Calrec EQ - PCB re-design - Prototype phase

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

sonicwarrior

Well-known member
Joined
Jul 25, 2006
Messages
266
Location
Cologne, Germany
Things started out in the synthesizerforum.de when someone
pointed out to the Gyraf Calrec EQ design.

As I'm in need of an equalizer I jumped on that and started a groupbuy for the not-so-easy-to-get components.

Then I've found out that Purusha makes also cases if there is enough interest, so I started with a FPD design from Frank (nrgrecording) and made some minor changes.
To be able to quickly compare the sound without a channel I definately wanted band-bypass switches which would also make center detent 10k pots obsolete so that you can operate smoother close to the midth point (which you IMO cannot with a center detent pot).
You can have the pot position everywhere and switch the band on and off to compare.
Here is how it goes:
http://www.groupdiy.com/index.php?topic=17603

Here is the thread with my FPD design:
http://www.groupdiy.com/index.php?topic=17863

And what Purusha made of it:
http://www.groupdiy.com/index.php?topic=18074

But when we both were working on the frontpanel design we
recognized that the PCB layout has some limitations:

1. I think the Omeg series PC16ECO pot with 3 pins is not as stable as the series BR16ECO:



In comparison the PC16ECO pot:


But it should be possible to use them as well.

The dual pots are the series PC2G16ECO which should be stable enough with 6 pins :



You can of course use compatible pots like the 16 mm metal can series from Omeg, or compatible Alpha pots.

2. The shelve switches - when PCB mounted - are located to close to the pots (Hi freq / Lo db).

3. The PCB set should be layout for the sequence
Lo, Lo-Mid, Hi-Mid, Hi in 19" rack format and not for a mixer layout like the original.

You can of course use the current layout mirrored but it's not perfect and may lead to confusion (and therefore errors) because of the mirroring.

4. Eventually:
Separate frontpanel and "backend" PCBs.
That would make the cantilever problem easier when mounting the PCB through the pots on the frontpanel.

5. Direct support for the band-bypass switches (but not PCB mounted because there is not enough place on the frontpanel for them).

6. Eventually:
Using a 10k lin pot with included push switch (for the band-bypass).

Pros:
- Frontpanel layout can be made for both people that want and people that don't want the band-bypass switches. The PCB should support both.

Contras:
- There is IMO no such switch (PCB mountable) from Omeg
(only pots with rotary switch for power on-off).

7. Holes for spacers if someone doesn't want to use SPUN switches and PCB mounted pots (Alpha/Omeg, don't know if there are others with 2.54 mm pin spacing and 100k dual rev-log values).
I personally think it's a pain to solder 4 pole switches in "free hand".
But spacers may also be used to have a more stable connection and to solve the cantilever problem.

8. A double sided PCB may be smaller and may make routing easier.

Edit: As you can see in my second post here: The first PCB of the re-design is already finished, so please no further comments about the 8 points listed here.
 
The first PCB (Hi/Hi Mid) is finished in a first draft.

I also redraw the schematic with numbers for all components that refer to the original schematic (but not everything was readable, so it was sometimes a bit of guessing the numbers):
[The pictures are a bit big that's why I included them as links]

http://www.sdiy.de/images/outboard/calrec-eq/Calrec-EQ-Gyraf-Re-Design-Gyraf-schematic-pre2.gif

I excluded the bypass switch because I needed more space and thought a separate bypass switch PCB would be more flexible for those who e.g. want to use a relais or so.
The main ground areas are not shown for more transparency (auto ground).

http://www.sdiy.de/images/outboard/calrec-eq/Calrec-EQ-Gyraf-Re-Design-Gyraf-Hi-PCB-pre1.gif

I implemented the following of my points:
1 (Mono pots with pot bracket for more stability when PCB is mounted using the pots)
2 (More place for shelve switches)
3 (Lo/Lo-mid/Hi-mid/Hi instead of vice versa)
5 (Support for band-bypass switches)
7 (Holes for spacers)

I rejected a double sided PCB because it wouldn't have helped much with the space.
Separated front & backend PCBs are simply bullshit here
(too much connections required)

Please let me know what you think.
This is my third PCB after the miniPSU and the Sidechain-Thrust-Filter for the Gssl (which is basically a clone from someone elses PCB) so there are maybe some faults in the design (maybe the 100Ns which are solo from + to -, instead of dual per IC in the Gyraf design).
 
[quote author="Michael Tibes"]I've just quickly browsed the circuit, why not use THAT input / output ICs if it's anyway redone?
[/quote]

I have not changed anything in the circuit compared to the Gyraf Calrec EQ. I have simply redrawn the schematic to have numbers for the components.
 
Hi There,

If space is the issue? Can we use these type of pots? Omeg in Uk are manufacturing these type of Pots.

displayimage.php


displayimage.php


displayimage.php


Sorry I dont know How to show pics here?

