• These commercial threads are for private transactions. diyAudio.com provides these forums for the convenience of our members, but makes no warranty nor assumes any responsibility. We do not vet any members, use of this facility is at your own risk. Customers can post any issues in those threads as long as it is done in a civil manner. All diyAudio rules about conduct apply and will be enforced.

Folsom EC7293: PVI Powered Frontend, 60/120w 8/4ohm

Jump to Latest
Destroyer OS

Destroyer OS

Member
Joined 2005
#1
Folsom EC7293: PVI Powered Frontend Amplifier, 60/120w 8/4ohm, 0.0005% THD

**Updated: Lower distortion with a slightly different board.
**Updated 2: Distortion is even lower than before, found a piece of steel was interfering with previous measurements. (thought it was tinned copper)

Alright everyone! We’ve got the redesign done.

EC7293-glow.jpg

(Image of prototype)

The EC7293 uses 2 chip amps in parallel for high power with a discrete error correction frontend which reduces distortion; and has very low memory distortion. High sensitivity makes it perfect for a variety of sources with 8ohm and 4ohm speakers.

Input impedance: 27k
Gain: 32db
Distortion 1K; 8ohm; full power w/Antek AS-3225: <0.0005% (immeasurable by us)

No Load

8ohm

4ohm

2ohm

Loopback


EC7293-Amp-GB-PREVIEW-1.png


Board size: 6.25x2.25"

*****There is not a mirrored left/right set because it would screw up the feedback layout*****

Discrete front end: 4 transistors

PVI: The frontend uses a photovoltaic power supply for galvanic isolation of the signal path. It uses a G4 LED bulb with a current source to disable the onboard driver of the G4.

Distortion: <0.0005%

High power: These are running in "modular" mode so that they can handle a 4 ohm load with some dips, while providing substantial power with a max output current of 20A. The Antek AS-3225 is a great fit for 60/120w. More power may be possible if you follow the guidelines of the TDA7293 datasheet and make provisions with higher voltage capacitors, and larger heatsinks.

Terminals: Happy to say we've found some that are based on a copper metallurgy, that accept decent sized wire, and don't strip too easily (I can torque them pretty good with screwdriver) that the boards are made to fit, and I recommend in the BOM.

Input Capacitor: This one is less sensitive to capacitors compared to the DIY7297. It has provisions for the same PHE426 and other caps. There is a slot-hole next to the pads so you can zip tie on larger caps and it has extra pads on the bottom to connect off board/long input capacitors.

CFC, smart PCB layout, & Copper Pours: Trace routing is optimized to improve performance and the power capacitors are placed for field cancellation (CFC); and copper pours are used for lower inductance.


dualpoleppv.png


Dualpole PSU PCB

Board Size: 4.25x3"

TO220 Diodes

CFC, smart PCB layout, & Copper Pours

Via holes for snap caps

RC and bypass: dampen the transformer & reduce noise


Suggested implementation with EC7293:

  • 18000uF to 24000uF per rail
  • Antek AS-3225 transformer
  • Keratherm insulators from DIYAudio store. (Chip tabs are 1.1” apart).
  • Heatsinks need to be larger than used on the 7297, nearly any enclosure with heatsinks for sides should work
  • 3U enclosure from DIYAudio store is a good fit. **
  • 2x Duapole's for dual mono with 2x transformers (Antek AS-2225)

**2U may work in dual mono with two smaller (shorter) transformers

Price & Amount needed to Print:

In order to print these up so I don't run out the day after, I need a commitment of 12x, two EC7293 boards and one PSU board. I'm hoping to see a lot more than 12x. To anyone that liked the DIY7297, this one is better, simply better.

EC7293+G4LEDR PCB $45ea (you'll need 2, to make a stereo amp)
DualPole PSU PCB $40ea

Shipping USA $8
Shipping International $16

And it will come with a BOM and some build details.

Paul Hirst 2,1 (1x)
s610adam 2,2 (1x)
Pi DE 2,1 (1x)
Roundtoit 2,1 (1x)
pistollero 2,2 (1x)
jmc207 2,1 (1x)
Danny23 2,2 (1x)
Domino99 2,1 (1x)
annamarykahn 2,1 (1x)
Click to expand...
 
Last edited:
#6
Oh my gosh! The documentation quality is so much better.
But, you really could have bragged up the output linearity instead of understating it.
Per each channel, four made-for-audio output fets in addition to an EC circuit!
Sure has clear and evident means-to-work.

Also, it is at fairly big scale, but if you wanted to know when you have enough, this will do it.
 
#8
danielwritesbac said:
Oh my gosh! The documentation quality is so much better.
But, you really could have bragged up the output linearity instead of understating it.
Per each channel, four made-for-audio output fets in addition to an EC circuit!
Sure has clear and evident means-to-work.

Also, it is at fairly big scale, but if you wanted to know when you have enough, this will do it.
Click to expand...

Well, I was trying not to use so much hyperbole this time... But it's true, it's a very good amplifier. I think people under estimate the value of memory distortion as well. The transistors are BC550's. I consider them to be extra special and near magical compared to other transistors I've used for things like phono.

I updated the graph to the 8ohm, I accidentally put up 4ohm. This one isn't as hard to read, either.
 
#14
Destroyer OS said:
Distortion 1K: <0.002%
Click to expand...

Neat amp.

Which measurement bandwidth did you use for your THD and THD+N measurements? It looks like 21.4 kHz. That means your measurements aren't valid past 5-6 kHz as you're not measuring much more than the fundamental and H2 at that point.

I'd also be interested in the THD+N vs output power and a measurement of the total noise (unweighted, 20 Hz - 20 kHz).

Tom
 
#17
You must log in or register to reply here.
Top Bottom