First official Bottlehead Quickie upgrade - the PJCCS

[Doc B]

PJ and I had a brainstorming session yesterday. One of the more fruitful storms was a neat constant current source circuit that PJ had cooked up. The most unique feature is that it is designed to compensate for the battery B+ supply voltage dropping over time as the batteries drain with use.

It's a fairly simple circuit that we easily fit onto the same SRC4S PC board that we use in the Foreplay III Extended upgrade kit. You just stuff a board with about 10 components, add 6 wires and install it in place of the 4K ohm plate resistors. The board is very lightweight so we just let the prototype hang by the connecting wires. There are mounting holes in the board, and I'm sure one could use one hole to attach the board to a standoff that could thread onto one of the tube socket mounting screws.

I had mentioned in another thread that I tried a standard C4S in the Quickie with no real improvement, and not the best bass. Stupid me! I was using a Quickie that had the Speco transformer mod for high impedance headphones, which is about 3dB down at 80Hz, as a line stage. And I was using some cheapo cables that seemed to make the light bass even worse. PJ also pointed out that the C4S will not compensate for the drop in B+ voltage as the battries drain, so it's not the best choice for this particular circuit.

With this new current source installed in a non-Speco'd Quickie and using decent cables there was a "Whoa!" reaction from all three of us when we compared against a stock Quickie. Gain was a little higher with the PJCCS in place so we compensated our listening level accordingly. At the same level as the stock Quickie the improvement in image size and location, and the improvement in resolution from the modded Quickie was very audible. I also noted that the bass seemed to show more harmonic structure. Since a battery power supply is dead quiet already this isn't due to the high power supply noise rejection of the CCS. So the sonic improvement seems due to the distortion being lower at any given output level.

This went together so well and so easily that it became an official kit during the cocktail hour last night. We will sell a kit of parts and simple one page instructions for $35.

 

[Grainger49]

First, a real tip-'o-the-hat to the crew for concocting a new product so quickly.

Reading the post I wondered if there was lubrication during the design stage but you mentioned that cocktail hour followed.  I guess I was wrong.

I'm going to smoke test a BH reg board this afternoon.   There will be more.

Congratulations on a truly "Quick Delivery."

 

[booangler]

Out Stinkin' Standing,

Can't wait to order another quickie and the board. Is there going to be a package deal?

Alan

 

[Len]

I was using a Quickie that had the Speco transformer mod for high impedance headphones, which is about 3dB down at 80Hz, as a line stage.

Tres cool!

You know I'm in. The Hammonds will go on the headphone amp.

I'll be using the PJCCS in my low pass filter Quickie that needs 500 ohm output impedance, so I'm using the Specos there too. Do you know of a high quality 4K:500 output transformer or have any plans for working with Mikey on one?

 

[Doc B]

Is there going to be a package deal?

\


The price is already so low that we make only a tiny profit on these as it is, and I don't think we are able to offer a lower price for the combination. But of course one would save on shipping by ordering them together. Give me a little time and I will get the order buttons up on the website. Meantime, if anyone would like to order a PJCCS kit call or email Eileen - 360-697-1936 or queen at bottlehead dot com. It will take a few days to get the assembly instructions completed and then we should be able to ship

 

[Dr. Toobz]

So, would these be an improvement over plate chokes? I'm using the Quickie with Specos as a headphone amp, and don't get the impression that current sources would make much of a difference if the Specos were still being used as OPT's. As a proper preamp, it sounds like they would. Correct?

 

[ashok]

Give me a little time and I will get the order buttons up on the website.

Hi Doc,

While you are adding the order button for the PJCCS, could you also put in a button for ordering the Seduction base with the Quickie?

Thank you,

Ashok

 

[Doc B]

could you also put in a button for ordering the Seduction base with the Quickie?

Yes, thanks for reminding me!

 

[Paul Joppa]

... I'm using the Specos there too. Do you know of a high quality 4K:500 output transformer or have any plans for working with Mikey on one?

Yes, I have asked Mike about the B7 parafeed line output, in either 5K or 8K primary, 500CT secondary (with a 125 ohm tap available). It would cover a wide range of headphones as well as being a decent driver of 500 ohm loads, balanced or single ended. If you have the bucks, this can even be stacked with 80% nickel laminations for extremely high inductance and low distortion.

I am working on a set of operating points and circuits for alternative plate loads and transformer outputs. Mostly it's just compiling the many ideas that have shown up here on the forum, and tweaking the operating conditions for them.

More on that when I have more to offer, but meanwhile it looks like the optimizing comes down to two special cases:

1) maximum power output per volt of input occurs when the load equals the plate resistance, i.e. a 4K plate load. Operating point is approximately 25v at 2mA (the other 11 volts is dropped in the plate load and the 1K cathode bias resistor.)

2) Maximum power output for a given plate voltage occurs when the load impedance is twice the plate resistance, i.e. an 8K load. At 25 volts as before, the plate current would be 1.2mA (cathode resistor about 2K)

 

[Doc B]

PJ asked me to mention that since this current source compensates for the B+ changing as the battery runs down PJCCS stands for the Paul Joppa Compensating Current Source.

