[Clutz]

I am contemplating building the gilmore dynalo headphone amplifier some time in the new year - particularly given all of the rave reviews that the Dynahi gets. I don't have the money to build a dynahi, so i figured a dynalo would be a good place to start.

But it also occurred to me, that perhaps I could modify the dynalo a bit. My idea is to use the dynalo as it stands for the left and right channel - but then seperately build an active ground channel (a la PPA). This would be sort of a compromise between the dynalo and the dynamid. Some people might ask why I don't just build a dynamid then.

Well, I don't really feel like spending twice as much money - esp. when I don't have any balanced sources or any balanced headphones. I also just thought this might be an interesting experiment.

Thoughts / comments?

Cheers,
Clutz

 

[Pars]

Well, since no one else has piped up, I'll take a stab at this. I'm not sure how much improvement, if any, an active ground would be on a Dynalo. But consider that you essentially would have to build up half of another Dynalo board in order to do this, unless you were going to do a different architecture for the active ground, such as buffered opamp or something. In addition you may either have to use another PSU or beef up the PSU to handle the additional channel. Just my 0.02.

 

[Garbz]

woopse missed this thread first time round otherwise would have answered

What you are proposing would work really well but I wonder what the actual point is. The Dynalo alone sounds fantastic. The only reason I scrapped it for a Dynamid is because I use fully balanced sources and my cdplayer is some 10m away from my amps.

The benefits would be there, you would keep the return currents off the circuit ground, however you really are not very far away form a Dynamid anyway.

You will have 3channels of amp, and a new powersupply already (Dynahi PSU can do it). Transistors are cheap the Jfets will also only set you back a few more dollars and all the PCB desgns here are stereo so that is taken care of. The onlything you really need to worry about is the 4ch volume control, and the extra effort of assembling and matching another channel.

 

[Clutz]

Quote:

Originally Posted by Pars Well, since no one else has piped up, I'll take a stab at this. I'm not sure how much improvement, if any, an active ground would be on a Dynalo. But consider that you essentially would have to build up half of another Dynalo board in order to do this, unless you were going to do a different architecture for the active ground, such as buffered opamp or something. In addition you may either have to use another PSU or beef up the PSU to handle the additional channel. Just my 0.02.

I'd use a buffered op-amp with a gain of 1 as an active ground channel - at least that was my first plan. Would it even be a advisable to use a half of another dynalo board for the ground channel? Isn't the gain on the dynalo board fixed by the choice of transistors?

Anyway, I think I may go ahead and do it. I'm not going to be building any more amps until the new year at this point - but I'm kind of interested in this idea.. The reason being is that one of the benefits PPA / M3 often advertised is the three channel topology.

 

[Clutz]

Quote:

Originally Posted by Garbz What you are proposing would work really well but I wonder what the actual point is.

Just doing it for fun really.

Quote:

The Dynalo alone sounds fantastic. The only reason I scrapped it for a Dynamid is because I use fully balanced sources and my cdplayer is some 10m away from my amps. The benefits would be there, you would keep the return currents off the circuit ground, however you really are not very far away form a Dynamid anyway. You will have 3channels of amp, and a new powersupply already (Dynahi PSU can do it). Transistors are cheap the Jfets will also only set you back a few more dollars and all the PCB desgns here are stereo so that is taken care of. The onlything you really need to worry about is the 4ch volume control, and the extra effort of assembling and matching another channel.

I could build a dynamid - but I only have single ended sources, and I don't really want to balance my headphones - so if I kept the output single ended then it would seem a bit unnecessary. Since I already have two boards - I already can pretty easily build a dynamid just by buying twice as many parts - I just thought this might be an interesting exercise.

With regards to the active ground channel - I was thinking of just using a buffered op-amp with the gain set to 1 - would that provide me with rather limited benefits?

 

[Pars]

Quote:

Originally Posted by Clutz I'd use a buffered op-amp with a gain of 1 as an active ground channel - at least that was my first plan. Would it even be a advisable to use a half of another dynalo board for the ground channel? Isn't the gain on the dynalo board fixed by the choice of transistors?

Ughhh, opamps

j/k

The gain on a Dynalo is set by the feedback resistors, typically a 10K and a 1K. I would think it would be unity gain stable, but never tried it.

As Garbz pointed out, the 4 channel pot would be a problem. But if the ground channel is at unitiy gain (buffer) I wouldn't think it would need to go thru the pot. That said, I've never built a balanced amp or contemplated one (yet) so I don't fully understand all of the ramifications.