Regards
Dinesh
 
[quote author="dinesh"]
Sorry I dont know How to show pics here?
[/quote]

It doesn't work because a file ending like ".gif" or ".jpg" is needed.

Concentric pots don't make real sense here and I will not change the re-design of the PCB for them. This PCB re-design was enough work.
 
[quote author="zitoune"]nice work !! :guinness: :guinness:[/quote]

Thanks, but let's wait until the prototype is ready.

I don't have the vaguest notion of what I'm doing there. :wink:
(But everyone started at zero)

And I changed the layout for a double sided PCB, so no home-etching possibitlity anymore, I'm sorry. But there are still the Gyraf files for that.

Reasons:

1. I could make the PCBs smaller.
2. Big groundplane on the top layer (the one with the silkscreen) which hopefully results in less vulnerability against hum and other noise especially on stereo versions.
Thread for that:
http://www.groupdiy.com/index.php?topic=21855
The groundplane seems to be not always a good idea.
3. Eases the routing very much (which also makes the PCBs to look better :wink:).
4. Another distinction to the Gyraf version as this was not intended as a competitor PCB to the Gyraf version although based on it.

Now I have to find a solution what to do with the overall bypass switch which I didn't packed on the PCBs so far as I thought about using an own PCB for that. But now with the decision for a double sided layout everything has changed and I think I'll bring it on the Hi/Hi Mid PCB as on the Gyraf version.

A second thing I'm currently thinking of is to combine everything to a long PCB instead of the separated PCBs.
 
[quote author="sonicwarrior"]
Now I have to find a solution what to do with the overall bypass switch which I didn't packed on the PCBs so far as I thought about using an own PCB for that.[/quote]

I have brought the bypass switch back to the Hi/Hi Mid PCB.

I've changed the holes to be all thru-plated and now the silkscreen lays above the solder pads. Corrected some of that yesterday but it's still not completed. Also I want to make the paths using 45 degree angles.
Because all solder pads are thru-plated I also decreased the size.
And then I also will try to have two 100n capacitors for each IC instead of the current one.

The Hi/Hi Mid PCB is now 170 mm x 85 mm and the Lo/Lo Mid PCB is 155 mm x 85 mm.
When I put them together to one PCB it would be slightly bigger because of the needed space between the Lo Mid and Hi Mid pots. Perhaps 340 mm x 85 mm.
I prefer to have 2 smaller PCBs. Didn't see a vote for the single PCB by now but the feedback is by hook or by crook a bit sparse.

In a batch of 50 PCB-Sets both PCBs will cost ca. 23 Euro together (@PCB-Pool, the company which made also made my first miniPSU batch)
but as I will order a prototype (professionally made like the batch but maybe without the silkscreen to reduce the costs) it will be slightly more.
 
Hi sonicwarrior,

many DIYers noticed that with Gyraf's redrawn Calrec schem, if you set the EQ to the flat position, you will have a 6dB gain with a balanced input signal (first opamp stage is an amplifier with the schem values)

If you want to fix it in the following schem,
http://www.sdiy.de/images/outboard/calrec-eq/Calrec-EQ-Gyraf-Re-Design-Gyraf-schematic-pre2.gif
you should change both R1 and R2 from 10k to 22k.

You will find many posts regarding this issue, especially in the Calrec Help thread.
 
Thanks.

Found it here:
http://www.groupdiy.com/index.php?topic=122085#122085

That's why I dislike having the errata spread over several forum threads instead of attaching it to the documentation.
 
Got the prototype boards on saturday.

Without silkscreen and solder mask. :sad:
I don't think I will order any more PCBs from Bilex-LP.
Also the drills would be ok for a self-drilled PCB but not
for a machine-made PCB.

Now I hope I have all the needed parts ... and, of course, that it works. :wink:
 
While not having much time at least I stuffed all resistors, IC sockets and some capacitors.

I used 8k2 for the 8k1 resistor. Hope that doesn't make a big difference. :?
 
I stuffed the whole stuff and have done a first power up yesterday without IC's and something smelled strange.

After powering down the smell was gone.

I then recognized that I've not used a heat sink on the LM317/LM337. As I don't have much experience I couldn't tell where the smell came from.

I'll try again today with a laboratory power supply.
 
News: No bad smell with heat sinks.

The Hi/Hi Mid PCB works fine but I have some problems with the other PCB.
Need to debug it when I have the time.
(It is the one with the 8k1 Resistor R87, maybe the 8k2 was not a good decision :? )
 
8200 is just ~1.2% "larger" than 8100.
This is not a problem.
Good luck
 
[quote author="syn"]Good luck[/quote]

Thanks! :thumb:

I didn't have the time for debugging by now (Cosmo groupbuy and other things).
But here is a picture of the working Hi/Hi-Mid PCB:

calreq-redesign_hi_PCB-prototype1.jpg
 
Back
Top