KG Balanced Dynahi build discussion thread

[luvdunhill]

Sadly not, they bounce around the long wires and suffer impact damage as they hit other molecules which makes them smaller. It's a real problem...

Using square wires seems to help with this.

[spritzer]

So says Tara Labs.  It also helps to make them in China but label them as made in the US.  Not sure why that matters but anything that helps must be a good thing... 

[mahdi8]

using modushop cases you should be able to find something that fits whether it's the offboard or onboard version

Thanks I'll check them out.

 

I'm considering the have the dynahi balanced with 2 unbalanced input and one balanced input with an balanced and unbalanced input switch.

 

The output would be one balanced and one unbalanced. for unbalanced I will wire the left to + board 1, ground to  - board 1, right wire to + board 2.

 

Will that work? or any better suggestion to have both balanced input and output configuration?

[kevin gilmore]

if  you are talking about the inputs, yes that is how its done.

 

don't short the outputs to ground

 

a stack of relays, or rotary switches is necessary to do the inputs right

[ang728]

Thanks I'll check them out.

 

I'm considering the have the dynahi balanced with 2 unbalanced input and one balanced input with an balanced and unbalanced input switch.

 

The output would be one balanced and one unbalanced. for unbalanced I will wire the left to + board 1, ground to  - board 1, right wire to + board 2.

 

Will that work? or any better suggestion to have both balanced input and output configuration?

 

for outputs, you can consider using XLR combo jack ,then you can have one balanced out and two single ended outputs

[Pars]

^ Don't think that would work. The combo jacks are 3-pin XLR and a TRS. You need 4-pin (minimum) XLR for balanced output (L+ / L- / R+ / R-). L- and R- cannot be tied together.

[mahdi8]

for outputs, you can consider using XLR combo jack ,then you can have one balanced out and two single ended outputs

 

^ Don't think that would work. The combo jacks are 3-pin XLR and a TRS. You need 4-pin (minimum) XLR for balanced output (L+ / L- / R+ / R-). L- and R- cannot be tied together.

Agree with Pars I can't see how using combo jack will work. Unless you have a balanced to unbalanced switch of some sort? but I think that will make it more complex.

 

How does ground gets wired in a standalone single ended dynahi? Is it a passthrough from the source ground?

[Pars]

^ It is wired to the analog ground of either the amp board or to the ground on the PSU (single ended dynahi).

[mahdi8]

^ It is wired to the analog ground of either the amp board or to the ground on the PSU (single ended dynahi).

Great so technically I can follow that way and just use one board when running single ended? I might put a switch so I can run 2 single ended input output the same time or running one balanced. That would be perfect. Might need to make a switchboard to handle it

[ang728]

Was guessing this method might work,just simply tap signals to trs contact but complexity for sure

from amb's beta22 page

[spritzer]

B22 and our Susy amps are very different. 

[ang728]

so this combo output wiring won't work on susy design?

I thought it will work on any amp if I tap R+ L+ G、R- L- G  to two TRS contacts to drive two headphones (feeding balanced signal for sure)

[Pars]

For single-ended, because of the fully differential input stage, L- and R- need to be grounded when using the SE input, and wired normally when using balanced input. You would need some kind of switch to do this. It had been suggested that a 4PDT switch be used, to switch all 4 signals between the RCA/XLR, and to ground the - signals when the RCAs are being used.

 

I'm not sure if you could get away with just using a DPDT to switch the - either to amp ground when using the RCA or to the XLRs in balanced mode. This would leave the + inputs connected to both the RCA and the XLR at all times.

[kevin gilmore]

you can give the amp a differential balanced input, and use the unbalanced output.

 

why you would want to do this i don't know.

[ang728]

Thanks for clarifying

 

some of my headphones is still sigle ended.

thinking it would be useful to have TRS out instead having two different build,just saving some flexibility 

[mahdi8]

For single-ended, because of the fully differential input stage, L- and R- need to be grounded when using the SE input, and wired normally when using balanced input. You would need some kind of switch to do this. It had been suggested that a 4PDT switch be used, to switch all 4 signals between the RCA/XLR, and to ground the - signals when the RCAs are being used.

