Re: BOM
C13 and C17 has the correct lead spacing according to BOM.
However, the pin are too thick to go through the PCB.
So is C5.
My PCB version is Rev 2.3
R43-R48 0.22 ohms resister need to be up-straight. It can not lay down to the PCB. So it fit. Just not ideal.
Most of the metal file capacitor are WIMA MKS4, they are not expensive among WIMA capacitor. For WIMA, MKP10 is way better. However, more expensive CAP does not guarantee better sound. You need to tune up per your taste. And the space on PCB won't fit anyway
C13 and C17 has the correct lead spacing according to BOM.
However, the pin are too thick to go through the PCB.
So is C5.
My PCB version is Rev 2.3
R43-R48 0.22 ohms resister need to be up-straight. It can not lay down to the PCB. So it fit. Just not ideal.
Most of the metal file capacitor are WIMA MKS4, they are not expensive among WIMA capacitor. For WIMA, MKP10 is way better. However, more expensive CAP does not guarantee better sound. You need to tune up per your taste. And the space on PCB won't fit anyway
One slight constrainment could be identical cabling for both left and right channels so that all left channel cables are identical length to all right channel cables.
Dual Mono is one precisely identical pair of monoblocs in one enclosure.
This will do: Recycled computer motherboard standoffs stacked double-height to allow sufficient air circulation under the board.
I read the build guide, the Frugalamp thread, the Mongrel thread, and given the parameters of Ostrippers preferences, probably frugal, then I went shopping. While doing so, I upgraded the capacitor selection to those likely to be signal worthy, decreased distortion, low esr and high efficiency. So, the choices are for inexpensive, not for cheap.
You may want to collect 5 different models of caps sized for input (c1) and NFB (c3) so that you can test drive and choose the most favorable. It is possible to use smaller C1 (3.3u and 4.7u both work) and larger C3 (plausible range from 220u to 680u). It is possible to use 220u for both C3 and C4 for deeper bass.
C4 can be a 1u, 0.47u or smaller electrolytic (the easiest way to treble bypass a large electrolytic), but otherwise C4 (treble booster) should be carefully chosen. On the BOM, I hope that I selected either a very tiny value electrolytic or a polyester dip cap since those are less likely to make a peak when used for arbitrary bypass. Instead of arbitrary bypass, which is like roulette, I think you might like to collect some options (several different values) to try at C4 for purposefully selecting a suitable treble bypass that blends in nicely.
I would certainly like to upgrade that BOM list to these outputs:
MJL or NJL
The core model numbers don't change, but MJL, NJL have packages that may be nicer for thermal management.
Also for avoiding ground loops, assuming that the left and right grounds have already met at the source, Once is enough, more is a ground loop, I'd spray this thing with some deoxit and use it for source selector: ROTP4P4 - 4 Pole 3 Position Rotary Switch Technical Data There are 4 conductors in an RCA cable, and a 4 pole switch can switch all 4 conductors for source selector. There is a 3 sources selector, zero ground loops. I believe that is suitable source selector for dual-mono and monobloc efforts.
Last edited:
Re: Cap selection for C3/C4 and C1
I would like to add my 2 cents for the cap selection
In addition to the value and type of capacitor, capacitor brand make significant difference as well.
1) For C1 / 4.7uF, WIMA has very good high while RIFA has very good low frequency. And large value electric cap has very smooth frequency response (although roll off at high frequency)
2) For C3 / 220uF, Nichicon KZ has very rich sound, Nichicon FW has very good high while ELNA Silmic II has very balance sound. C3 only need 10V rating though.
3) And don't forget the brand of power supply cap (40000uF per rail). It change the sound a lot.
This is what I am planning to start after powering up successfully :
a. C1 with 100uF electrical cap from Nichicon VR series. It is cheap and has normal sounding without any high-end cap sound signature
b. I left C4 and other 0.1uF cap empty (C14, C10, C16 and C12) first
c. Without using 0.1uF for the power supply decoupling, the high frequency impedance may be higher and may cause oscillation. If that occurs, I have to install some 0.1uF to begin with.
d. I will add 0.1uF one (ore a pair for both rail) at a time to determine whether the sound is better or worse. Then go through all caps that left empty.
e. I also left C6 empty. I think C6 is to bypass the noise created by ZENA diode. Since I am not using ZENA, thus no need for cap. The cap may tweek the sound which makes tuning more difficult.
