I dont remember if I got the scheme of the servo somwhere or I reverse-engineered the assembly. A thing seems strange to me: perhaps a resistor is missing berween the OpAmp output and the inverting input. Mine are mounted in the amplifier and I do not want to dismount in order to check. Could any of you do the job, please?
You're right! My values are wrong.
I don't know why I reported those wrong values.
However, the only things I changed are in brown and red colors in the schema. I could not check original values on my servos because they are mounted and not accessible that easy; I checked the pictures you can find by Googling them.
In the next days I have to fix a couple of fans inside the cabinet because, when closed, the temperature starts becoming hot; so I will be able to check the assembly.
Rather hear it without the servos first. If you don't like it without you are not likely to like it with. Even the Dartzeel's manual warns the servo is a sonic compromise.
I thought the Servos were designed by a Chinese company, KaisayaHIFI?
Dartzeel NHB 108 Ultimate Upgrade Module for hifi amplifier T0431|dartzeel|module - AliExpress
Kaisaya are using the LF3537 opamp
Dartzeel rather added the Supercapacitors in version B to compensate for DC drift, not the servos.
Dartzeel NHB 108 Ultimate Upgrade Module for hifi amplifier T0431|dartzeel|module - AliExpress
Kaisaya are using the LF3537 opamp
Dartzeel rather added the Supercapacitors in version B to compensate for DC drift, not the servos.
Last edited:
I hope this is an experimental assembly. DO NOT TRUST THE METALLIZED HOLE as a conductor! I saw you soldered a wire on top of a pad allowing the current flow thru the metallized hole. Do not trust it, it could burn and bring with it some other stuff. I'm used shorting all pass-thru holes.
And let us know!
Version A starts life with manual offset adjustment. Dealers get tired of complaining customers with blinking eyes. Emergency supercap "upgrade" is introduced to solve the issue. The supercap obviously does not compensate for diamond buffer drift but IME this is not a problem.
Version B introduces a servo. The few customers who can actually hear start wining that although the eyes don't blink, the sound now sucks. Optional downgrade to Version A is possible and cheap. I have no idea if this means back to supercaps or back to manual labour.
It is easy to understand why people who have shelled out the equivalent of a serviceable car are a bit reluctant to mess with screwdrivers and pots. It is less understandable why supposedly enthusiastic diyers feel the need for a servo.
Supercaps certainly worked well in the experimental Model 2 discussed earlier on. Offset was stable and low. Sound was poor but there may have been other reasons too.
Crazy thing, I cannot believe that finding a solution for the midpoint is impossible! All amplifiers have a control to set the midpoint.
They say they cannot avoid to intercept the signal path but you all know that a solution can be arranged to minimize the interference under human perception. A better filter will do for sure. Nobody can deny this without denying the whole Hi-Fi world. The best amplifier would be speaking in a loud voice.
Some minutes ago I finished servicing my WHA-216, in my lab, adding a couple of fan inside. Then I connected it to the lab speakers (awful indeed) and they started sounding incredibilly good (Sheffield Lab Lincoln Mayorga Vol III). The amplifier I have in my lab is a LJM L20 with some light mods, its sound is not a miracle but decent. But the WHA-216 really pulverized the L20. Not for power, that scares me to raise the volume (I was able to burn a woofer recently) but for sweetness and absolute control of the speakers. I was in a hurry so I was not able to dedicate the right time to listen carefully but, usually, the first impression is the right one.
So the servo cannot be that bad, at least with my mod. I cannot imagine how sound quality could improve removing the servo. One thing for sure, I really hate to have the midpoint floating, and it floats a lot. I think selecting carefully the 2N5xx1 and coupling them thermically would be a good idea to limit midpoint excursion. I will do on the next assembly. I want to experiment carefully and without any hurry, on the NHB-108 because it is very worth it. And even because I have a double 60+60Vac, 1KVA transformer looking for a job.
I was a follower of Bartolomeo Aloia (Italian engineer) and watched at all amplifiers he proposed: they have no total feedback (some - the elder ones - have a very low fb) and have the servo connected as a phase inverter. So, on those amplifiers, you must use an operational or equivalent circuitry to invert the phase. On the contrary, in the NBH-108 the servo is used in phase because it is connected to negative input (emitter of first MJE1503x transistors). So, why do you want to use an active circuitry? Perhaps you need some voltage gain to generate the due correction signal? To be verified.
By the way, dear DEOUNE, I could check my assembly: the resistors on the servo are 1MOhm and 33KOhm, no 100KOhm, no 10KOhm. I confirm I was wrong.
The history of the 108 goes like this:
In the first 108 you could only adjust DC drift once in awhile.
In the next version 108B it got the Neutrik inputTransformers and to choose, servo/no servo.
Later, I think it was around 2010-11, it got the Supercap and the servo/no servo got disabled.
The manual on the internet is not updated!
Last edited:
I don't understand. I think you have seen the sensor conductor from the servo? I can't possible see there could be a problem with that. Orange or green conductor.
You are certainly correct. Thanks for setting this right.
The supercap allows Version A to be retrofitted, which is of course not so easy with the servo. It is still interesting to know which mod makes the sound worse: the supercap or the servo. My bet will be on the supercap.