Ciao Marcello,
welcome back on this topic after a couple of years.
I hoped to use this way only to thanks you, but I had trouble on your project....
http://www.diyaudio.com/forums/showthread.php?s=&threadid=104585
BTW: I have one amp working marvellous!
Ottimo lavoro e Grazie 1000!
Peppe
welcome back on this topic after a couple of years.
I hoped to use this way only to thanks you, but I had trouble on your project....
http://www.diyaudio.com/forums/showthread.php?s=&threadid=104585
BTW: I have one amp working marvellous!
Ottimo lavoro e Grazie 1000!
Peppe
Any boards available??
Just wondering if anyone knows of a source for the PCBs for Marcellos amp? I'm thinking of building this amp and would prefer not to roll my own boards (it's been 30 years!) but from what I can see the group buys finished ages ago.
Oh well, if I roll my own I'll incorporate a simple relay so I can bypass the headphone amp input onto my speaker poweramp. Thinking of rigging it so that if the headphone amp is powered up, the bypass is disconnected, but when power is off, the normally closed contacts will feed to a pair of RCAs on the back and through to my power amp.........anything wrong with this approach?
Just wondering if anyone knows of a source for the PCBs for Marcellos amp? I'm thinking of building this amp and would prefer not to roll my own boards (it's been 30 years!) but from what I can see the group buys finished ages ago.
Oh well, if I roll my own I'll incorporate a simple relay so I can bypass the headphone amp input onto my speaker poweramp. Thinking of rigging it so that if the headphone amp is powered up, the bypass is disconnected, but when power is off, the normally closed contacts will feed to a pair of RCAs on the back and through to my power amp.........anything wrong with this approach?
thanks for the prompt reply James.
I'll think about it for a while first. The modified board by Digi01 appeals to me more as it includes PS and mods on a single board......but I may go with a pair of the original boards offered by A'af as it looks like at least they are available. I'd be interested in knowing how your project shapes up.
I'll think about it for a while first. The modified board by Digi01 appeals to me more as it includes PS and mods on a single board......but I may go with a pair of the original boards offered by A'af as it looks like at least they are available. I'd be interested in knowing how your project shapes up.
Zen H.ph amp
Been looking at doing one of these for awhile now and have "run-in" my AKG 701s , got some of those Lovaltec power Jfets and chasing parts for a variation of the "Lightspeed" passive vol control to add to the kits when it eventually arrives - I expect REALLY good results from this one, even without the balanced version or "negative" GND wire buffer stage idea (Meitner).
One thing I've found is that typical standard power supplies for h.ph amps "really suck", so nowadays I generally start with a decent "starter" system (EI transformer, C-R-C, then seperate CMx, onto fairly simple series/shunt regs with decent caps) and then get a bit fancy "tuning things up a bit" later on (snubbers, filters, bipassing caps,wiring, etc.)
Lot of fun, these things - very direct - extremely useful too.
A first class head amp and phones is mandatory for building dacs & variations, preamps, etc - saves an enormous amount of time/effort.
Been looking at doing one of these for awhile now and have "run-in" my AKG 701s , got some of those Lovaltec power Jfets and chasing parts for a variation of the "Lightspeed" passive vol control to add to the kits when it eventually arrives - I expect REALLY good results from this one, even without the balanced version or "negative" GND wire buffer stage idea (Meitner).
One thing I've found is that typical standard power supplies for h.ph amps "really suck", so nowadays I generally start with a decent "starter" system (EI transformer, C-R-C, then seperate CMx, onto fairly simple series/shunt regs with decent caps) and then get a bit fancy "tuning things up a bit" later on (snubbers, filters, bipassing caps,wiring, etc.)
Lot of fun, these things - very direct - extremely useful too.
A first class head amp and phones is mandatory for building dacs & variations, preamps, etc - saves an enormous amount of time/effort.
Thanks for the tip, Marcello - much appreciated - and a timely reminder to include the turn-on delay cct that was forgotten!
