Need a pair of eyes on a spectra sonics 101 vero layout

[pH]

Finally got around to tracking down the problem. Like I said, there was no voltage flowing through the circuit, checked all resistor values which were ok so I started pulling transistors to test with my multimeter and found one of the 2N4248s was shot. Just got some NOS replacements and dropped one in last night without even testing for hfe and everything works fine. I played around with these on drums and really liked what I heard.
Paul

 

[Whoops]

Finally got around to tracking down the problem. Like I said, there was no voltage flowing through the circuit, checked all resistor values which were ok so I started pulling transistors to test with my multimeter and found one of the 2N4248s was shot. Just got some NOS replacements and dropped one in last night without even testing for hfe and everything works fine. I played around with these on drums and really liked what I heard.
Paul

Nice one,
happy that you manage to solve the problem

 

[craigmorris74]

I saw numerous references to the diodes in the 101/110 circuit.  Are those the built in peak limiter that the 610 manual refers to the `01 having?

Thanks,
Craig

 

[mutterd]

I saw numerous references to the diodes in the 101/110 circuit.  Are those the built in peak limiter that the 610 manual refers to the `01 having?

Thanks,
Craig

No - the diodes are used to create a voltage divider with R4 (4.99k) between +24V and 0V for biasing one of the transistors of the 2nd stage...

the 101 limiting mentioned in the 610 docs comes from running more than -22dbm into the 101 so you are up against the +18dbm max output of the card... once you hit that point no matter what you put in - you'll never exceeded that output level.

the magic comes in that a properly tuned Spectra 101 starts to symmetrically clip right at +18dbm - they are amazingly consistent - and because of the transient response and overload recover you can run pretty far into the red before things start to really sound like your in the red...

 

[craigmorris74]

No - the diodes are used to create a voltage divider with R4 (4.99k) between +24V and 0V for biasing one of the transistors of the 2nd stage...

the 101 limiting mentioned in the 610 docs comes from running more than -22dbm into the 101 so you are up against the +18dbm max output of the card... once you hit that point no matter what you put in - you'll never exceeded that output level.

the magic comes in that a properly tuned Spectra 101 starts to symmetrically clip right at +18dbm - they are amazingly consistent - and because of the transient response and overload recover you can run pretty far into the red before things start to really sound like your in the red...

Thanks for the reply.  That was the answer I was looking for.

Craig

 

[groselicain]

Maybe a little OT, but has anyone been in touch with Richard recently? I paid for membership to his site and emailed him about schematics to no avail.

Thanks.

 

[Whoops]

the magic comes in that a properly tuned Spectra 101 starts to symmetrically clip

Hello Mutterd,
What would define a "properly tuned 101"? How can you fine tune this circuit?

Thanks

 

[mutterd]

By "fine tuned" I mean 2 things:

1. the card must be factory set.

its a complex little thing, the 101, and I have bought hundreds and hundreds of them, and the only time I have seen things go wrong is when people start swapping out components...

with the exception of the input and output electrolytics, every component in a vintage 101 or 110 is a precision component and was confirmed in circuit on the bench, there are no accidents here, you can't swap out a transistor here or there, or swap an electrolytic for one of the tantalums - it just doesn't work...

now that being said - i have crates of these things and unless people have gotten creative with component substitution they all still work and meet original factory spec - once the in and out 'lytics are replaced.

2. the card must be properly powered and properly grounded.

GROUNDING IS KING IN THE WORLD OF SPECTRA - I can't say this enough - there are a few absolute rules, and the rules MUST be followed. To start, you cant run wires from random points on the card to some star ground, you can't have significant impedances between the junction point of input - and output -, and you have to have a quiet B+ rail - better than -60dB...

now if those conditions are met - these cards are shockingly consistent - and you can drive right up to +18 with even clipping and in listening tests, nothing sounds like that... and transient response - forget about it - nothing comes close.

hope this helps.
T.

 

[Whoops]

2. the card must be properly powered and properly grounded.

GROUNDING IS KING IN THE WORLD OF SPECTRA - I can't say this enough - there are a few absolute rules, and the rules MUST be followed. To start, you cant run wires from random points on the card to some star ground, you can't have significant impedances between the junction point of input - and output -, and you have to have a quiet B+ rail - better than -60dB...

Thank you.

Do you  have any info or diagrams on how to properly ground this boards?

The symmetrically cliping at +18dbm is also true to the 110A cards?

 

[MidnightArrakis]

I think the input pad as described in the JLM schematic is not a good idea. It's used before the transformer, a dual gang Pot will have too much tolerance in each gang and will mess around with the common mode rejection. Pots can have tolerance from 10% to 30%.
But let me know if someone has a different view on this.

You will be better with this kind of arrangement (Richard Guy recommendation):

Since both positions "1" and "2" of the shown switch do not have anything connected to their terminals, I then do not understand how -- Switch Position "2" -- offers a -6dB pad of the signal. What am I missing?

 

[MidnightArrakis]

component side

Whoops: Do you still have your Spectra-Sonics 110A circuit boards? If so, could you possibly create and send me a detailed mechanical drawing of all of the important-to-have dimensions? This would include the curved radii dimension on either side of the edge-fingers and the exact dimension from the bottom of the PCB to the centerline of the bottom edge-finger copper. I have read somewhere else on here that the edge-fingers are on a 0.156" spacing, so that's covered. Also, about an hour ago I found a source for the 1N461 diodes here in the States. Since this distributor specializes in hard-to-find components, I put in an RFQ request for all of the 110A semiconductors in quantities of 100-each. Being the July 4th weekend here in the U.S., I probably won't receive a response until the middle of next week sometime.

THANKS!!! for all of your help!!!

/

 

[Whoops]

Hi,
Yes I have two 110A boards.
I can take dimensions no worries.
I have some scans of the boards which you can use to print and home etch some pcbs.
It’s really simple PCB, very easy to home etch and DIY.
Send me a pm explaining the dimensions you need

As for the 1N461 diodes, I doubt those are critical for the circuit, you can probably use many replacements instead of them.
Anyway you can also find the 1N461 in Ebay