| lgreen |
Speaking of Gainclones, has anyone ever tried to clone the JOB Circuit?
Here is a link to their website:
http://www.jobsys.com/products.htm
I know that the 47 Labs Gaincard is expensive and has garnered some good reviews so a clone of the same circuit is very interesting. But come on, given the chance to clone the JOB circuit used in the Goldmund Millenium amplifier (and other Goldmund amps), priced at $35,000, one would think that this would be a worthwile effort as well.
link to goldmund products using the JOB circ.
http://www.goldmund.com/old/Millennium/intro.html
http://www.goldmund.com/V4/catalog2...ultimate/29.htm
Plus the JOB is just as small as the GainCard, so it doesn't look too tough from a complexity viewpoint. Am I missing something? Does anyone have the circuit? A photo of the insides? Any JOBclones out there? |
|
|
| greg7 |
| The big question is, what is the Job circuit, anyway? I see no mention of IC's, nor to the internals pictures tell much. The Millenium link mentioned something suggesting IC's but who knows. |
|
|
| Schaef |
If you ask me, judging from the picture of the circuit, I would throw out a wild guess of some class-D circuit. This isn't based on any hard facts or data, but simply from looking at the picture of the board.
I see two power chips per channel (don't know what they are, transistors, power op-amps, something else totally different, whatever) very few components, and a central heatsink. Its the central heatsink that has me thinking a class-D. If I recall from, who was it? National or Analog Devices that sells the class-D chipset, that the controller chip could get hot. The biggest argument against this theory is the lack of chokes, but they could easily be on the bottom of the board. (Also the relatively low visible component count, but again, only the top of the board is visible)
Am I completely off my rocker with this thought? |
|
|
| mlloyd1 |
I'm thinking the JOB amp design just a very simple, highly tweaked/tuned, fairly conventional circuit. They hide it because they don't want you to figure that out. ;)
Bryston power amp designs products are often held up as an example of a fairly conventional circuit that has been tuned and tweaked well (I'm not so sure about "conventional" after looking at the output stage). You can go to the Brsyton web site and see the schematic. However, I have yet to see or hear about any Brsyton clones. And, folks seem to like the products.
Class D? I really doubt it. Did you really want us to tell you if you're off your rocker :clown:
So, who has one to cut open, take pictures and report? :devilr:
mlloyd1
| quote: | Originally posted by Schaef
If you ask me, judging from the picture of the circuit, I would throw out a wild guess of some class-D circuit.
....
Am I completely off my rocker with this thought? [/B] |
|
|
|
| lgreen |
Here is the image. I don't know that I would call it a switching amp, but it does look like there might be some stuff on the back of the board.
 |
|
|
| greg7 |
| What's the big heatsink in the middle of the board for -- regulators? Otherwise it looks like a typical discrete amplifier, probably MOSFET given Goldmund's heritage and the fact that MOSFET amplifiers can be made very simple. I know Redgum's amplifiers, for example, came right out of a circuit book and have only 2 gain stages and minimal feedback (from the horses mouth). They've tweaked and refined it, of course. |
|
|
| jam |
My guess is that it is a comp. current feedback amplifier.
Jam |
|
|
| audionline |
where is the circuit?
Thank you....... |
|
|
| lgreen |
| no one has the circuit, I think that's the point of the post! |
|
|
| CheffDeGaar |
Well, as far as I recall, the job circuit has been patented by a group of students, and licensed to Goldmund. From what I've read for yonks, Job seems to be a particular topology, very fast they say ;). It's always encapsulated in a small metallic case and potted. Seems to me that it has been produced in several versions, one for preamps, and others for amps, that should withstand higher currents/voltages (hence a heatsinked case as shown). Maybe these versions also include VAS ??. From photos I've seen in magazines years ago, the preamp version has a "normal" look (no heatsinks/fins).
But I've never never seen a schematic of it... :bawling: The other thing approching (but with other performances/figures) are Marantz's HDAM modules,and scheamtics can be found...
Sorry not to help at all :(
Edit : ANd it's absolutely not Class D. To be chosen between A/AB/B, but not more ;) |
|
|
| jam |
Cheff,
Can you get a copy of the patent for us? The bias of the amp seems very low for one with a mosfet output stage.
Regards,
Jam |
|
|
| Elso Kwak |
Hi , My best guess is that it is a potted module comprised of SMD parts with a yellowish cooler on top driving MOSFETs or bipolar output transistors....
I can no longer stand that hilarious high end talk on Goldmund's and Job's website. Became to sceptical after all those years....:mad: |
|
|
| jam |
Elso,
It is not my favorite type of sound , bit on the cool and dry side (maybe they need more bias?) but they make a great chassis.