 

[Clutz]

Quote:

Originally Posted by Pars Ughhh, opamps j/k The gain on a Dynalo is set by the feedback resistors, typically a 10K and a 1K. I would think it would be unity gain stable, but never tried it. As Garbz pointed out, the 4 channel pot would be a problem. But if the ground channel is at unitiy gain (buffer) I wouldn't think it would need to go thru the pot. That said, I've never built a balanced amp or contemplated one (yet) so I don't fully understand all of the ramifications.

Yes. At times like this I am reminded of the things I would say to my undergraduate students if they made a lame assed comment like "cannot set the gain of the dynalo" - esp. when it would've been oh so easy for me to find out for myself...

Well - maybe what I'll do is this - build the dynalo - hear what it sounds like. Listen to it, smile and enjoy it. Then try building a buffered op-amp as a ground channel. Ideally, connect it up so that I can easily switch between the active and passive ground circuits using a DPDT switch. Listen to it, smile and enjoy it. One of the benefits of this is that it would easily fit into a Hammond 1455N160x case.

Then, build a ground channel using one of the channels on my other board - setting the gain to one... and giving it a try.. sit back, listen, and enjoy.

and then finally.. build the dynamid.

 

[Teerawit]

Quote:

Originally Posted by Clutz I could build a dynamid - but I only have single ended sources, and I don't really want to balance my headphones - so if I kept the output single ended then it would seem a bit unnecessary. Since I already have two boards - I already can pretty easily build a dynamid just by buying twice as many parts - I just thought this might be an interesting exercise.

Can't you just build it in dual-mono then?

 

[Clutz]

Quote:

Originally Posted by Teerawit Can't you just build it in dual-mono then?

I suppose so- but if I did that, I'd end up having to convert my headphones to dual mono.. And anyway, I guess what I'm interested in is whether or not adding an active ground circuit would improve the dynalo much. The answer seems to be, yes - just depends on whether it's worth it. I'm going to build a dynalo anyway- so I might as well try building it, and then try adding the active ground channel too...

Anyway, I won't have the time to do this until at least Feb. I have another project to do in January and I'm going to be really busy in December with work and the Holidays.

Thanks all!

 

[Garbz]

Bing. This board really needs a little lightbulb smiley. Oh wait found it.

Anyway consider building a balanced bridged dynalo but not a dynamid which is what I was talking about and you were too. Dynamid takes the balanced input and amplifies it through 4 discrete channels. But the original design however showed another way to get this. Basically you get all the benefits of balanced config, but with a single ended source. Best of all it's cheaper then building a 3 channel dynalo like you sugested.

The idea is simple. The original shematic showed the points to tap it at:

The schematic is easy the points beside the jfet go to the output transistor cascode. However the points A* and B* will be inverted from the primary output. So... add another cascode stage in there to the inverted side of the jfets:

And there it is an extra 10 transistors and 13 resisotrs. no need for extra jfets or any of the other nonsense required for a balanced dynamid. It takes a single ended input and runs a fully balanced output. The ground returns are not only isolated from the headphones but are amped as well to provide the usual balanced shinannigans 2x voltage 2x current 4x power, this may even run on a dynalo PSU given proper heatsinking.

Consider that. Very easy to implement and very cheap.

 

[Clutz]

Quote:

Originally Posted by Garbz And there it is an extra 10 transistors and 13 resisotrs. no need for extra jfets or any of the other nonsense required for a balanced dynamid. It takes a single ended input and runs a fully balanced output. The ground returns are not only isolated from the headphones but are amped as well to provide the usual balanced shinannigans 2x voltage 2x current 4x power, this may even run on a dynalo PSU given proper heatsinking. Consider that. Very easy to implement and very cheap.

I think I now know what my next project is. I hope I can power it with a dynalo PSU with proper heatsinking - cause I'm not sure I want to go and build a dyanhi PSU. I've got to run oiff to work - but I'm going to take a closer look at this later today (during those hours in the day where I'm in my office, getting paid and yet somehow procrastinating).

Thanks Garbz!

 

[Clutz]

Edit.

 

[Garbz]

double post 11 1/2 hours appart? Someone leaves there computer on while at work

 

[Clutz]

Quote:

Originally Posted by Garbz double post 11 1/2 hours appart? Someone leaves there computer on while at work

Computers... can.. turn off? I always new I could restart them.. but shutting them off just seems mean.

Oops.