Its not that terribly bad. I ordered Spencer DAC kit and soldered everything by myself (excluding SMD parts). Worked just fine. But thats just PCB with all needed elements etc. Im afraid of collecting all parts by myself, matching transistors etc.
Dont need any chassis for now. Im searching best solution for me - Ill think about chassis when i find it.
All I know is to go DIY path as many commercial products or projects dont have proper application. I like sound of Hifimans (I tried a lot of different HP). Regarding headamp i only compared it to Woo Audio WA 6 SE and it was better. Not mention to other chinese clones of BCL. One of friends who listened to it at meeting said its class A+ and he really heard massive stack of equipment.
On the other hand if we only look at measurements shouldnt O2 amp be the best in the world theoretically? Only guessing, maybe Im wrong.
800 USD is too much for testing
PSU shouldnt be problem. I use 2x12V accumulators for headamp
Friend also makes me 2 Salas PSU.
Dont need any chassis for now. Im searching best solution for me - Ill think about chassis when i find it.
All I know is to go DIY path as many commercial products or projects dont have proper application. I like sound of Hifimans (I tried a lot of different HP). Regarding headamp i only compared it to Woo Audio WA 6 SE and it was better. Not mention to other chinese clones of BCL. One of friends who listened to it at meeting said its class A+ and he really heard massive stack of equipment.
On the other hand if we only look at measurements shouldnt O2 amp be the best in the world theoretically? Only guessing, maybe Im wrong.
800 USD is too much for testing
PSU shouldnt be problem. I use 2x12V accumulators for headamp
Friend also makes me 2 Salas PSU.
Last edited:
"drives 32 Ohms" isn't too informative
what's the bias current/Class A output current range of the Marsh amp? (a JLH outupt stage delivers less than 2x the bias in push-pull due to Q beta mismatch and modulation)
the HE-5LE are way low sensitivity @87.5 dB/mW compared to say Grado 32 Ohm headphones @ 98 dB/mW and requires more current and V to reach the same SPL as typical near 100 dB/mW headphones - >10x more power
what's the bias current/Class A output current range of the Marsh amp? (a JLH outupt stage delivers less than 2x the bias in push-pull due to Q beta mismatch and modulation)
the HE-5LE are way low sensitivity @87.5 dB/mW compared to say Grado 32 Ohm headphones @ 98 dB/mW and requires more current and V to reach the same SPL as typical near 100 dB/mW headphones - >10x more power
Last edited:
Ah I noticed theres review in english language already. I have now that headamp. Its IC based but again: are there any real advantages of using discrete designs when both solutions are theoretically well implemented?
High Fidelity
High Fidelity
The Marsh headphone-amp is a Class-A/B output stage (see attached topology schematic from the article abstract), the circuit values might have to be adjusted to "crank-up" the standing bias current. In my own experiences with my HiFiMan HE-6's, I've found that we really need about 200mA bias-current in a Class-A/B output-stage to make the planar-magnetics really "come into their own" from a sonics standpoint.
As I mentioned, I haven't had an opportunity to read Richard Marsh's article on his headphone-amplifier in Linear Audio, but I would assume if we run with at least +/- 24VDC power-supply rails and 200mA bias-current in the output stage, we should be in good shape for driving the HiFiMan beasties (Richard noted that he himself uses this circuit with HiFiMan planar-dynamics...).
Attachments
Its not that terribly bad. I ordered Spencer DAC kit and soldered everything by myself (excluding SMD parts). Worked just fine. But thats just PCB with all needed elements etc. Im afraid of collecting all parts by myself, matching transistors etc.
Dont need any chassis for now. Im searching best solution for me - Ill think about chassis when i find it.
All I know is to go DIY path as many commercial products or projects dont have proper application. I like sound of Hifimans (I tried a lot of different HP). Regarding headamp i only compared it to Woo Audio WA 6 SE and it was better. Not mention to other chinese clones of BCL. One of friends who listened to it at meeting said its class A+ and he really heard massive stack of equipment.
On the other hand if we only look at measurements shouldnt O2 amp be the best in the world theoretically? Only guessing, maybe Im wrong.
800 USD is too much for testing
PSU shouldnt be problem. I use 2x12V accumulators for headamp
Friend also makes me 2 Salas PSU.
The problem with pursuing a DIY solution is that it's pretty difficult to find an existing implementation to audition prior to committing to constructing your own instance. I constructed a half-dozen prototypes before I built my current prototype that I briefely described earlier (John Curl-designed JC-80 as the core engine, Erno Borbely-designed all-FET shunt power-supply regulator, and relay-based 128-step attenuator). I'll likely build Richard Marsh's headphone-amplifier PCB's and insert them into my current prototype for an audition; that's what DIY audio is all about, after all...
Without the ability to audition most of the circuits, you'll eventually reach a point where you'll just have to commit to a given circuit, build it and see what you think...
