Ryssen said:
Yes, the minimum is the INPUT board. As well, you can use the input board without component to "extract" signals if you want to play yourself with input and output signals.
First, you can wire 4 independant cables with RCA from the small boards... That's all and the rest of the DEQ features will still work.
INPUT using original DEQ XLRs:
- unsolder C26, C17, C19 and C28 on the analog IO board.
- wire the input for left and right, and connect it to the IO board:
+ connect left wire on positive hole of C26 (pin 2 of left XLR)
+ connect right wire on positive hole of C19 (pin 2 of right XLR)
- if no OUTPUT board, strap J4 as indicated on the schematics.
OUTPUT using original DEQ XLRs:
More complex. Need to cut PCB traces around the relay, or remove SMD resistors.
- remove R9, R40, R45, R42, R43 and R10 (left output)
- remove R11, R48, R53, R50, R51 and R12 (right output)
- connect left input on one of the left resistor trace (XLR side)
- connect right input on one of the right resistor trace (XLR side)
Doing this, the relay will still be in the circuit (bypass mode when unit off).
On my web site, on the "capacitor mods" page, you have a table with all caps used in the DEQ.
BUT never tested yet and if my board works... 😉 Need to wait 2 weeks at minimum to receive the PCBs.
Stef...
Stephane where do you get your boards made? Are they reasonably priced? Just curious; I get mine from Olimex in Bulgaria for 30 euro's for a eurocard, and I can mix and match within the 160*100cm and they will panelize and cut for free.
Jan Didden
Jan Didden
janneman said:
Olimex too. Eurocard size and panels. Job is well done and price are correct (if no option). And he will not print your file if error as some other. It's ideal for a very small quantity.
I've not found lower price for so small quantity and panels.
Stef...
janneman said:
It was not a nice idea to ask this question. I've a problem with Olimex and my PCBs. On these PCBs, I used automatic ground plane on bottom side, and my CAD software (Sprint Layout) generated composite layers and negative plots (G36/G37 commands from RS-274-X) that Olimex doesn't support...
Currently no way to print them.
Googling on this, I've found that Olimex do not support these protocols. I'm not the first to have the problem. Sprint Layout, KiCad and gEDA do ground plane like this. Eagle not.
I've to find another suppliers. Sh...t.
Stef...
Are you sure?
This is what the Olimex website says:
"GERBERS for Copper, Soldermask and Component print layers:
RS274-X - also know as EXTENDED Gerber format, with embedded apreture info, any aperture sizes above 8 mils is acceptable".
Seems that maybe here is another issue. Ask Tsvetan at support@olimex.com about this; he's very helpfull.
Jan Didden
This is what the Olimex website says:
"GERBERS for Copper, Soldermask and Component print layers:
RS274-X - also know as EXTENDED Gerber format, with embedded apreture info, any aperture sizes above 8 mils is acceptable".
Seems that maybe here is another issue. Ask Tsvetan at support@olimex.com about this; he's very helpfull.
Jan Didden
janneman said:
That Tsvetan who do not want to process my files... In short words, he say that my cad software is bullsh... The tech guy from Abacom is in vacations and I've not got a formal answer from them. May be I made an error designing my current pcb but I already designed tens of pcb with Sprint without problem using Olimex. But, this is the first time that I use automatic copper pour plane.
I can send you his answer in PM if you want.
Stef...
Despaired.
100nF in the supply should not be adequate. The datasheet suggest an additional 10uF tantalum cap to prevent low frequency noise.
4real said:
No change.
I also tried to add 1nF between in/out. The curve is similar that the one on the output circuit when I try an OPA627 instead of an OPA2134.
Input impedance?
Stef...
Maybe you will find some more info here: http://focus.ti.com/lit/an/sloa054d/sloa054d.pdf
Besides: where did you get these noise figures from? I'm guessing one of the two outputs? What if you join them to single ended? Could it be that the noise is canceling itself out?
Besides: where did you get these noise figures from? I'm guessing one of the two outputs? What if you join them to single ended? Could it be that the noise is canceling itself out?
I think what we see is 'normal' shot noise and is basically the same for all opamps. And noise will not cancel if you go from balanced to single ended. Noise is uncorrelated, it is not that at a certain time the noise is pos on one input and the same negative on the other. Rather, it is random amd it cancels vectorially, so you get 3dB less total noise if you combine two equal level sources.