I've been optomising my F3 amp and can't see any reason not to adapt it to your Cct - be interesting to see if the F3 current gain characteristics work as well with the lower Idss for the h.phones.
I've been optomising my F3 amp and can't see any reason not to adapt it to your Cct - be interesting to see if the F3 current gain characteristics work as well with the lower Idss for the h.phones.
I am working on optimizing my Zen headphone amp, for my headphones (Beyerdynamics DT880) The first thing I think of trying is adjusting the resistors R7, R9, R14 to basically half their original value. Wouldn't that bump the -3dB point from 20kHz to about 40-50kHz? Just to make sure, I got it right
Steen
Ps. The next thing could be adding a Jfet input, along the lines of this one in the link here:
http://www.passlabs.com/np/FORBLUES.pdf
Steen
Ps. The next thing could be adding a Jfet input, along the lines of this one in the link here:
http://www.passlabs.com/np/FORBLUES.pdf
Sorry Kimbo - last reply cut short - mucking about with the vB Code thingie and ......
Not too keen on your combination signal/ouput relay idea - might work tho.
I can help out with some matched fets, although no shortage o/seas and most of the other parts are okay - not sure what your phones are.
I would encourage you to beef up the power supply a lot, so seperate supply pcb is a BONUS, in my opinion - just more work!
The standard Marchello cct (and Terry Aben's) is a pretty good unit and it upgrades rather well, I understand. Using the Jfet and Cascode adaptation of the F3 should give a dramatic result but no doubt, a few problems.
Steen,
Good to hear from you - did you continue with that Jfet cascode headphone Cct that Grey started awhile ago?
Incidently, something for you to try out - add an 0.1R resistor in series to one of those 100nF blue styrene caps and add them as a bipass across a 220uF electro such as the C2 in the F3 - sound a bit silly, but works a treat.
Regards ... James
Not too keen on your combination signal/ouput relay idea - might work tho.
I can help out with some matched fets, although no shortage o/seas and most of the other parts are okay - not sure what your phones are.
I would encourage you to beef up the power supply a lot, so seperate supply pcb is a BONUS, in my opinion - just more work!
The standard Marchello cct (and Terry Aben's) is a pretty good unit and it upgrades rather well, I understand. Using the Jfet and Cascode adaptation of the F3 should give a dramatic result but no doubt, a few problems.
Steen,
Good to hear from you - did you continue with that Jfet cascode headphone Cct that Grey started awhile ago?
Incidently, something for you to try out - add an 0.1R resistor in series to one of those 100nF blue styrene caps and add them as a bipass across a 220uF electro such as the C2 in the F3 - sound a bit silly, but works a treat.
Regards ... James
jameshillj said:
James,
I did build the DAO JFET Follower. (EUVL's work) I am rebuilding it now with another supply that might be good (better?) (CLC without active regulation) I thought it would be a good idea to try and make the Zen produce a little more topend details, if possible.
The amp deserves another shot , since it is an overall very charming and musical amp. The C2 is not in the standard Zen circuit, but i will try to use your bypass technique on the output caps which are 2*1000uF Panasonic FC. Those blue styrene's might work wonders, thanks a lot
Do you recommend the 0,1r at the output caps also?For reference, i attached the DAO JFET Follower.
Attachments
Ah! that's what it's called. DOA Jfet follower - is there a thread on this one? Must look it up.
Yes about adding a 0R1 in series with 100nF (0.1 ohm + 0.1uF!)across the 2 x 1000uF o/p panasonics - can try 0.15 or other tosee differences. The bipass thread mentions moving cap resonance freqs up (with just bipass cap) but producing excessive ringing that the small R seems to dampen - non tecnical I'm afraid.
Incidently, perhaps a cap across that 8volt zener may take down the noise floor a bit.
Can you use the 610 fets or are these ones better?
For now
Yes about adding a 0R1 in series with 100nF (0.1 ohm + 0.1uF!)across the 2 x 1000uF o/p panasonics - can try 0.15 or other tosee differences. The bipass thread mentions moving cap resonance freqs up (with just bipass cap) but producing excessive ringing that the small R seems to dampen - non tecnical I'm afraid.
