QUAD33 REGULATED PSU

  • No one logged in.

14-Jun-2019 01:37 PM

LOWTHERMAN62

LOWTHERMAN62

Posts: 50

Hi All,

I have upgraded one of my 33 preamps to a regulated PSU using an L7815cv regulator with 2 diodes in4002, a 2200 35v cap and a 10uf 35v cap in the PSU. Changed R300 to a wire link, R305-306 to 133k and R315-316 to 6k5 as per the blog information. Sound very clear except  for  a low level hum on both channels. This is constant low level and is the same level not affected by the volume control. I have treble checked all components but have drawn a blank. Before the conversion the preamp was quiet. i have checked for ac accross green and red and it is zero. Note dada revision has recently been done. Regards  

  Reply

14-Jun-2019 10:15 PM

LOWTHERMAN62

LOWTHERMAN62

Posts: 50

Hi again.

Just to rule out any interference taht i have read about on the forum I have checked and plugged in my 2nd 33 which has had the full dada revision including the 16volt zener conversion. This  soonds clean with just a hint of noise at full volume, no hum whatsoever. Note I have NOT revised the gain mods on any of the two units. Any thoughts would be much appreciated. Regards.

  Reply

15-Jun-2019 06:22 AM

EJP

EJP

Posts: 1617

This could be inadequate PSU voltage going into the regulator. Measure across the 2200uF capacitor. You need at least 18V to provide the 7815 with enough headroom. If you don't have it, change to a 7812 regulator and put back the resistors the way they were.

I always use a 12V regulator in this modification, to be sure of enough headroom. The measured difference between 12V and 15V running is insignificant.

  Reply

15-Jun-2019 11:01 AM

LOWTHERMAN62

LOWTHERMAN62

Posts: 50

Thanks for your reply EJP,

I have taken the following measurements

Accrross 2200 cap 6.03 volts dc. 0.029 volts ac. Accross red and green on psu board 4.98 volts dc. 0.010 volts ac. This obviously is not correct. Any thoughts. Regards. 

  Reply

15-Jun-2019 11:55 AM

EJP

EJP

Posts: 1617

None of it is right. Check all your work. Diodes should have the cathode mark inboard. The 2200uF capacitor could be back to front. Remove the 7815 and see what the input voltage is. Check how you inserted the 7815 against the blog article. Check the insertion and position of the 10uF capacitor. Remove all the removable boards and check everything again: if it comes good, put them back one by one until it goes bad, then the last board you replaced is the culprit.
  Reply

15-Jun-2019 01:12 PM

LOWTHERMAN62

LOWTHERMAN62

Posts: 50

Hi EJP, Checked all and you were correct. Cap the wrong way round.  Problem is we are dealing with a mirror back to front image on the blog so my honest mistake.Have fixed and now all good. Going to spend some time listening but initial impressions are excellent. Thank you so much for your help in this matter . I t might be a good idea if this could be revised to show a picture or schematic of the regulated connversion from the component side if the board. Kind regards Stewart. 

  Reply

27-Jun-2019 04:58 PM

LOWTHERMAN62

LOWTHERMAN62

Posts: 50

Hello all, I am about to convert my older 33 to a regulated psu. On taking voltage redings from transformer out i am only getting 17.01 volts ac. The question is will the LM2937et-15 have enough voltage to work properly and avoid hum.

Regards Stewart.

  Reply

27-Jun-2019 08:03 PM

LOWTHERMAN62

LOWTHERMAN62

Posts: 50

Sorry for the omission. The 33 in question has had the full dada revision done about 2 years ago. and works perfectly with the 16volt zener update. Recently i have replaced the volume pot and suppressor with a da da supplied replacement. Regards Stewart

  Reply

28-Jun-2019 04:31 AM

EJP

EJP

Posts: 1617

I would say that's marginal, but if you already have the 15V regulator you may as well try it. I think you will probably have to drop back to 12V.
  Reply

28-Jun-2019 09:10 AM

LOWTHERMAN62

LOWTHERMAN62

Posts: 50

Thanks EJP. I have all three types so will try 15v first and report back. Regards Stewart