 

[booangler]

PJ asked me to mention that since this current source compensates for the B+ changing as the battery runs down PJCCS stands for the Paul Joppa Compensating Current Source.

All Hail​

The Paul Joppa Compensating Current Source


I can't wait to try it... come on Hanukkah

Alan

 

[John Swenson]

Hi PJ,
I'm not quite sure of the reason for a special CCS here. A regular CCS should automatically provide the needed voltage drop necessary to present the plate with the needed voltage for the OP. For a standard C4S the only thing I can think of that would cause any change depending on the B+ would be the current in the LED string that would make a slight change in the voltage across the current set resistor which would of course cause a small change in the plate current.

Of course you might want to go with a completely different CCS design in order to get lower minimum voltage drop across the CCS, but even so I still don't see why a regular CCS design won't work.

Not to disparage the design, I just don't see why a regular CCS design wouldn't work.

What am I missing?

John S.

 

[Paul Joppa]

When the battery is fresh, the B+ is 36 volts and the plate voltage is about 28 volts, bias voltage 2 volts at 2mA plate current. That leaves 8 volts compliance for the plate to swing  positive and negative. 

But an alkaline cell voltage drops linearly as it is discharged, with end-of-life voltage about 2/3 of the initial voltage, i.e. 24 volts for the Quickie B+. By the time the battery drops to 29 or 30 volts, i.e. half its energy has been used, the remaining compliance of 1-2 volts is not even enough to keep the current source running, much less allow for output voltage swings.

By dropping the current to 1.3mA or so when the battery drops to 24 volts, the plate voltage will drop, to maybe 18 volts. That leaves 6 volts compliance and the preamp continues to work.

Of course a regular current source will work well if you use NiMH or NiCd batteries, which have a fairly constant voltage throughout their discharge cycle.

 

[Dr. Toobz]

Is it worth bothering with these if we've already loaded the plates with chokes, though? (I mean, in place of the chokes - are CCS's better for sound quality?).

 

[Koda]

Here is another "what if" question.

Is it worth bothering with these if we've already loaded the plates with chokes, though? (I mean, in place of the chokes - are CCS's better for sound quality?).

What if Nicad's were used for power, would there be an advantage of the PJCCS over a choke?
I can almost hear the response of "try it and let us know" but if it is the same and I already have the chokes in place...
  Thanks, and I really do apreciate the help and education everyone provides!
              K.

 

[Doc B]

I guess the best you can do at this point is compare my description of what I hear against what you hear, as I have not tried the chokes. I heard much better definition of notes and the image became much more clear and realistic in terms of proportion and sense of depth. A hunch is that it is probably doing more good stuff since the impedance of the choke is lower than the current source, and it appears that what we are hearing is primarily due to a lowering of distortion at any given level vs. the stock setup - a result of the high impedance load.

I have been forming a theory over the past 15 years that a lot of folks prefer a particular change to a tube circuit because it may help the bottom end or the image or whatever and doesn't change the treble, and part of the reason they went to tubes was to smooth out an edgy digital setup. IME a CCS will make the treble more quick and concise than a plate choke. Since a lot of listeners do use digital front ends, that change in the treble definition that I consider an improvement may not be to their taste because it may tend not to help downplay any digital mid-treble edge coming from upstream.

That is perhaps a circuitous route to say that I think a CCS often does more good for the sound than a choke, and that any problems that are upstream should be addressed upstream rather than trying to iron them out them after the fact. BUT I'm also saying YMMV.

 

[Paul Joppa]

...
I can almost hear the response of "try it and let us know" ...

I see I may have been too consistent in my responses lately ...  :^)  It's not an experiment I have done. But here's one I did do:

Some years ago I built an experimental parafeed preamp with a 6BL7. I had some tapped chokes on hand, whose basic inductance at the tap was about what I think is right for those impedances. end to end, I could get 20% more inductance. So I added a second choke per channel and wired it all up to a switch - I could select 100%, 120%, 220%, and 240% of the "right" inductance. Sure enough, every increase of inductance was an improvement in the sound. Doc B wanted to swap in a C4S, but that would have changed the operating point (plate voltage specifically), making it hard to interpret the result reliably, and I never got around to resolving that issue.

I conclude more impedance is better. Now the PJCCS is not a cascode (it would rob too much of the limited available compliance voltage) so it is not as high an impedance as a C4S. I don't know the Early voltage of the MJE350, but assuming it's around 50v (typical for transistors) the dynamic impedance would be 25K ohms. A 150 henry choke is 19K at 20Hz, 190K at 200Hz, and 1.9Meg at 2kHz unless its capacitance is dominant at the frequency. So it's possible the choke would sound better than the current source.

 

[Doc B]

OK folks, the manual is getting written today and the order button is ready to push!

http://www.bottlehead.com/store.php?crn=220&rn=441&action=show_detail

 

[booangler]

Hot Damn!!!

 

[Doc B]

Got behind a little, but I finished the PJCCS manual yesterday and tomorrow we will be getting out all the PJCCSs that have been ordered so far. Thanks for those orders!