PRR

Re: Bandwidth problem
« Reply #20 on: December 25, 2018, 10:06:20 PM »
Abbey dislikes the linear scale on *frequency*.

Bothers my eyes too. I see the dB scale on amplitude, and don't understand dB up but linear sideways.

But it looks like a 20KHz high-cut as commonly used for digital audio.


JAY X

Re: Bandwidth problem
« Reply #21 on: December 26, 2018, 01:05:52 PM »
Hi!

Less good news...: I have 10.80mv ripple p-p in the PSU.
That's about -40dbu noise, and modulates the audio band.. :(

The reservoir caps are 1000uf 35v. Maybe i have to increase them?...rebuild the PSU?

Jay x


abbey road d enfer

Re: Bandwidth problem
« Reply #22 on: December 26, 2018, 01:11:08 PM »
Hi!

Less good news...: I have 10.80mv ripple p-p in the PSU.
That's about -40dbu noise, and modulates the audio band.. :(

The reservoir caps are 1000uf 35v. Maybe i have to increase them?...rebuild the PSU?

Jay x
How much is the current draw?
1000uF is huge if you draw 1mA, but it's small if you draw 10 A.
Are there regulators? How can we evaluate the PSRR (Power Supply Rejection Ratio) of your circuit without a schemo?
You need you to help us help you...
Who's right or wrong is irrelevant. What matters is what's right or wrong.
Star ground is for electricians.

JAY X

Re: Bandwidth problem
« Reply #23 on: December 26, 2018, 01:37:57 PM »
A PSU picture.

JAY X

Re: Bandwidth problem
« Reply #24 on: December 26, 2018, 01:46:31 PM »
A bit too Big the picture...sorry!.
Tomorrow i Will post a New one.

In any case, for such ripple, the current draw must be 635ma! For 7 opamps..doesn't make sense!.

abbey road d enfer

Re: Bandwidth problem
« Reply #25 on: December 26, 2018, 01:55:51 PM »
I have no problem visualizing your pic, but you don't show the characteristics of the transformer. Maybe the voltage is a little weak so the regulators are starved...? What is the voltage at the input of the regs? What is the current draw?
Who's right or wrong is irrelevant. What matters is what's right or wrong.
Star ground is for electricians.

JAY X

Re: Bandwidth problem
« Reply #26 on: December 27, 2018, 09:45:06 AM »
Hi!

Transformer is 18V AC.

Voltages at regulators inputs:
+25,8v and -25,9v.

I Also placed two 220uf/35v capacitors at output rails, before all opamps.

I did not measure current, because i would need to cut supply rails...I will try later.

Jay x

JAY X

Re: Bandwidth problem
« Reply #27 on: December 29, 2018, 01:29:28 PM »
Hi!

Current draw is 31ma/rail aprox.

But i think,the main problem is the PSU. I think the joemeek vc3q, and the RNC compresor, both have voltage doublers PSU, and less ripple.

Attached ripple image output Signal [email protected]

JAY X

Re: Bandwidth problem
« Reply #28 on: December 29, 2018, 01:32:07 PM »
Another picture when there is no Signal at output.

Re: Bandwidth problem
« Reply #29 on: December 29, 2018, 02:48:37 PM »
Less good news...: I have 10.80mv ripple p-p in the PSU.
That's about -40dbu noise, and modulates the audio band.. :(

The reservoir caps are 1000uf 35v. Maybe i have to increase them?...rebuild the PSU?
Are you sure that's "real" hum and not just a probing / measurement problem? -40dBu is huge given the regulators and the PSU circuitry you're using. Or, the ground arrangement in your regulator circuit is very sub-optimal - if the regulators are "standing" on top of hum, that'll make it into the PSU rail outputs unattenuated.


benb

Re: Bandwidth problem
« Reply #30 on: December 29, 2018, 07:57:06 PM »
Yeah, that hum is a sine wave, most likely something not grounded right. Increasing the reservoir capacitor size won't change this. Look directly across the cap, if there's anything there, it's a sawtooth wave. Don't even worry about the amplitude as long as the lower peak is above 20V, the regulator fixes that.

