« on: October 05, 2018, 08:19:01 PM »

In regards to the (2)output 5k trim pots and 5k resistors on the Cavendish board, it's  recommended in the installation instructions to remove the 7.5k's, there are no 7.5k's on my GSSL board, maybe they changed the values on a revision ? I recently just purchased the GSSL board from pcgrinder.
Where exactly are these resistors on the board. I read in a post it's the 15k accross pins 1,and 2 on the 5532(see attached PDF) do I jumper them or just remove them?
The post also mentioned to change the 5k trim, and 5k resistor to 10k.

Tony Wisniewski


Re: Cavendish
« Reply #1 on: October 05, 2018, 08:47:26 PM »
I'm not familiar with the mod.

You could always reach out to Expat

 or message member Rochey here if you don't get any answer....;u=451

Re: Cavendish
« Reply #2 on: October 05, 2018, 09:18:18 PM »
Thanks for the links

Re: Cavendish
« Reply #3 on: October 06, 2018, 03:46:20 PM »
Got the cavendish assembled placed the 4- op amps in and fired it up  there was distortion and very little volume.
I swapped the pins on the 5532 as per instructions. Then I un-swapped  the pins on the left channel and the distortion was gone on that channel,
but volume was low, and the 5k trim pot seem to increase the volume but very little.
I know the opamps work because I pulled them out of a working unit. I did notice there is a 7th pin on the cavendish opamp socket for trim, but on the opamp itself there is none.

Re: Cavendish
« Reply #4 on: October 07, 2018, 09:00:59 AM »
Has anyone even built one of these? Expat Audio a little guidance would be helpful.


Re: Cavendish
« Reply #5 on: October 07, 2018, 07:15:03 PM »
The resistance values are to set the gain to unity throughout. Since some people use 'just the hot pin' in single-ended/unbalanced applications, they get 6dB less gain... with a transformer you can't use just one pin, you HAVE to use both, so the negative feedback resistance of 15k will double the gain (output stage gain is deirectly proportional to negative feedback resistance, so double the resistance = double the gain, or 6dB higher signal level).

The practical upshot is that a 7.5K TOTAL negative feedback resistance should give you unity gain at no gain reduction. -To allow for small tolerance variations, having a trim pot fitted makes it easy to dial in unity gain precisely. -Since you should only need a few dB of range across the trim pot, instead of putting a 15k trim pot in circuit, where midway position (theoretically 7.5kΩ) "should be about right", we suggest a 5k fixed resistor, and a 5k trimpot, where -again- mid position should be "about right" but the difference is that the range of adjustment is greatly reduced, which makes the trim pot far less "fussy" about the tiniest adjustment 'overshooting' setup.

What happens with the output stage is basically the 5532 on the main board is replaced by the discrete op-amp, the output is fed to a transformer for balancing (in classic API fashion) and the feedback resistance (which sets the gain) is also moved off the mian board. -You DID  remove the feedback resistor on the main board, -right? -If not, your gain would be expected to be quite low.

"A waist is a terrible thing to mind"
Quote from: PRR
Ah, but that was 1999; we don't party like that any more.

Re: Cavendish
« Reply #6 on: October 07, 2018, 07:49:51 PM »
Hi Keef

Thanks for the info can you indicate where the feedback resistor is located on the attached file.


Re: Cavendish
« Reply #7 on: October 07, 2018, 07:52:27 PM »
Do you remove it from the circuit and jumper ?

Re: Cavendish
« Reply #8 on: October 08, 2018, 07:39:10 PM »
I removed the right chan. 15k across the 5532 see attached schematic, someone said this is the feedback resistor that should be removed for the cavendish.
Well I don't think so, it totally cut out the left chan.
I wish someone can help me out, I am using the gar2520 opamps.
I am getting signal through at a decent volume now, but it is slightly distorted,  and there is a slight hum when at Idle. Everything is grounded properly.


Re: Cavendish
« Reply #9 on: October 08, 2018, 09:33:53 PM »
Maybe post some pictures of everything you have going on???

Re: Cavendish
« Reply #10 on: October 08, 2018, 11:36:31 PM »
Good Idea I will do that tomorrow

Re: Cavendish
« Reply #11 on: October 09, 2018, 03:08:33 PM »
Here is the picture of the board, the interface wiring from the GSSL to the Cavendish is a rats nest, but once all the problems are resolved I will clean that up, It is wired correctly.  FYI  the GSSL was a working unit before this mod with no issues.
I did remove the (2) 15k's across the 5532, (see picture). Before I did this I was getting output on both channels, It was almost acceptable, but did have a slight bit of distortion when you cranked it up a little. I also had a low hum output I think 60hz at idle.
When I removed the 15k resistors the output dropped down to just about nothing with distortion.
The Op-amps are functional I tested them in another unit.
If you noticed on the picture I just used the pins on the Cavendish that are connected to the inputs/outputs. The +/-15vdc supply is wired to the test points 1 thru 8 then to the GSSL board +-/15vdc supply.


Re: Cavendish
« Reply #12 on: October 09, 2018, 03:28:06 PM »

I'm guessing your Cavendish is newer version PG 1.0 since you mentioned swapping pins???

Is this a retrofit on an already working GSSL??

Re: Cavendish
« Reply #13 on: October 09, 2018, 03:28:48 PM »
I Just jumpered the (15k's) and signal came back, it's about the same amplitude but less distortion.

Re: Cavendish
« Reply #14 on: October 09, 2018, 03:29:38 PM »
Yes the GSSL was working fine


Re: Cavendish
« Reply #15 on: October 09, 2018, 03:38:05 PM »
I'm not familiar with these so hopefully someone can chime in. It's a bit confusing trying to figure out why you didn't terminate /use both board's 8 pin op amp connections?? I think you explained this but I haven't studied further yet....
« Last Edit: October 09, 2018, 03:41:23 PM by scott2000 »

Re: Cavendish
« Reply #16 on: October 09, 2018, 04:01:39 PM »
Almost half the pins on the Cavendish board are not used, so why connect them to the GSSL board.
« Last Edit: October 10, 2018, 12:39:33 AM by TonyW »

Re: Cavendish
« Reply #17 on: October 09, 2018, 04:51:12 PM »
So can anyone tell me if you are suppose to short the (2)15k's instead as an open?
It sure seems to me by shorting them it brings my GSSL into usable except for the somewhat low output (maybe its not so low). Is there a way to test the output with a 1khz signal.


Re: Cavendish
« Reply #18 on: October 09, 2018, 05:05:52 PM »
it would specify that most likely.... So,  shouldn't at least one of those 5532 spaces be fully used in order to get one channel working? The schematic shows one full opamp for one channel.....

Sorry if I'm confused......
« Last Edit: October 10, 2018, 09:40:56 PM by scott2000 »


Re: Cavendish
« Reply #19 on: October 10, 2018, 02:43:49 PM »
Maybe take the GSSL schematic and draw on it what you have going on .  I don't see a schematic for the Cavendish which would be great too.....

Maybe with this, you could post in the lab to see if anyone can help if you don't figure it out by then......

Or even try posting in a Gssl build thread...... I remember seeing some mention of this mod somewhere there I'm sure......