  Reply

28-Jun-2019 09:14 AM

EJP

EJP

Posts: 1617

Let us know what the input voltage is to the regulator after you fit it. Measure at the top of the 1000uF capacitor. Then measure AC at the output, i.e. top of the 10uF capacitor.
  Reply

28-Jun-2019 09:29 AM

LOWTHERMAN62

LOWTHERMAN62

Posts: 50

Can the 10Uf cap location be changed to the old zener position with negative facing towards the centre of the board as this is the position if the tracks are traced back. This would maybe be a better solution for placement of this cap? I seem to recall a blog which stated the cap in this position but with the same polarity as the diodes?Any thoughts?.Regards Stewart

  Reply

28-Jun-2019 10:03 AM

Joost Plugge

Joost Plugge

Posts: 967

See my blog post on 

FRIDAY, AUGUST 10, 2007

I used a low drop 15V regulator, you can't use a normal 3 pin regulator, as Esmond already mentioned. I think in that blog post you will find all your questions answered.

Joost

  Reply

28-Jun-2019 10:55 AM

EJP

EJP

Posts: 1617

See my comment to Joost's blog. The 10uF capacitor should indeed be where the zener was, but you have the polarity back to front. Positive towards the centre of the PCB, or 'up' when it is installed back into place.
  Reply

28-Jun-2019 11:04 AM

LOWTHERMAN62

LOWTHERMAN62

Posts: 50

Thanks EJP. The regulator that i am planning to use is an LM2937ET-15 in a to220 package which is 3 pins? Much the same configuration as the 7812? Regards Stewart

  Reply

28-Jun-2019 11:51 AM

EJP

EJP

Posts: 1617

Check the pinout for yourself. There is no substitute.
  Reply

28-Jun-2019 10:11 PM

LOWTHERMAN62

LOWTHERMAN62

Posts: 50

Hi EJP. I have fitted the LM2937ET-15 and reinstalled the PSU . Note following readings are with all removable cards out.Top of 2200uF cap 29.28 VDC

Top of 10uF cap 0.088 VAC and 14.99 VDC

Red Green from PSU 14.99 VDC and 0.086 VAC .

Substituting for a7812 I get the following readings:- 

Top of 2200uF cap 29.92VDC

Top of 10uF cap 12.02VDCand 0.002VAC. I havent had time to amend the board to suit 15volt psu, but listening to the 12 volt regulated supply seems defined, is very clear, low noise at full volume, with a little more defined base than I remember (all subjective) 

Not sure about the AC on the 15 volt supply?Regards Stewart 




  Reply

29-Jun-2019 04:31 AM

EJP

EJP

Posts: 1617

The unloaded DC is of no interest. I need the input DC as loaded by all the boards, and the AC ripple at the output of the regulator.
  Reply

29-Jun-2019 08:16 AM

Joost Plugge

Joost Plugge

Posts: 967

It will work, the dropout voltage is 1V max with a load of 500mA. I tested it long ago. The 12V is safer, but the 15V lowdrop will work. I even tested the 33 with higher external voltages. The distortion readings will not change, the noise is dominant. I did not check the subjective performance, because then anything is possible, see all the HiFi Magazines. 

Joost

  Reply

29-Jun-2019 09:34 AM

EJP

EJP

Posts: 1617

... subject to the usual caveats about line voltage, transformer variation, etc.
  Reply

29-Jun-2019 02:32 PM

FBatista

FBatista

Posts: 125

Hello,

I too tested the mentioned power supply topologies with the 33.

The DADA 16v zener one, is the most straight forward, easier to apply and had no drawbacks.

The use of regulators implies the annoying initial buzz when powered on. The 33 has no delay circuit switching the output to overcome this aspect, like the 34 has, for instance). I've tried both the 7815 and LDO regulators.

Soundwise none of the above has any audible effect, except for the phono input, benefiting with the use of a regulated PS. If you don't use this input, you're better using the 16V zener and removing both internal phono preamp and adapter boards. this will save you 0,1V (mine went from 15,85V to 15,96V) - sound very litle but not, once compared with the limited trafo the 33 originaly has.

Nevermind the use of bigger or toroidal transformers inside the 33 chassis. They will certainly radiate over the slope filter coils. Only use them if encapsulated with a good thick sheet of metal, like the original 33.

