Symmetrical FET compressor.

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Gustav

Well-known member
Joined
Jun 4, 2004
Messages
2,271
Location
DK
Full disclosure - I will be making self etch files and the BOM available if I get this up and running, but  this is part of my commercial endeavors.


I plugged the 1178 schematic into Eagle today and made some minor changes. I think most bases are covered, and I will be stacking all the pots and switches and leave out the stereo link switch, so this is a "stereo only" project.

I think the meter VU buffer circuit and switchable modes is overkill, and I never noticed any added THD in VU mode on the 1176, so I modified this - mostly from intuition based on former project+whats already in the circuit (I dont really understand how the GR driver works either to be honest - I am used to seeing a dedicated component for the meter and a mirror for the GR reading, but not so here. Its tied back all the way to the dual FET!?).

Does the metering circuit I cut it down to make any sense?

Any other comments before I clean up the schematic and start laying out a board? (I read about Jim William's Opamp mods, but to be honest, I don't really want to go in that direction)

1178 - http://www.jblproservice.com/pdf/Vintage%20JBL-UREI%20Electronics/UREI-1178%20manual.pdf (last pages)

My bastard (You can just read one side of the schematic and ignore my artsy presentation of the symmetry) - https://pcbgrinder.com/download/Data_Sheets/Symmetrical.pdf

Gustav
 
I would like to build one of these. Is the feed to the meter opamp on the wrong side of the cap on your shematic?? Its from ground now..??
(should be on the R48 side)
 
Joechris said:
I would like to build one of these. Is the feed to the meter opamp on the wrong side of the cap on your shematic?? Its from ground now..??
(should be on the R48 side)

Thanks so much! Ill write that down and keep it for when I check for "typos". I'd like some feedback on the general direction before I clean everything up and go over it ten times though.



Gustav
 
No substitute for empirical testing, which is what Ill be doing.

Schem - https://pcbgrinder.com/WIP/PTS.pdf

Cleaned it up a little more, but still needs some work.

Board layout - https://pcbgrinder.com/WIP/PT.gif

Trying something different for me. Gonna be interesting to see how it turns out.

Front panel. - https://pcbgrinder.com/WIP/PTF.pdf

As an experiment, Ill be doing the front panel layout in Eagle and make it as a PCB. I only put in the holes and dimensions so far...

- Still waiting for a reply from Alpha and Omeg Re double stacked pots.
- Still considering options for the dual FET
- Still looking for two grayhill switches at a reasonable price. ;)

Feel free to use the schem or anything else you might find use for.

Gustav
 
I haven't looked through anything but wanted to comment ...

Gustav said:
. . . and I will be stacking all the pots and switches and leave out the stereo link switch, so this is a "stereo only" project.

How hard would it be to leave the switch?  Most days I leave it unlinked.  I have NO problem with stacked controls though.  I just like the variations....

As for the greyhill switches ... what do you need?  I may have some spares that I could post to you if needed as I would love to try and get a board ....  ;D

Michael
 
MHanson said:
I haven't looked through anything but wanted to comment ...

Gustav said:
. . . and I will be stacking all the pots and switches and leave out the stereo link switch, so this is a "stereo only" project.

How hard would it be to leave the switch?  Most days I leave it unlinked.  I have NO problem with stacked controls though.  I just like the variations....

As for the greyhill switches ... what do you need?  I may have some spares that I could post to you if needed as I would love to try and get a board ....  ;D

Michael

You can insert the switch if you like - its pretty simple, and I put in a resistor to mark the point.

Trying to clean it up some more, hopefully to a point where I can start thinking about  single sided routing. Also checking the schematic for errors a few more times (Found one yesterday).

Still no more than 80% sure about the metering circuit, but I'll find out soon enough.

Gustav
 
culteousness1 said:
Gustav,

how about adding DOA footprint as an option for IN- and OUTPUT- amps?
Just like Igor did in his F76...

Best,
Carsten

I might  do a general purpose translator board at some point, but wont add the footprints here.

Output opamp?

Gustav
 
Added footprint for the original Dual FET.

I also decided to ditch the Grayhill connectors (20+ euros a piece here) and make a footprint for an angled connector card for standard Lorlins.

latest board update attached...

