[amb]

tintin47, yes, the turn-on surge current can be very high, not only due to the transformer initializing its magnetic field, but also from charging the PSU's bulk caps. A relay is a good idea to prevent the switch contacts from wearing out prematurely. The following diagram should speak for itself about how to wire it up. An AC 120V relay is what you'd need, for a "normal" β22/σ22 (i.e., no unusually large transformers) I would recommend a relay that has at least 10A contacts.

The one we used for the krmathis group build σ22 is a Omron G4B series (Mouser 653-G4B-112T-AC120).

Oops, see the next post for the schematic.

 

[amb]

 

[tintin47]

Thanks much. Seeing it in schematic form makes a lot of sense.

 

[boinger]

I have the happy news of saying that my beta is perfectly fine and i was just being somewhat of a large idiot... It is singing quite well i must say well the one channel i tested... Just have to finish the final wiring and i should be good to go ....

i made the silly mistake of assuming i didnt need to wire the signal ground to the input because I thought that since the headphones didnt need to be wired with the signal ground i assumed that i didnt have to do that...

so after not being able to figure out what i had done wrong i just decided to wire it and check it out... and low and behold all my distortion issues were gone....

*smacks face*

 

[johnwmclean]

Well that is good news bolinger, congrats. I'm glad it hasn't ended up as some ugly doorstop. lol.

Beefys got some word eating to do now. hehehehe

 

[wink]

johnwmclean said:
Quote:

Beefys got some word eating to do now. hehehehe

.

Gee, I hope they're tasty. I hesitate to suggest like beef jerky.
The inuendos alone are staggering.
I think we should leave the poor dude alone in light of his good SOUND advice in most of his posts...

 

[Beefy]

I stand by my position. I'm glad it is fixed, but this guy shouldn't have done a B22 as his first build.

 

[johnwmclean]

Quote:

Originally Posted by Beefy /img/forum/go_quote.gif I stand by my position. I'm glad it is fixed, but this guy shouldn't have done a B22 as his first build.

I’m with you on that Beefy, the earlier comments regarding mains power were very worrying to say the least. Some of my future builds will require working with much higher voltages, I’m currently researching all aspects of safety working under those conditions - to tell you the truth it’s kind of scared me off

 

[boinger]

You know i learned quite a lot from the initial postings and i did a lot of research before commiting to it... was just concerned for my own safety etc.. i have done work on lots of electronics before mainly computers i just didnt know the techincal details etc. i figure rather ask then getting hurt... this was my first audio diy build but i dont regret building the beta22 its much better for me to have done this than done something else and not afford it later i dont have an infinite supply of money.

i just believe that if you do your homework and research it well you should be able to gauge whether or not your comfortable doing such a project or not. I am by no means an idiot that was asking for step by step instructions on how to set it up etc. for the most part i got most of it working initially on my own. Aside for me blowing q25 i think it was due to me poking through the soldermask on r9.

 

[luvdunhill]

Ti:

Given unlimited heatsinking for the pass devices, how much current can the sigma22 provide before the fact that the rectifier diodes aren't heat sinked (and are in relative close proximity) becomes an issue?

 

[Medel]

I just "finished" my 3-channel today (about 2 songs ago)!

Remarkably everything worked first time. I had one issue where I keep doing work late at night and read my transformer pinout wrong a few weeks ago on my sigma build, but I caught it even before my fuse blew. I only have my Senn 555s from my e-drum here tonight, but they already sound amazing. I finished it just in time for a week long vaca I'd planned of doing nothing but reading and listening to my 702s on it. I'm so excited! It still needs a volume knob and feet on both cases and it's done. I'd like to clean up the wiring since it's awful, but that would require not listening to it for a bit and that probably won't happen for a while. Thank you all so much for the community here. It was definitely a confidence builder that there was help if I needed it, and Ti's website could not be better! I'll get pics up soon enough in the galleries. Cheers!
-Mike

 

[amb]

Quote:

Originally Posted by luvdunhill /img/forum/go_quote.gif Given unlimited heatsinking for the pass devices, how much current can the sigma22 provide before the fact that the rectifier diodes aren't heat sinked (and are in relative close proximity) becomes an issue?

You need to qualify what "how much current" means. Is it continuous duty? Or is it short duration peaks? If peaks, then is it repetitive or not? If repetitive, what's the duty cycle?

Assuming continuous, the current limitation will most likely still be limited by the MOSFETs' thermal resistance (and any isolation pads/thermal compound used) even if you use huge heatsink(s). The total thermal resistance is the sum of the device's own thermal resistance, isolation pad/thermal compound thermal resistance, and heatsink thermal resistance. At some point as you increase the size/efficiency of the heatsink, its thermal resistance becomes a small portion of that sum and the MOSFET's own thermal resistance becomes dominant. This is a factor in the computation of the junction temperature, which is outlined in the "board and heatsinks" section of the σ22 website. You can plug in a very low number for Rsa in the equation and see for yourself.

