Xmas Amp - Dibya's TDA7293 by Jhofland

Jump to Latest
#907
Hi Jesper,

Jhofland replied to me and I’ll post his response here. Do you have C2 bypassed with a jumper?

I assume that R3 is tied to ground by substituting a short across C2. Other than that, I'd measure the output of the servo op amp, and a careful measurement of the DC voltage at the + and - inputs of the 7493. They both ought to be about the same and darned close to zero volts. I would be very careful measuring these nodes though since stray signals will be amplified. In other words, measure these nodes as a last resort.
Click to expand...
 
#908
xrk971 said:
Hi Jesper,

Jhofland replied to me and I’ll post his response here. Do you have C2 bypassed with a jumper?
Click to expand...
That's nice. -Glad you care to help XRK!

I also did look at the schematic, and checked that it was the correct TL071 i had.
Regarding if C2 & C17 both or only one of them have to be jumped, it's doesent really matter, as long as one of them is the R3 is connected to GND.
I have C2 shorted, and C17 i did cut the short, before I looked at the schematic 😉

Well, output on TL071 (pin6) = 0.455vdc
VCC- to GND = -15.13vdc
VCC+ to GND = +15.13vdc
IN+ to GND = 0.2vdc
IN- to GND = 0.2vdc

Offset today, with input shorted = 103mVdc
Output from PSU = +23.4 / -23.4

Jesper...

Schem_new.PNG


TL071BCP.PNG
 
#909
Do you think you have a genuine TDA7293?

Other places to look for DC offset are small solder bridges or leftover flux making contact between pins that are very close together. I recently fired up a 40W class A amp based and one channel had 25v offset while the other was fine. Looking carefully at the solder joints I noticed a small bridge - almost inperceptible on the input stage. Reflowed the joint and used some wick to clean up fixed the issue.
 
  • Like
Reactions: lykkedk
#910
xrk971 said:
Do you think you have a genuine TDA7293?

Other places to look for DC offset are small solder bridges or leftover flux making contact between pins that are very close together. I recently fired up a 40W class A amp based and one channel had 25v offset while the other was fine. Looking carefully at the solder joints I noticed a small bridge - almost inperceptible on the input stage. Reflowed the joint and used some wick to clean up fixed the issue.
Click to expand...
Hi...

I bought them 7293 from Mouser, so i sure think they are real, no fakes. TDA7293, but offcause i can never be 100% sure.
I also think they are real, when playing music without any problems on my shop, but again you could be right ?

I must admit 🙄 that i did not clean the PCB with iso. alcohol as i allway's do, I will do, and also carefully look through the PCB in my microscope.

I'll be back within the weekend.

😎

Jesper.
 
#912
Well... having a hard time realising this is actually true 🙂

Cleaning and removing flux etc... actually lowered the offset to 18mV channel L and 6mV channel R, and when reducing the voltage on the variac i see a small rising in the offset, so i should really wait until i have the right trafo now... Please stop Jesper now 😵

Well, output on TL071 (pin6) = 0.455vdc

IN+ to GND = 0.2vdc
IN- to GND = 0.2vdc
Click to expand...

All of them are low now, the IN+ / IN- are within 20mV and the output (pin 6) around 130mV now!

I do think everything is as it should be.
I will rapport back when i have +/-37vdc rails ready.

Jesper.
 
  • Like
Reactions: xrk971
#913
Glad I could help. Are tidy and neat solder joints an extension of a tidy and neat bench top. 🙂

My bench is a mess but my solder joints are neater, thank goodness. That last solder bridge was not mine but a friend who helped assemble the amp for me.

Toothbrush and mineral spirits paint thinner first (that dissolves rosin flux better), then isopropyl alcohol and toothbrush. Some people put the board at this point under running warm water and use dish detergent and brush for final step then air dry.

Here is latest photo of my bench while debugging that solder bridge. Fixed it - DC offset went from 25V to 1mV. Rails are +/-30v. Class A with 1.7A bias current flowing. It’s a warm amp.

1674163362082.jpeg


I have a whole thread about what’s on my bench and how messy it is. Also, others post their bench photos there too.
 
#914
Hi!
I have a possibly silly question.
I have seen that the commonly used values in the R1/R3 position are 20-22k/680-750R. But I came across a connection where the creator used values of 68k/2k or 3.9k in the R1/R3 position. Can higher than normal resistance values cause problems? Also, what parameters change in this setting of the TDA7294? Unfortunately, I don't know how to use simulators (LTSpice, TINA).
Thanks for the help! D'
 
#915
The TDA7293 is spec'ed for a typical input bias current of 0.3 microamps, 1 microamp max. That current goes through the parallel equivalent of the input resistors and feedback resistors and will cause an offset.. The higher the values, the larger the offset.
 
#920
What I wanted to say, basic design (equal R1 and R2) places the offset in the proximity of zero, then the purpose of the servo is there to push the offset even closer to the zero.
Initially, have been puzzled with the poor performance of the servo but by replacing the R15 with that of 10K, as suggested here, did the magic. The servo performs well now.
Though, I have a couple of problems to solve, despite using quality components and caring assembly. I still have slight HF boost and still measure slightly increased THD values. I hope to find proper medicine for this somewhere in this thread.
Thanks guys for the great thread and especially the design team for supplying us with great amplifier.
 
  • Like
Reactions: lykkedk
You must log in or register to reply here.
Top Bottom