Sry... missed the explanation.
What I did was the following :
In calculator ::
Then I calibrated REW, and took the measurement's with resistors attached to input jack.
I then adjusted only the RL field, to have the same values as thoose I measured.
I then assume that RL must be the input impedance.
That's why I called it BackCalculate 🤔
But again, I'am not sure it can be done this way?
What I did was the following :
In calculator ::
- Input voltage : E.g. [2.8V]
- Input R1 : E.g. [100K]
- Input R2 : E.g. [27K]
- Let RL be as the unknown faktor []
Then I calibrated REW, and took the measurement's with resistors attached to input jack.
I then adjusted only the RL field, to have the same values as thoose I measured.
I then assume that RL must be the input impedance.
That's why I called it BackCalculate 🤔
But again, I'am not sure it can be done this way?
Last edited:
Well, I'm not a specialist so I try to keep it simple. First I would calibrate the soundcard for dB full scale and measure the Vrms it takes at the input for that. The voltage divider should always deliver this voltage. And its total resistance R1+R2 should be significantly higher than the source and significantly lower than RL. Then both source and load should stay more or less out of the equation. So, if you want to measure an amp on a 8 ohm load at 1W and atenuate 2,83V to something like the half, you could start with a 10-20k pot to find the resistance ratio and then replace with fixed values.
Yes... you are right.
I see also that a lot of attenuator's are done with 10k/20k resistor's. - Never thought of this :
So, I will try it out but I don't really believe that it would make much difference.
I wish I was able to construct a real autoranger, giving output's within +/- 0.5dB before clipping, and even be able to be programmed to the ref. voltage of wish.
But this is another story 🙂
Jesper.
I see also that a lot of attenuator's are done with 10k/20k resistor's. - Never thought of this :
And it sound's correct!
So, I will try it out but I don't really believe that it would make much difference.
I wish I was able to construct a real autoranger, giving output's within +/- 0.5dB before clipping, and even be able to be programmed to the ref. voltage of wish.
But this is another story 🙂
Jesper.
I have been looking for an autoranger recently, but there is a shortage of this kind of equipment right now!
Both the LinearAudio autoranger and the Cosmos Scaler aren't availible at all nowhere 😕
BUT diy is also fun --- well a lot more fun, than just to buy; but I'am not sure I can success with such a challenging project? - At least not without some help.
There is this old thread on diyaudio : HERE
@H713 was making an Arduino/cpu based autoranger, but the thread died.
I'am tempted to ask @phofman (which was also contributing in that thread), if he know's how the sensing for the autoranging is working on the schematic attached here, from the 2. last post in the thread.
I really cannot figure out how the sensing connection's to the Arduino is working. But that said, I do belive I'am capable of creating the program to control the relay's and writing information on some display.
Well... I'am going down the rabbit hole now 🙂
Jesper.
Both the LinearAudio autoranger and the Cosmos Scaler aren't availible at all nowhere 😕
BUT diy is also fun --- well a lot more fun, than just to buy; but I'am not sure I can success with such a challenging project? - At least not without some help.
There is this old thread on diyaudio : HERE
@H713 was making an Arduino/cpu based autoranger, but the thread died.
I'am tempted to ask @phofman (which was also contributing in that thread), if he know's how the sensing for the autoranging is working on the schematic attached here, from the 2. last post in the thread.
I really cannot figure out how the sensing connection's to the Arduino is working. But that said, I do belive I'am capable of creating the program to control the relay's and writing information on some display.
Well... I'am going down the rabbit hole now 🙂
Jesper.
It's been a while since my last post, but I finally finished my UMC202HD-based audio analyzer. I will share pictures, schematics, etc. later, probably sometime in July. In the weeks ahead, I will not have time.
And now something completely different:
If you haven't noticed yet: in the latest issue of the Elektor magazine, a webinar on Affordable Audio Measurements was announced, to which Alfred Rosenkränzer also contributes. You can register by clicking on the link below.
https://elektor.clickmeeting.com/test-measurement/register
Jan
And now something completely different:
If you haven't noticed yet: in the latest issue of the Elektor magazine, a webinar on Affordable Audio Measurements was announced, to which Alfred Rosenkränzer also contributes. You can register by clicking on the link below.
https://elektor.clickmeeting.com/test-measurement/register
Jan
New driver for UMC202HD (2023-05-22):
https://mediadl.musictribe.com/download/software/behringer/UMC/BEHRINGER_UMC_Driver-V5.57.0.zip
https://mediadl.musictribe.com/download/software/behringer/UMC/BEHRINGER_UMC_Driver-V5.57.0.zip
Maybe I not have gone through all of the pages on this thread carefully, but I didn't see where you cut and rewire resistor connections. Isn't this the reason for AD8694 burnout?
I can't help. I haven't installed it, I just provided it as information.
I am using driver version 5.3. Version 5.51 was reporting a missing driver for the subdevice in the device manager (if I remember correctly).
I am using driver version 5.3. Version 5.51 was reporting a missing driver for the subdevice in the device manager (if I remember correctly).
No, I didn't rewire the resistors at AD8694 output. I removed the 10k shunt resistor and substituted the in series 1k for 100R. They stayed outside the feedback loop. Taking the 100R inside the loop helps to keep low the output impedance but it won't take the opamp to any advantage with regard to phase margin. 100R proved too low for AD8694. Maybe with different feedback resistor and capacitor it could do better. For the record, I tried a "transplant" pcb with AD8694 configured in the text book scheme with 100R inside the loop. I didn't insist until break down but it was clear that it was unhappy. However, those were my tests with limited knowledge and resources.
I repeated the ADC buffer mod from jp8 and it works great with AD8694!
before mod (THD 0.0052%, noise shelf about -135 dB)
after mod (THD 0.00048% and noise shelf about -160 dB)
...and phase
before mod (THD 0.0052%, noise shelf about -135 dB)
after mod (THD 0.00048% and noise shelf about -160 dB)
...and phase
Last edited:
It also worked nice with what I did but it didn't last much. 🙂 It would be interesting to see how it does with this scheme.
I think I've found the cause of the unstable input buffer, and it's not the opamp
This is due to two extra 1uF capacitors that are soldered directly to the output line of the VCOM buffer (IC16:B).
I think you are referring to the oscillating Vcom buffer which Alfred mentioned in post #553. I had exactly the same issue. Alfred solved the instabillity by adding a 47 uf cap on the op amp output. I chose to carefully lift pins 6 and 7 and solde a 100 Ohm 0402 or 0603 resistor between these two pins and the PCB solderpads of these pins. A small blob of solder connects pins 6, 7 and one end of the resistor. The resistor isolates the poor op amp from its capacitive load and prevents the oscillations. One of the many design flaws solved...I think I've found the cause of the unstable input buffer, and it's not the opamp
This is due to two extra 1uF capacitors that are soldered directly to the output line of the VCOM buffer (IC16:B).
Jan