Would it be possible to omit the OPA and connect the power amp directly to the output of the BA-3 line stage?
I connected my HD-800 to the BA-3 line stage.
I know this isn't ideal, but the HD-800's 300 ohm impedance will do.
And the sound is absolutely phenomenal.
I've heard the HD-800 for a long time and on many amps, but this beats them all by far.
So I could imagine that BA-3 and Whammy power amp might be the final solution for me...
(BA-3 Gain 8x 100mA idle current)
I connected my HD-800 to the BA-3 line stage.
I know this isn't ideal, but the HD-800's 300 ohm impedance will do.
And the sound is absolutely phenomenal.
I've heard the HD-800 for a long time and on many amps, but this beats them all by far.
So I could imagine that BA-3 and Whammy power amp might be the final solution for me...
(BA-3 Gain 8x 100mA idle current)
Attachments
All you need is to boost the current... the gain can even be 0.95...even 0.9 is perfectly fine, which is coincidently the gain of many such buffers/followers...
And YES, I get simply amazing results and absolute bliss when I connect my HD650s straight to my May DAC XLR outs... I am now looking to upgrade the headphones... the HD800s are on special at the moment 🙂
And to answer your question.... I can not see any reason why you shouldn't try it.... the Vcc/Vee can even be much higher because there's no more OPAmp limitation... which is usually + and - 15V with good OPAmps. With regard to this, I have already suggested in this thread that the WHAMMY supply should be separated into one for output MOSFETs that can easily be + and - 24V, and LT3045/3094 for OPAmps (up to + and - 15V), giving each section what it truly deserves...
And YES, I get simply amazing results and absolute bliss when I connect my HD650s straight to my May DAC XLR outs... I am now looking to upgrade the headphones... the HD800s are on special at the moment 🙂
And to answer your question.... I can not see any reason why you shouldn't try it.... the Vcc/Vee can even be much higher because there's no more OPAmp limitation... which is usually + and - 15V with good OPAmps. With regard to this, I have already suggested in this thread that the WHAMMY supply should be separated into one for output MOSFETs that can easily be + and - 24V, and LT3045/3094 for OPAmps (up to + and - 15V), giving each section what it truly deserves...
Last edited:
So in principle I could connect it like this?
Why is the BA-3's 8x gain too much?
The last amplifier stage of the WHAMMY doesn't add any additional gain, does it?
Sorry for the questions, unfortunately I don't have enough knowledge... 😡
Why is the BA-3's 8x gain too much?
The last amplifier stage of the WHAMMY doesn't add any additional gain, does it?
Sorry for the questions, unfortunately I don't have enough knowledge... 😡
yes, you can connect it like that, in principle, but there is one thing critical and missing
original schematic is having OP not just for gain duty but also for servo action, keeping output at 0 DC Offset
without OP, you need to introduce at least passive solution, in form of resistor-trimpot-resistor, to adjust OS input node voltage level
now, it'll never be so stable as original
one of the ways to solve that problem is to introduce output capacitor
so, when you change something, expecting sudden rain of Mana from Heaven, there is a whole new bunch of compromises you need to make ......... and then you realize that there are some schematics where constructor guy really did his job, taking care of everything - project goal in specific context
now - gain - original WHAMMY is having gain in range aroundish 5*
and even that is too much, when one is having source with today's proper voltage level, resulting in volume pot at max 9Hrs, while eyes popped out of sockets
OK, back to start - what are your reasons to combine BA3FE with WHAMMY OS?
in other words - why intention to ditch out original WHAMMY FE?
original schematic is having OP not just for gain duty but also for servo action, keeping output at 0 DC Offset
without OP, you need to introduce at least passive solution, in form of resistor-trimpot-resistor, to adjust OS input node voltage level
now, it'll never be so stable as original
one of the ways to solve that problem is to introduce output capacitor
so, when you change something, expecting sudden rain of Mana from Heaven, there is a whole new bunch of compromises you need to make ......... and then you realize that there are some schematics where constructor guy really did his job, taking care of everything - project goal in specific context
now - gain - original WHAMMY is having gain in range aroundish 5*
and even that is too much, when one is having source with today's proper voltage level, resulting in volume pot at max 9Hrs, while eyes popped out of sockets
OK, back to start - what are your reasons to combine BA3FE with WHAMMY OS?
in other words - why intention to ditch out original WHAMMY FE?
Zen Mod, thank you so much for the explanation.
It was just an idea since BA-3 + HD-800 sounds so great.
But it would need a Buffer to work with lower impedance Headphones ...
And so I thougt 'why not use WHAMMY' as a buffer.