Other than that, the rectifier diodes and power transformer are the other limiting factors. Assuming you use MUR820 rectifiers, the forward drop (Vf) is slightly less than 1V, the instantaneous power dissipated on each of them is the current through it multiplied by Vf. But, keep in mind that the current through the rectifier is not continuous, but AC pulses, so the average power is much less. These are rated at 8A, and can pass much higher peak currents. In normal use situations they will not need heatsinking because we'd have reached the thermal limit of the MOSFETs first.

If the load is pulsive in nature rather than continuous, then σ22 can deliver a lot of peak output current, well beyond the calculations based on continuous duty. How much is "a lot"? Well, it depends on the nature of the load current.

 

[luvdunhill]

Quote:

Originally Posted by amb /img/forum/go_quote.gif You need to qualify what "how much current" means. Is it continuous duty? Or is it short duration peaks? If peaks, then is it repetitive or not? If repetitive, what's the duty cycle? Assuming continuous, the current limitation will most likely still be limited by the MOSFETs' thermal resistance (and any isolation pads/thermal compound used) even if you use huge heatsink(s). The total thermal resistance is the sum of the device's own thermal resistance, isolation pad/thermal compound thermal resistance, and heatsink thermal resistance. At some point as you increase the size/efficiency of the heatsink, its thermal resistance becomes a small portion of that sum and the MOSFET's own thermal resistance becomes dominant. This is a factor in the computation of the junction temperature, which is outlined in the "board and heatsinks" section of the σ22 website. You can plug in a very low number for Rsa in the equation and see for yourself. Other than that, the rectifier diodes and power transformer are the other limiting factors. Assuming you use MUR820 rectifiers, the forward drop (Vf) is slightly less than 1V, the instantaneous power dissipated on each of them is the current through it multiplied by Vf. But, keep in mind that the current through the rectifier is not continuous, but AC pulses, so the average power is much less. These are rated at 8A, and can pass much higher peak currents. In normal use situations they will not need heatsinking because we'd have reached the thermal limit of the MOSFETs first. If the load is pulsive in nature rather than continuous, then σ22 can deliver a lot of peak output current, well beyond the calculations based on continuous duty. How much is "a lot"? Well, it depends on the nature of the load current.

Yes, fair point. I was thinking along the lines of use in a Class A power amplifier. For example, the popular First Watt amplifiers by Nelson Pass all use 300VA transformers. He specifies bridges rated at 35A and experience shows that these bridges need to be mounted to thick base, or the heat sinks themselves to handle the current. I just couldn't imagine that the rectifiers in a sigma22 with a similar sized toroid that's being pushed towards it limit of 300VA would be all that happy, having dealt with this particular scenario. I'm not at all suggesting such a use, but it seems that there is an upper limit and I just wanted to get an idea what it was.

 

[Anonanimal]

Quote:

Originally Posted by Medel /img/forum/go_quote.gif I just "finished" my 3-channel today (about 2 songs ago)! Remarkably everything worked first time. I had one issue where I keep doing work late at night and read my transformer pinout wrong a few weeks ago on my sigma build, but I caught it even before my fuse blew. I only have my Senn 555s from my e-drum here tonight, but they already sound amazing. I finished it just in time for a week long vaca I'd planned of doing nothing but reading and listening to my 702s on it. I'm so excited! It still needs a volume knob and feet on both cases and it's done. I'd like to clean up the wiring since it's awful, but that would require not listening to it for a bit and that probably won't happen for a while. Thank you all so much for the community here. It was definitely a confidence builder that there was help if I needed it, and Ti's website could not be better! I'll get pics up soon enough in the galleries. Cheers! -Mike

Congrats! I'm happy to hear the build went well and that you're enjoying your new amp.

 

[Magsy]

Can somebody explain the calculation for gain please?

I think I need something like 6-8 in balanced, so 3-4 per board. These options are not listed on the website and I'm not sure I can just extrapolate the figures to get 3 or 4.

I have the components for 2 (4 overall) but I am constantly at 12-1 o clock on my M cubed, which has a gain of 11 and I fear 4 won't be enough. My source is 2v unbal or 4v bal.

11 is too much on my M cubed but a gain of 4 is almost a third and if that means it'll produce a third of the voulme it won't be enough. I am aware the pot is not linear but I'm too stupid to work it all out. I would like to use all of my attenuators travel with a click or two spare.

Thanks in advance... :]