in series with the power supply there is a resistance of 5 watts, then restoring with the zener and then finally using the 317, less noise etc etc
the first 4 transistors are crucial for the amplifier, so they must have a very high-performance network
the offset varies from 15 to 26 mv
the BCM53DS chip for Q 1, 2, 3 and 4, two chips, manages to minimize the offset and makes everything very easy
PS
we used the BCM53DS chip, but there is something better: that we had
with the 2 chips,
the 4 pairs of finals with the R from 0.22,
the 4 drivers the same as the finals,
the correct polarizations of the 4 pairs,
the cascaded stabilized feeds for driver and input,
the skotty diodes
the entrance cap Mundorf instead of the two Wima red fetuses
we can no longer speak of dartzeel
the first 4 transistors are crucial for the amplifier, so they must have a very high-performance network
the offset varies from 15 to 26 mv
the BCM53DS chip for Q 1, 2, 3 and 4, two chips, manages to minimize the offset and makes everything very easy
PS
we used the BCM53DS chip, but there is something better: that we had
with the 2 chips,
the 4 pairs of finals with the R from 0.22,
the 4 drivers the same as the finals,
the correct polarizations of the 4 pairs,
the cascaded stabilized feeds for driver and input,
the skotty diodes
the entrance cap Mundorf instead of the two Wima red fetuses
we can no longer speak of dartzeel
Last edited:
it's convenient to start from the point where others have arrived, but the DIY this is.
The scheme in question is intrinsically interesting and must be understood for what it can give.
But an open loop will never have a dynamic and a control on the low and medium-low as the pattern I posted.
Accepting these indisputable inputs let me post that the dartzeel scheme can be varied with very interesting listening results.
today the dartzeel scheme can benefit from dual SMD chips due to the extremely difficult absolute level input stage which, at the time, either did not exist or were never considered for some strange reason
sometimes I am led to think that some, not all, a few, or many, hiend builders fill the container with numbers of useless pieces
The scheme in question is intrinsically interesting and must be understood for what it can give.
But an open loop will never have a dynamic and a control on the low and medium-low as the pattern I posted.
Accepting these indisputable inputs let me post that the dartzeel scheme can be varied with very interesting listening results.
today the dartzeel scheme can benefit from dual SMD chips due to the extremely difficult absolute level input stage which, at the time, either did not exist or were never considered for some strange reason
sometimes I am led to think that some, not all, a few, or many, hiend builders fill the container with numbers of useless pieces
This is great and thanks for sharing, but it looks like a departure form the original topology. It might measure better than the original and have less drift, but not sure how much of the Dartzeel sound it will have.
Last edited:
the Dartzeel sound is that of the original appliances with the merits and defects typical of the circulation of the brand.
If one loves the sound Dartzeel does not look for an alternative way to the known scheme.
Buying the Chinese card hoping that it will come close to the original is there, but it must be revised as best as possible (absolute BJT selection) also to preserve all the limitations that characterize the original.
Use 6 x Vishay 22,000 micro 63 volt caps (and this one of the 63 volts I think is an error: certainly 100 volts) per channel, even if using only 2 x cap Vishay from 22,000 63 volts changes nothing: it changes using 2 x Vishay cap of 22,000 x 100 volts.
Instead this review keeps the basic DNA of the scheme limiting its defects, NOT canceling them:
a comparison with the original Dartzeel is not correct, but makes the revisited immediately in all the known audiophile parameters
If one loves the sound Dartzeel does not look for an alternative way to the known scheme.
Buying the Chinese card hoping that it will come close to the original is there, but it must be revised as best as possible (absolute BJT selection) also to preserve all the limitations that characterize the original.
Use 6 x Vishay 22,000 micro 63 volt caps (and this one of the 63 volts I think is an error: certainly 100 volts) per channel, even if using only 2 x cap Vishay from 22,000 63 volts changes nothing: it changes using 2 x Vishay cap of 22,000 x 100 volts.
Instead this review keeps the basic DNA of the scheme limiting its defects, NOT canceling them:
a comparison with the original Dartzeel is not correct, but makes the revisited immediately in all the known audiophile parameters
also the other scheme I posted is superior to both Goldmund and Norma and also to Spectral.