Gustav
 

Attachments

  • Spass.pdf
    537.6 KB · Views: 91
- A good thing to implement when you have plenty space like this, would be a star grounding, taking ground from each "section" to a common point - e.g. at the input signal connector (this is where I usually have it). You could consider the pots one "section"
- Larger pads for the trimmers and for the alt-fet
- possibly option for 5x10 trimmers too (these are more common than the ones you use)

Looking at the schematic, I worry a bit about the lack of AC decoupling at the top (the "input") of the input and output level potentiometers. They are driven directly from outputs of 5532's, and there's bound to be a significant DC offset here. In my book, this is a sure way to scratchy potentiometers. I would prefer to have a 100uF/25V between opamp output and pot top, but ymmw [maybe google 1178 service/help requests to see if this is a problem in real world?]

+ maybe add a number of test points (extra pads, differently shaped) on top pcb layer for easier error tracing? E.g. after input amps, after preamps, after line-amp, at psu voltages, at sidechain inputs (IC7D-8D-pin14), at sidechain output (R44-C37), and at source of fets (to verify Q-bias level)

Jakob E.

Edit: Googling "urei 1178 scratchy" gives quite some hits. I think the added 100u's would be a good thing - purists can always just bridge the spots in the pcb, if they're willing to change pots regularly  :)
 
Thanks for all the suggestions.

I was planning on adding silk screen markers for supply voltage when I had a final routing finished, but I will consider actual test points.

I tried working in as much as I could today - will do some more work on it during the weekend.

Gustav
 

Attachments

  • almost.gif
    almost.gif
    809.6 KB · Views: 145
Short:

LL5402 NEEDS at least 1.45mm drills, ideally 1.5mm. And possibly bigger pads for secure soldering.

I still think a "modular" star ground would do you more good than groundplane fill.

Jakob E.
 
gyraf said:
Short:

LL5402 NEEDS at least 1.45mm drills, ideally 1.5mm. And possibly bigger pads for secure soldering.

I still think a "modular" star ground would do you more good than groundplane fill.

Jakob E.

I tried making "half" of a star ground to the pots, the PSU section and the in and outputs to see how it would affect routing before doing anything more elaborate, but I misread and drew them with reference to the power input - The pour is not connected to ground, and Ill correct the reference point mistake, draw them to the audio input instead and work on getting everything included before having a board made.

Gustav
 
I may be mistaken, but shouldn't R3 in your differential input be 10k? The current value would double value of one input, leaving you with very bad CMRR.
 
Scott ODell said:
I may be mistaken, but shouldn't R3 in your differential input be 10k? The current value would double value of one input, leaving you with very bad CMRR.

Thanks so much for taking some time to take a look.

I already found that particular error. (check the silk screen on the board posted before this). I cant remember all of them, but I also found a few other value errors and one major layout error (ops). On the last 4 passes, it seemed like everything was good to go schematic-wise.

Schem is not "publication friendly" right now. Its a bit of a mess of tags going to pin connectors/pins on switches etc, but Ill clean it up when I have given this board layout a shot (wether it works or not). My main concern now is the switches. I think I am still cross eyed from concentrating on those connectors yesterday!

Here is the latest PCB layout (both top and bottom), and when I've added a few measuring points and checked with the CAM engineer at the factory (for feedback on general layout issues), I will have the board made and see if it works as intended.

https://pcbgrinder.com/download/FF78.pdf

I dont mind shipping out a prototype or two when they are done, if someone wants to give it a go.

Gustav
 
Such a cool idea, a stereo FET Grinder in a 1 rack unit space.
As usual I don't have anything to contribute 'cause I'm no tech engineer,
but I really like all the nice stuff you make.

Best
Magnus
 
Populating the PCB.....

I'm a few parts short, and I'll try to find an original FET....

Gustav
 

Attachments

  • jk.gif
    jk.gif
    287.9 KB · Views: 192
Wow! 

You ARE quick!

Is the filtering strong enough on the power supply?  To me, power supplies always seem to be the weak points ... just asking.  :D

Are those single turn trimmers you are using?  Why not utilize multi-turn trimmers?

Regard less, it looks great!  Keep it up!

Michael
 
Supply filtering sizing would depend on current draw. We're below 100mA at +/-15V - and rule-of-thumb says 1000-2000uF per Amp.

As for trimmers, multiturn are great where a lot of precision or a very-large trimming area is important - in the 1178 we need neither. And single-turn are much easier to work.

Jakob E.
 
Back
Top