Incidently, perhaps a cap across that 8volt zener may take down the noise floor a bit.
Can you use the 610 fets or are these ones better?
For now
James, thanks for your kind offer...I may take you up on it once my boards arrive. I agree that the separate boards are better, as I will beef up the supply as you suggest and this gives me a bit more flexibility.
This is my first attempt at a headphone amp so I have nothing to compare it with except the phones output from my NAD701......so I am really hoping to be pleasantly surprised. I don’t have much of a budget …..the missus has only just returned from the states ….so I got her to bring back a pair of Audio Technica ATH-AD700 to replace the 30y+ old crap pair of Sennheisers I had lying around. Pretty cheap but apparently good value for the price. I understand they have about the same impedance as the Grados that Marcello was using so I don’t expect any probs driving them.
As I need to use the same CD player and turntable for the family stereo, I'll persevere with adding a second relay to bypass the amp when the power is off......it will be a drag to fool around with leads all the time. If it gives me grief I'll disconnect it.
My circuit theory is pretty rudimentary (but I used to put together the odd kit amp wayback in the 70’s, yes I’m that old) so I’m just really sitting on the sidelines here and learning heaps with each new post from you guys…..keep up the great work.
I’ll probably be keeping the circuit pretty standard for the moment as most reports are pretty glowing with the amp as it stands, but I’m very interested in what you finally put together……but sometime later…who knows…I may follow your lead.
This is my first attempt at a headphone amp so I have nothing to compare it with except the phones output from my NAD701......so I am really hoping to be pleasantly surprised. I don’t have much of a budget …..the missus has only just returned from the states ….so I got her to bring back a pair of Audio Technica ATH-AD700 to replace the 30y+ old crap pair of Sennheisers I had lying around. Pretty cheap but apparently good value for the price. I understand they have about the same impedance as the Grados that Marcello was using so I don’t expect any probs driving them.
As I need to use the same CD player and turntable for the family stereo, I'll persevere with adding a second relay to bypass the amp when the power is off......it will be a drag to fool around with leads all the time. If it gives me grief I'll disconnect it.
My circuit theory is pretty rudimentary (but I used to put together the odd kit amp wayback in the 70’s, yes I’m that old) so I’m just really sitting on the sidelines here and learning heaps with each new post from you guys…..keep up the great work.
I’ll probably be keeping the circuit pretty standard for the moment as most reports are pretty glowing with the amp as it stands, but I’m very interested in what you finally put together……but sometime later…who knows…I may follow your lead.
Hi James,
Great to contact you the other day...thought I'd post my question here so you can share with other members.....hope that's OK
I've been pouring through the 'lightspeed' thread and certainly in its standard form, feeding anything below 50kohm input imp is not recommended. As this headphone amp has a 35k imp for -3db at 20khz, it doesn't look easy to bump up the imp without reducing the bandwidth. How are you going to approach the problem.......modding the amp or modding the lightspeed?
The more I read about the lightspeed the more appealing it is so I'm keen to hear your ideas.
Cheers,
Kim
Great to contact you the other day...thought I'd post my question here so you can share with other members.....hope that's OK
I've been pouring through the 'lightspeed' thread and certainly in its standard form, feeding anything below 50kohm input imp is not recommended. As this headphone amp has a 35k imp for -3db at 20khz, it doesn't look easy to bump up the imp without reducing the bandwidth. How are you going to approach the problem.......modding the amp or modding the lightspeed?
The more I read about the lightspeed the more appealing it is so I'm keen to hear your ideas.
Cheers,
Kim
Ah Kimbo,
The Lightspeed impedance load problem once again!
Headamp input gets a buffer - at present, thinking of similar to the Jfet D1 o/p buffer (Spencer's), or possibly a variation of the surprising NS10 preamp.