That is the reason sometimes you see a lot of parallel small-signal transistors in a phono preamp input. Each doubling of the number of devices lowers the noise by 3dB. With 8 device parallel you get 9dB reduction.
Jan Didden
That is the reason sometimes you see a lot of parallel small-signal transistors in a phono preamp input. Each doubling of the number of devices lowers the noise by 3dB. With 8 device parallel you get 9dB reduction.
Jan Didden
I tried changing input impedance and this change the curve a lot. Less impedance = less noise. Seems normal.
But in fact if we switch to the total noise figure we realize that the levels are so small that this problem is not really a problem.
I'll send new curve later for the OUTPUT board with 3 different op-amp. Very interesting.
Stef...
ps: I finally found an agreement with Olimex to have the proto boards printed without CRC check.
But in fact if we switch to the total noise figure we realize that the levels are so small that this problem is not really a problem.
An externally hosted image should be here but it was not working when we last tested it.
I'll send new curve later for the OUTPUT board with 3 different op-amp. Very interesting.
Stef...
ps: I finally found an agreement with Olimex to have the proto boards printed without CRC check.
3 differents noise curves using the output design with same component values. The last one is a manual compilation.
I'm looking to understand the OPA2134 curve compared to others.
OPA627
OPA2134
LM4562 (LME49860)
Just to have an idea as the curves not use the same scale.
I'm looking to understand the OPA2134 curve compared to others.
OPA627
An externally hosted image should be here but it was not working when we last tested it.
OPA2134
An externally hosted image should be here but it was not working when we last tested it.
LM4562 (LME49860)
An externally hosted image should be here but it was not working when we last tested it.
Just to have an idea as the curves not use the same scale.
An externally hosted image should be here but it was not working when we last tested it.
stef1777 said:
Stephane,
There may be a large difference between the models for the various opamps. Especially noise is often modelled very cursory or not at all.
janneman said:
You're true, the best will be my scope and my ears when I will have assembled a prototype... 😉
.
Hi Folks!
I'm back on this. I finally received my prototype boards.
First the good news. Both boards works.
The Input board fit well and the components values seems fine. I just changed R9 to 39K to have 2.4v for Vocm. For the moment, I just tested the analog features. I've to test the others DEQ features (DIG, mic, aux, etc.) to verify if all still work.
Level seems ok. I have -20dB on the RTA for 500mV sinus at input.
The input board inside the DEQ2496 with the output DAC signal bypassed with 4 jumpers.
The output board is more problematic. I first done a mistake calculating the total used height. I forgot to count the black female connector. The board don't fit, I can't close the box... I've put the board on the side but it's still too high. I'll have to modify the PCB. I also found a small mistake with ground plane.
The output level is not good. For 500mV at input, I got only 160mV at output using the output board. I've to add 10dB in the Utility menu to get the original 500mV.
Signal look fine on the scope for 100, 1K and 10K sinus.
I'm back on this. I finally received my prototype boards.
First the good news. Both boards works.
The Input board fit well and the components values seems fine. I just changed R9 to 39K to have 2.4v for Vocm. For the moment, I just tested the analog features. I've to test the others DEQ features (DIG, mic, aux, etc.) to verify if all still work.
Level seems ok. I have -20dB on the RTA for 500mV sinus at input.
The input board inside the DEQ2496 with the output DAC signal bypassed with 4 jumpers.
The output board is more problematic. I first done a mistake calculating the total used height. I forgot to count the black female connector. The board don't fit, I can't close the box... I've put the board on the side but it's still too high. I'll have to modify the PCB. I also found a small mistake with ground plane.
The output level is not good. For 500mV at input, I got only 160mV at output using the output board. I've to add 10dB in the Utility menu to get the original 500mV.
Signal look fine on the scope for 100, 1K and 10K sinus.
you're doing good work so far, THe input board works, and a few bugs to iron out.
Have you listened to music yet?
Have you listened to music yet?
Pulse-R said:
Not yet. I'v to adjust some components first. My priority is to fine tune the input board. Later, I'll look to fine tune the DAC output board. I will not use it in fact as I've a pair of UcD180.
Jan if you pass here: Are the OPA1632 hot in your analog DCX board? Mine are around 45°C minimum box open.
.