Any comment ? Any thought ?
I really like to see how other people perform the sound tuning.
This is probably the last mile between DIY and high-end manufacture for the parts selection.
I would like to add my 2 cents for the cap selection
In addition to the value and type of capacitor, capacitor brand make significant difference as well.
1) For C1 / 4.7uF, WIMA has very good high while RIFA has very good low frequency. And large value electric cap has very smooth frequency response (although roll off at high frequency)
2) For C3 / 220uF, Nichicon KZ has very rich sound, Nichicon FW has very good high while ELNA Silmic II has very balance sound. C3 only need 10V rating though.
3) And don't forget the brand of power supply cap (40000uF per rail). It change the sound a lot.
This is what I am planning to start after powering up successfully :
a. C1 with 100uF electrical cap from Nichicon VR series. It is cheap and has normal sounding without any high-end cap sound signature
b. I left C4 and other 0.1uF cap empty (C14, C10, C16 and C12) first
c. Without using 0.1uF for the power supply decoupling, the high frequency impedance may be higher and may cause oscillation. If that occurs, I have to install some 0.1uF to begin with.
d. I will add 0.1uF one (ore a pair for both rail) at a time to determine whether the sound is better or worse. Then go through all caps that left empty.
e. I also left C6 empty. I think C6 is to bypass the noise created by ZENA diode. Since I am not using ZENA, thus no need for cap. The cap may tweek the sound which makes tuning more difficult.
Any comment ? Any thought ?
I really like to see how other people perform the sound tuning.
This is probably the last mile between DIY and high-end manufacture for the parts selection.
Epic good clue. These will work for amplifier board power circuit needs. By that description, it should be practically perfect. Do any of your preferred (tested) samples of Nichicon FW have voltage and capacitance suitable for Honey Badger's C11, C13, C15, C17?
Instead of 100n versus omission, here's a different method:
The better quality electrolytic caps usually need a smaller size bypass cap, probably polyester is easy. I believe that you can literally eavesdrop on the situation by installing samples of both the electrolytic and the bypass cap, temporarily at C3, C4 for test and then proceed to choose an attractive value for bypass cap. The goal is to avoid peaks/ringing. I'll bet that you never select 100n if you had to listen to a ringing peak. That's because arbitrary 100n bypass is a poor choice that is usually better than nothing. It won't be difficult to do better than that. Whatever is cap selection works well as NFB cap is also probably good models for the power circuit caps too (just keep it simple 'cause you can't hear RF). In my opinion, these caps are all quite important and need specific (non-ringing, non-peakish, non-abrupt, non-arbitrary) bypass. The little polyester bypass cap's value is specific and different per each different model of electrolytic cap, so ideal bypass cap value can't be predicted by a schematic.
Well, that was weird, but I think it is better than omission.
Last edited:
I agree. Quality large caps can be costly, so we don't have the opportunity to try different brands, as a manufacturer would do. Is there a review of the cap types used by more exotic brands in their power supplies anywhere ?
I would like to predict for you that OsTripper's answer to exotic caps in the power supply reservoir would be paralleled panasonic or paralleled nichicon. It takes good caps, not weird caps.
In my opinion, the excellent design shown, will defeat/reduce the difference between ordinary versus high end caps on the power board. Amplifiers are mostly sensitive to any cap on the amplifier board; and, Honey Badger's power supply reservoir isn't on the amplifier board.
However. . .
Input cap, NFB caps and amplifier board power caps are plausible spots to apply some zoot because they are on the amplifier board.
In my opinion, the excellent design shown, will defeat/reduce the difference between ordinary versus high end caps on the power board. Amplifiers are mostly sensitive to any cap on the amplifier board; and, Honey Badger's power supply reservoir isn't on the amplifier board.
However. . .
Input cap, NFB caps and amplifier board power caps are plausible spots to apply some zoot because they are on the amplifier board.
It appears you have the deluxe chassis which has holes drilled and tapped into the heatsinks to match the mounting holes in the boards and the output devices at the correct distance in relation to the boards. The idea is to have them mounted (vertical) right onto the heatsinks using standoffs. Sure makes building an amp a whole lot easier!