NP mentioned publishing an article soon that may also include a finished design, if we get lucky
The F3 amp has lower Zin (>10kR) and needs a buffer to be usable with the Lightspeed, so there will be some different circuits appearing soon.
Steen,
Sorry for not getting back to you.
- the Styrene caps as a bipass are a "treble enhancement device", even with 0.1 ohm!
I use bipass caps all over the place, but especially across electrolytics, even B.Gates, and that little series resistor is a "fussy bugger" but when it's right, it's quite obvious. Same thing for ICs, particularly the better ones.
The cap are a general purpose 0.1 - 0.15uF/50v Wima mkp-10 propylene caps with a series 0.1ohm - 0.25ohm (0R1 - 0R25) SMD as bipass to the 220 - 470uF electros that I tend to use often.
I tried some Eromaks (0.12uF/400v FKF) and got a different sound so ........
Also tried the [0.1 + 0R1] bipass across the 15,000uf power caps but found a better result when added a 220uF (Rubicon KZ) in parallel to the big cap and then added the little wonder [0.1uF + 0R1] unit across both electros - something about raising cap resonances - the improvement in clarity is startling and surprise, surprise - in the base too!
I'm thinking about a "mini F3" but with a fixed (not active) current source for a headamp running about 250 mA/ channel off maybe +30 volt rail - perhaps somebody with access to a sim program could run it up to give some approx starting values, yes?
The 216 styrene caps are rather bright as bipass caps - haven't found the correct series R, I think - perhaps you may try it and see, yes?
System? I use a (1541A) multibit Dac in non-O/S, but the obtained similar results with the Denon 2900 o/s, up sampling dac feeding the F3 amp via the Pass X2.5 preamp.
Hope this is of use.
The Lightspeed impedance load problem once again!
Headamp input gets a buffer - at present, thinking of similar to the Jfet D1 o/p buffer (Spencer's), or possibly a variation of the surprising NS10 preamp.
NP mentioned publishing an article soon that may also include a finished design, if we get lucky
The F3 amp has lower Zin (>10kR) and needs a buffer to be usable with the Lightspeed, so there will be some different circuits appearing soon.
Steen,
Sorry for not getting back to you.
- the Styrene caps as a bipass are a "treble enhancement device", even with 0.1 ohm!
I use bipass caps all over the place, but especially across electrolytics, even B.Gates, and that little series resistor is a "fussy bugger" but when it's right, it's quite obvious. Same thing for ICs, particularly the better ones.
The cap are a general purpose 0.1 - 0.15uF/50v Wima mkp-10 propylene caps with a series 0.1ohm - 0.25ohm (0R1 - 0R25) SMD as bipass to the 220 - 470uF electros that I tend to use often.
I tried some Eromaks (0.12uF/400v FKF) and got a different sound so ........
Also tried the [0.1 + 0R1] bipass across the 15,000uf power caps but found a better result when added a 220uF (Rubicon KZ) in parallel to the big cap and then added the little wonder [0.1uF + 0R1] unit across both electros - something about raising cap resonances - the improvement in clarity is startling and surprise, surprise - in the base too!
I'm thinking about a "mini F3" but with a fixed (not active) current source for a headamp running about 250 mA/ channel off maybe +30 volt rail - perhaps somebody with access to a sim program could run it up to give some approx starting values, yes?
The 216 styrene caps are rather bright as bipass caps - haven't found the correct series R, I think - perhaps you may try it and see, yes?
System? I use a (1541A) multibit Dac in non-O/S, but the obtained similar results with the Denon 2900 o/s, up sampling dac feeding the F3 amp via the Pass X2.5 preamp.
Hope this is of use.
I've been playing around trying to give a slight boost to the treble to offset the drop caused by increasing the input impedance. (I need min 50k if I was to use a lightspeed instead of a regular pot)
My circuit theory is pretty rudimentary, so my idea below is probably full of holes......but hey.....I'm here to learn...so don't be shy with your criticism.
