KG Balanced Dynahi build discussion thread

[UFN]

19 hours ago, Pars said:

I'm just looking for the best inexpensive device tester for validating devices. DY294? The Atlas DCA (though it is 4x the price of the 294, and won't do higher voltages).

Not sure about the DCA75, but FWIW I have both the smaller DCA55 and the DY294. The DCA55 is brilliant but the DY294 has a couple of features that the DCA55 hasn't. First of all it has the HV breakdown test for (potentially) spotting fake HV-transistiors and secondly it can measure BJTs at different base currents. The DCA55 only tests BJTs at 2.5mA collector current which honestly isn't much if it's a driver- or powertransistor.

So, my conclusion would be that I actually need both  If I had to only get one it would probably be the DCA55 though, and the DCA75 might have even more features.

 

//UFN

[ti5002000]

i just came to left a photo of my Dynahi and also to thank all those who help specially Kevin Gilmore ,the man himself,for all the patience with me ( and even when he didn´t have it) and all his clarifications. Without  that now i don´t have the Dynahi to show.

About the amplifier himself it is truly neutral in my opinion .you change a thing in sound chain ,and you notice the change.Fantastic.

Also the sound balance is a bliss .My HD800 with my Chord 2Qute are at his best.

FANTASTIC REALLY.

[ang728]

54 minutes ago, ti5002000 said:

i just came to left a photo of my Dynahi and also to thank all those who help specially Kevin Gilmore ,the man himself,for all the patience with me ( and even when he didn´t have it) and all his clarifications. Without  that now i don´t have the Dynahi to show.

About the amplifier himself it is truly neutral in my opinion .you change a thing in sound chain ,and you notice the change.Fantastic.

Also the sound balance is a bliss .My HD800 with my Chord 2Qute are at his best.

FANTASTIC REALLY.

 

Nice!! gald to see you tackle all the problems

[ti5002000]

18 minutes ago, ang728 said:

 

Nice!! gald to see you tackle all the problems

The noise floor is still there but i will try to change the gain next week ,to see if it will solve it.

But already sounds great.

[sbelyo]

Out of these two layouts, what looks best to you all?  This is the PSU chassis for my Dynahi.

 

Edited December 28, 2016 by sbelyo

[nopants]

the latter

[ang728]

latter for sure 

[sbelyo]

I agree

 

[Pars]

Does anyone have a BOM for this (SS Dynahi)? The one that Vortex posted in the 1st post is blank, and I suspect for the Toshiba boards anyhow. Lil Knights also was for Toshiba parts.

Edited March 4, 2017 by Pars

[jose]

Componente		Unidades/2ch				Ref. Mouser
20	R	64				71-RN60D-F-20
50		4				71-RN60D-F-49.9
100		12				71-RN60D-F-100
300		16				71-RN60D-F-301
500		4				71-RN60D-F-499
1 K		4				71-RN60D-F-1.0K
1 M		4				71-RN60D-F-1M
10 K		8				71-RN60D-F-10K
200 K		4				71-RN60D-F-200K
3,3 K		8				71-RN60D-F-3.32K
30 K		4				71-RN60D-F-30.1K
5 K		14				71-RN60D-F-4.99K
OPA445		2				OPA445AU
MPSW56		14				MPSW56RLRPG
MPSW06		14				MPSW06G
THAT 340		2				887-340P14-U
10K	Trim	8				652-3299W-1-103
0,1 uF	Cap.	4				FK26X7R2E104K Or C320C104K1R5TA7303
0,47 uF		2				MKP4F034704D00JSSD
0,2 uF		4				FK28C0G1H020C
1000 uF/35 V	Electrolítico	12				UKW1V102MHD
	Led	4				351-5001
	Disipadores	16				529802B02500G

[Pars]

Thanks all, got a couple of BOMs now.

You guys don't like putting the output devices in your BOMs, eh?

 

EDIT: also, can MJF15030/31s be used in place of the MJE15030/31? Specs seem the same with the exception of rated for 36W instead of 50W.

