[Brinkman]

I was looking at building a simple CMoy headphone amp (for op amp rolling and headphone use), but would really like a balanced circuit to drive my XLRed AKG K701s. Since I almost exclusively listen to vinyl, I realized phono equalization was worth inclusion.

So I ditched the simple CMoy concept and would now like to build a balanced Bugle. Using a pair of the balanced input op amp circuits here (3/4 of the way down the page, with the green smiley next to it), could I then use two parallel Bugle boards each handling LEFT POS & NEG and RIGHT POS & NEG respectively?

Also, could each POS and NEG per channel share the same EQ network (R13 & C2 and R14 & C3) rather than doubling the EQ components and referencing them to ground?

Thanks for any help.

Ben

[poty]

At the beginning I would say that I'm not in favor of balanced amplifying, but I know you do not agree with my point , so I'll continue...

So I ditched the simple CMoy concept and would now like to build a balanced Bugle. Using a pair of the balanced input op amp circuits here (3/4 of the way down the page, with the green smiley next to it), could I then use two parallel Bugle boards each handling LEFT POS & NEG and RIGHT POS & NEG respectively?

Yes, you can. I haven't actualy understood you reference to tubecad.com article, but if you got only the idea from the article - it's OK. The pair of Bugles can be used without any added circuits in "balanced" design only with minor modifications (around EQ circuits).

Also, could each POS and NEG per channel share the same EQ network (R13 & C2 and R14 & C3) rather than doubling the EQ components and referencing them to ground?

Yes, the idea is in the reffered article too. You can connect the EQ circuits between right sides of R10 and R8 of corresponding POS and NEG channels.

[Brinkman]

Thanks Poty.

The combined EQ concept came from a study of the Trumpet schematic.

I'm going to attempt an upload of the schematic I came up with. Power supply will be +/-15V.

In the meanwhile, I'll have to wait and see what Jim thinks about using his circuit this way.

It's amazing how much a simple idea can get away from you.

[Brinkman]

Okay,

Turns out the extra front end will not be needed. I sent my schematic to Jim and he made some suggestions. I will post a schematic when all looks good.

Waiting to see what power supply I will have to use...

[poty]

Turns out the extra front end will not be needed.

Exactly what I told you in my post.

Waiting to see what power supply I will have to use...

What is the problem? The power supply for the Bugle is available from te hagtech.

[jcg0322]

I have three Bugle power supply kits left.

Just in case you are interseted,

Thanks,
Jack

[Brinkman]

Exactly what I told you in my post.What is the problem? The power supply for the Bugle is available from te hagtech.

Poty,

Do you have any access to schematic design programs? I'm using Designlite for Macs and my 30-day trial ended after one day.

I'm not sure the Bugle power supply can power a balanced Bugle @ +/-15V. Then again, I'm not sure it can't until I know the current requirements for such a design.

I'm also not too sure how close the design is to completion until I've heard a working "prototype."

I've been too busy to contact Jim in the last few days. Perhaps after some more homework I can forward a better design to Jim and learn from his feedback some more.

[poty]

Do you have any access to schematic design programs? I'm using Designlite for Macs and my 30-day trial ended after one day.

Sorry, I'm using a PC. And I think not very known program RusPlan.

I'm not sure the Bugle power supply can power a balanced Bugle @ +/-15V. Then again, I'm not sure it can't until I know the current requirements for such a design.

The Bugle Kit Manual helps us again. On the page 9 we can find: "typical battery draw is roughly equal to the idle current" and "total of 30mA idle current" with "The stock opamps". The Bugle power supply could bear 100 mA, so it can easily power two Bugles with "the stock opamps".

I'm also not too sure how close the design is to completion until I've heard a working "prototype."

You can easily do it even with only one Bugle. Just use the two channels as a mono channel in balanced design and - voila - you have a prototype!
I think the following picture have enough changes to get from one bugle a mono channel with balanced design:

[Brinkman]

The Bugle Kit Manual helps us again. On the page 9 we can find: "typical battery draw is roughly equal to the idle current" and "total of 30mA idle current" with "The stock opamps". The Bugle power supply could bear 100 mA, so it can easily power two Bugles with "the stock opamps".

The sound just heard was me slapping my forehead!
What a relief. I already have one assembled...

