AT2380v1 // 2 Channels digitally controlled stepped attenuator

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Did you include auto-tracking? Did the calibration work? I am struggling quite a bit with mine, the 2 trimmers aren't linear, but jump around quite a bit... Hence I never get both adjustments right.


Best regards,

hallodeletue



Yes, channel tracking enabled. Haven't started the calibration yet - traveling soon and might be 1-2wks before I can get to it and report back.


BK
 
Thanks Frex, again.

Just, I completed the board and test as pdf described.

AT2380_diy.JPG

PT13 & PT14 is OK
PT15 is not, it reaches until 3.3V not 5V.

I am looking the area of LT317, resistor values are OK.
D19, D20 are OK
LT317.jpg

The only part that I don't use the BOM List was the D1 MBR0540, I use a LT8716 B140 that I had.
From what I looked the difference is to the IF 0.5 vs 1A.
 
Hello,

It is curious to measure 3.3V on P13 if there is no higher voltage on PT15.
Can you measure voltage on X6, and across R64 (you must have 1.25V) ?
What did you use as voltage source ?
About the diode, i don't know LT8716 and i don't find datasheet.
Can you give a link to look this ?

Frex
 
Frex, thanks for help.
In a couple of hours I will check again all the points and everything that to suggest me.

It was a used diode, with LT8716 there is no any info but with the B140 it seems to be like this: https://www.diodes.com/assets/Datasheets/ds13002.pdf

As a voltage source, I used a Rigol DP832 lab psu with 50mA current limit. At 5V of input, the current was 50mA about.
 
Last edited:
After 3 hours from last soldering, the voltage was OK!
I upload the fw of CPLD and I done and the last work, shield and display fiberglass.

It seems OK.

AT2380_diy2.JPG
AT2380_diy3.JPG

The leds aren't light on, probably I used wrong type of led or I have a different legs series (- + - or Cathode-Anode-Cathode). I checked them with the multi-meter diode test and lights on normally.
 
Thanks.

I do step by step all the checks that are required.

Just, I completed the "Attenuation levels test".
I used the Rigol DP832 as DC Source at 10Vdc (10.001-10.002), a Fluke45 as a multimeter.
The results are:

Channel -1 - 2
Attenuation(dB) - Vmeasured
0 - 9.99 - 9.99
0.5 - 9.38 - 9.44
1 - 8.81 - 8.91
2 - 7.78 - 7.94
4 - 6.12 - 6.31
8 - 3.81 - 3.94
16 - 1.54 - 1.57
32 - 0.248 - 0.249
64 - 64.5 - 64.3

I am wondering, why the channel2 goes awesome until 16dB and at the low limit at the 32db and the hi limit at the 64db, but the channel1 is always at the left limit or under of this!
I used 0.1% resistors in array, except R17,37/R39,40 that I used 1% but I matched them to 0.1%.
 
Hello,

Your results seems very good.
On witch channel do you have connected the tracking mode ?
Do you check input voltage each time before measuring output voltage ?
Do you have others things that your DVM connected at the output ?

About the 64dB step, precise "64" dB attenuation is more difficult to get
because of relay contact contribution.
If you require more precise value, you can adjust R39/R40 value by doing
a measurement and calculate then the new value to reach real 64 dB step.
(SO in your results, i suppose that for 64 you values are 6.45mV and 6.43mV ! )

Regards.

FRex
 
Hi,
The tracking mode is on Channel 1, this is the channel with the low limit values at the Attenuation Level Test.

The Rigol DP832 PSU has a monitor that measures all the time the output (all the time the Voltage Output was 10.001-10.002).
The only thing at the Output was the Fluke45 via a 50 Ohm bnc cable. The parameteres was identical at the measuring of two outputs.

The precision resistors (0.1%) was bought from Mouser, but I don't make a measurement (Ohm) before to solder them.
I am afraid that a value is wrong at the Channel-1.
Is there a possibility this is causes from tracking mode?

...SO in your results, i suppose that for 64 you values are 6.45mV and 6.43mV ! ...

Yes, typo error!
 
Last edited:
Today, I make a real test with RTX6001

*) The generator produce each time a sine of -3dB (bnc), to Channel-1 and Channel-2.
*) The AT2380 attenuates by -3.5dB the signal and the output goes to RTX6001 input (xlr).
The cables are,
Generator -> AT2380 = bnc to bnc
AT2380 -> RTX6001 = adaptor bnc to rca and rca to xlr (mode6 on RaneNote Guide, with the shield to the air on RCA)

I see that the Channel-2, this is not tracking (it had a very good result at the Attenuator Level Test(ALT)), goes awesome.
I see that the Channel-1, this with tracking (it had moderate result on ALT) have a ground loop issues.

I have enclosed the bottom of my AT2380 for this reason.
Frex, what you suggest about this?
 

Attachments

  • bottom.JPG
    bottom.JPG
    320.5 KB · Views: 148
  • Sig-3dB, Att-3.5dB Channel1.jpg
    Sig-3dB, Att-3.5dB Channel1.jpg
    225.2 KB · Views: 145
  • Sig-3dB, Att-3.5dB Channel2.jpg
    Sig-3dB, Att-3.5dB Channel2.jpg
    207.4 KB · Views: 148
Hello,

First yes, the tracking channel is loaded with the differential amplifier impedance.
So because output impedance of attenuator move from 0 to 300 Ohms, there is little
contribution on the overall attenuation.
If greater gain accuracy is required, it is possible to choose higher resistors value
for the differential amplifier (U8).

So about the ground noise you measure on CH1 ;
What is the power source used ? (Always DP832 lab PSU ?)
Does the enclosure is closed ?
Is there only one channel connected to RTX6001 at same time ?
What is the signal source ? The RTX generator ?
Regards.

Frex
 
Hello,

First yes, the tracking channel is loaded with the differential amplifier impedance.
So because output impedance of attenuator move from 0 to 300 Ohms, there is little contribution on the overall attenuation.
If greater gain accuracy is required, it is possible to choose higher resistors value for the differential amplifier (U8).

Then, the small deviation that observed on Channel 1, is normal under these circumstances. I see now.
Are you talking for R73,75 about?


So about the ground noise you measure on CH1 ;
What is the power source used ? (Always DP832 lab PSU ?)
Does the enclosure is closed ?
Is there only one channel connected to RTX6001 at same time ?
What is the signal source ? The RTX generator ?
- Yes, always with the DP832
- The rear with the main case, not the upper and front
- Yes, each time one channel connected
- RTX was the signal generator
 
It sure that there is gnd issue noise on Channel-1.
Just I done the following.
With the pcb out of case, return it up_side_down and take a measurement on Channel-1 with the same parameters as before.
Of course the noise now is a little bigger and the THD jumps from 0.00020 to 0.00027%, but just I bridged the gnd of channel-1 with the gnd of face with a screwdriver, there was a little improvement to 0.00018%
 

Attachments

  • open1, no joined.jpg
    open1, no joined.jpg
    239.7 KB · Views: 131
  • open1, screw joined.jpg
    open1, screw joined.jpg
    235.5 KB · Views: 132
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.