API clone

[Soeren_DK]

I have bought two Eckmiller W86a eq moduls some time ago. I thought I could use them as master eq but attempt and no sound through them I looked into the schematics and found out that they need some kind of a preamp.

Therefore I desided to build two API clones and put them in the case as well. I found a schematic a friend had modulated and modifyed to my needs.
I am using a LL1528 input transformer and a Cinemag CMOQ-2S output transformer. As for the DOA I was so lucky to get two Jensen 990 from the same friend who helped me with the schematic. I have also removed the 48v option.


Then I tried to draw some PCB layout but I'm not sure how to put the ground link around the board.
Someone here that maybe could help me looking into the layout so I'm absolutly sure that it will work. 

Cheers
Soeren 

 

[Soeren_DK]

the schematic

 

[Infernal_Death]

The 10uf 50V on the V+ supply rail seems to be reversed in polarity.
+ should be at the diode / V+.

Flo

 

[Soeren_DK]

Thanks Fio...

That's right.. . I have changed the layout in the first post.

Cheers
Soeren

 

[[silent:arts]]

for the 990 (if it is C) I would go +/-24V.
I don't really understand the Ground question ???
(your PCB layouts allows a ground plane)

 

[Infernal_Death]

Yes it's not that hard to lay out a ground plane in your layout.
Also i would move the 470uf/25V a bit to the top and right. This way you have
a bit more space between the 470uf and the 1000uf. Makes it easier to use different
sizes of caps and you got enough space on the board to do so.

Flo

 

[Soeren_DK]

for the 990 (if it is C) I would go +/-24V.
I don't really understand the Ground question ???
(your PCB layouts allows a ground plane)

Sorry for the poor language :-[
What I ment was that I was afraid of groundloops. I don't know anything about designing pcb and I remember the problems with the G1176 design..
I have added a pic of the DOA but I don't know if its a C version. The only text I can see is 990...

Yes it's not that hard to lay out a ground plane in your layout.
Also i would move the 470uf/25V a bit to the top and right. This way you have
a bit more space between the 470uf and the 1000uf. Makes it easier to use different
sizes of caps and you got enough space on the board to do so.

I change the layout in the first post

Cheers
Soeren

 

[Infernal_Death]

Actually the opamp you have is a DIY Version of the 990 (Mnats did that version if i remember correctly). The kits and boards were/are sold by PTownKid here.
It's not a original Jensen, but that doesn't mean it doesn't sound good.

The way it is now with the 470uf you can mount more caps, handy thing if you have enough space on the board.

 

[Soeren_DK]

Thanks for the info.  ;D
Do I still need to give et +-24V insted of the 18V
I have changed the size of the board in the first post so its not that crowed any more.

The way it is now with the 470uf you can mount more caps, handy thing if you have enough space on the board.

Will it be a good idéer to mount more caps?

Cheers
Soeren

 

[rodabod]

Do I still need to give et +-24V insted of the 18V

That would give you more headroom. Not necessary though unless you want to drive very high levels. The 990 opamp will probably be quitest with an input transformer set for 1:2 or so, but you can experiment. If you need the extra 6dB gain, the 990 should be happy driving an output transformer set to 1:2 ratio. An input pad might be an idea too.

 

[mitsos]

RE: operating voltage, you should check R10 and R11 on the opamp. For 24V you need 62K, for 18V, 42.2K and for 15V, 33.2K, according to the schematic on John Hardy's site.  The 990C which works on a range of voltages is a modified circuit available only in potted form from John Hardy (as far as I know). The mnats board (which you apparently have) is based on the original 990 circuit. Anyway, when I ran mine (which had 62K in R10 and R11) at 18V, all I got was static. I'm not sure what would happen if you went the other way (higher voltage with resistors meant for lower voltage).

 

[Ptownkid]

yeah, you need to know what voltages it was built for, here's the info...

http://www.geocities.com/m_natsume/JE-990

 

[Soeren_DK]

According to this site http://samengstrom.com/nxl/2020/6_band_resistor_color_code_page.en.html

the resistor R10 and R11 is 61.9kOhm soo with peace in mind I can give it 24V.
I build a Mnats PSU with a LM317 and a LM337 on it so it will be no problem to increase the voltage.

Cheers
Soeren

 

[Soeren_DK]

How to prevent groundloop?