Last edited:
Hi all
I have a problem with bias setting, i checked and follow the guide...dmm on tp1 and tp2, set voltage on R30 about 20mv, wait.for a few minutes for thermal stablize recheck again after half hour, tp1 - tp2 become 25.4mv. Turned it back to 20mv. check again for another 1 hour, tp1-tp2 become 27.7mv
It happen always all the time when i checked and tried to setting bias.
Basicly it always moving up slowly but sure.
Thanks
I have a problem with bias setting, i checked and follow the guide...dmm on tp1 and tp2, set voltage on R30 about 20mv, wait.for a few minutes for thermal stablize recheck again after half hour, tp1 - tp2 become 25.4mv. Turned it back to 20mv. check again for another 1 hour, tp1-tp2 become 27.7mv
It happen always all the time when i checked and tried to setting bias.
Basicly it always moving up slowly but sure.
Thanks
Greetings to all and congratulations for providing this project to the diy community.
I have received the boards and I am beginning to populate them. I have an issue though with the jumpers and particularly those at the driver stage transistors Q14 & Q15. The jumper from the transistors base to C18, C19 have no soldering pads. Does this mean that the wire will be soldered to the capacitor and then directly on the transistor base?
regards
I have received the boards and I am beginning to populate them. I have an issue though with the jumpers and particularly those at the driver stage transistors Q14 & Q15. The jumper from the transistors base to C18, C19 have no soldering pads. Does this mean that the wire will be soldered to the capacitor and then directly on the transistor base?
regards
4 ohms with 63vdc rails
Hi All.
Making monoblock amps with a similar circuit, but only 2 instead of 3 OPT's.
I have a pair of 300VA 45-0-45vac transformers(63vdc rails, measured), and will be using two each of the beefier ON Semi MJL4281-4302 transistors, mounted on large 350mmx150mmx50mm heatsinks (Conrad MF35-151.5).
My speakers are rated at 4ohms (Dynaudio Focus), and therefore present a lower impedance than most.
Any concerns or cautions?
Happy Holidays!
-Chas
Hi All.
Making monoblock amps with a similar circuit, but only 2 instead of 3 OPT's.
I have a pair of 300VA 45-0-45vac transformers(63vdc rails, measured), and will be using two each of the beefier ON Semi MJL4281-4302 transistors, mounted on large 350mmx150mmx50mm heatsinks (Conrad MF35-151.5).
My speakers are rated at 4ohms (Dynaudio Focus), and therefore present a lower impedance than most.
Any concerns or cautions?
Happy Holidays!
-Chas
This is the case with one of my AX's (larger badger -4 op devices). It finally does stabilize , but it takes longer. I did build a smaller amp like the badger and
extended the Vbe out on a short piece of flat cable to attach it directly to one of the output devices. This shortened the stabilization time considerably.
But both amps are thermally stable ... try to use an all plastic device for the Vbe .. use either a mje340 or 2sa3503 , the differences in Hfe/internal silicon structure will create a different temperature coefficient.
To go insane , one could even use a 2 device Vbe brought out to one of the output devices with a cable. This would not require reworking the board as one could design a small "daughtercard" with the 2 devices and a couple passive component quite easily. Read Doug Self's "Audio Power Amplifier Design Handbook" - thermal compensation and thermal dynamics... This is just a suggestion as I tried this for use with a "inadequate" heatsink - and it worked !The single vbe should be sufficient with most builds.
OS
Thanks ostripper for your reply
Im using KSC1503, did try change one KSC1503 to another KSC1503, nothing change. Removed mica insulation on VBE and changed to thermal pad and still nothin changed. then i put KSC1503 directly on heatsink without mica or thermal pad, voltage on TP1 - TP2 still moving up. And finally i put VBE on top of Q12 and it does stabilize.
Now i set R30 to 44mv across TP1-TP2 and so far it stable, but i havent check it again since i went out for holiday since Christmas and just back home.
Thanks
Last edited:
use 3 pair @60+ V rails!!
Dynaudio focus's are sealed. Big impedance dips near the Fc of the bass driver might make the amp see 2-3R on occasion. These dips are small and close to resonance but with a strange source (my virtual synths) , I have seen my amp
load down.
This might possibly be a "doomsday" scenario , but it CAN happen.
BTW , what dynaudio (340/360/260)... nice speakers !!
OS
that is why it is important that when you do the bias adjustments, the heatsinks are already heated up and you set your bias accordingly...