New API 525 build thread.

[jplebre]

might have been me but:

- In/Out pots: they are labelled as CW being ground. it looks like as if turning CW on the schematic would smash the i/p against the threshold and would increase make up gain. The way it's suggested on the picture Peter posted (the one I downloaded anyway) it says ground pads (top 2 vias) are ground and are CW, whilst the bottom 2 are CCW.

this will make the pot perform the other way around (maximum make up gain is ccw, turning left.

Call me old fashioned, but I preferred to reverse this for both pots

Oh, and you really NEED that log 1k. I had put linear ones just to get things going and the useful range is from about 9-11o'clock. Then the threshold goes BUUUUUM!!!!

- Attack pot: I may try a 500k pot here. On the scope it seems as if towards the end of the travel the release starts getting severely affected (some weird release curves). That might make the pot more usefull as well? (glad I've installed a molex connector for the attack pot will make swapping a breeze

 

[qmp audio]

I'm a little confused as to what the second value represents in the BOM.  e.g.: C1 250UF/6V (470UF/16V).

 

[jplebre]

first value is original, the (value) is the one that Peter had at hand and put on his prototype

 

[qmp audio]

thanks jplbre.  Can anyone point me in the right direction for Q1 & Q2 stands?  There is a note that they must be on stands... but unsure to what that means.  Thanks

 

[jplebre]

something like this will do http://www.mouser.com/ds/2/273/123N-58560.pdf

 

[qmp audio]

Thank again jplebre.

http://www.mouser.com/ds/2/244/SSL-LX3059IGW%20R-92825.pdf

I've chosen the above main pcb LED (pdf).  If i understand correctly, the LED forward voltage is 2.6V max and 20mA, and the supply is 16V, so the proper RLED would be 680Ω.

or if im doing this correctly:

~680Ω = ( 16V - 2.6V ) / 20mA.

Just hoping to verify this and make sure i have the #s right... learning.

 

[jplebre]

I got a much brighter LED, I limited it to 10mA and it still blinds everyone in the room w a 1.8k resistor.
You might be fine though https://www.rapidonline.com/pdf/56-0346.pdf 30mcd vs 5000mcd

if you go for 680ohms go for a 1/2w (dissipating ca. 260mW to 280mW)

 

[Toure14]

Peter,

Anymore kits left?

 

[jplebre]

give him a shout here: http://www.lazpro.com/

He's a good fellow replies to email quicker than through here

 

[Toure14]

Thanks

 

[jplebre]

Hi guys

one thing. I was having problems with the attack control.
I found  removing the pot and shorting would be pretty similar to the original, but I couldn't get the same results.

So, I was thinking, and I found out that what I think we should be seeing would be this.

Doing that extra connection would allow the connection to be exactly as the original in fully CCW position (eg. C3 seeing both CV and R5 without going through the 1M pot) and then at any point of the pot C3 will only see half of the potentiometer, which should regulate the rate of charge.

Basically, shorting clockwise lug/cable and wiper. Takes 2 seconds to do either on pot or PCB and in my case the attack now behaves really well.

Let me know if this help anyone, or if you foresee problems with this. Ran a whole bunch of material and so far so good.

Anyone thinks that the LED's are too slow? (eg. almost no GR shown at very fast, some then shows a lot, then less... seems like it can't pick up enough for the first half of the wiper as it's too fast, and then starts tracking correctly).

Cheers
J

 

[qmp audio]

Will bc550cta and bc560cta work?  They look similar except the 150mhz fT

 

[jplebre]

what are you trying to replace?

 

[qmp audio]

what are you trying to replace?

op amp transistors.

In the meantime, before i finish the opamps,  I filled my 2 units w/ known gar2520s that i have.  Unfortunately both of my units do not function correctly.  They both act identical, so whatever i did, i did the same to both of them.

The unit passes audio when in off mode ( but about 9db down... maybe not calibrated )
When i turn on.  There is barely any signal being passed.  I can hear it but its very quiet.  Like the signal is going to ground perhaps.  Im second guessing my transformer wiring.  Can anyone verify if this is correct or wrong?  particularly the sidechain.

2623 http://www.classicapi.com/catalog/images/gallery/2623/2623-1-specs.pdf

 

[jplebre]

I used 550C for Q3, Q6, Q8 and 560C for Q4, Q5, Q7

as for the transformers they are wired differently. O/p is in series, but sidechain is not.

Also, mind the pinout. you currently have in-phase (dot side on the transformer schem available at classic api) on the minus, and vice versa.
I can't remember the PCB printout for the sidechain but it seems odd to me you might want to check that out. The o/p one seemed fine only out of phase

 

[qmp audio]

Finally got the parts to finish the laz2520s,  and finished my pair!  i did the attack mod as suggested by jplebre on one module to compare to the unmodded versions.  Yes i feel the modded version has a much better response to it... so added the mod to the 2nd unit.

However... as mentioned above,  the meter seems to be scaled by the attack pot.... attack full CCW,  the meter stays on 0.  The meter goes to full scale as the attack pot is turned clockwise.  When full clockwise the meter spans the full bar when driven.

The meter on the on-modded unit barely moved at all in any attack position ( but i could clearly hear lots of compression ).

Interesting,  something to look into further. 

Love these though,  they sound great.

 

[901]

The meter on the on-modded unit barely moved at all in any attack position ( but i could clearly hear lots of compression ).

You need to adjust R55 for meter calibration. Did you execute the API initialization procedure?

 

[qmp audio]

The meter on the on-modded unit barely moved at all in any attack position ( but i could clearly hear lots of compression ).

You need to adjust R55 for meter calibration. Did you execute the API initialization procedure?

Yes, calibrated. 

 

[jplebre]

The meter on the on-modded unit barely moved at all in any attack position ( but i could clearly hear lots of compression ).

Interesting,  something to look into further. 

Love these though,  they sound great.

Thank you for reporting back. Always good to have confirmation
I felt the attack was not responding as it should and more importantly not getting fast enough. The De-esser function was almost useless in some scenarios.


The meter, I didn't see anything on the spreadsheet regarding time response for common application. But meter functionality is restored once the attack time is set slightly slower.

Is R55 simply adjusting the current the meter sees?

 

[peter purpose]

Is R55 simply adjusting the current the meter sees?

Absolutely no idea.