Edited March 5, 2017 by Pars

[ti5002000]

19 hours ago, Pars said:

Thanks all, got a couple of BOMs now.

You guys don't like putting the output devices in your BOMs, eh?

 

EDIT: also, can MJF15030/31s be used in place of the MJE15030/31? Specs seem the same with the exception of rated for 36W instead of 50W.

They don't the put the output devices because without it, it seems cheaper to build

[wink]

.... and they make good preamps........   

[jose]

11 hours ago, Pars said:

Thanks all, got a couple of BOMs now.

You guys don't like putting the output devices in your BOMs, eh?

 

EDIT: also, can MJF15030/31s be used in place of the MJE15030/31? Specs seem the same with the exception of rated for 36W instead of 50W.

Oh no Pars. Au contraire !!!

I forgot add this reference because I bought for my KSA5 too. I used MJF15030/15031 but now If I did the new thing, I would choose MJE because amplifier gives off a lot of heat.

 

[kevin gilmore]

the mjf are plastic parts, so you can use metal screws, you still need a thermal isolator

the mje parts are exposed tab, so you need to use insulators on the screws

[gepardcv]

I'm having trouble biasing a Dynahi. Board is ssdynahibalsmtm.zip from around a year ago, with the feedback resistors changed to 140k and the compensation caps to 3pF in an attempt to lower the gain. The highest bias I can manage is just around 0.42V. If I get it higher across one set of devices (the outer four on each side), it plummets to 0V on the other set (the inner eight). 0.42V is the best I managed so they're all the same (give or take a couple mV), with offset hovering at 30mV (I could probably get it lower, but it didn't seem worthwhile trying yet).

Haven't tried my other board yet, but I imagine it'll be the same story, since I used the same batch of hFE matched transistors.

I was thinking of replacing the 500ohm resistors next to the outer trimpots with 300ohm or 390ohm parts (I think they're R1 and R4 on the circuit diagram), since I have both values lying around, and try to get 0.75V bias. Is that a good idea?

[gepardcv]

One day last year, while planning my Carbon build, I thought it would be neat to have a dynamic headphone amplifier in a matching case. The Dynahi suggested itself, though I wanted something a little less overkill. Then it occured to me that I might be able to stuff a Dynahi and a Dynalo into one case, and so make an amplifier capable of driving any headphone. So I made this, which I called the DDCA (Dynalo-Dynahi Combination Amplifier): one set of inputs feeding two amps, with separate output protection circuits, separate volume pots, three GRLV power supplies, and one transformer. It was a 3D puzzle to fit everything in (if I were to make another one, I'd position the Dynahi boards right-side up, turned out there is enough clearance that way and it would have simplified assembly) and the wiring took forever (plus keeping it perfectly tidy is beyond my skill).

I lowered the gain on both the Dynalo and the Dynahi, since on most amps, I have the volume at 10 o'clock or less, and I wanted to use more of the volume control. Have to admit: the RK50 on the Carbon spoiled me, and the TKD 601 pots here don't live up to it. Not bad, though.

Like I posted a couple of days ago, I couldn't bias the Dynahi section as high as I wanted, but even at 0.40V across the 20ohm output device resistors it runs fairly warm, so I decided to leave it alone for now. I also need to get the top anodized black. Maybe add some speaker binding posts... wonder how much of a speaker load I can drive from the Dynahi?

Both sections sound great to me, though I don't have an HE-6 or K1000 to really put the Dynahi through its paces.

As usual — many thanks to @kevin gilmore and @spritzer for all the designs and boards. This wouldn't exist without you guys. Thanks to @vilts for making matching titanium knobs, thanks to @Zashoomin for getting extra sand with me so we could make matched sets, and thanks to everyone here for group buys and all the help and support.

[Pars]

Interesting build! Nice to see someone do one single chassis also. Appears to be some beefy input wiring!