This sums up my experience with the 33 around this subject.

I like it a lot - exactly the way it sounds - and use one (zener 16v and no phono boards configuration) everyday in my desk with a 303 and FM3.

Great words from Joost regarding his experiences on sound, distortion and external PS. You should definitely consider his post.

Best,

  Reply

29-Jun-2019 03:55 PM

EJP

EJP

Posts: 1617

My experience is otherwise. I both hear and measure considerably lower THD+N and better channel separation with a regulator in place compared to the zener-based circuit.

The Phono Adapter board doesn't draw any current and doesn't need to be removed in the scenario you mention.

  Reply

29-Jun-2019 10:03 PM

LOWTHERMAN62

LOWTHERMAN62

Posts: 50

Hello EJP, I have now reinstalled the lm2937et-15, 10uF cap replaced to zener position polarity as advised, 2200uf cap and IN4002 diodes.All boards installed and measurements as follows:-

accross 2200uF cap 20.82vdc 0.00 vac

red green on psu 15.00vdc 0.00vac

pin 1 input regulator 20.82vdc  0.00vac

pin 3 output regulator 15.00vdc  0.00vac

  Reply

01-Jul-2019 04:42 AM

EJP

EJP

Posts: 1617

These are all very good numbers, but you should measure AC ripple in mV, on the 200mV scale if you have one.
  Reply

01-Jul-2019 10:01 AM

LOWTHERMAN62

LOWTHERMAN62

Posts: 50

 As advised and measured on my fluke mVsetting (autorange) 

Red green on PSU 0.001mVAC

Pin 3 output of regulator 0.001mVac 15.00vDC

These look good.

  Reply

09-Jul-2019 10:26 AM

FBatista

FBatista

Posts: 125

That's great EJP! I wouldn't mind trying it again. In your experience did it sort out the crosstalk between channels?


LOWTHERMAN62 > What do you reckon regarding the update on your 33? Are you pleased with it? Did you sort the initial buzz from the regulator?

Best to all,

FBatista

  Reply

09-Jul-2019 03:06 PM

LOWTHERMAN62

LOWTHERMAN62

Posts: 50

Hi All, Foolow up to the ldo regulator update. As you can see numbers are all very good . Have had a lot of time listening and this is now a great little preamplifier. I have another with the 16V zener update ans subjectively i do think that the LDO is better. I observe the sound to be precise and clear,  bottom end is defined and soundstage is good. I am very pleased with this update and would recommend this . It is not difficult but double check orientation of the caps etc as i did make a mistake despite being (I thought ) methodical. There is still a very slight low level switch on sound that lasts for a fraction of a second . PS I used a 100uF cap instead of the 10Uf  output Also I have 2 spare 15 volt LDO regulators if anyone wants one . Regards Stewart 

  Reply

11-Jul-2019 06:45 AM

EJP

EJP

Posts: 1617

Yes, crosstalk was improved, along with stereo imaging as I said above and several other things. Cymbals went from smears to point sources. You can reduce either the volume or the length of the initial hum by increasing or reducing the input capacitor. I normally use 1000uF but occasionally 2200uF (softer and longer) if the customer wants less noise.
  Reply

11-Jul-2019 06:55 PM

LOWTHERMAN62

LOWTHERMAN62

Posts: 50

Hi EJP,

 Can you tell me why there is that slight initial noise on switch on, I dont find it a problem but at the same time it would be great to get rid of or minimise. Is there anything else that can be done to improve this? Regards Stewart.

  Reply

12-Jul-2019 01:50 AM

EJP

EJP

Posts: 1617

The regulator can't regulate until the input voltage reaches its dropout voltage, which in the case of an LM7815 is 17-18V. Hence the concern above about the input voltage (loaded). Below the dropout voltage it just lets the 50/60Hz hum straight through. A bigger capacitor on the input side shunts more of the hum to ground, but takes longer to charge up to the dropout voltage. I accidentally used a 10uF input capacitor once and the hum was really short but really loud.

The only cure is a delayed output muting circuit like in the 34, 44 etc. I tried a really big capacitor like 1000uF on the output side where the 10uF one is in the blog, but it didn't help at all.

  Reply