Now I'm building the API clone together with my passive Eckmiller Eq moduls but I can't figure out how the ground should be connected.
The Eckmiller moduls have a ground tip and the API also have it.

I have uploaded a PDF with the drawings.
I know its a little difficult quation. :-[ ???

Cheers
Soeren

Edit: New drawing

 

[EZ81]

Hi Soeren,
all these transformers make things easier.

I'd try to do the layout/grounding as below, maybe some other guys can chime in and correct:

1) Connecting all XLR 1 pins directly to the case. These green lines can be considered antennas broadcasting the HF crap that was picked up by the cable shields to the inside of the box. The cable shield is an extension of the metal box, so it should be connected this way.

2) The XLRF1 connection to the API's circuit ground is truly evil. The HF crap mentioned above (and ground loop currents through the cable shield) will flow through the circuit, causing a voltage drop across the ground traces, possibly introducing noise.

(google "pin 1 problem" for 1) and 2) )

3)  There are three kinds of "ground"  in the API circuit:
- power supply ground (0V input from power supply, gnd connections of PS filter caps, 990 gnd pin)
- output ground (gnd lead of output transformer)
- input voltage reference ground (gnd of 1000u/16V cap,  pin 7 of input transformer)
These three "grounds" should ideally be connected in one place. You do not want power supply ripple current flowing through the preamp ground traces, possibly creating a tiny voltage drop between the bottom of the 1000u cap and pin 7 of the input transformer, as this voltage drop would be amplified with the full gain of the opamp. The star ground principle also applies for ground planes, you want to keep the big currents out of the low voltage/high gain areas of the circuit.

4) The 0V from the power supply could be connected to chassis ground at the power input of one of the APIs or at the power supply PCB output.
5) I'd be inclined to leave the 0V pin of the Eckmillers  floating and connect the GND pin to the chassis, but i'm not sure...

Matthias

 

[gemini86]

I was under the impression it is better to only have one connection from ground to case, to prevent ground loops.

??? (Was I misinformed?)

 

[Soeren_DK]

Hi Soeren,
all these transformers make things easier.

I'd try to do the layout/grounding as below, maybe some other guys can chime in and correct:

1) Connecting all XLR 1 pins directly to the case. These green lines can be considered antennas broadcasting the HF crap that was picked up by the cable shields to the inside of the box. The cable shield is an extension of the metal box, so it should be connected this way.

2) The XLRF1 connection to the API's circuit ground is truly evil. The HF crap mentioned above (and ground loop currents through the cable shield) will flow through the circuit, causing a voltage drop across the ground traces, possibly introducing noise.

(google "pin 1 problem" for 1) and 2) )

3)  There are three kinds of "ground"  in the API circuit:
- power supply ground (0V input from power supply, gnd connections of PS filter caps, 990 gnd pin)
- output ground (gnd lead of output transformer)
- input voltage reference ground (gnd of 1000u/16V cap,  pin 7 of input transformer)
These three "grounds" should ideally be connected in one place. You do not want power supply ripple current flowing through the preamp ground traces, possibly creating a tiny voltage drop between the bottom of the 1000u cap and pin 7 of the input transformer, as this voltage drop would be amplified with the full gain of the opamp. The star ground principle also applies for ground planes, you want to keep the big currents out of the low voltage/high gain areas of the circuit.

4) The 0V from the power supply could be connected to chassis ground at the power input of one of the APIs or at the power supply PCB output.
5) I'd be inclined to leave the 0V pin of the Eckmillers  floating and connect the GND pin to the chassis, but i'm not sure...

Matthias

Hi Matthias.
Thansk a lot for the comprehensiv answer.
I don't fully understand part 3.
Could I maybe ask you for a simple paint drawing of the hole ground circuit?

Cheers

 

[ppa]

the schematic

using a 990 it's could be change the 20k resistor with a 10k type and than the 200 ohm resistor and 22k pot with 100 ohm res and 10k pot, to have less metallic sound. The pot could be the same 22k.
The only problem could be that 990 needs the OLI output isolator that's an RL network to avoid oscillations at low gain.  Regarding 24V supply, should be change in the schematic the cinemag configuration from 1:2 to 1:1 putting the primaries in series configuration.

Pier Paolo.