KG Balanced Dynahi build discussion thread

[jose]

The problem with using two PSUs is precisely the differential between the two PSU grounds. It´s a pain especially if you use a volume pot in your amp... I know quite well.

I usually use a star "loop" (connect all the ground in this point), the check, if you have problems  like noises, etc... disconnect the ground from one PSU and check it again. If this does not work for you, you should try different combinations of doughs. It is time consuming and quite tedious.

I have read several posts where using ONE PSU is preferred for this reason.

 

[Pars]

Thanks Jose. I'm not having ground loop problems, but because of this differential in the grounds between the 2 PSUs (with one ground missing from the pot to the amp board) I was having the protector board trip (it was grounded to the other amp board). It doesn't look like the grounds of the amp boards or PSUs are tied to the chassis at any point either. The only ground connection between the R and L channels was at the pot. Reconnecting the ground on the pot seems to clear up the problem completely. I think I will run a ground wire between the output grounds on both PSUs.

[jose]

If that cable fixed the problem, leave it like that. 🙂 

[penmarker]

Started this project on 2018 I believe. Was a penny pincher because of budgets so cut a lot of corners. Ultimately this isn’t a project you’d want to cut corners. Finally went around finishing it this week because my area got into lockdown and I’m desperate to get myself preoccupied. Ordered some parts and went ahead to redo and finish it. 
 

Initially I built it in one case but the transformer was inducing noise into the board. With the whole PSU (including small 12v transformer for channel selector) rehoused in a separate chassis it’s silent. The amp is being powered by a single GRLV for now. PSU output devices are bolted to the chassis heatsink and only gets somewhat warm. 
 

The amp bottom right obviously has canted headphone/XLR outputs  because I used a hand drill to make mounting holes and made a mistake with the angle. 
 

The dynahi makes a great sounding preamp too. 
 

One question I have is I got some C3881/A1349 to replace my noisy THAT340, and after storage for a year or so they got distorted. These were bought from a reputable seller in diyaudio. Anyone has any idea what happened?

[starcat]

On 3/21/2021 at 5:22 AM, mwl168 said:

I tie the ground of the PSU together when I do a dual mono amp - one wire that goes from the output ground of one PSU to that of the the other one and then, from there, a single wire to the chassis. I do the same thing when tying the LV and HV PSU ground together. So far this has worked well for me.  But I typically do not install a volume pot on my headphone amps. 

Michael, why not running the ground wire from each of the PSUs directly to the star ground (instead of the serial connection of PSUs)?

[Pars]

On 7/18/2021 at 2:00 AM, penmarker said:

Started this project on 2018 I believe. Was a penny pincher because of budgets so cut a lot of corners. Ultimately this isn’t a project you’d want to cut corners. Finally went around finishing it this week because my area got into lockdown and I’m desperate to get myself preoccupied. Ordered some parts and went ahead to redo and finish it. 
 

Initially I built it in one case but the transformer was inducing noise into the board. With the whole PSU (including small 12v transformer for channel selector) rehoused in a separate chassis it’s silent. The amp is being powered by a single GRLV for now. PSU output devices are bolted to the chassis heatsink and only gets somewhat warm. 
 

The amp bottom right obviously has canted headphone/XLR outputs  because I used a hand drill to make mounting holes and made a mistake with the angle. 
 

The dynahi makes a great sounding preamp too. 
 

One question I have is I got some C3881/A1349 to replace my noisy THAT340, and after storage for a year or so they got distorted. These were bought from a reputable seller in diyaudio. Anyone has any idea what happened?

 

Did you remove or re-install these after bringing the amp out of storage? From your pic above, pin 1 doesn't seem to be clearly marked, so the potential for putting these in backwards is there. Also, have you tried it with the THAT340s?

[penmarker]

11 hours ago, Pars said:

Did you remove or re-install these after bringing the amp out of storage? From your pic above, pin 1 doesn't seem to be clearly marked, so the potential for putting these in backwards is there. Also, have you tried it with the THAT340s?

Nah these were left on during storage. I don’t think I’ve put them in backwards, will they work if they’re put in backwards?

Just checked I can’t see the pin1 marking on these. Hmmm yeah it’s possible I put them in backwards. Also the pins has some green colour like copper tarnishing. It came off by rubbing.

The THAT340’s are working fine. Though they’re noisy with some minor intermittent frying sound. With my low sensitivity headphones and as a pre out the noise isn’t audible. 

[Pars]

No, they won't work correctly backwards, as it basically flips the N's and P's, and those will be in backwards as well (emitters/collectors). Barring having a "1" marking, the pad for pin 1 should be square vs. round for the others. From what I recall, putting a similar JFET module in backwards resulted in nasty DC offset. I never tried to see if any sound came through or not. The seller of diyaudio of these  does provide genuine devices. Also, it should be noted that a similar adapter for dual JFETs (2SJ109/2SK389) will not work for bipolars (SA1349/SC3381) as the pinouts are different, even though they use the same device package. Yours are marked for the bipolars, so should be good.

Is the distortion in both channels?

[penmarker]

I've inserted the SA1349/SC3381 correctly this time, the square pad is less visible now after filled with solder but can still be identified. They sound fine, although they do have the same slight scratchy noise sound the THAT340 developed. It's intermittent and I only notice it in the right channel for now.

