[EchoWars]

Hi guys! Very nice to have a forum here for Jim's stuff.

Anyway, to the point. Just finished the electrical part of the Piccolo build, and except for some brain-fade that made me install the LED backwards (which was a mini-disaster, what with having to desolder the battery box and get the LED out...love those through-plated holes ) everything looks pretty decent for the bench check. Only...

...the negative supply, when powered by the wall-wart (have not tried batteries yet) is only about -3.86V as it exits the TC7662B, and drops to about -2.81 by the time it gets to the 'far side' of R20. I doubt if this is a problem, was just surprised to see it off by 200mV. Positive supply looks great, BTW.

I have done the 'pin 1' trick on the TC7662B, but the scope on pin 2 or 4 or 7 shows a frequency of only about 3.5KHz. I suppose that the high-frequency cannot be confirmed externally?

And...reading the data sheet for the TC7662B, I find:

The on-board oscillator operates at a nominal frequency of 10kHz. Frequency is increased to 35kHz when pin 1 is connected to V+, allowing the use of smaller external capacitors.

I see nothing in the data sheet about connecting pin 1 to ground...seems it is supposed to go to V+ if you want a higher frequency, and should be shorted to pin 8 instead. Someone (Jim?) want to clarify?

Lastly, DC offset. Certainly most phono stages that this is meant to drive are capacitively coupled (BJT's). But, my favorite stages are FET input. The Piccolo shows offset of about 12mV on one channel, and 22mV on the other. Now, FET input phono stages generally have a 1K~5K resistor in series with the input...would it be correct to assume that the small amount of DC generated at the output of the Piccolo is dropped across this resistor? Just asking, since there's more DC at the output of the Piccolo than 5x the RMS output of most MM phono carts (at least on one channel)...

Its been fun...and once things are straightened out and I have it mounted in the box, I'll post a few pics.

[hagtech]

it is supposed to go to V+

Holy inversion, Batman!  You're absolutely correct.  Gads, I can't believe I still make these kinds of mistakes.  I even remember checking this out on a scope to make sure all was ok (and then promptly forgot what I did).  Instead of dyslexia, I must have polarexia.  My entire career has been plagued with inversion errors.

Anyway, thanks for catching this!  I will update manual.  Yes, the trick is to connect pin 1 to pin 8. 

jh

[hagtech]

As far as dc offset, yes, the PICCOLO can have a relatively large up to 20mV dc offset (per LSK389 matching spec).  This can be a problem for dc coupled phonostages.  No problem with tube stages.  I need to add a warning to the manual and web page.  The BUGLE has the same thing.  I do a dc cancellation using a coupling cap, but then buffer that with a low offset stage.  This limits the sonic impact of the capacitor (having it do less work) resulting in lower smear.

jh

[EchoWars]

Looks like I'm gonna remove the battery pack one more time again...

RE: DC offset. Any reason why I should not replace the 221 ohm R10 with a 500 ohm Bourns 25-turn trimpot to allow me to zero out offset? It'll be a tight squeeze, but it's possible. I'd love to see offset sitting at about a millivolt or less.

So I assume that the -2.8V measured at the gates of Q2A & B is not an issue...?

Also meant to mention that the height of the front panel means that the DC wall-wart jack could not make contact with the PC board. I had to clip off the 'thin' part of the solder lugs on the jack, drill a tiny hole in the 'stub' of the lugs, and soldered on short sections of solid-core copper wire stripped from some Cat-5 cable I had. This allowed me to firmly mount the jack onto the lid, and feed the wires into the board & solder 'em. Worked out fine...

Jim, I think you've done a wonderful job here. Much appreciated.

[EchoWars]

Got the battery pack back out again...this time it lifted the pad from the '-' side, on the side of the board opposite the components. Pad looks fine on the component side, so I don't think this is an issue.

Shorted pin 1 to pin 8. Powering with the wall-wart, the square wave at pin 2 and 4 went from about 3.5KHz to about 18.8KHz. Pin 7 went from about 3.8KHz previously, to about 24KHz after. Frequency still isn't up to Microchip's specs, but it's a lot higher, and hopefully that's a good thing. 

Output voltages from the 7662 dropped. The voltages on either side of R19 and R20 are now 7.2V and 6.87V for the positive side, and -2.94 and -2.61 for the negative side. But offset was cut in half (likely due to the new current value through the Q3B current source). The channel that was 22mV dropped to about 12mV, and the channel that was at 12mV dropped to about 5.5mV. I'd still consider changing R10 to a multiturn trimpot.