Superior both technically and listening, in all the parameters dear to audiophiles.
is a scheme derived from Hitachi, just like the other 3.
Spectral comes from Hitachi, Goldmund as well, and Norma comes from Spectral, anyone sees it.
Goldmund has drivers in TO 92, in my opinion an error, Spectral and Norma in TO 220.
All three have the same identical and well-made entry stage.
Norma has in addition an emitter follower, at the time with the dual fetal J 109 Toshiba, now obsolete, derived from Electrocompaniet.
These evidences lead the DIY to be able to obtain circuits always modern and with a final yield superior to all the mentioned brands.
For 40 years, Spectral has always used the same circuitry that varies according to the production of some fet and bjt.
Norma has stabilized driver and input stages.
Norma has very low filter caps.
The modern DIY, I repeat, while keeping the same basic circuitry, can do better without too much effort
Superior both technically and listening, in all the parameters dear to audiophiles.
is a scheme derived from Hitachi, just like the other 3.
Spectral comes from Hitachi, Goldmund as well, and Norma comes from Spectral, anyone sees it.
Goldmund has drivers in TO 92, in my opinion an error, Spectral and Norma in TO 220.
All three have the same identical and well-made entry stage.
Norma has in addition an emitter follower, at the time with the dual fetal J 109 Toshiba, now obsolete, derived from Electrocompaniet.
These evidences lead the DIY to be able to obtain circuits always modern and with a final yield superior to all the mentioned brands.
For 40 years, Spectral has always used the same circuitry that varies according to the production of some fet and bjt.
Norma has stabilized driver and input stages.
Norma has very low filter caps.
The modern DIY, I repeat, while keeping the same basic circuitry, can do better without too much effort
we have the forklift and we are redoing the printed material because we only found the dual fet just a few days ago to replace the unfindable 2N 5566
We did the Spectral 50 and also the Standard 8.7 B
redo the printed material to use the dual chips mentioned
as soon as they are made and mounted, I will definitely post them
We did the Spectral 50 and also the Standard 8.7 B
redo the printed material to use the dual chips mentioned
as soon as they are made and mounted, I will definitely post them
as always, one must choose which direction to go.
the Chinese card, I call it this way, goes in the direction of fascination and interpretation, like the valves, therefore with all the pros and cons of valves and open loop schemes.
my reinterpretations greatly diminish the limits of the scheme, but more than that it is impossible, except to distort it: but does it make sense?
I would say no
In the other direction the second scheme goes, complete, has everything, speed, fidelity to engraving, low range, dynamics, articulation, but never could have the magic of the Chinese card that can be defined a splendid my sound.
Important is the combined final amplifiers: the Chinese card is finicky, the second scheme, instead, is a running train.
PS
to note that in the second scheme there is not a single compensation, not bad, indeed
I add that the Chinese card is also very much affected by the presence of the supercap and the input cap which, in this sheet, is a very bad cap: Wima MKP 10, a sonorous infamy
with the second scheme, if I use his pre, I can play by inserting a cap at the AC output and playing until you find a cap that you like, but we're in my total sound
the Chinese card, I call it this way, goes in the direction of fascination and interpretation, like the valves, therefore with all the pros and cons of valves and open loop schemes.
my reinterpretations greatly diminish the limits of the scheme, but more than that it is impossible, except to distort it: but does it make sense?
I would say no
In the other direction the second scheme goes, complete, has everything, speed, fidelity to engraving, low range, dynamics, articulation, but never could have the magic of the Chinese card that can be defined a splendid my sound.
Important is the combined final amplifiers: the Chinese card is finicky, the second scheme, instead, is a running train.
PS
to note that in the second scheme there is not a single compensation, not bad, indeed
I add that the Chinese card is also very much affected by the presence of the supercap and the input cap which, in this sheet, is a very bad cap: Wima MKP 10, a sonorous infamy
with the second scheme, if I use his pre, I can play by inserting a cap at the AC output and playing until you find a cap that you like, but we're in my total sound
Last edited:
Of course. I chose the route of no feedback cap, matched transistrs and BG-N at the input and the result more than justifies the effort. Ok, also have Mlytics in the PS which obviously helps a bit
But is it not time to move on?
