would be the dominant source of harmonics anyway. I suppose the noise is radiated from your power trafo, but a good build should not be vulnerable to that.

I saw a review paper (can’t seem to find it) where the thermal conductivity between mica, silicone, Keratherm, and aluminum oxide were compared for transistor heat sink insulators. The ceramic was better than all of them. I can attest that the pad transfers 150w of dissipation from a TO-264 IXYS to a CPU cooler just fine continuously for days on days. Keeping the transistor body at 75C. Look at the thermal conductivity of aluminum oxide (circa 28 W/m K). Mine are 1mm thickness and with a thin layer of thermal grease, work well for Class A amps and so works splendidly for Class AB. They are quite durable and inexpensive too compared to Keratherm. You should try them. I use them exclusively now on my discrete Class A amps.
Click to expand...

Your link gives a thermal conductivity of 28W/mK (another page gave 20W/mK). Keratherm is 6.5W/mK. So for a given thermal resistance, the alumina can be 4.3 times thicker. Since the Keratherm is 0.25mm thick, an equivalent Alumina insulator would be about 1mm. So I am surprised by the alumina if the claims are honest. It makes me wonder why no one ever used them before.
 
Last edited:
#262
The alumina thermal conductivity varies. They give it as a range on a materials property website. 20W/mK to 30W/mK - depends on the grade perhaps. Let’s assume about half at 25W/mK. Keratherm is 6.3W/mK and 4x thinner so about 25W/mK for equivalent 1mm thick Al2O3. But if you have the 30W/mK grade, and use - 0.5mm thick one, it’s more than 2x better. Plus, it won’t melt of it gets really hot - which can happen. I always get the 1mm thick stuff because it’s durable. I think you can get it 0.5mm thick and even thinner. As long as it does not break its reusable. It won’t get pin prick holes from burrs. It has been used for electronics for a long time. Well known refractory material with high electrical insulation properties.

Oxide Ceramics – Aluminum Oxide (Al2O3)

Properties of Alumina/Aluminum Oxide (Al2O3)

Very good electrical insulation (1x1014 to 1x1015 Ωcm)
Moderate to extremely high mechanical strength (300 to 630 MPa)
Very high compressive strength (2,000 to 4,000 MPa)
High hardness (15 to 19 GPa)
Moderate thermal conductivity (20 to 30 W/mK)
High corrosion and wear resistance
Good gliding properties
Low density (3.75 to 3.95 g/cm3)
Operating temperature without mechanical load 1,000 to 1,500°C.
Bioinert and food compatible
Click to expand...

I think high power industrial electronics and IGBTs and big MOSFETs for power switching use it.
 
Last edited:
#264
I can see a good argument for spring clips and thin alumina insulators.

I would assume the transistor is long dead before a sil-pad starts to melt. What is the scenario where you saw this? Have you seen failures due to Kapton or something else melting or softening under reasonable operating conditions?
 
#265
MaxReboBand / 2 boards / USA
as8912 / 2 boards / USA *Cancelled*
ahtllc / 2 boards / USA *Paid Twice!!*
Flamethrower1 /2 boards/ USA *Paid*
Signal Lost / 2 boards / USA *Paid*
diy.tiger / 2 boards / NZ *Paid*
MShipmanPE/ 2 boards / USA *Paid*
Angchuck / 2 boards / Singapore *Paid*
Francois G / 2 boards / USA *Paid*
Asuslover /2 boards /USA *Paid*
astromo / 2 boards / Australia *Paid*
Kay Pirinha / 2 boards / Germany *Paid*
Sixto / 2 boards / USA *Paid*
Tubesguy / 2 boards / USA *Paid*
Hicoco / 2 boards / France *Paid*
Kokanee / 2 boards / Canada *Paid*
chermann/ 2 boards / Austria *Paid*
twocents/ 2 boards / USA *Paid*
s610adam / 2 boards / USA *Paid*
hochopeper/ 2 boards / Australia
Jlevro / 2 boards / USA *Paid*
nautibuoy / 2 boards / UK *Paid* (plus 2 * All Cees PCBs and 2 * RTR SSR Speaker Protection Gen2 (Etsy order 1908473317) - all paid)
invaderzim / 2 boards / USA *Paid*
kp93300/ 2 boards/ Malaysia *Paid*
comicrcc / 2 boards / USA *Paid*
Legis31 / 2 boards / USA *moved from waiting list*

Waiting List:

Pandpliers/ 2 boards/ USA
DarpMalone / 2 Boards / Canada
turion64 / 2 boards / USA
 