Maybe its time to try out the K170/J74. The noise floor on the SA1349/SC3381 is similar or slightly lower than the THAT340 and I can hear it on sensitive cans. How does yours sound?

I'm very glad I built the Dynahi, they sound better than everything else I've listened before.

[Pars]

I use JFETs, which to me sound markedly superior to bipolars in the front end.

I'm not so sure your problem lies with the front end. Is the scratchiness there noise there without changing volume, etc.? What are you using for a volume pot? Does the source feeding this have any dc offset present? As a dc coupled amp, it will not like that, and will amplify any dc offset present on the input.

[penmarker]

I'm using a TKD4CP and the scratchiness had been there even before I used a volume pot, when I was using the pre out from my DAC. There is a small offset drift within the low 10's mV, but it goes from positive to negative. There's no distortion from the volume pot, the TKD is pretty smooth.

I don't think any of my sources has any DC offset (Hugo V1, Matrix Mini-i). This doesn't happen with my other headphone amps.

Since the opamp is in there as a servo, could changing the opamps help with the drifting offset? I'm half expecting it to be the opamp because I noticed the offset doesn't change with and without the opamp.

[Pars]

You could try replacing the opamp. Maybe something is wrong with that servo?

Your other headphone amps could very well be ac coupled (cap on the input), so would not be effected by dc offset on the source. Is the dc offset you mention with a DAC connected and powered up? You might take a look at that.

[mdr30]

Finally completed my Dynahi, an on/off project (like some of them) for five years. Bought the set of boards from Samsie in December 2015, so I thought I'd just let him know that they eventually came of use!

Turned out to be a relatively uncomplicated build, except for a couple of resistors around the THAT340 with too many zeros, discovered by MLA.

Also a problem with the -30V side on one of the PSUs. Turned out to be a malfunctioning rectifier. A BC556 popped, but the supply worked once replaced. For a week or so, then the 556 popped again. After the third time I scratched my head, but some sparks from one of the surrounding caps told me it had taken some beating from the faulty rectifier. Replaced that and the DN2540 and now it has played without incidents for a couple of weeks.

Sounds very nice and has become my go-to-amp, with some competition from a KSA5 I took over from MLA recently. Spoiled for choice.

 

 

[penmarker]

On 7/30/2021 at 2:13 AM, Pars said:

You could try replacing the opamp. Maybe something is wrong with that servo?

Your other headphone amps could very well be ac coupled (cap on the input), so would not be effected by dc offset on the source. Is the dc offset you mention with a DAC connected and powered up? You might take a look at that.

I switched opamps with no effect, then I switched out my SA1349/SC3381 between L and R and I can hear that the scratchy sound is now on the left channel. It is mostly audible while the amp is going up to temp, but it's hard to reproduce and its long periods between noises. Seem to be random. The other channel also have that scratchy sound but at a lower level, I can hear that on sensitive headphones.

Well at least now I know what's the problem now.

[wudilmy]

Has someone done square wave test for ss dynahi? Was that good at high frequenz?

[Four Kneez]

I was able to pick up an old Dynahi a few months back. Even though it still operates just fine, the chassis is pretty beat up and the internals are quite a mess, so it's mostly been gathering dust on the shelf since it arrived. However, I'd like to start using it again, and I think it deserves to be turned into something nice. Would anyone here possibly be open to repairing it for me? I'd want to work on it myself, but I'm currently swamped with work and am also coming up on a career change. So, I won't have enough time to sit down and really focus on it until well into 2022 at the earliest.

Edit: Someone is helping me work on it.

Edited September 14, 2021 by Four Kneez

[wudilmy]

Does someone know if the compensation capacitor is less than 2pf(1.5pf) for 20x gain, will it cause instability of the amp?

I have bought 2pf capacitors with 0.5pf tolerance and the value measured is 1.5pf😂.

Thanks a lot. 

[wudilmy]

Another question, if i replace the that340 with k170/j74, are there some resistors should also be changed or keep all other components the same?

Thanks for answer.😀

[Pars]

I think your comp. capacitors should be fine, but I don't know for sure. On the THAT340 replacement, no changes required. Just make sure you put them in correctly (N-JFET for NPN and P-JFET for PNP).

[wudilmy]

12 hours ago, Pars said:

I think your comp. capacitors should be fine, but I don't know for sure. On the THAT340 replacement, no changes required. Just make sure you put them in correctly (N-JFET for NPN and P-JFET for PNP).

Thank you for your answer😄, then that will be easy.

[JoaMat]

Circled 50R resistors seem to have been 100R in earlier versions of Dynahi.

What might the reason for the change be?

[kevin gilmore]

no idea, have to look where those resistors go. if those are the input resistors, anything over 50 ohms is fine.

[JoaMat]

They are R23/34 in schematic:

dynahibal1.pdf

Edited March 20, 2022 by JoaMat

[kevin gilmore]

those resistors are output stage bias set.

[JoaMat]

Senior HeadCase member MLA has lent me his Dynahi. A week listening to it have convinced me that one actually can listen to non-electrostatic headphones.

So here is my first attempt...

... a balanced something with lsj689 and lsk489. And as always everything I make is based on a Kevin design. In this case on the famous Dynahi.

Edited March 21, 2022 by JoaMat