 

I'm not sure if you could get away with just using a DPDT to switch the - either to amp ground when using the RCA or to the XLRs in balanced mode. This would leave the + inputs connected to both the RCA and the XLR at all times.

I might give 4PDT Switch a try. looks promising

[mahdi8]

Building is running well but still quite a bit to go. managed to get a large case to put the amplifier boards and the power supply boards but I still keep the transformer separate in a smaller case. Was going to have 2 of the large case (bought both of them actually) but decided it's too big. Got a 4PDT switch also so will give it a try. Anybody think if there is any issue running output of single ended and balanced simultaneously? I will only put the 4pdt switch on the input. or should I put it in both input and output?

 

https://flic.kr/p/s6KQxM

https://flic.kr/p/rrd45N

https://flic.kr/p/rrpkdg

[Zashoomin]

I think that your khozmo pot is a 50k ohm one.  I believe you want 10k if I am not mistaken.  

[Pars]

The outputs should be fine (SE and balanced), no switch there. The differential input requires the - signal to be grounded in the case of SE, so that is why you need a switch or something there. On the Pass Aleph P preamp I had, which has both SE and balanced inputs, shorting plugs were provided which needed to be inserted into the XLRs on channels where you were running SE input on. The XLR->RCA adapters do the same thing.

[mahdi8]

I think that your khozmo pot is a 50k ohm one. I believe you want 10k if I am not mistaken.

It is the 50k, I am continuing a abandoned project so that's what I got. Anyone think this will be a problem? Prefer to use this or the existing pot would be a waste.

In another matter the initial owner says he planned to have it to be 8x gain in balanced. But I rechecked all the resistor in the board are default value. Shouldn't I change the input resistor to be 50k to have 8x gain? That way it matches the volume pot?

Or can I replace the series resistor in the pot from 50k to 10k and leave the shunt resistor alone

[mahdi8]

Got a reply from Khozmo manufacturer looks like replacing the series resistor to 10k is the go

[luvdunhill]

Pretty impressive engineering these guys have done there! How user friendly?

[mahdi8]

Big set back on my build. Was doing the voltage balance and offset and manage to get around 40mv and 50mv, but can't get lower the servo seems to bring the voltage up every time I reduce the trimpot, so decided to try on a throwaway headphone. There was quite a high noise floor which sound like a ground loop. So then I change the ground in the pot from power supply ground to the source xlr ground. The ground him dissapear but there is still a high noise floor.

Couple second later one of the capacitor blows. Turns out the guy I got the board from populated them with a 25v caps rather than 35v caps written on the board.

I think this was where the noise was coming from. What do you guys thinks?

Really hoping none of the transistor gets fried. Because that would be a real pain to source. Caps are easy. Will replace all the caps and put in the correct voltage.

[mahdi8]

More progress now. I haven't found the 55k resistor so I haven't replace the gain. But all the caps has been replaced with a proper 470uf 35v caps.

On default 500 ohm resistor that paraller to the trimpot the bias was stuck at 200mV so I changed it to 680 ohm. Adjust the bias the best currently I can get is 78mV for + and 88mV for - on the 20ohm parallel resistor. Output offset is now at 5mV. I thought it would be good enough to retest the hiss.

The power supply is using a test DC PSU set at 30V just to make sure the hiss is not from the sigma22. Trim pot is not installed so I still leave the 1k resistor shorted to ground that I used to do the bias set up. Hook up the test headphone and can hear the hiss noise. Can't figure out where its from. Any suggestions on how to troubleshoot this? Its not connected to any source or pot so that shouldn't be an issue. And its connected to a a clean DC test power supply so it won't be PSU issue.

[Pars]

Input grounded? Or do you have a volume pot installed and turned down all the way? If not, ground the inputs.