In regard to the schematic, here are excerpts from Jim's email exchanges to me, cobbled together:

"The 392R (R3 & R5) do not need to go to ground. You have to be very careful with input wires. They need to be well shielded.

You can lower the input series R.  If MC, make them very low.  They add noise.

The EQ has to change.  You really have two EQ RCs in series per side.  That is, the 220n and 1.43k is ok to ground, but not to the other side.  Impedance is off by 2.  You either change the values (2 pair down to 1 pair) or change the driving impedance.  Two 220n in series in 110nF, and the resistance is 2.86.

Same for the 10nF EQ.

I think you could also change the 13ks [in series with the RC EQ] to 6.49ks. Or change the RC in the RC.  One or the other.  Not both.

Note also the 8.45k output resistors are only correct if the load on the other side of coupling caps is 47k."

To complicate matters, I was hoping to add a pair of crossfeed networks between the 220nF series capacitors and U3 (crossfeeding L+ & R+ and L- & R-, respectively). The schematic for the kits I purchased is here. This is a passive network, so I need to make sure the gain is enough to compensate for power losses through the network.

I also came across some low-capacitance (13pF/ft) XLR cables from Blue Jeans Cable. I plan on wiring them to my turntable to replace the low-capacitance RCAs I installed a while back. You can see them here.

Thanks for taking the time to draw up a schematic. If I can revise mine I will figure out how to post it here.

Regards,
Ben

[poty]

"The 392R (R3 & R5) do not need to go to ground. You have to be very careful with input wires. They need to be well shielded."

Yes, agree. You can easily connect ONE resistor to another "halves" (for example, the grounded side of R5 - to "signal" part of R23).
The same thing could be done with R17 and R35.

"You can lower the input series R.  If MC, make them very low.  They add noise."

I think I would delete them instead of lovering.

The EQ has to change.  You really have two EQ RCs in series per side.  That is, the 220n and 1.43k is ok to ground, but not to the other side.  Impedance is off by 2.  You either change the values (2 pair down to 1 pair) or change the driving impedance.  Two 220n in series in 110nF, and the resistance is 2.86.
Same for the 10nF EQ."

The sentence is rather puzzled me. Maybe it is my bad English again. I think, the changes I draw in my scheme is correct. But this definitely needs another opinion.

I think you could also change the 13ks [in series with the RC EQ] to 6.49ks. Or change the RC in the RC.  One or the other.  Not both. "

Agree. It was my mistake not to include this in the changes.

Note also the 8.45k output resistors are only correct if the load on the other side of coupling caps is 47k.[/i]"

As soon as we change the values of R15 and R33 - it's OK.

To complicate matters, I was hoping to add a pair of crossfeed networks ...

Sorry, this part needs some time to understand. I hope I'll be able to do that.

I also came across some low-capacitance (13pF/ft) XLR cables from Blue Jeans Cable. I plan on wiring them to my turntable to replace the low-capacitance RCAs I installed a while back. You can see them here.

It's a great thing! It should be worth trying then the balanced configuration.

[Brinkman]

Poty,

Why did you add the 47K resistor in parallel to the first EQ network (R14 & C2)? Isn't this affecting the EQ?

[poty]

Poty,

Why did you add the 47K resistor in parallel to the first EQ network (R14 & C2)? Isn't this affecting the EQ?

No, it is not affect the EQ, but it affects gain. It was my wrong assumptions. I changed the scheme in accordanve with our last conversation.

[Brinkman]

So Jim had a little more input in regards to a headphone-specific Balanced Bugle application:

After first stage, should be 6k49 ohms.  After second stage, 4.24k.  These need to be cut in half too.  The following 47k needs to be split into two 24k to ground (this is the dc reference for output stage).

100 ohms on input ok.  They're only for spark protection and to cut a little RF from demodulating on parasitics.

I think this is good to go.  Make sure you have access to parts to change later in case we missed something.

You can change last stage to another opamp.  Output ohms should drop to 10 or less.  Then you need probably 220uF (not nF) in series.  Assume a 32 ohm load here, caps need to be big.  They will also impact sound.  I usually just go without, hoping the dc offset is acceptably small.

jh

I have a .pdf of the schematic, which looks very similar to the schematic Poty drew up above except it's geared toward headphone use. PM me if anyone's interested in a copy of it and I'll email it to you, as I cannot upload the .pdf to my gallery for some reason.