#266
MaxReboBand / 2 boards / USA
as8912 / 2 boards / USA *Cancelled*
ahtllc / 2 boards / USA *Paid Twice!!* Etsy order 1868983988 & Etsy order 1914458647 (as PM'd on Etsy)
Flamethrower1 /2 boards/ USA *Paid*
Signal Lost / 2 boards / USA *Paid*
diy.tiger / 2 boards / NZ *Paid*
MShipmanPE/ 2 boards / USA *Paid*
Angchuck / 2 boards / Singapore *Paid*
Francois G / 2 boards / USA *Paid*
Asuslover /2 boards /USA *Paid*
astromo / 2 boards / Australia *Paid*
Kay Pirinha / 2 boards / Germany *Paid*
Sixto / 2 boards / USA *Paid*
Tubesguy / 2 boards / USA *Paid*
Hicoco / 2 boards / France *Paid*
Kokanee / 2 boards / Canada *Paid*
chermann/ 2 boards / Austria *Paid*
twocents/ 2 boards / USA *Paid*
s610adam / 2 boards / USA *Paid*
hochopeper/ 2 boards / Australia
Jlevro / 2 boards / USA *Paid*
nautibuoy / 2 boards / UK *Paid* (plus 2 * All Cees PCBs and 2 * RTR SSR Speaker Protection Gen2 (Etsy order 1908473317) - all paid)
invaderzim / 2 boards / USA *Paid*
kp93300/ 2 boards/ Malaysia *Paid*
comicrcc / 2 boards / USA *Paid*
Legis31 / 2 boards / USA *moved from waiting list*

Waiting List:

Pandpliers/ 2 boards/ USA
DarpMalone / 2 Boards / Canada
turion64 / 2 boards / USA
 
#267
MaxReboBand / 2 boards / USA
as8912 / 2 boards / USA *Cancelled*
ahtllc / 2 boards / USA *Paid Twice!!* Etsy order 1868983988 & Etsy order 1914458647 (as PM'd on Etsy)
Flamethrower1 /2 boards/ USA *Paid*
Signal Lost / 2 boards / USA *Paid*
diy.tiger / 2 boards / NZ *Paid* (plus 2x All Cee's PCBs and 2x RTR SSR Speaker Protection Gen2 (Etsy order #1904430160)
MShipmanPE/ 2 boards / USA *Paid*
Angchuck / 2 boards / Singapore *Paid*
Francois G / 2 boards / USA *Paid*
Asuslover /2 boards /USA *Paid*
astromo / 2 boards / Australia *Paid*
Kay Pirinha / 2 boards / Germany *Paid*
Sixto / 2 boards / USA *Paid*
Tubesguy / 2 boards / USA *Paid*
Hicoco / 2 boards / France *Paid*
Kokanee / 2 boards / Canada *Paid*
chermann/ 2 boards / Austria *Paid*
twocents/ 2 boards / USA *Paid*
s610adam / 2 boards / USA *Paid*
hochopeper/ 2 boards / Australia
Jlevro / 2 boards / USA *Paid*
nautibuoy / 2 boards / UK *Paid* (plus 2 * All Cees PCBs and 2 * RTR SSR Speaker Protection Gen2 (Etsy order 1908473317) - all paid)
invaderzim / 2 boards / USA *Paid*
kp93300/ 2 boards/ Malaysia *Paid*
comicrcc / 2 boards / USA *Paid*
Legis31 / 2 boards / USA *moved from waiting list*

Waiting List:

Pandpliers/ 2 boards/ USA
DarpMalone / 2 Boards / Canada
turion64 / 2 boards / USA
 
#269
keantoken said:
I can see a good argument for spring clips and thin alumina insulators.

I would assume the transistor is long dead before a sil-pad starts to melt. What is the scenario where you saw this? Have you seen failures due to Kapton or something else melting or softening under reasonable operating conditions?
Click to expand...

I had both silpad and Kapton punctured by a tiny sharp burr on the heatsink tapped hole. One case shorted and blew out the outputs and drivers on the amp. Sure, I was young and careless and always check heatsink for no-continuity now before power up. On the big Class A that dissipates 150w per TO-264 MOSFET, I don’t even want to try a silpad- it’s like sticking a soldering iron on it and expecting the heat to dissipate. It might work, but I have more confidence in alumina as it’s a 1000C capable refractory.
 
#270
xrk971 said:
I had both silpad and Kapton punctured by a tiny sharp burr on the heatsink tapped hole. One case shorted and blew out the outputs and drivers on the amp. Sure, I was young and careless and always check heatsink for no-continuity now before power up. On the big Class A that dissipates 150w per TO-264 MOSFET, I don’t even want to try a silpad- it’s like sticking a soldering iron on it and expecting the heat to dissipate. It might work, but I have more confidence in alumina as it’s a 1000C capable refractory.
Click to expand...

