Hi all,
I built a linear supply module to replace the stock SMPS in my DCX2496.
For the +/- 15v I used a 19Vct 30VA (this transformer I recycled from a Primare CD20 cdplayer) tx to supply to LM317/337.
For the 8.7V I used another tap from the Primare tx (10.8v ac) to supply into an LM7808 and have a 0.7v diode to lift the ground to get 8.7vdc.
For the 5V and 3.3v I used a separate transformer of 10VA 12v to supply into separately rectified bridge feeding LT1083ct-5 and LT1085ct-3.3.
When I connected everything (loaded) into the DCX I got +/-15v, 7.7v, 4.93v, 3.27v. All the regulators are only warm to touch and nothing is burning hot which means no shorting or unwanted oscillations.
Now comes to problem. Only the green background light display is on, the DCX does not boot up at all.
When I replaced with the ori SMPS supply, the DCX boots up normally.
I also pay special attention to ensure the gound likes of the analog and digital are separated by a 0.1uF cap.
Where am I wrong?
I built a linear supply module to replace the stock SMPS in my DCX2496.
For the +/- 15v I used a 19Vct 30VA (this transformer I recycled from a Primare CD20 cdplayer) tx to supply to LM317/337.
For the 8.7V I used another tap from the Primare tx (10.8v ac) to supply into an LM7808 and have a 0.7v diode to lift the ground to get 8.7vdc.
For the 5V and 3.3v I used a separate transformer of 10VA 12v to supply into separately rectified bridge feeding LT1083ct-5 and LT1085ct-3.3.
When I connected everything (loaded) into the DCX I got +/-15v, 7.7v, 4.93v, 3.27v. All the regulators are only warm to touch and nothing is burning hot which means no shorting or unwanted oscillations.
Now comes to problem. Only the green background light display is on, the DCX does not boot up at all.
When I replaced with the ori SMPS supply, the DCX boots up normally.
I also pay special attention to ensure the gound likes of the analog and digital are separated by a 0.1uF cap.
Where am I wrong?
Hi Will,
What do you mean by the above? Is the diode in series with the 7808 ref pin (gnd pin)?
What do you mean by separating the analog/digital grounds?
jd
Hi Jan,
For the Analog supply since I was a bit worried that 8V supply might not be sufficient and I took the easy way out, I installed a diode between on the ground pin to make it 8.7v supply to the Analog portion.
For the analog/digital gounds I just mimic this circuit:
http://www.awdiy.com/uploads/pdf/DEQ2496-PSU-DIY-v2-1.0.pdf
Since my DCX cannot boot up I believe the issue is related to the 5V and 3.3V supplies right?
Hi Will,
I first have to ask: What motivated you to replace the stock SMPS?
Also, your "7.7V" was a typographical error, I assume.
And why not just use an adjustable regulator to get the 8.7 Volts?
And I guess I don't understand what you are actually accomplishing with the 100 nF between the analog and digital grounds. Maybe you could just leave them separate, instead of joining them that way. A lot might then depend on the exact layout, and on whether or not (and from where to where) you connected the second 100 nF shown in the circuit in the PDF link you gave, which went from DGND to "Ground", near the one from AGND to DGND that you mentioned using.
So maybe you even changed too many things at one time and now it's more difficult to know where a problem might have been introduced.
Anyway, assuming your voltmeter is calibrated and also reads 0v when the leads are shorted together, why are the 5.0v and the 3.3v rails slightly lower than spec? Can they supply enough current? Do you know the specs for the maximum current needed for each supply rail? What about the voltage ripple specs?
Assuming you know what you're doing and know all of the risks, if you had some sufficient dummy loads and were careful with grounds etc, maybe you could try using the original SMPS supply for all of the rails except one at a time from your power supply (or probably two at a time in the case of the +/- 15V).
It would also be interesting to look at each rail with an oscilloscope, for each supply.
Cheers,
Tom
I first have to ask: What motivated you to replace the stock SMPS?
Also, your "7.7V" was a typographical error, I assume.
And why not just use an adjustable regulator to get the 8.7 Volts?
And I guess I don't understand what you are actually accomplishing with the 100 nF between the analog and digital grounds. Maybe you could just leave them separate, instead of joining them that way. A lot might then depend on the exact layout, and on whether or not (and from where to where) you connected the second 100 nF shown in the circuit in the PDF link you gave, which went from DGND to "Ground", near the one from AGND to DGND that you mentioned using.
So maybe you even changed too many things at one time and now it's more difficult to know where a problem might have been introduced.