JAY X

Re: Bandwidth problem
« Reply #31 on: December 30, 2018, 08:44:23 AM »
Attached is my PSU layout. Made in one layer PCB.

The yellow trace is a link from the single point Ground, to the rest if the circuit.

Jay x

JAY X

Re: Bandwidth problem
« Reply #32 on: December 30, 2018, 02:17:47 PM »
Hi!

I checked the solderings, and re-did some jumpers. Also i replaced the filter caps. Now, vp-p is 5,7mv. 926uv RMS.

Still too high. I will try a C-R-C filter, but i don't expect much improvement.

Jay x

JAY X

Re: Bandwidth problem
« Reply #33 on: December 31, 2018, 03:35:47 AM »
Hi!

last night it occured me that i could add snubber caps in parallel with the diodes: 10 or 100nf 100v .
Also i could add an x2 cap in parallel with the  AC input socket, but i'm not sure about this.


In advance, ¡¡Happy new year!!

Jay x

abbey road d enfer

Re: Bandwidth problem
« Reply #34 on: January 01, 2019, 02:16:17 AM »
Another picture when there is no Signal at output.
Then what is it you're measuring? The output of your circuit? What's in it? If it has 100dB gain you shouldn't worry about it... :o
Can you measure the actual ripple on the rails?
Who's right or wrong is irrelevant. What matters is what's right or wrong.
Star ground is for electricians.

JAY X

Re: Bandwidth problem
« Reply #35 on: January 08, 2019, 10:52:33 AM »
Hi!

Sorry for the lack of communication last days. ::)  At home i have no scope to measure the ripple, but i will try this week at the lab of the institute where i studied... 8)

In any case i have build two C-R-C filters:   470uf-22R-470uf  and  1000uf-22R-1000uf. I will test both filters.

Jay x


JAY X

Re: Bandwidth problem
« Reply #36 on: January 09, 2019, 12:34:41 PM »
Hi!

I was surfing the net and  a question arose my mind... ¿does it make sense to use an AC input filter in voltage doubler circuits?

like this one:   https://uk.rs-online.com/web/p/power-line-filters/4932989/

Jay x


abbey road d enfer

Re: Bandwidth problem
« Reply #37 on: January 10, 2019, 04:54:02 PM »
Hi!

I was surfing the net and  a question arose my mind... ¿does it make sense to use an AC input filter in voltage doubler circuits?

like this one:   https://uk.rs-online.com/web/p/power-line-filters/4932989/

Jay x
No more no less than with other types of rectifiers.
Who's right or wrong is irrelevant. What matters is what's right or wrong.
Star ground is for electricians.

JAY X

Re: Bandwidth problem
« Reply #38 on: January 11, 2019, 01:18:31 PM »
No more no less than with other types of rectifiers.

I suspected that!!.  But as always...I start designing from a schematic...and forget to add all those bits... ::).

Today i tested the PSU ripple, and well, with 1000uf/22 Ohm, the ripple was lower, but i measured the frequuency of ripple, and is about  22mhz. So maybe i need smaller bypass caps and perhaps 22mh chokes.

In any case, I will post the measurements as a Word file attachement ASAP.

Jay x.

radardoug

Re: Bandwidth problem
« Reply #39 on: January 11, 2019, 04:51:56 PM »
That doesn't make any sense. You are saying the frequency of the ripple is 22mHz? 22 milliherz? Please try and be precise.
Otherwise its very difficult for us to help you.
Anyway, the ripple frequency in a linear mains supply will be at or a multiple of the mains frequency, so 50 Hz, 100Hz, 150 Hz.
Or the equivalents for 60 Hz supply.
In a switchmode supply you would see ripple at the switching frequency, normally 10s of kHz.


 

Related Topics

  Subject / Started by Replies Last post
16 Replies
4184 Views
Last post July 05, 2005, 09:28:18 AM
by SSLtech
39 Replies
4880 Views
Last post July 16, 2008, 02:20:10 PM
by Svart
4 Replies
1325 Views
Last post April 03, 2012, 12:26:49 PM
by ruffrecords
4 Replies
1175 Views
Last post January 29, 2013, 02:54:29 AM
by PRR