So, what is wrong with inserting a cap in the middle of the feedback to remove some hf.....I figure that bumping up R7 (51K), R9 (360K) & R14 (270K), splitting R9 and adding a 0.08nF to earth results in a -3db at 17.5KHz and -4db at 20kHz. Which is not that far off the original spec of ~ -3db at 20kHz. Will I be altering the operating points too much?
(of course a whooping great inductor in series with R9 will do it but it would need to be bigger than the amp!)
My circuit theory is pretty rudimentary, so my idea below is probably full of holes......but hey.....I'm here to learn...so don't be shy with your criticism.
So, what is wrong with inserting a cap in the middle of the feedback to remove some hf.....I figure that bumping up R7 (51K), R9 (360K) & R14 (270K), splitting R9 and adding a 0.08nF to earth results in a -3db at 17.5KHz and -4db at 20kHz. Which is not that far off the original spec of ~ -3db at 20kHz. Will I be altering the operating points too much?
(of course a whooping great inductor in series with R9 will do it but it would need to be bigger than the amp!)
Attachments
Iv'e been thinking about this issue of input impedance for sometime and have just reread Nelson's 'Zen' series of papers, in particular 'Zen variations part 4 - The Penultimate Zen'.
http://www.passdiy.com/pdf/zen-v4.pdf
"For many audio sources, this input impedance is simply
too low to give optimal performance. The ideal input impedance
would be something up around 47 K ohms. We will address this fault here and now...."
And he then details how to go about adding a simple buffer. This appears to me to be the best approach if I wanted to pursue the Lightspeed. I suspect that my scheme in the post above would increase distortion too much.
However after much deliberation (not to mention the fact that so many people love this headphone amp just as it is) I think I will pretty much keep it standard and use a normal pot for volume control. However, James, you may wish to try this.....I'd be very interested in the results especially if you hook it up to your lightspeed,
I am in two minds about the power supply. For the last week I have been planning on using a pair of LM338s for a regulated supply, possibly based on Tangent's 'Tread' http://www.tangentsoft.net/elec/tread/
or even 'Steps'
http://www.tangentsoft.net/elec/teps/
.....but had finally decided to return to the simple original Marcello supply (with perhaps a slight increase in the caps).........
.......but rats! I know I shouldn't have re-read Nelsons 'Zen' papers! I have just been checking out his 'Zen Variations - Part 3 - Active Supply Regulation.' So I'm thinking that if Nelson thinks regulation is a good idea for the zens, perhaps I should reconsider. I know that Metalman says that "A separate P/S voltage regulator for each channel made a whopping difference" Has anyone else made a change from the original to regulated and noticed a difference?
http://www.passdiy.com/pdf/zen-v4.pdf
"For many audio sources, this input impedance is simply
too low to give optimal performance. The ideal input impedance
would be something up around 47 K ohms. We will address this fault here and now...."
And he then details how to go about adding a simple buffer. This appears to me to be the best approach if I wanted to pursue the Lightspeed. I suspect that my scheme in the post above would increase distortion too much.
However after much deliberation (not to mention the fact that so many people love this headphone amp just as it is) I think I will pretty much keep it standard and use a normal pot for volume control. However, James, you may wish to try this.....I'd be very interested in the results especially if you hook it up to your lightspeed,
I am in two minds about the power supply. For the last week I have been planning on using a pair of LM338s for a regulated supply, possibly based on Tangent's 'Tread' http://www.tangentsoft.net/elec/tread/
or even 'Steps'
http://www.tangentsoft.net/elec/teps/
.....but had finally decided to return to the simple original Marcello supply (with perhaps a slight increase in the caps).........
.......but rats! I know I shouldn't have re-read Nelsons 'Zen' papers! I have just been checking out his 'Zen Variations - Part 3 - Active Supply Regulation.' So I'm thinking that if Nelson thinks regulation is a good idea for the zens, perhaps I should reconsider. I know that Metalman says that "A separate P/S voltage regulator for each channel made a whopping difference" Has anyone else made a change from the original to regulated and noticed a difference?