That’s good that you ran simulations to determine the DC setpoints for the various transistor combinations.
Regarding the different types of voltage regulators. You probably won’t hear it unless trying to pick up very very low level detail using high sensitivity (110+dB/mW), the 4uV LDO’s will give a blacker background. This will also be visible as shorter “grass” at the bottom of an FFT. Standard regulator will have taller “grass”.
If you add heatsinks, and boost to 120mA bias current - you can drive those lower impedance headphones very cleanly.
Regarding the different types of voltage regulators. You probably won’t hear it unless trying to pick up very very low level detail using high sensitivity (110+dB/mW), the 4uV LDO’s will give a blacker background. This will also be visible as shorter “grass” at the bottom of an FFT. Standard regulator will have taller “grass”.
If you add heatsinks, and boost to 120mA bias current - you can drive those lower impedance headphones very cleanly.
What sort of "packages" to you intend to sell?
1.) Completely assembled Amps, including case & battery
2.) Completely assembled Amps, with case, but no battery
3.) Assembled boards, all components, battery (+ file for 3D printing a case)
4.) Assembled boards, only SMT components, no battery (+ BOM for other components, & file for 3D printing a case)
5.) Board, BOM, Case
6.) ???
1.) Completely assembled Amps, including case & battery
2.) Completely assembled Amps, with case, but no battery
3.) Assembled boards, all components, battery (+ file for 3D printing a case)
4.) Assembled boards, only SMT components, no battery (+ BOM for other components, & file for 3D printing a case)
5.) Board, BOM, Case
6.) ???
Catching back up on DIY - Last three weeks have been hectic, and the next few likely will be crazy. I haven't held an iron in a few weeks. Withdrawal is setting in 
First on my list is to get back to the Hakuin work and try out the new fuel gauge circuit.
Sorry to have been away for a bit.
Hope everyone is happy and healthy!
First on my list is to get back to the Hakuin work and try out the new fuel gauge circuit.
Sorry to have been away for a bit.
Hope everyone is happy and healthy!
Hakuin DB's for Yarra verification build
I just got done soldering this up. Will have to wait before applying power and listening as my Yarra bourd is being used with Melbourne for something else at present. I may just have to solder wires to it. I found out what the resistor changes need to be for the 2SA1837.
New resistor values for 2SK209BL with 2SA1837:
These will sell for the usual price of the Yarra DB's (please check on my shop). I can provide 2SA1837's as an option.
I just got done soldering this up. Will have to wait before applying power and listening as my Yarra bourd is being used with Melbourne for something else at present. I may just have to solder wires to it. I found out what the resistor changes need to be for the 2SA1837.
New resistor values for 2SK209BL with 2SA1837:
These will sell for the usual price of the Yarra DB's (please check on my shop). I can provide 2SA1837's as an option.
Attachments
Last edited:
Good to see that I calculated the same values when I was playing with the Sim
Good deal then, I know this will work when I connect it up.
I guess you did not even need heatsinks. I will be running at 24v so maybe a sink might be needed.
I have a very small DC-DC step up to 24v good for 400mA that I might try. I also have Meanwell bricks that are mains 110/220vac to 24vdc.
Hakuin DB's Verification HPA Build
The Hakuin DB's for the Yarra have now been verified to work and it sounds great. Another headphone amp/preamp on a plank. Using a DC-DC step up and MOSFET cap multiplier (for slow ramp up) to get about 24.7v from a 12v linear wall wart. Getting exactly 24.0v at the amp after the on-board CRCRC. Bias current is 76mA and 14.3v at the collector of the 2SA1837. I did not have any 300ohm resistors so using 270ohms 2w, so the gain is a little less than nominal 11.6dB, but that's ok. Testing with 250ohm Beyer DT880's right now. Will test with the much more sensitive ESS 422H's later. There is still room to increas the bias current and maybe drop voltage to 18v. 2x130ohms (R111/112) and 22R (R113) would work well for circa 160mA bias current.
The Hakuin DB's for the Yarra have now been verified to work and it sounds great. Another headphone amp/preamp on a plank. Using a DC-DC step up and MOSFET cap multiplier (for slow ramp up) to get about 24.7v from a 12v linear wall wart. Getting exactly 24.0v at the amp after the on-board CRCRC. Bias current is 76mA and 14.3v at the collector of the 2SA1837. I did not have any 300ohm resistors so using 270ohms 2w, so the gain is a little less than nominal 11.6dB, but that's ok. Testing with 250ohm Beyer DT880's right now. Will test with the much more sensitive ESS 422H's later. There is still room to increas the bias current and maybe drop voltage to 18v. 2x130ohms (R111/112) and 22R (R113) would work well for circa 160mA bias current.
Attachments
Hi xrk971!
I am undecided about which amplifier build. This or your variant of the Melbourne Class A.
I have a Sennheiser HD598 and I my DAC is a Behringer UMC202HD.
I love the semplicity of your Hakuin, but how it is respect the Melbourne?
For the PCB I don't know what doing. I am able to design the pcb by myself and I love to do it because I am Elettronic Engineer. But at the end your design it is very good and probably I will buy your PCB to support you. I should only check the shipping cost for the Switzerland.
Do you give also the bill of materials?
Questions: Do you have the simulation file for LTSPICE? I would like to try to simulate with the BJT MJL21193 that seems to be a really good
Thank you very much!
Last edited:
Hi Bambinz,
The Hakuin fits the Melbourne/Yarra format daughterboard. It is not a variant of the Melbourne topology which is based on the Aksa Lender topology.
For a headphone amp, I prefer the sound of the Hakuin - I believe it has the highest resolution. If you have lower impedance low sensitivity headphones the Melbourne has more power capability to drive.
Yes, the BOM is already posted somewhere but will he provided if you can’t find it. I’ll post the LTspice soon - the 2SK209 or 2SK170 and 2SA1837 combo is going to be hard to beat as those are some of the finest transistors that Toshiba has ever made for audio, ever.
The Hakuin fits the Melbourne/Yarra format daughterboard. It is not a variant of the Melbourne topology which is based on the Aksa Lender topology.
For a headphone amp, I prefer the sound of the Hakuin - I believe it has the highest resolution. If you have lower impedance low sensitivity headphones the Melbourne has more power capability to drive.
Yes, the BOM is already posted somewhere but will he provided if you can’t find it. I’ll post the LTspice soon - the 2SK209 or 2SK170 and 2SA1837 combo is going to be hard to beat as those are some of the finest transistors that Toshiba has ever made for audio, ever.
Here is the Hakuin LTSpice file. It has an approximate model for the 2SK209GR based on tweaks of Vto of the BL model to get Idss similar to what I measured. This was the approach recommended by Keantoken, who has since published a revised 2SK209GR model, but I have not had time to try it with the revised model. From what I have seen, the current model predicts the DC setpoints close enough for the BL version of the 2SK209.