In post #205 Gaborbela shared the reversed circuit of the big brother. Apparently the 108mk2 is pretty much the same minus the parallel outputs.
So, goodbye Diamond buffer, hello Standard follower with a bias spreader
The updated circuit is certainly simpler and despite the supercaps is reported to sound even better than the mk1.
It is very easy to adapt it for a single output pair and lower PS voltage, but i am thinking of building it as shown. Perhaps just reducing the voltage to below 80V, so that standard PS capacitors can be used.
After living with the 108 mk1 for a while i decided that despite the extraordinary midrange clarity, some extra driving oomph can only be welcome. Of course some of the clarity is likely to be lost in the parallel transistors, but perhaps this will be a worthwhile trade. One of my speaker sets simply refuses to play with the single transistor version.
Of course. I chose the route of no feedback cap, matched transistrs and BG-N at the input and the result more than justifies the effort. Ok, also have Mlytics in the PS which obviously helps a bit
But is it not time to move on?
In post #205 Gaborbela shared the reversed circuit of the big brother. Apparently the 108mk2 is pretty much the same minus the parallel outputs.
So, goodbye Diamond buffer, hello Standard follower with a bias spreader
The updated circuit is certainly simpler and despite the supercaps is reported to sound even better than the mk1.
It is very easy to adapt it for a single output pair and lower PS voltage, but i am thinking of building it as shown. Perhaps just reducing the voltage to below 80V, so that standard PS capacitors can be used.
After living with the 108 mk1 for a while i decided that despite the extraordinary midrange clarity, some extra driving oomph can only be welcome. Of course some of the clarity is likely to be lost in the parallel transistors, but perhaps this will be a worthwhile trade. One of my speaker sets simply refuses to play with the single transistor version.
Great post.
I decided about 25 years ago that specialized amplifiers is the way to go for me. I love the 108 mk1's mids. They don't scream at me like other amplifiers do. So I want to keep it. Maybe use a different amplifier for bass or highs etc. This is why I never went with single full range drivers
Of course. I chose the route of no feedback cap, matched transistrs and BG-N at the input and the result more than justifies the effort. Ok, also have Mlytics in the PS which obviously helps a bit
But is it not time to move on?
In post #205 Gaborbela shared the reversed circuit of the big brother. Apparently the 108mk2 is pretty much the same minus the parallel outputs.
So, goodbye Diamond buffer, hello Standard follower with a bias spreader
The updated circuit is certainly simpler and despite the supercaps is reported to sound even better than the mk1.
It is very easy to adapt it for a single output pair and lower PS voltage, but i am thinking of building it as shown. Perhaps just reducing the voltage to below 80V, so that standard PS capacitors can be used.
After living with the 108 mk1 for a while i decided that despite the extraordinary midrange clarity, some extra driving oomph can only be welcome. Of course some of the clarity is likely to be lost in the parallel transistors, but perhaps this will be a worthwhile trade. One of my speaker sets simply refuses to play with the single transistor version.
Of course. I chose the route of no feedback cap, matched transistrs and BG-N at the input and the result more than justifies the effort. Ok, also have Mlytics in the PS which obviously helps a bit
But is it not time to move on?
In post #205 Gaborbela shared the reversed circuit of the big brother. Apparently the 108mk2 is pretty much the same minus the parallel outputs.
So, goodbye Diamond buffer, hello Standard follower with a bias spreader
The updated circuit is certainly simpler and despite the supercaps is reported to sound even better than the mk1.
It is very easy to adapt it for a single output pair and lower PS voltage, but i am thinking of building it as shown. Perhaps just reducing the voltage to below 80V, so that standard PS capacitors can be used.
After living with the 108 mk1 for a while i decided that despite the extraordinary midrange clarity, some extra driving oomph can only be welcome. Of course some of the clarity is likely to be lost in the parallel transistors, but perhaps this will be a worthwhile trade. One of my speaker sets simply refuses to play with the single transistor version.
Great post.
I decided about 25 years ago that specialized amplifiers is the way to go for me. I love the 108 mk1's mids. They don't scream at me like other amplifiers do. So I want to keep it. Maybe use a different amplifier for bass or highs etc. This is why I never went with single full range drivers
there is
It is important to have clear ideas
the forums for this serve