Anyway, assuming your voltmeter is calibrated and also reads 0v when the leads are shorted together, why are the 5.0v and the 3.3v rails slightly lower than spec? Can they supply enough current? Do you know the specs for the maximum current needed for each supply rail? What about the voltage ripple specs?
Assuming you know what you're doing and know all of the risks, if you had some sufficient dummy loads and were careful with grounds etc, maybe you could try using the original SMPS supply for all of the rails except one at a time from your power supply (or probably two at a time in the case of the +/- 15V).
It would also be interesting to look at each rail with an oscilloscope, for each supply.
Cheers,
Tom
I agree with Tom above that the cap between the two gnds doesn't do anything usefull; the grounds should of course be connected but in a controlled way. On my replacement linear supply board I do it just at the output connector.
Have you tried just to jumper the two grounds to see if that solves it?
Tom's idea of successively replacing each supply might work, but it also might introduce problems when the supploes come up with a delay between each other.
jd
Thanks for the replies guys.
Yes the A and D ground separated by the cap does not make sense according to the AWDIY website and I have grouped them together albeit without success.
I got a feeling perhaps it's the 5V and 3.3V supply sag (or my transformer don't have enough juice) because I'm measuring 5.01V and 3.32V on the SMPS supply wherelse using the LT1083-5 I'm getting 4.92V and 3.27V which is slightly lower. Is the DCX really that sensitive?
Let me try to use a larger VA transformer and see.
Yes the A and D ground separated by the cap does not make sense according to the AWDIY website and I have grouped them together albeit without success.
I got a feeling perhaps it's the 5V and 3.3V supply sag (or my transformer don't have enough juice) because I'm measuring 5.01V and 3.32V on the SMPS supply wherelse using the LT1083-5 I'm getting 4.92V and 3.27V which is slightly lower. Is the DCX really that sensitive?
Let me try to use a larger VA transformer and see.
I would be surprised if that would solve it.
Are you absolutely sure of all the voltages on the mainboard and front panel boards? Not just on the supply board, but after the connections to the main board?
What do you see on the front panel display when you switch the unit on?
jd
I measured the voltages on the 7pin white connector at the DCX board - using a sharp probe. Replacing the linear supply with its original SMPS and re-measuring at the same place gives almost similar voltage albeit higher than the linear supply.
Using the linear supply I don't even get a welcome screen of "DCX2496"
With the SMPS it's a breeze to boot up.
Every regulator on the linear supply namely LT1085ct-3.3, LT1083ct-5, LM7808, LM317/337 has its own bridge rectifier using discrete MUR860 - way overkill but that's how I like it.
A possible cause is that the supplies come on slightly delayed or in the wrong order. You could try with a single rectifier, to make sure they all come up at the same time.
Is there a large difference in capacitors after the rectifiers? I guess all rectifiers work from the same secondary? If so, just temporarily shorting all rectifier outputs together ensures that they start up the same time.
jd
That's good input I will take note on this. Currently the 5V and 3.3V are connected to the same AC winding but have their own bridge, and the smoothing caps are the same values. I will try to tap the 3.3V off the 5V supply and see if it works. Thanks.
Thanks Jan !
I made the changes of using a more powerful 12V ac transformer feeding a single bridge supplying all 3 regulators of the 8V,5V and 3.3V and it WORKED !
That means it WAS indeed a power up timing issue. Such crude wiring of mine even without proper smoothing caps at the regulator's output and yet it WORKED!
Perhaps on my previous attempt of lumping the 5V and 3.3V together, I should have connected supply to the 5V and 3.3V regulator in parallel instead of cascading the 3.3V regulator after the 5V regulator. Let me re-try that again.
However, I would preferably like to isolate the 8V analog supply away from the digital 5V and 3.3V.
More experiments to perform, but at least I know what works and what don't.
Thanks again.
I made the changes of using a more powerful 12V ac transformer feeding a single bridge supplying all 3 regulators of the 8V,5V and 3.3V and it WORKED !
That means it WAS indeed a power up timing issue. Such crude wiring of mine even without proper smoothing caps at the regulator's output and yet it WORKED!
Perhaps on my previous attempt of lumping the 5V and 3.3V together, I should have connected supply to the 5V and 3.3V regulator in parallel instead of cascading the 3.3V regulator after the 5V regulator. Let me re-try that again.
However, I would preferably like to isolate the 8V analog supply away from the digital 5V and 3.3V.
More experiments to perform, but at least I know what works and what don't.
Thanks again.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.