Also noticed that the Piccolo isn't terribly well-behaved at power-down WRT DC offset. Power off results in over 500mV of offset till the caps bleed off. Turning the unit off while connected to a phono input seems to be ill-advised, and worrisome for FET input phono stages. Even with power off on the receving phono stage, I'm concerned about 1/2 a volt forward biasing a FET gate and taking it out (or is this a tempest in a teacup?).

[hagtech]

Sure, you can try trimming out dc.  I might look into this more myself.

jh

[hagtech]

Ok, I ran some experiments on my own PICCOLO.  There is a slight device mismatch in that I operate the two at different drain voltages (4V versus 8V), which caused a higher shift in Vgs than I anticipated.  One way to tweak this in is to add about 10 ohms to R10.  Change them both from 221 ohms to roughly 232 ohms.  At the moment, I am just putting a spare 10 ohm resistor in series with the 221.  Should be a way to add trimpots for almost perfect nulling (at a particular temperature), if you are ambitious.

jh

[EchoWars]

I gots some 500 ohm Bourns multiturn pots on the way.   I do believe I can squeeze them in there. Would have been real easy to stick them on the back of the PC board, but the battery holder is there.

When this one is completed, I may see who's interested in auditioning it at the Audiokarma forums (which is where I can usually be found), and build a second one that stays here at home. On this second Piccolo, I'd move the RCA jacks to the ends of the box instead of the top, and mount the battery holder in the bottom of the box with a removable connector...which would allow me plenty of room to put an offset trimpot on the back of the board.

Might be able to get you a little exposure doing that.

[amandarae]

it is supposed to go to V+

Holy inversion, Batman!  You're absolutely correct.  Gads, I can't believe I still make these kinds of mistakes.  I even remember checking this out on a scope to make sure all was ok (and then promptly forgot what I did).  Instead of dyslexia, I must have polarexia.  My entire career has been plagued with inversion errors.

Anyway, thanks for catching this!  I will update manual.  Yes, the trick is to connect pin 1 to pin 8. 

jh

Same is true for the TC7662A?

regards,

Abe

[EchoWars]

The TC7662A does not have the ability to boost the internal clock frequency like the 7662B, and the data sheet shows pin 1 as 'NC' (no connection), so I certainly would not connect it to anything (on some IC's [I dunno about this one] 'NC' pins are connected to the device substrate, and it may be a bad thing to connect it to V+).

Personally, I can't think of a reason where I'd use the older 7662A.

[amandarae]

The TC7662A does not have the ability to boost the internal clock frequency like the 7662B, and the data sheet shows pin 1 as 'NC' (no connection), so I certainly would not connect it to anything (on some IC's [I dunno about this one] 'NC' pins are connected to the device substrate, and it may be a bad thing to connect it to V+).

Personally, I can't think of a reason where I'd use the older 7662A.

That was the part listed on the manual.  I built mine bone stock and I have been using it for many months now.

regards,

Abe

[EchoWars]

If it's workin' for you, then you're stylin'.

[EchoWars]

Updated image for the corrected Piccolo PDF.



Others may note that the corrected PDF reads:

If you do not build in the recommended stock chassis configuration, you can reduce the potential for high frequency noise contamination by forcing the dc-dc converter to operate at a higher frequency. Connecting pin 1 of U1 to pin 8 does this. Add a jumper on bottom of board (only works with the TC7662BCPA).

...but the image is gone in the newest PDF, so I offer one here. 

Anyway, I still don't have the trimpots to replace R10 (L & R) to try and zero out the offset. I needed to wait till I had enough parts to make an order worthwhile, but should have that stuff in before the weekend is out.

[EchoWars]

Parts came in today, and since there's obviously such rapt attention focused on this mod ( ), I thought I'd update the thread.



Here's the Piccolo with two 500 ohm / 25-turn Bourns trimpots replacing R10 (L & R). Because R11 sits so close to it, and because I ordered the version of the pot with the offset center lead, AND because the 100µf 25V caps block access to the adjuster on the right side, AND because the battery holder prevents mounting the trimpots on the backside of the board, this is how it was done...with one adjustment screw facing one way and the other in a reversed direction. With the top of the box mounted, a long skinny screwdriver is necessary to get to the adjusting screw on the right side, but it's perfectly do-able. Offset easily adjusts to 0.00mV and is perfectly stable (but will change slightly [a mV or two] when changing from battery to wallwart).

If I build another one (and that's a good possibility), I'll use a different battery box that is remotely mounted so the adjust pots can be mounted under the PC board. Or...maybe Jim will change the design.