Regards,
Jam |
|
|
| Elso Kwak |
| quote: | Originally posted by jam
Elso,
It is not my favorite type of sound , bit on the cool and dry side (maybe they need more bias?) but they make a great chassis.
Regards,
Jam |
Hi Jam, Don't be afraid. I am not a cloner, trying to build something original. Not easy with a poweramp..... |
|
|
| lumanauw |
| In post #5, why is the output transistor's leg is so long (this makes inductance?) and put ferrite bead on leg1 and leg3? What is the purpose of these things? |
|
|
| ionomolo |
The ferrite bead makes me think its mosfet for sure, and probably class AB. The output devices look like 2SK1058 & its girlfriend whose number i don't remember. If it was class D it would probably be made with ir mosfets. Under the heatsink... I don't expect a discrete input stage and the new drivers are too new, what's under the heatsink? New technology delivering 5 MHz... wow, sure those who have played with the OPA635 will perceive the increase in speed :smash:...
I bet there is an opamp there, it just fits (i don't see an input stage with 4 input pair transistors + 2 Vas transistors + 4 driver transistors there (Assuming no ccs load of either vas or input stage and no current mirrors for the input stage, that would climb to 16 transistors.)
EDIT: 0.003 % seventh and tenth harmonic?
That's more than 1% THD with (n-1)^2 weighting! Terrible! I don't know how to do worse... well we could have 10% tenth harmonic... |
|
|
| lumanauw |
How is this JOB sounds? Anyone has heard this module?
Goldmund use this technology for their amps? |
|
|
| banana |
| quote: | Originally posted by Elso Kwak
Hi , My best guess is that it is a potted module comprised of SMD parts with a yellowish cooler on top driving MOSFETs or bipolar output transistors....
I can no longer stand that hilarious high end talk on Goldmund's and Job's website. Became to sceptical after all those years....:mad: |
I was thinking the same, smd module board underleaf yellow heatsink......... until I opened my old SRA and checked closely.
That's just a cover, on top on some TO92 transistors.
The newer Goldmund 390 uses full SMD parts and no longer have this mysterious heatsink. |
|
|
| lumanauw |
Hi, banana,
How do you describe the sound of this amp? |
|
|
| ionomolo |
I have looked at the "expensive" version and i found it to be a textbook example of what people should never buy! Lots of expensive patches to solve problems that the designer should have never created. Perhaps the most laugh-provoking is the use of a pentium processor as "On screen control for AC line polarity, DC offset detection, Oscillation detection"
AC line polarity... i will quote from memory Douglas Shelf "audio signal is AC, so cables cannot be directional as long as 2 + 2 cannot equal five. Anyone prepared to belive this nonsense will never design an audio amplifier"
DC offset detection... wouldn't it be better to have an amplifier without DC offset?
Oscillation detection: This implies that the amplifier will oscillate. What does the oscillation detector? Show on-screen "The thing you payed 70K$ for is oscillating. Pity." ? |
|
|
| ionomolo |
| quote: | Originally posted by banana
I was thinking the same, smd module board underleaf yellow heatsink......... until I opened my old SRA and checked closely.
That's just a cover, on top on some TO92 transistors.
The newer Goldmund 390 uses full SMD parts and no longer have this mysterious heatsink. |
So no opamp there? I would like to know what's that new thing. I don't expect any sort of bjt to be able to react in picoseconds. |
|
|
| banana |
Hi lumanauw,
Mind is the entry level SRA, bought in year 2000, which only cost USD$1500 at that time. Not a bad deal for the price. Sound is neutral. Modern transistor style, no super warmth tube voice.
It uses the first generation JOB circuit, together with under size power supply, which can only provide 50W 8ohm and 75W 4ohm. Dynamic compression kicks in early in medium loud volume. Good for soft music only. Maybe better match with high efficiency speaker.
I also listened to my friend's JOB150. Dynamic is improved significantly. Treble and clarity is about the same as my SRA.
From a diyer's point of view, the same level of quality can be achieved by a standard non-inverting LM3886. Now I switched to Ucd400, much better in every aspect IMO. |
|
|
| banana |
I have some objective measurement for the 50W SRA.
THD vs Power, 8 and 4ohm load. Single channel driven. |
|
|
| banana |
FFT 8ohm 1W, both channel driven.
You can see the LF noise due to supply ripple.
Not bad at all, consider it's only 2000uF per rail.
Residue noise is about 250uV (20KHz bandwidth). |
|
|
| lumanauw |
Interesting, your FFT is cleaner than the one on JOB's website :D| quote: | | together with under size power supply |
I also noticed this. Their bridge rectifier (and cap bank value) is relatively small. |
|
|
|
