I completed the amp just before the holiday break, and buttoned up temporarily with the borrowed back panel. Here are some build notes that may be helpful:
- LED leads routed under baseplate, with slack to facilitate install/uninstall. 18AWG segment to create spring-loading against front panel
- No lock-washers along bottom of heatsink rail -- they'll interfere with baseplate
- Tape front panel screws to flexible screwdriver to facilitate final assembly
- Tape baseplate-to-heatsink-rail counter-sinking screws to bottom of HS rail to facilitate mounting to baseplate
- Beware front-to-back orientation of HS frame. Threaded mounting holes are for rear-panel attachment (yes, I got it wrong bot this and the last time I did this too).
- Apply electrical tape to cover exposed pads on ingress-power termination board. (I routed wiring underneath, which leaves topside open for contact.)
- Note that it's easy to strip the threads on the PSU boards' aluminum standoffs. Would favor steel standoffs next time
- Create binding posts for chassis ground attachment points with 16mm M3 screws attached to baseplate. Then no need to open bottom plate to attach/remove grounding lead(s)
@ekobi
Nice video of DBT power up! What switch are you using there? Is that using a relay circuit at all, or AC mains going through directly?
Disregard my question. I see now it's a 20A switch doing direct AC mains switching from Aliexpress from an earlier post. Nice!
Nice video of DBT power up! What switch are you using there? Is that using a relay circuit at all, or AC mains going through directly?
Disregard my question. I see now it's a 20A switch doing direct AC mains switching from Aliexpress from an earlier post. Nice!
Last edited:
Back panel from SendCutSend arrived right after the new year, and I wasted no time swapping it in, and restoring the F5m to service. It might not be easy to see from these pix, but both the input and output audio connectors are about a centimeter further outboard than before, which greatly helps with the tight inner clearance.
Also closed up the bottom of the new amp, attaching feet with 16mm M3 screws and lock washers.
My very first listening impressions of my newly-built Aleph Jzm w/ dual-mono supplies were ... not favorable :<
To be fair, it's over 25 years old, but I was still very surprised to see an AudioQuest cable fail. The good news is that, once I fixed the cable, the Aleph J sparng to life quite dramatically. With the F5m back up and running, it's been fun listening to both and comparing them.
Aleph J is about 18dB louder than the F5m in my setup. The Benchmark HPA4 has both single-ended and balanced outputs; according to the manual "Balanced outputs are 15.8 dB hotter than the unbalanced outputs", so that accounts for most of the difference.
Also closed up the bottom of the new amp, attaching feet with 16mm M3 screws and lock washers.
My very first listening impressions of my newly-built Aleph Jzm w/ dual-mono supplies were ... not favorable :<
- Big, wide mids and highs
- Highs very bright, quickly fatiguing "at volume" (I didn't write down what levels I was listening at, but probably betweem 50 and 60dB SPL, which is the high end of my normal listening range).
- Missing texture in mid bass. Can't hear "crunch" or "air" around bass notes.
- Lower bass very solid, but not sharp
To be fair, it's over 25 years old, but I was still very surprised to see an AudioQuest cable fail. The good news is that, once I fixed the cable, the Aleph J sparng to life quite dramatically. With the F5m back up and running, it's been fun listening to both and comparing them.
Aleph J is about 18dB louder than the F5m in my setup. The Benchmark HPA4 has both single-ended and balanced outputs; according to the manual "Balanced outputs are 15.8 dB hotter than the unbalanced outputs", so that accounts for most of the difference.
- Aleph J presents big, airy soundstage, similar to F5m
- At 60dB SPL, the Aleph J sounds amazingly good: very sharp image, crystal clear highs (but no longer fatiguing)
- Upper-mids and highs are more pronounced with the Aleph J, soundstage is more forward than the F5m
- Aleph J has phatter bass, but not as coherent as F5m
- At late-night listening levels (30dB SPL) the Aleph J becomes unengaging; the F5m remains very engaging even at 30dB SPL
if you want fair comparison between amps, drive them in same mode from your preamp - so SE
I'm expecting that with "Balanced outputs are 15.8 dB hotter" you're in region of too muuuuuuuuuch signal attenuation ditto at vol pot
in other words - you're killing it way before it even reach any amplification stage (eve in preamp itself), whatever amp put on end; F4 is more for that
read this for more beans: https://www.diyaudio.com/community/threads/what-is-gain-structure.186018/
edit: both F5 (whichever iteration) and Aleph J ( whichever iteration) are hardly demanding more than 6db of preamp gain, with nowadays sources
any db above that is waste of signal, done with attenuation
take that as useful grain of salt, where grain of salt is +/-3db worth
I'm expecting that with "Balanced outputs are 15.8 dB hotter" you're in region of too muuuuuuuuuch signal attenuation ditto at vol pot
in other words - you're killing it way before it even reach any amplification stage (eve in preamp itself), whatever amp put on end; F4 is more for that
read this for more beans: https://www.diyaudio.com/community/threads/what-is-gain-structure.186018/
edit: both F5 (whichever iteration) and Aleph J ( whichever iteration) are hardly demanding more than 6db of preamp gain, with nowadays sources
any db above that is waste of signal, done with attenuation
take that as useful grain of salt, where grain of salt is +/-3db worth
Last edited:
Agree 100%, and I'll do that here at some point. I wired up my build for balanced only, so I need to figure out if I can get away without reconfiguring beyond installing J1.
