a peak at a TDA7293 amp

[quekky]

(amp completed by CGH)

for more pics, visit here. description on the amp are there as well




I designed the PI7293 amp some time back, and only got the pcbs during mid-dec. It is a parallel inverting TDA7293 (hence the name), running 3 TDA7293 in parallel

A friend of mine, CGH, was already interested in this amp before I got the pcbs. As I was very busy at work and had to complete my preamp (which i'll post my description later next week), he started with the project first.

He completed the essential circult a few weeks ago, just b4 before NOISing(singapore), but the casing was sortof unfinish yet at that time. During then, he had chance to enjoy the amp for a day.

a day later, just before the NOISing, what happen was CGH acidentally shocked something to something on one of the boards, while doing something, he thought the power was off then. it causes fire on the ICs. we still cant figure out what was shock, and it isnt cheap to try that again.

He had to get a new pcb from me and re-built the channel that was burnt. he only managed to complete it just after chinese new year (last friday)

we only had gainclone to compare to. my gainclone kept clipping at not-so-loud volumes, while the PI7293 just push the drivers effortlessly at louder volume. well, it's unfair to compare the tonal, etc. when 1 amp kept clipping...., but so far we all like the sound of this amp

schematic and other info to come into my website soon next week. this is just a sneak preview

[millwood]

Originally posted by quekky
for more pics, visit here. description on the amp are there as well

I particularly liked the Pentium class fan'd heatsink, just like what I used on my JLH, very effective, .

tda7293 is very popular in the far east diy circle, and some people actually preferred it to the National chips.

It kind of validated the assertion that a) they are pretty much the same; and b) there is nothing unique about the NSC chips.

[millwood]

I might be wrong on this but i thought 7293 is one of those chips with MOSFET on the output.

[Alex S]

ah, a Linn Klimax Twin (US$9000) clone minus the SMPS...

[quekky]

Originally posted by Alex S
ah, a Linn Klimax Twin (US$9000) clone minus the SMPS...

not exactly a clone. before i do the pcb i didnt know that linn uses 3 TDA7293. moreover we didnt know the schemtic of linn's twin

however i guess the graphs should be more or less the same

[daniel goh]

Quekky...

But we knew the klimax twin uses 3 7293, that how the pi7293 ended up having the exact same number of tda7293...

[carlosfm]

Linn has amps with 1, with 2 and with 3 TDA7293s.
One particular Linn amp uses 2 TDA7293s per channel in master-slave mode, it seams like a copy of what's in the datasheet.

[mb]

Really nice job guys!

A real pity it didn't make it to the gathering. Were you there? One of the GCs there clipped pretty badly with a pair of VR-1s, at fairly moderate levels, so we steered clear of GCs for the rest of the session. I don't recall any other amps (ranged from several AKSAs to 100W Mosfet amp and a Tripath based amp) doing so.

Should try to meet up in the coming months.

[quekky]

Originally posted by daniel goh
Quekky...

But we knew the klimax twin uses 3 7293, that how the pi7293 ended up having the exact same number of tda7293...

u only told me that after the 1st prototype (should i post pics of the prototype as well?)

[daniel goh]

oh yes! please do! That one had 2 7293s because I had only 2 to spare.

Just in case you guys are wondering, I'm the Dgah you see on the board silkscreening. Quekky did the board design, and did a pretty good job at it. I did the trivial stuff like deciding resistor values, how many ICs to use, and *****ing Quekky about the various layouts that came before before the board you all see here
CGH built the first complete amplifier from our design.

Btw, If you all look at the picture on CGH's amp closely, you would notice those dark green, wirewound resistors close to the TDA7293s. Those are ballast resistors, and were included to test the current sharing between the ICs.

Turned out that the datasheet is right (duh!) and the current is good enough to not require those resistors. Subsequently, they were shorted out by jumpers underneath the boards.

So if anyone else would like to built a paralleled 7293, feel free to leave them resistors out...

Oh MB, we arrive late at the gathering. Missed most of the amps you mentioned, except for the AKSAs that drove the linkwitzlab orions.

Our gainclones clip at moderate levels too. The fact that we ran them from 18Vac transformers probably didn't help as well.