That’s why I always countersink my heatsink holes, at least on the flat, polished side, after drilling. I even polish the area (where the chip will oppose) with fine steel wool. I’m even thinking of using nylon screws and nuts for tightening down the audio chip.

Call me paranoid...

Cheers,

Pete
 
#271
That's what I do now too - I use a larger drill bit and give it a few twists with my fingers to deburr the threaded entry hole. Then some fine sandpaper to smooth it over. It's important that the surface be flat as a bump might crack the ceramic spacer.

I should have said "young, careless, and dumb" 🙂

New List, Maxreboband and Hochopeper moved from main list to waiting list for no payment, Pandpliers and Darpmalone moved to main list:

ahtllc / 2 boards / USA *Paid Twice!!* Etsy order 1868983988 & Etsy order 1914458647 (as PM'd on Etsy)
Flamethrower1 /2 boards/ USA *Paid*
Signal Lost / 2 boards / USA *Paid*
diy.tiger / 2 boards / NZ *Paid* (plus 2x All Cee's PCBs and 2x RTR SSR Speaker Protection Gen2 (Etsy order #1904430160)
MShipmanPE/ 2 boards / USA *Paid*
Angchuck / 2 boards / Singapore *Paid*
Francois G / 2 boards / USA *Paid*
Asuslover /2 boards /USA *Paid*
astromo / 2 boards / Australia *Paid*
Kay Pirinha / 2 boards / Germany *Paid*
Sixto / 2 boards / USA *Paid*
Tubesguy / 2 boards / USA *Paid*
Hicoco / 2 boards / France *Paid*
Kokanee / 2 boards / Canada *Paid* *Combine with Etsy order*
chermann/ 2 boards / Austria *Paid*
twocents/ 2 boards / USA *Paid*
s610adam / 2 boards / USA *Paid*
Jlevro / 2 boards / USA *Paid*
nautibuoy / 2 boards / UK *Paid* (plus 2 * All Cees PCBs and 2 * RTR SSR Speaker Protection Gen2 (Etsy order 1908473317) - all paid)
invaderzim / 2 boards / USA *Paid*
kp93300/ 2 boards/ Malaysia *Paid*
comicrcc / 2 boards / USA *Paid*
Legis31 / 2 boards / USA *Paid*
Pandpliers/ 2 boards/ USA *moved from waiting list*
DarpMalone / 2 Boards / Canada *moved from waiting list*

Waiting List:
turion64 / 2 boards / USA
MaxReboBand / 2 boards / USA *moved from main list for no payment*
hochopeper/ 2 boards / Australia *moved from main list for no payment*
Click to expand...

Again, if I made a mistake with your payment status or missed you completely, please let me know and feel free to update the above list. It is preferable that you update the list for me if I missed something.
Thanks,
X
 
Last edited:
#273
Hi X,

I have looked at my recent purchases from your Etsy location, and I am not expecting any other shipment except the Xmas amp pcbs. I have received the 2 All Cees pcbs (thank you). Please check on your end so I do not receive someone else's purchase.

MM
 
#274
Ok Myles thanks for letting me know. So only the Xmas amp left to send you.

ahtllc / 2 boards / USA *Paid Twice!!* Etsy order 1868983988 & Etsy order 1914458647 (as PM'd on Etsy)
Flamethrower1 /2 boards/ USA *Paid*
Signal Lost / 2 boards / USA *Paid*
diy.tiger / 2 boards / NZ *Paid* (plus 2x All Cee's PCBs and 2x RTR SSR Speaker Protection Gen2 (Etsy order #1904430160)
MShipmanPE/ 2 boards / USA *Paid*
Angchuck / 2 boards / Singapore *Paid*
Francois G / 2 boards / USA *Paid*
Asuslover /2 boards /USA *Paid*
astromo / 2 boards / Australia *Paid*
Kay Pirinha / 2 boards / Germany *Paid*
Sixto / 2 boards / USA *Paid*
Tubesguy / 2 boards / USA *Paid*
Hicoco / 2 boards / France *Paid*
Kokanee / 2 boards / Canada *Paid*
chermann/ 2 boards / Austria *Paid*
twocents/ 2 boards / USA *Paid*
s610adam / 2 boards / USA *Paid*
Jlevro / 2 boards / USA *Paid*
nautibuoy / 2 boards / UK *Paid* (plus 2 * All Cees PCBs and 2 * RTR SSR Speaker Protection Gen2 (Etsy order 1908473317) - all paid)
invaderzim / 2 boards / USA *Paid*
kp93300/ 2 boards/ Malaysia *Paid*
comicrcc / 2 boards / USA *Paid*
Legis31 / 2 boards / USA *Paid*
Pandpliers/ 2 boards/ USA *moved from waiting list*
DarpMalone / 2 Boards / Canada *moved from waiting list*

