John Curl's Blowtorch preamplifier part II

[scott wurcer]

Interesting.

Funny typo in Table 7.

Yea, the package sucks for DIY too. I managed to get one down and found 25mV of offset so the match is so-so.

 

[bcarso]

Yea, the package sucks for DIY too. I managed to get one down and found 25mV of offset so the match is so-so.

With the pinchoff voltage thus and high Idss they might be o.k. for cascoding---except in that package the max dissipation is pretty small.

 

[gpapag]

Harold S. Black, Modulation Theory, Van Nostrand, 1953. Pre-ISBN days

Thank you Brad

George

 

[EUVL]

My problem with the PMBFJ620 is the high Idss, coupled with low Yfs.
So if you try to degenerate to get the bias down to reasonable level, you end up losing more gm.
Plus dissipation is limited.

Then I rather use matched BF862.


Patrick

 

[dvv]

These are well distributed and cheap, at least worth a look. Same AM radio reference app. so gm is good and noise probably pretty good. All 6 leads brought out, but micro SOT only.

http://www.nxp.com/documents/data_sheet/PMBFJ620.pdf

Good tip, thank you Scott.

 

[EUVL]

I thought I mentioned thee before, but since we are on the topic :

CPH6904 (6 pin)
www.onsemi.com/pub/Collateral/ENA1966-D.PDF

MCH5908 (5 pin)
www.onsemi.com/pub/Collateral/ENA1218-D.PDF


Patrick

 

[31697B]

Blues -- get: The Breeze.

An appreciation of J.J.Cale.


Travel .... never go with a group... even if you dont know how or where you are going exactly.... is my motto. One time, recently, I got excited when I found a way to get to XIAN, China. There wasnt a way from Bangkok.... so I went up north to Chang Mai and there was a plane to XIAN from there...... I as so excited, I bought tickets as soon as I landed in C.Mai and took off again. Only one problem, I did not think about...... dropping into the middle of China without a Visa can get you a long term in a prison in some remote, unknown place. But it worked out great and I got a private tour of the region by the daughter of Xian's chief of police. She showed places no one outside China had seen and was an expert in history. One of the best times I ever had - which could not be had without just 'going for it' on your own.


THx-RNMarsh

 

[gerhard]

These are well distributed and cheap, at least worth a look. Same AM radio reference app. so gm is good and noise probably pretty good. All 6 leads brought out, but micro SOT only.

6nV/sqrt Hz typ @ 100Hz.
It consists of 2 chips that are isolated and sit on their gate pads,
as should be expected. I just cut one open.

Why can't they do that with selected BF862 ???

I also just found a piece of MMBF5460LT1G tape which is one of the few surviving
p-channels. But 60 to 115 nV/sqrtHz. That hurts.

THE GOOD NEWS:
I just got my copy of Hill's Art Of Electronics release III
That cold & rainy Easter weekend is saved!

Gerhard

 

[Esperado]

An appreciation of J.J.Cale.
Travel .... never go with a group...

He obviously apply this principle to his musical life: A different city with different musicians for each tune. And he visibly get bored pretty fast by their presence: each tune is very short ;-)

 

[bogdan_borko]

Any other NEW ideas to make better amps and preamps?
In all seriousness, we are still looking for different ways to make a better sounding amp or preamp. Many of us have found that exotic approaches, such as using feedforward and feedback together, can LOWER distortion, allowing one to start with a higher distortion gain stage, fixing it with advanced feedback techniques, and while measurably commendable, don't necessary improve the sound quality, and in fact, might compromise it.

John, when you were designing Parasound Halo JC 1, why didn`t you start as making it as linear as you could, in open loop (without global feedback loop), and then raise the gain and close the feedback loop, I get the feeling that you relied on the fb loop from the start or am I wrong? BTW you could get that distortion figures for THX or whatever with much less GNFB and better design...
Why don`t you experiment with topologies a little bit, your last amp is same as your first, just with little bit upgraded schematic and new parts.

 

[jan.didden]

Slightly OT (if that is possible here):

I need to get a stripper for my rework kynar wirewrap wire. Wire outside diameter with insulation is 0.4mm, the wire itself is 0.2mm.

Do I get a 0.4mm stripper or a 0.2mm stripper? I looked at a bunch of datasheets but they do not tell what the specs refer to.

Thanks, Jan

 

[rayma]

I need to get a stripper for my rework kynar wirewrap wire. Wire outside diameter
with insulation is 0.4mm, the wire itself is 0.2mm. Do I get a 0.4mm stripper or a 0.2mm stripper?

Go by the wire size, 0.2mm.

