To previous builder how hot does your case get. For me get as hot as 50 degrees C

for signal input, unbalanced you put hot to amplifier input +

you put ground to signal ground and input -

 

for output you wire hot to amplifier out +

and you wire ground to signal ground

 

anything else, and bad things happen

Thanks I'll try that.

BTW been using the amp with the case close for an hour it does get quite hot the caae is barely touchable now. The case is sealed full aluminum with no ventilation. Looks like I need to drill holes for ventilation. Or is it okay to leave it as it is. Considering the full aluminum case will act like a big heatsink. The case is 3u rack case but I am putting 2 amp board and 2 sigma22 board in it. Soo its packed. I put the transformer in another case.

Will the temperature hike effect the bias and offset? Can't really check it with the case closed

holes, lots of holes

top and bottom

Top and bottom? So top only not enough? I'm thinking 20 holes around 6mm possition parallel with the heatsinks would that be enough?

has to suck the heat from somewhere 

the heated heatsinks cause airflow

A high power class A amp in a sealed box is a very bad idea so holes and more holes in those holes. 

drill away it is

try not to do it hennyo style!

not sure what hennyo style is

 

BTW Looks like a protector board or delay board is really necessary with my build on switch on and off the offset jumps to 1V for 1 second before it stabilize to 10-15mV after 3 second

for signal input, unbalanced you put hot to amplifier input +

you put ground to signal ground and input -

 

for output you wire hot to amplifier out +

and you wire ground to signal ground

 

anything else, and bad things happen

 

 

Yes, I use a similar scheme. The sound is very good but I have a little HUM on the left side when I use the input unbalanced/ output unbalanced.

If I use a protector board the only cut of point is 0.5v. Isn't that too high already? Does anyone know any design that cut off below 0.5v? What is the safe cut off point for headphone?

Try AMB's e12. I'm using it in almost every build of beta22, and will use for dynahi as well.

To my opinion, 0.5v is too large and may damage some headphones, while others will survive. Anyway, I wouldn't risk my collection of 1000-1500USD headphones. I want the amp to be as most universal as possible.

Edited July 9, 201510 yr by Helium

I think kevin mentioned elsewhere the e12 doesn't work

you can change one resistor and the setpoint will be .25v

or anything else you want.

 

the e12 does not protect against dc offsets at the imput

causing differential output voltage that can be significant

but adds up to zero and then won't trip the relay

Wouldn't two e12s work for balanced; 1 for the + signals and 1 for the - signals? On amb's site, he only states that 1 e12 per PSU should be used, which I'm not quite sure I understand the rationale for. I never cared for the virtual ground on the e12 board, but understand the need for it given some of the amps that this is intended for.

 

I agree that your protector board would be better and more economical.

the problem with the e12 is the common summing.

 

so even if you do 2 boards, and one hot output is +2v and the other

hot output is -2v, the result is zero. if you think that can't happen with

some of these whacky dacs, think again

 

you really do need a pair of comparators on each of the 4 lines

Edited July 9, 201510 yr by kevin gilmore

you can change one resistor and the setpoint will be .25v

or anything else you want.

 

the e12 does not protect against dc offsets at the imput

causing differential output voltage that can be significant

but adds up to zero and then won't trip the relay

Referencing you balanced protector board http://gilmore.chem.northwestern.edu/boards/protect3.zipwhich resistor should I change to reduce the trip point to 0.25V?

On amb's site, he only states that 1 e12 per PSU should be used, which I'm not quite sure I understand the rationale for.

It's because e12's virtual ground is referenced to PS ground. Imagine you are using 2 sigma22 and only 1 e12, which is referenced to one sigma22. Even if you built both sigma for the same voltage, due to parts tolerance the DC voltage of both will not be identical. Even 0.1v difference of sigmas output will make e12 false trigger because it will think that DC offset is 0.1v+amp's native DC offset.

the 1k resistor in series with the 2 x 11k resistors.

 

change to 500 ohms

 

and I never liked virtual ground anything.

just a bad idea.

Edited July 10, 201510 yr by kevin gilmore

The modified e12 boards that I laid out and home etched had a PSU ground connection on them and this was what I used for the comparator ground, not the virtual ground. The vgnd bounced around too much, particularly before he went to the rail to rail relay.

 

I understand the explanation of why 2 boards may not work as well, so thanks for that. One thing I hadn't considered with the summing.

So I'm still trying to figure out where the ground noise is coming from. My setup is 2 case: metal case housing amp and power supply and plastic case housing the transformer. DC power plug, trs jack are isolated from case ground. The volume pot are not isolated from case ground. Should the rca be isolated or connected to the case ground?

RCA Should be isolated from metal case

Does anyone know where if there is grounding guide write up for the Susy dynahi? Or this b22 grounding guide will work with the dynahi?http://www.amb.org/audio/beta22/wiring.shtml

If I want to decrease bias do I need to decrease the 500ohm resistors in parallel with the 10k pots?

you increase the resistance, not decrease it.