Waiting List:
turion64 / 2 boards / USA
MaxReboBand / 2 boards / USA *moved from main list for no payment*
hochopeper/ 2 boards / Australia *moved from main list for no payment*
Click to expand...
 
#275
Just adding my etsy order # as a reminder, since others are doing that...

xrk971 said:
ahtllc / 2 boards / USA *Paid Twice!!* Etsy order 1868983988 & Etsy order 1914458647 (as PM'd on Etsy)
Flamethrower1 /2 boards/ USA *Paid*
Signal Lost / 2 boards / USA *Paid*
diy.tiger / 2 boards / NZ *Paid* (plus 2x All Cee's PCBs and 2x RTR SSR Speaker Protection Gen2 (Etsy order #1904430160)
MShipmanPE/ 2 boards / USA *Paid*
Angchuck / 2 boards / Singapore *Paid*
Francois G / 2 boards / USA *Paid*
Asuslover /2 boards /USA *Paid*
astromo / 2 boards / Australia *Paid*
Kay Pirinha / 2 boards / Germany *Paid*
Sixto / 2 boards / USA *Paid* (+ 2 BTSB buffer boards, Etsy order #1921899516, no rush - combine shipping if possible)
Tubesguy / 2 boards / USA *Paid*
Hicoco / 2 boards / France *Paid*
Kokanee / 2 boards / Canada *Paid*
chermann/ 2 boards / Austria *Paid*
twocents/ 2 boards / USA *Paid*
s610adam / 2 boards / USA *Paid*
Jlevro / 2 boards / USA *Paid*
nautibuoy / 2 boards / UK *Paid* (plus 2 * All Cees PCBs and 2 * RTR SSR Speaker Protection Gen2 (Etsy order 1908473317) - all paid)
invaderzim / 2 boards / USA *Paid*
kp93300/ 2 boards/ Malaysia *Paid*
comicrcc / 2 boards / USA *Paid*
Legis31 / 2 boards / USA *Paid*
Pandpliers/ 2 boards/ USA *moved from waiting list*
DarpMalone / 2 Boards / Canada *moved from waiting list*

Waiting List:
turion64 / 2 boards / USA
MaxReboBand / 2 boards / USA *moved from main list for no payment*
hochopeper/ 2 boards / Australia *moved from main list for no payment*
Click to expand...
 
#277
Hi X and other builders,

I’m looking at the schematic and which components would you think might have the greatest impact on the incoming/outgoing signal paths, and therefore possibly sound? Yes, I realize the board’s design is extremely important as well ..;>)

C1/101/102, R4, R1, R9, L1
Secondarily ...R20, R3?
If implementing DC servo...R19, R16, R15

Cheers,

Pete
 
#278
Hi Pete,
You are correct. Those are all “signal path” components. Use quality metal thin film resistors and you will be ok. As Dibya pointed out, and spec’d in notes on schematic and BOM, the gain setting resistors R1 and R3 should be carbon film for best harmonic profile of dominant H2 vs H3. The path through DC servo less important as it is filtered to 0.3Hz DC. The output Thiele network 10 ohms is also important as is the inductor. Air core pure copper magnet wire can’t be beat.

904725d1608921278-xmas-amp-dibyas-tda7293-jhofland-tda7293-amp-rev1-1-schematic-main-amp-jpg


In a way, R2/8/19 are important as they shunt the signal to ground as low pass filters. They should be metal thin film and definitely not metal oxide thick film and never wire wound.
 
Last edited:
#279
Many of you will notice that you got a shipment notice email saying your PCBs are on their way. I sent some of you a PM or email asking for your address and or DIYA screen name.

Some of you whom have the shipment tied to an Etsy order may have to wait a bit longer as the SSR’s cannot ship yet because I ran out of Molex KK connectors to include in the shipment.
 
#280
Hi X. Thank you for putting so much work into this.

I received a notification via your Etsy store that my speaker protection boards have shipped - it included a note to say the Xmas amp boards are included but didn't mention the All Cees PCBs?

Sorry to add to your burden.
 
You must log in or register to reply here.
Top Bottom