Good luck!
just for clarity, Nelson contracted the now bankrupt Semisouth to make him a SIT all of his own to his specifications, the transistor used in that amp linked is a standard sic power JFET, not a SIT, different things. I do hope that kickstarter guy managed to buy his jfets before Semisouth went belly up last week, there really isnt an easy substitute except a mosfet*,which i'm sure wouldnt go down well with the backers of the project. (*at least if it was using the depletion mode Semisouth parts, they were fairly easily dropped into many projects that use Mosfets with a simple change of bias),
Zibra,
Static Induction Transistors (SiT's) are a relatively new genre of solid-state gain devices which offer transfer-function characteristics similar to those of triode vacuum-tube devices. As you noted, Nelson Pass is among the first noted audio designer to invest in the production of suitable SiT's for audio-reproduction purposes. These devices cannot be just substituted for more conventional solid-state devices on a "drop-in" basis; instead, the circuit must be tailored to the devices in order to exploit their "special" characteristics...
Richard Marsh is a retired key researcher in High Energy Physics R&D at Lawrence Livermore National Laboratory and a published member of IEEE and AES. He recently published his "diamond" topology-based headphone amplifier design in Volume 3 of Jan Didden's "Linear Audio" bookzine (www.linearaudio.net - home). I'm a bit behind on my audio-design journal reading, so I haven't read the article as yet, but the abstract illustrates a simple/elegant topology which could be built in a "bridged" configuration to easily drive balanced-drive headphones such as the HiFiMan planar magnetics...
Just my $0.02's worth...
Last edited:
Thank you very much
I found that Im listening at max 100-105dB probably 90% of time. Does that mean I should have amp which should have 500mW free of distortion output power to properly drive even HE6 in that case? How does that compare to demanding classical music where instruments can reach 110dB and some symphonics can have 120dB peaks? Do I loose anything by listening at lower volumes in that case?
I see a lot of threads where people connect these HP to speaker taps and say HE6 sound much better. Is it false from theoretical point of view?
I found that Im listening at max 100-105dB probably 90% of time. Does that mean I should have amp which should have 500mW free of distortion output power to properly drive even HE6 in that case? How does that compare to demanding classical music where instruments can reach 110dB and some symphonics can have 120dB peaks? Do I loose anything by listening at lower volumes in that case?
I see a lot of threads where people connect these HP to speaker taps and say HE6 sound much better. Is it false from theoretical point of view?
The Panda headphone amp kit is very good with senn HD600 and hifman planars are supposed to be similar but better than those, the panda also has more power than most headphone amps. There are of course mods/tips that will get more out of a build:
Have transformer away from amp.
If you have the know how setup a seperate PSU for each channel using the RC resistors as the input points.
Use a polystyrene cap for the feedback miller cap rather than the ceramics.
Liberally add bypass caps to the electrolytic psu caps, it really helps with this amp.
Use MJE15030/31 family rather than tip41/42, I suggested this ages ago and with the new built edition they have labelled those spots on the pcb as tip41/mje15030.
Fancy volume control of your choice, it comes with an Alps blue velvet so that can be improved upon.
Listen for a month and then consider tuning with different caps and premium resistors, input and feedback resitor being the important ones.
Match the jfets if you have the equipement and inclination.
Have transformer away from amp.
If you have the know how setup a seperate PSU for each channel using the RC resistors as the input points.
Use a polystyrene cap for the feedback miller cap rather than the ceramics.
Liberally add bypass caps to the electrolytic psu caps, it really helps with this amp.
Use MJE15030/31 family rather than tip41/42, I suggested this ages ago and with the new built edition they have labelled those spots on the pcb as tip41/mje15030.
Fancy volume control of your choice, it comes with an Alps blue velvet so that can be improved upon.
Listen for a month and then consider tuning with different caps and premium resistors, input and feedback resitor being the important ones.
Match the jfets if you have the equipement and inclination.
Someone else was loking for the same here :
http://www.diyaudio.com/forums/head...lass-zgf-headphone-amplifier.html#post3293956
Patrick
http://www.diyaudio.com/forums/head...lass-zgf-headphone-amplifier.html#post3293956
Patrick
Search DIYaudio and Google for ExtremeA a balanced ClassA amp.
Consider building a balanced Class_A amp like the "ExtremeA" that can easily drive balanced headphones like the HiFiMan or modest power demand speakers.
If you are not using battery power, it seems smart to build ONE balanced DIY amp that can be used with both headphones and speakers. YEP... about 20 pounds with 10 watts of ClassA power.... HOT HOT HOT... ''Till death do us part" Give it to your kids
It is possible to use dual JFETs for the ExtremeA input stage instead of bipolar transistors in the schematic if you have a funky audio source with irratic output impedance.
Consider building a balanced Class_A amp like the "ExtremeA" that can easily drive balanced headphones like the HiFiMan or modest power demand speakers.
If you are not using battery power, it seems smart to build ONE balanced DIY amp that can be used with both headphones and speakers. YEP... about 20 pounds with 10 watts of ClassA power.... HOT HOT HOT... ''Till death do us part" Give it to your kids
It is possible to use dual JFETs for the ExtremeA input stage instead of bipolar transistors in the schematic if you have a funky audio source with irratic output impedance.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.