Ovation e-Amp is Completed

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
www.hifisonix.com
Joined 2003
Paid Member
My 180W per channel e-Amp is finished. There's a .pdf file you can download to read about by design effort on the first page of my website.

:)
 

Attachments

  • IMGP6752.jpg
    IMGP6752.jpg
    218.8 KB · Views: 1,305
www.hifisonix.com
Joined 2003
Paid Member
Heres some more pictures. This amp weighs 30Kg's (my earlier Ovation 250 came in at 38Kg's) but this one is actually mechanically a lot stronger/more solid.

The second picture shows the e-Amp on the left, and right foreground the Ovation 250.

I did some more sound tests this weekend and the sonics are very good. I need to spend some time comparing the different feedback modes over the coming weeks. Its currently set up for TMC and wide band feedback.

I am pretty pleased with the noise as well - absolutely silent (inputs unplugged) and ear right up against the speaker bass, mid and treble units.
No hiss and zero hum.
 

Attachments

  • IMGP6751.jpg
    IMGP6751.jpg
    253.4 KB · Views: 1,221
  • IMGP6756.jpg
    IMGP6756.jpg
    139.4 KB · Views: 1,164
Hi Andrew,
excellent and very detailed paper on your e-amp. I noticed that page 18 was blank, was this intentional? The quality of the build is world class. Very well done, anyone who owns it will be very proud of you, the designer!!
 
Last edited:
Member
Joined 2010
Paid Member
WoW! What a paper accompanying the pics.- That will ring a lot of bells on this forum at least.
I'm a little puzzled about the choice of mounting the Vbe sensor transistor. You describe it variously as "loosely coupled" early in the paper and "tight" later when detailing it but the assembly shows a fair gap between the sensor and collector lead. I guess you would need to take such an approach with a commercial product but for your own use, perhaps an E-line package device would allow direct contact by say, inserting through a hole from the top side of the PCB?

That's no criticism of what looks to be an absolutely awsome and well researched project. Congratulations and errr....you wouldn;t have a spare lying about would you? :D
 
www.hifisonix.com
Joined 2003
Paid Member
Sorry, no spares lying about Ian!

I need to check out what you found and clarify it when I update the doc. I am afraid I looked at that doc so many times I no longer see the wood for the trees. I expect there'll be quite a few typo's I ll need to correct. Thanks for the comments and pointing that out.

:)
 
Member
Joined 2010
Paid Member
Having reread the relevant pages 6 & 39, a few times, I think I've misread "tight" for "light". Apologies, but I'm still curious about the final thermal coupling scheme described for the output transistor.

Re: PDF files. I was increasingly finding this problem on some some sites a few months ago but Adobe Reader (also a free download) seemed to fix that. Worth a try?
 
Last edited:
dadod, unfortunately I have not listened enough to say at this stage which type of feedback I prefer. I will need to live with the amp for a few months, let my ear tune in and then change to see which one I like. I can only comment after that. Maybe there is no difference ;-)

True enough. What about PCB layout, it would be very useful(for me at least) to see how you did it?
dado
 
www.hifisonix.com
Joined 2003
Paid Member
Ian,
basically what I found is that EF3's are quite hard to thermally compensate. You can get them stable over a narrow temperature range, but once the heatsink heats up, you get errors - in my case, the Iq increased significantly. The only way to get a reasonable setting was to set the Iq very low at ambient (now you know why some amps specify very low Iq at ambient) and then rely on the amp warming up so that after some time Iq settled at about the optimum value. On my Ovation Amp, I used a conventional 2 transistor circuit, but ended up having to fiddle with the amount of thermal grease between the sensor and one of the output transistors. Under quiescent conditions it is stable, but if It runs at higher power levels (so 80-100W) it is not stable enough.

To get round this I used a NTC on the e-Amp which has only a moderate effect at ambient, but at higher temperatures the resistance drops very quickly. If you look at page 40 and 41 you can see a detailed description of how this approach was calibrated during the design phase. With this scheme, Iq is set at two points over the operating temperature (thats why I called it two point thermal compensation) and this is much more stable than the usual single point approach.

That said, EF2's which are much more common in bipolar amps, are a lot easier to thermally compensate, so you can usually get away with standard Vbe spreader circuits.
 
Last edited:
Ian,
basically what I found is that EF3's are quite hard to thermally compensate. You can get them stable over a narrow temperature range, but once the heatsink heats up, you get errors - in my case, the Iq increased significantly. The only way to get a reasonable setting was to set the Iq very low at ambient (now you know why some amps specify very low Iq at ambient) and then rely on the amp warming up so that after some time Iq settled at about the optimum value. On my Ovation Amp, I used a conventional 2 transistor circuit, but ended up having to fiddle with the amount of thermal grease between the sensor and one of the output transistors. Under quiescent conditions it is stable, but if It runs at higher power levels (so 80-100W) it is not stable enough.

To get round this I used a NTC on the e-Amp which has only a moderate effect at ambient, but at higher temperatures the resistance drops very quickly. If you look at page 40 and 41 you can see a detailed description of how this approach was calibrated during the design phase. With this scheme, Iq is set at two points over the operating temperature (thats why I called it two point thermal compensation) and this is much more stable than the usual single point approach.

That said, EF2's which are much more common in bipolar amps, are a lot easier to thermally compensate, so you can usually get away with standard Vbe spreader circuits.

My experience is that you have to separate thermal compensation of the drivers from thermal comensation of the output transistors. Use two separate transistors thermaly coupled one to the driver separate heatsink and other one to the main heatsink, look attachment(simulated only). In mine TT amp http://www.diyaudio.com/forums/solid-state/182554-thermaltrak-tmc-amp-10.html#post3031504 I used TT diodes to compensate output transistors only and Vbe multiplier transistor to compensate drivers only. A predrivers is not necessary to compensate in mine opinion.
dado
 

Attachments

  • Vbe multiplier.jpg
    Vbe multiplier.jpg
    102 KB · Views: 628
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.