I have spent a fair amount of time listening to the ACA monoblocks in balanced mode, including some direct A/B with this amp. Even though they're easily bested by the Aleph J overall, the ACAs retain their seductive sweetness at very low listening levels.
The gain structure article is a great reference, thank you. FWIW, my DAC emits a 4.2VRMS full-scale signal via balanced outputs. The preamp has excellent dynamic range, and uses stepped attenuators for gain control. Your point about needing to attenuate the input signal is well made -- I reckon it's down in the tens of millivolts for low-volume listening (I don't have access to lab gear these days, so can't readily measure this). Single-ended operation is going to boost that by a few dB. It'll be interesting to see if and how different it sounds. Stay tuned.
use either RCA to XLR cable (tie pin1 & pin3), or RCA to XLR adaptor
as I said - test (you did) itself is totally flawed, but you did learn something
so, positive outcome
proper scientific reasoning, step in positive direction
What??? The test was to determine how well I like the sound of the amp I just built. My conclusion is that it performs relatively poorly under certain conditions, compared to two other amp designs I built. Not seeing how that makes for a "totally flawed" test.
care to examine facts I laid in front of you?
theme is not to measure which one's is bigger, theme is to give you understanding and way to have better sound
try it, draw your own conclusions
test is flawed simply because you got excessive gain in system ....... anyhow, re-read what I wrote
theme is not to measure which one's is bigger, theme is to give you understanding and way to have better sound
try it, draw your own conclusions
test is flawed simply because you got excessive gain in system ....... anyhow, re-read what I wrote
As anticipated, switching to single-ended inputs didn't really change my impressions of the Aleph Jzm. Perhaps slightly better pressence at super low volumes, but honestly can't say for sure. Definitely wants to be played at least moderately loudly to sound it's best in my setup. This episode did get me thinking about gain structure more explicitly, though, and even more interested in an F4 build.
Moving along, I ported the F5m boards to this dual-mono PSU chassis last night. Took a couple of hours to build new harnesses, and about an hour a side to remove and replace the amp boards. It was fiddly, and sometimes frustrating, to get everyting into place without dismantling the chassis. Very grateful for the bendy drive handle that came with my birthday iFixit kit.
Dim bulb testing and power-up were uneventful. I had originally set the bias voltages to 650mV, dissipating around 70 watts per channel, and decided to keep them there for now. Unsurprisingly, the 400mm 3U chassis runs just cooler than "uncomfortably hot" with this level of dissipation, compared to "too hot to touch for long" with the smaller, 300mm 3U chassis.
Got a delightful late-night session in before bed, listening to Malcolm Jiyane Tree-O's first album, Umdali. His bass player, Ayanda Zalekile, is amazing -- lyrical and deft, and yet understated throughout his bandmates' solos (track 2 has been one of my go-to A/B auditioning pieces).
Moving along, I ported the F5m boards to this dual-mono PSU chassis last night. Took a couple of hours to build new harnesses, and about an hour a side to remove and replace the amp boards. It was fiddly, and sometimes frustrating, to get everyting into place without dismantling the chassis. Very grateful for the bendy drive handle that came with my birthday iFixit kit.
Dim bulb testing and power-up were uneventful. I had originally set the bias voltages to 650mV, dissipating around 70 watts per channel, and decided to keep them there for now. Unsurprisingly, the 400mm 3U chassis runs just cooler than "uncomfortably hot" with this level of dissipation, compared to "too hot to touch for long" with the smaller, 300mm 3U chassis.
Got a delightful late-night session in before bed, listening to Malcolm Jiyane Tree-O's first album, Umdali. His bass player, Ayanda Zalekile, is amazing -- lyrical and deft, and yet understated throughout his bandmates' solos (track 2 has been one of my go-to A/B auditioning pieces).
Attachments
now, it would be interesting to compare THD Spectra of these two amps, and you'll know exactly what your preference is
If you really want to bake your noodle, make outboard linear power supplies for your ACAs.
😈
So I had to go look this one up -- never heard this expression! Like the idea of beefier power supplies for the ACA, although I did promise my better half that I was going to start shedding amps "real soon now." Sigh.