[Skooby]

On 4/10/2017 at 11:28 AM, gepardcv said:

The highest bias I can manage is just around 0.42V. If I get it higher across one set of devices (the outer four on each side), it plummets to 0V on the other set (the inner eight). 0.42V is the best I managed so they're all the same (give or take a couple mV), with offset hovering at 30mV (I could probably get it lower, but it didn't seem worthwhile trying yet)

I was thinking of replacing the 500ohm resistors next to the outer trimpots with 300ohm or 390ohm parts (I think they're R1 and R4 on the circuit diagram), since I have both values lying around, and try to get 0.75V bias. Is that a good idea?

Super build!

I am having the same problem of biasing my Dynahi (ver1.5).  I can only manage to get one set (inboard or outboard) of transistors to 0.75V but zero for the other; around 0.402V for both sets.  Would like to know a solution to this problem too.

Also, on low impedance/high sensitive cans (D7K notably), I hear quite a bit of floor noise (running at 0.4V bias).  But no noise for high impedance cans, LCD3/HD800.  The way it is?  MiniDynalo were dead quiet though...   

[gepardcv]

Yes, the Dynahi has a higher noise floor than the Dynalo. Not an unusual price for much higher gain. Just don't use the Dynahi with sensitive headphones, it's not meant to drive them and carelessness with the volume knob might damage the drivers (not to mention your hearing). It has a lower noise floor than a few much-less-powerful commercial amps I heard, which hiss like a snake with the HD600 and HD800.

[jose]

29 minutes ago, gepardcv said:

Yes, the Dynahi has a higher noise floor than the Dynalo. Not an unusual price for much higher gain. Just don't use the Dynahi with sensitive headphones

 

On ‎12‎/‎04‎/‎2017 at 7:04 AM, Skooby said:

Also, on low impedance/high sensitive cans (D7K notably), I hear quite a bit of floor noise (running at 0.4V bias).  But no noise for high impedance cans, LCD3/HD800.  The way it is?  MiniDynalo were dead quiet though...   

I have the same problem on my Dynahi (1.4 ver). I changed both 340P14, changed the gain... but the noise floor is stell there.

I only use it with high impedance headphones

[Skooby]

Alright, thanks guys for clarifying the high noise floor.  That dual Dynalo/hi combo is a great solution for that problem, nicely thought out!  LCD-3 definitely sounds better on the Dynahi.

On 4/10/2017 at 11:28 AM, gepardcv said:

I was thinking of replacing the 500ohm resistors next to the outer trimpots with 300ohm or 390ohm parts (I think they're R1 and R4 on the circuit diagram), since I have both values lying around, and try to get 0.75V bias. Is that a good idea?

So took on this idea last night, I strapped 750R parallel to R1/500R (300R equiv.) and adjusted RV1, I was able to get the bias up to 0.75V.  The resultant resistance across R1 was about 250R.  Look back at the schematic, the 1K/R50 and R51 are there to keep the resistance from getting below 333R.  So, thinking of shorting R50/51 out;  just have to remember to set RV1/2 to max before power up.  No problem right?           

 

[gepardcv]

Are you trying to avoid the 750ohm and 500ohm parallel rig, or replacing the resistor with a single 300ohm part? I personally wouldn't short the R50/R51 because I'm not sure what might fry if the trimpot fails or if I slip while adjusting the bias and offset.

Or are you trying to bias higher than 0.75V? I wouldn't do that, either. How hot does the amp run with the 0.75V you have now? My 0.40V passes the "can I hold my hand on it for a while" test, but runs way hotter than my Carbon, and I don't want it to run hotter than that.

[Skooby]

Good point...didn't think about safety issue; being a cheapass, I was just thought to save some resistors for other projects

The 500R is nestled pretty tight in against the pot and 5K, so 1K is easier to replace/mod with.  Will work out what needs to get lower floor (~200R or so?).

I'll run it at 0.75V bias; the heat was ok, just warm, not hot.  Though, my heat sink is smaller than yours, so may be keep it low (will test it out and decide).  Yours is so beautiful, mine is so bland and ugly!   So jelly...

 

    

[Pars]

Having the transformer(s) that close to the audio boards/pot probably isn't helping your noise floor at all. AC wiring also appears to be right over the input wiring to the left side of the pic.