 

[john curl]

Bogdan, thanks for your input. I do try to make a specific design as linear as possible, always. However, I don't 'cut and try' and 10 years ago, I did not bother with Spice simulation, either.
I use a design topology that I first developed over 40 years ago, and just make it bigger and better.
Parasound had already made a number of amp models following this basic topology, with good success, most of the time. However, we found that two other areas of design had to be seriously included to make a really optimum design. First, was upgrades in parts quality, including caps, wire, and connectors, diodes, etc.
At the same time, we wanted an optimum layout from my former Vendetta and then CTC business partner, to do the circuit layout. We had to put a lot of energy into just getting what we got done, and I deliberately stuck with an 'improved' version of the HCA-3500, to see what these changes in layout and parts quality would give us.
I have done other amps that would cost even more than the JC-1, if produced today, but Parasound was not set up to pay me enough to develop them, although I had developed 3 working models in the past for other manufacturers, so we stuck with the standard topology.

 

[cliffforrest]

I have used one of these for many years:

WIRE STRIPPER AND CUTTER ST-100 SERIES OK INDUSTRIES / RS 543-771 | eBay

 

[scott wurcer]

Slightly OT (if that is possible here):

I need to get a stripper for my rework kynar wirewrap wire. Wire outside diameter with insulation is 0.4mm, the wire itself is 0.2mm.

Do I get a 0.4mm stripper or a 0.2mm stripper? I looked at a bunch of datasheets but they do not tell what the specs refer to.

Thanks, Jan

I have an adjustable one (no longer listed) similar to Knipex, works great on small wires but pricey (you have to pay for German engineering).

 

[ostripper]

So many JC-2 clones .....

I had lots of iterations of this most copied circuit. I suppose the fact so
many Chinese kit makers copy is the ultimate "flattery".

I could not get the basic JC-2 below .01% THD20. The Jfets and it's utter
simplicity might be the reason for such a "following".

Going from the 200W amps to a class A line-stages was fun. I could run
everything hotter , use low Vceo/low noise semi's .... and have a class A
output. 😎😎

Symmetric is the way to go ... the leach/hawksford circuit did 2ppm-20K !
(below)
Sub -PPM at 1k - a nice 5X "wire with gain".
A Jfet cascode did the same (almost). Jfet models are crap - still <5ppm.
The hawksford was the real change that made the "bottom drop out"- THD
wise.

I'm going to build it ! (any ideas - never built line stages , just big amps)

OS

 

[brianco]

Any other new ideas on proving that the earth is round?

Is it?

A guy attempted top prove the same contention to me whereas I stuck to my guns, as it were, on the fact that it is an oblate spheroid...........I gave up when he admitted to being a subjectivist!


🙄

 

[dvv]

When carried on too far, subjectivism can produce some funny effects. For example, a lot of local DIYers have come to believe that current feedback is the not just THE thing, if it's not current NFB, then it can't sound good. They forgot real fast that until a few months ago, they were swearing on their voltage feedback gear and hand products.

 

[Max Headroom]

I'm going to build it ! (any ideas - never built line stages , just big amps)
OS

Ok, so how does the sim report with variable source impedance...ie, connected to the wiper of a volume pot, say 5k, 10k, 20k, 50k at differing pot positions ?.
How 'bad' does the 0.01% JC-2 actually sound in practice ?.

Dan.

 

[31697B]

I had lots of iterations of this most copied circuit. I suppose the fact so
many Chinese kit makers copy is the ultimate "flattery".

I could not get the basic JC-2 below .01% THD20.

Symmetric is the way to go ...

A Jfet cascode did the same (almost). Jfet models are crap - still <5ppm.
The hawksford was the real change that made the "bottom drop out"- THD
wise.

I'm going to build it ! (any ideas - never built line stages , just big amps)

OS

Some thoughts from following along your journey in topologies ---- one thing seems never got resolved was the Slew Rate and what differences it made to subjective results.... using very low thd (<-100dB) and same topology. High SR vs minimum (.5v/v - 1v/v).

jFET models are crap and so are many compliment jFEt parts available. A real problem for symmetrical topologies..... causing circuits to get more complicated for low distortion.

Never- the-less, selection of jFETs in the simple circuits is absolutely a requirement for lowest distortion in practice. Relatively simple circuits have produced -120dB levels (measured) with selection and trim. This isnt practical for mass production but it is for DIY or limited production HiEnd construction products.

Since, you, myself and others have shown you can get 'The Numbers' in many forms of circuitry, the circuitry/topology with the fewest parts while maintaining highest performance levels, wins IMO.


THx-RNMarsh