Discretely built, without OPA and without feedback.
But it might be a dumb idea.
And yes, I was afraid that further adjustments would be necessary ...
It was just an idea since BA-3 + HD-800 sounds so great.
But it would need a Buffer to work with lower impedance Headphones ...
And so I thougt 'why not use WHAMMY' as a buffer.
Discretely built, without OPA and without feedback.
But it might be a dumb idea.
And yes, I was afraid that further adjustments would be necessary ...
Hmmm... why don't you try to lift the two resistors at the point "x" off the PCB, and see what happens with the DC offset before you give up on your idea? If it doesn't drift too much, you could always go ahead... just place a 1000uF cap in series with your headphones and put a bleeder resistor in parallel with the output.
Yes I will try that once the order has arrived and the circuit is set up.
I'm sorry I didn't mean to contaminate the thread...
I might start a new one "Discrete headphone buffer for BA-3 without OP, no feedback" 🙂
I'm sorry I didn't mean to contaminate the thread...
I might start a new one "Discrete headphone buffer for BA-3 without OP, no feedback" 🙂
@SomekPoland I got the Sparkos pro opamps. They fit fine within the enclosure. You need two risers from them. You need these two risers even if you are using the SS3602 to clear the neighboring caps. I know of Sparkos opamps. I saw a YouTube video where a guy said he put these in his DAC and made it better. It turned out my DAC had sockets and the SS3602 transformed the performance of my DAC. So, when I was going to do this Whammy, I knew I wanted to use their opamps. I actually had the SS3602 first, but it had high DC offset of 140mV on one channel and 90mV on the other. Right now, I am getting 0.5mV on one and 0.8mV on the other. I understood the SS2590 required more current so I was unsure if the power supply could do it. Thank you for helping me with that.
Attachments
@Chris Jones KS As I mentioned in the post above, my experience with the Sparkos was in a DAC. I just assumed they will perform better. I got these to deal with the high DC offset I could do nothing about in the SS3602. I can't say if these are better than the 3602 in this amp.
This is the big problem, the offset.
Looking at the circuit and it should be possible to easily configure the circuit to do what you want while retaining the opamp as a servo (you must use a suitable opamp though which is invariably a FET device). An OPA or TL0 type would be perfect.
The input impedance is low hence the need for a larger input coupling cap. The 0.22uF can be added either on top or below the chip.
DC offset is essentially zero using the opamp like this.
The gain is less than '1' due to the attenuation of the input signal by the all the resistive bias networks.
That looks great, thanks a lot !
Will try this. Would use my stock FQP3P20/FQP3N30 (unmatched).
I have OPA 2132 and 2137. Which would be better?
BA3 Frontend already has 10uF output cap.
So I could omit C1 ?
If you want, the posts can be moved to a new thread.
"Discrete Class-A headphone buffer for BA3-FE, no feedback"
or similar ...
Will try this. Would use my stock FQP3P20/FQP3N30 (unmatched).
I have OPA 2132 and 2137. Which would be better?
BA3 Frontend already has 10uF output cap.
So I could omit C1 ?
If you want, the posts can be moved to a new thread.
"Discrete Class-A headphone buffer for BA3-FE, no feedback"
or similar ...
Last edited:
Only omit C1 if there is no DC present on the DAC output. The servo would cope with around -/+40 mv DAC offset and still give 0.00 volts at the Whammy output.
I would think either opamp would be fine. The OPA2137 mentions 'precision integrator' as a usage and in a way that is what a servo is.
I would think either opamp would be fine. The OPA2137 mentions 'precision integrator' as a usage and in a way that is what a servo is.
🙂 Sounds like a plan.
Here's the sim file for those wanting to play. All default models, click and run.
Here's the sim file for those wanting to play. All default models, click and run.
This thread has been split from here - https://www.diyaudio.com/community/threads/whammy-pass-diy-headphone-amp-guide.317803/
I agree with your observation so I installed 7815,7915 which are connected to the opamp circuit. I drilled two holes in the non-functional fins of the torroid to hold the little PCB with regulators. The wiring is not pretty but it works with 22v supplying the MOSFETs. The photo shows the diodes missing, I have recently re-installed them for 17v operation. Given the power supply design, with +-26vdc on the regulator inputs, I think there should be 2 diodes in series for 19vdc to the MOSFETs. The main reason for that is to reduce the regulator stress. They get very hot with the cover closed.
The Whammy is a fine kit but the hidden ground layer in the PCB makes component change difficult. It needs professional extraction tools and I only have handheld tools which are usually adequate.
