A New Take on the Classic Pass Labs D1 with an ESS Dac

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi,

tested the Borbely I/V (pasive resistor conversion + discrete OP-amp gain stage) against a JockoHomo-sructure (grounded Base, very similar to Owen´s stage but with active ccs and lower bias currents) and modified Hawksford structures. All of them sounded clearly better than OP-amp stages, but the Hawksford came last. The Borebely and the Jocko were nearly on par, with the Jocko beeing preferred more of the time. The Jocko remained unbeaten in the mid-highs while the Borbely showed more weight and slam in the bass.
The Jocko improved in the bass after careful tweaking of supply paths.
Still though the Jocko remained rather a conoisseur´s circuit, the Borbely rather the Wham-Bam-thank-You-Mam circuit.
I expect Owen´scircuit to be sonically at last on par with the Jocko, but with better measurement value.

jauu
Calvin
 
Member
Joined 2009
Paid Member
Hi,

tested the Borbely I/V (pasive resistor conversion + discrete OP-amp gain stage) against a JockoHomo-sructure (grounded Base, very similar to Owen´s stage but with active ccs and lower bias currents) and modified Hawksford structures. All of them sounded clearly better than OP-amp stages, but the Hawksford came last. The Borebely and the Jocko were nearly on par, with the Jocko beeing preferred more of the time. The Jocko remained unbeaten in the mid-highs while the Borbely showed more weight and slam in the bass.
The Jocko improved in the bass after careful tweaking of supply paths.
Still though the Jocko remained rather a conoisseur´s circuit, the Borbely rather the Wham-Bam-thank-You-Mam circuit.
I expect Owen´scircuit to be sonically at last on par with the Jocko, but with better measurement value.

jauu
Calvin

Hi Calvin & thanks, could you post Jocko schematics?
 
Hi,

there have been posted unnumerous incarnations of the Jocko here already. Some using bipolartransistors, some using MOSFETs, some using JFETs......I won´t add another one.

What became obvious was that the lower the parts count within the signal path/circuit the more authentical or pleasing it sounded. The drawback of this is usually inferior measurement values especially THD-wise. The highcurrent-highvoltage Jocko or Owens D1B1 (name them as You wish) seem to break the rule. Still though show Sims that if You add a simple active current source towards the negative rail in Owen´s schematic that the already low distortion values slightly decrease. It will be up to an listening evaluation though if this mod is worth beeing implemented.

jauu
Calvin
 
Last edited:
Member
Joined 2009
Paid Member
Hi,

there have been posted unnumerous incarnations of the Jocko here already. Some using bipolartransistors, some using MOSFETs, some using JFETs......I won´t add another one.

What became obvious was that the lower the parts count within the signal path/circuit the more authentical or pleasing it sounded. The drawback of this is usually inferior measurement values especially THD-wise. The highcurrent-highvoltage Jocko or Owens D1B1 (name them as You wish) seem to break the rule. Still though show Sims that if You add a simple active current source towards the negative rail in Owen´s schematic that the already low distortion values slightly decrease. It will be up to an listening evaluation though if this mod is worth beeing implemented.

jauu
Calvin

Hi Calvin, if you won't want to do a public post could you send me a PM with your own schematic used.

Thanks in advance.
Felipe
 
Member
Joined 2009
Paid Member
Hi,

there have been posted unnumerous incarnations of the Jocko here already. Some using bipolartransistors, some using MOSFETs, some using JFETs......I won´t add another one.

What became obvious was that the lower the parts count within the signal path/circuit the more authentical or pleasing it sounded. The drawback of this is usually inferior measurement values especially THD-wise. The highcurrent-highvoltage Jocko or Owens D1B1 (name them as You wish) seem to break the rule. Still though show Sims that if You add a simple active current source towards the negative rail in Owen´s schematic that the already low distortion values slightly decrease. It will be up to an listening evaluation though if this mod is worth beeing implemented.

jauu
Calvin


Hi Calvin, to know more things about your comparisons, please be so kind to let me know: are you using his own all fets discrete reg. for Erno Borbely I/V & about PSU are you using Borbely his own with chokes & 10.000uF 63V Jensen 4 poles caps?

An externally hosted image should be here but it was not working when we last tested it.
 
Hi,

I tried a own design, based on the Jocko using bipolar transistors and and more advanced current sources and dimensioned for the PCM1794A DAC working in Mono mode. Sound was absolutely great, size was tiny, but THD was worse than I liked it to be. It ran on 12V dual supplies. Probabely similar to Owen´s stage I assume there was not enough voltage overhead.
The Borbely was connected to a good SMPS, maybe some caps between, I don´t know any more. Because beeing of a fed-back-OPamp-topology and the sheer size of the PCB ruled out further investigations into the matter.

jauu
Calvin
 

opc

Member
Joined 2004
Paid Member
Hi Paul,

Lauret is absolutely correct in that the 37 volts is a maximum Vin to Vout number. The issue is more clearly explained on the LM117 datasheet:

"Besides replacing fixed regulators, the LM117 is useful in a
wide variety of other applications. Since the regulator is “floating”
and sees only the input-to-output differential voltage,
supplies of several hundred volts can be regulated as long as
the maximum input to output differential is not exceeded, i.e.,
avoid short-circuiting the output."

What that means is that you could regulate say 60VDC down to 50VDC, but you can't regulate 60VDC down to 10VDC. You could potentially use these even in a tube circuit to drop say 350VDC down to 320VDC, but I would imagine that without any other regulation on the input voltage, you could easily exceed the Vin to Vout limit under high-line conditions.

I haven't had any issues with mine at all, and it has been running for quite some time now.

Regards,
Owen
 
If you stick with the mosfets and working conditions indicated by Owen, this D1 variant has a much lower input impedance. Thus this circuit should work better than the Zen I/V in conjunction with dac chips that are very sensitive to the load impedance. This is the case of the ES9018. The drawback is a much greater power demand.
 
Last edited:
isnt the zen I/V the new bad boy in town, which would make this i/V a thing of the past?

not at all, only if you wish to blindly follow trends. the D1 as owen has implemented it is a FAR better match for the sabre, as described above. measured real world performance is also considerably better.

mosfets have much higher transconductance than JFETS, especially when you pump up the voltage; this suits the sabre down to the ground.

I wil try out tyhe zen IV also, but going even on best case modeled scenarios for the JFET stage, the measured performance of the D1 is superior
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.