Dynahi help

[NRD]

I've finished building the Dynahi and got it powered up but need help with bias and dc offset. I read AMB's comments in the build thread where he mentioned using two trim pots per board, but I also came across SFT's comments of being able to use one. Here is the thing, I got the DC offset down from -100mv in the left channel to ~1mv and from -75mv in the right channel to ~1mv (all this with no servo installed) BUT the offset seems to oscillate/vary (especially in the left channel) anywhere from -8mv to +4mv even after the thing has been on for 10min.

Q1) That said, would using two pots per board keep things more stable? If so, what value across pins 2 and 3 of the pots should I start with (I don't know if I should start at max pot resistance in each pot or min)? I'd prefer to try to get away with using one pot, since I'd have to go out an buy two more, but if it really is a more stable config, I'd rather go that route.



Q2) I get some hum depending on the position the volume pot is at. If it is at 50% it is dead silent (with 16ohm earphones too!), at 100% (i.e. max volume)dead silent until I plug my ipod into the RCA jacks then I get hum, at 0% I get slight hum regardless of the use of RCA's. Any suggestions? Ground scheme is as follows: all grounds tie together at one point at the chassis, but the RCA jacks themselves are grounded to the chassis on the back chassis panel, from which they connect to the board's "signal input" ground node. Also, fyi to prevent pulling any noise, I am not using the power outlet ground

 

[amb]

Quote:

Originally Posted by NRD /img/forum/go_quote.gif I got the DC offset down from -100mv in the left channel to ~1mv and from -75mv in the right channel to ~1mv (all this with no servo installed) BUT the offset seems to oscillate/vary (especially in the left channel) anywhere from -8mv to +4mv even after the thing has been on for 10min.

The DC offset may drift to temperature variations without the DC servo. -8mV to +4mV isn't great, but is not really out of whack either.

Quote:

Q1) That said, would using two pots per board keep things more stable?

No, but having two trimpots per board not only lets you adjust for minimum DC offset, but it also allows you to tune the overall bias. This may be necessary if your input stage CCS LEDs are not exactly the "right" voltage. Just a little off and your output stage could be flowing a lot less or more current than the designed value.

Quote:

If so, what value across pins 2 and 3 of the pots should I start with (I don't know if I should start at max pot resistance in each pot or min)? I'd prefer to try to get away with using one pot, since I'd have to go out an buy two more, but if it really is a more stable config, I'd rather go that route.

Like I said, having two pots is not going to make it more "stable", it just gives you a bit more bias adjustability. In the dynahi, the input stage bias influences the bias of the VAS stage as well as the output stage, so it makes sense to have that adjustability.

As for DC offset stability, once you put in the DC servo opamp things should get a lot better.

Quote:

Q2) I get some hum depending on the position the volume pot is at. If it is at 50% it is dead silent (with 16ohm earphones too!), at 100% (i.e. max volume)dead silent until I plug my ipod into the RCA jacks then I get hum, at 0% I get slight hum regardless of the use of RCA's. Any suggestions?

Try adding a input series resistor after the pot, I used a 475 ohm, but anything up to 2.21K is ok. This resistor should have been on the PCB in the first place. See my dynahi schematic for that resistor.

Quote:

Ground scheme is as follows: all grounds tie together at one point at the chassis, but the RCA jacks themselves are grounded to the chassis on the back chassis panel, from which they connect to the board's "signal input" ground node. Also, fyi to prevent pulling any noise, I am not using the power outlet ground

You may have a small ground loop. Since the amp's ground is already tied to the chassis at the star point, having an additional connection at the input RCA jacks to the chassis creates multiple paths. It may or may not be a problem. But you might want to try isolating the RCA jacks to see if it makes any difference. If you don't want to isolate the jacks, the alternative is to try decoupling the star ground point from the chassis. It would still be a star ground, just not tied to the chassis there.

If the power transformer and IEC are in the same case as your amp, then I highly recommend that you connect the AC ground to the chassis for safety. Use a ground loop breaker between the chassis and the amp's signal ground.

I suggest reading my posts in the headwize dynahi amp thread.

 

[NRD]

Thanks AMB. One more quick question, I measured the voltage across the 20ohm output resistors and got ~0.85v, meaning ~85ma. Is this ok or too much bias? I tried to use two 10k pots but couldn't get it down, I guess I need to get bigger pots, but I want to listen sooo bad I hope it is ok

 

[amb]

Quote:

Originally Posted by NRD /img/forum/go_quote.gif Thanks AMB. One more quick question, I measured the voltage across the 20ohm output resistors and got ~0.85v, meaning ~85ma. Is this ok or too much bias? I tried to use two 10k pots but couldn't get it down, I guess I need to get bigger pots, but I want to listen sooo bad I hope it is ok

The designed value should be around 75mV but 85mV is not too far off. It means that the output transistors will run a little hotter but if your heatsink is big enough then it should be ok.

If you want to turn down the bias a bit, don't substitute with higher value trimpots. Keep them at 10K, but swap the 500 ohm resistors (in parallel with the trimpots) with 620 ohm or 680 ohm. This will then allow you to get the proper range of adjustment.

 

[lcp12345]

Where can I get the Dynahi PCB else other than Dan? I have got no reply from him for a month. Do I email to wrong address? Is email of Dan is dan@djgardner.com? Thanks for any response.