[f] AMB on Facebook

johan851's build log

beta22

johan851's build log

Postby johan851 » November 20th, 2019, 10:43 pm

Hi all,

I'm beginning the purchasing process for my β22 and I'd like to keep a build log. I hope that this will prove both useful to others and help keep me motivated to finish. I originally started kicking this idea around a couple of years ago, and with a furniture reconfiguration that limits to me to small, full range speakers, I thought that an amp that could handle both my new near field speakers and headphones would be fun to have.

Previous projects:
  • M3 + STEPS PSU (circa 2010)
  • LM3875 gainclone
  • γ1 lite + Mini3
  • γ2 + σ25/σ78

I decided on a build similar in layout to the reference build, which has a single σ22 and three β22 boards for L/R + active ground. AMB explains the benefits of this configuration in this very informative post. This configuration is scaled down from my original four-channel dual mono plan (with option for future balanced input) because that snowballed into more complexity than I was excited to handle. You can see my thought process here and here.

One breakthrough came in finding the Mini Dissipante case options, specifically this one. The substantial external heatsinks allow the power supply to scale up to around 3A per rail, which should be plenty for speaker output. Mounting the ground channel on the other heatsink allows using that as the ground for even extra light speaker loads (maybe ~10w per channel). With an easy internal reconfiguration the speakers could also be grounded on the passive grounds, giving me a few more options in the future. There's sufficient height for 2.5" tall onboard heatsinks on the amp boards. I haven't decided whether to use +/- 30V or +/- 27V rails yet, but the 27V rails would give me extra confidence in the heat dissipation numbers.

The 330mm W x 200mm D x 120mm H enclosure meets my size goals for on-desk use. The fit, however, is quite tight. There's less than an inch of clearance for the amp boards depth, and the boards will be placed close together as well. Panasonic FR capacitors are used for most electrolytics due to the excellent ESR numbers, but the larger capacity spots on the power supply and amp channels use Nichicon UPM caps (low profile) to increase the space I have to work with. These have a trivially higher ESR and slightly lower rated lifespan. The transformer (200VA) will go in a separate steel chassis under the desk with an AC umbilical to the σ22.

I crunched all the numbers for fit, and put together the following rough panel layouts (to scale) to ensure that I'd be able to fit all required jacks. I'll eventually send the front panel to Front Panel Express, which I realized is just 20-30 minutes drive from my house. Convenient!

Front (volume, output selector, 4-pin XLR out, 1/4" TRS, jacks are Neutrik D-mount):
β22 Panel Layout (front).jpg


Back (binding posts and RCA out, RCA in, umbilical in):
β22 Panel Layout (back).jpg


I created a rough sketch of the internal wiring to confirm the Molex connectors necessary:
IMG_20191120_212750493.jpg


And finally, a complete BOM.

I've ordered knobs and potentiometers so far. I had to call around a bit for pots, as the 10mm thick panel on the Mini Dissipante is a pain to mount things on, and the 25mm shaft version of the Alps blue velvet was required. I found them at PartsConnexion if anyone is curious.
You do not have the required permissions to view the files attached to this post.
johan851
 
Posts: 76
Likes: 1 post
Liked in: 0 post
Joined: June 4th, 2010, 10:57 am
Location: Seattle, WA
Country: United States (us)

Re: johan851's build log

Postby amb » November 21st, 2019, 12:39 am

Great post.

Your consideration of wiring speakers for passive ground is a good one.

The 3-channel concept works its best on headphones anyway because the TRS plug is a 3-wire system. The supply current cancellation benefit in a 3-channel system holds up only while all channels are operating in class A, which may not be true when driving speakers (due to much higher load currents).

Running speakers with passive ground also avoids overloading the ground channel amplifier completely.
User avatar
amb
Site Admin
 
Posts: 10379
Likes: 175 posts
Liked in: 482 posts
Joined: March 29th, 2010, 9:49 pm
Location: Sunnyvale, CA
Country: United States (us)

Re: johan851's build log

Postby johan851 » November 22nd, 2019, 8:26 pm

I'll concede that there's a bit of "what do I do with this other heatsink" happening here. :) I am curious to see if the difference is noticeable to my ears. I'll be using Molex connectors to make things easy to assemble and service, so switching from active to passive ground will be an easy unplug and plug.
johan851
 
Posts: 76
Likes: 1 post
Liked in: 0 post
Joined: June 4th, 2010, 10:57 am
Location: Seattle, WA
Country: United States (us)

Re: johan851's build log

Postby johan851 » December 3rd, 2019, 4:52 pm

I've settled on +/- 30V rails and I'm taking a closer look at transformers before I order. I was planning on a 200VA 2x28VAC Antek unit, but it looks like a 2x30VAC will give me just a bit more headroom if the AC mains happen to be a little low.

It looks to me like a 160VA transformer might be sufficient, however, and the 160VA 2x30V Avel Y23 transformer (Y236505) is quite a bit smaller than the Antek, which would allow for a smaller transformer case. Thoughts on transformer sizing?

I see common "rule of thumb" recommendations that the transformer VA rating should be anywhere from 1-4x the maximum output power. Using 18Wrms per channel into 8ohms from the specifications, 36Wrms total, 36 x 4 = 144, so 160VA is plenty by even the super conservative 4x rule of thumb estimate.

Looking at it another way, the per-secondary current spec on the Y236505 is 2.67A. I can't think of a situation where the amp would actually draw anywhere near that continuously. Say I used 4ohm speakers and assumed a 30Vpp sine wave at the output - that's 15V peak / 4ohms = 3.75A peak, 2.65A RMS per channel. 2.65A RMS per channel is 5.3A RMS for both, each rail gets half the sine wave cycle, 2.65A RMS per rail. That's still under the 2.67A transformer secondary rating. And that kind of output is not at all realistic, so if the transformer could supply that kind of load I can't imagine it being the bottleneck.

I don't know how to account for any kind of peak demand, but I assume that's what the input capacitor bank is for.

Regulation on the 160VA transformer is 6.7%, so at full load expect the secondaries to drop to around 28VAC each. The sigma22 needs 30V + 4V (Vgs) + 2V (rectifier) = 36V. 28VAC / 0.707 = 39VAC, so there's still plenty of headroom there at full load.

Am I missing something? Is there anything to be gained here from a bigger transformer?
johan851
 
Posts: 76
Likes: 1 post
Liked in: 0 post
Joined: June 4th, 2010, 10:57 am
Location: Seattle, WA
Country: United States (us)

Re: johan851's build log

Postby amb » December 3rd, 2019, 4:57 pm

160VA is good enough.
User avatar
amb
Site Admin
 
Posts: 10379
Likes: 175 posts
Liked in: 482 posts
Joined: March 29th, 2010, 9:49 pm
Location: Sunnyvale, CA
Country: United States (us)

Re: johan851's build log

Postby johan851 » December 3rd, 2019, 5:12 pm

Thanks! Is current consumption plus verifying a tolerable voltage drop an appropriate way to size more generally?
johan851
 
Posts: 76
Likes: 1 post
Liked in: 0 post
Joined: June 4th, 2010, 10:57 am
Location: Seattle, WA
Country: United States (us)

Re: johan851's build log

Postby amb » December 3rd, 2019, 10:44 pm

Also thermal loss.
User avatar
amb
Site Admin
 
Posts: 10379
Likes: 175 posts
Liked in: 482 posts
Joined: March 29th, 2010, 9:49 pm
Location: Sunnyvale, CA
Country: United States (us)


Return to β22 amplifier

Who is online

Users browsing this forum: No registered users and 0 guests

Quick Links

AMB audio DIY resources
AMB audio shop

cron