34: First upgrade success!

  • No one logged in.

03-Mar-2019 02:24 AM

Fabian

Fabian

Posts: 72

Back in August 2015 I've been reporting about the successful revision of my two Quad 306 power amps.http://www.dadaelectronics.eu/ForumRetrieve.aspx?TopicID=485738

 Back then I also mentioned that I had ordered DADA kits for overhauling my two Quad 34 pre-amps. One regular kit (which at that time already came with the LME49710 and LME49720 opamps. One "deluxe" kit from EJP, DADA Australia, additionally containing 5 Maxim MAX4066CPD switches.Today, not even four years later, I can confirm that I finally managed to revise the first of my two 34s. :D

 

Of course, I started with the standard kit to gaining some practice and applied it to the (seemingly) older unit. I applied a total of 4 ceramic decoupling caps around IC9 and IC10, between pins 3 and 4 and pins 3 and 7. Firstly, just to be on the safe side regarding oscillation. Secondly, because the caps did come with the standard kit. :) Placing and soldering them has been just as tricky as I had suspected.

 

To tell the truth, I don't even know if the revision was successful, yet. I managed to complete all the soldering work, but did not test the pre-amp. It's simply to late right now. However I was still in the mood to give some feedback here. I will let you know shortly if the amp works as intended. Here are some comments about the my observations and the optional steps I did perform so far.

 

I did replace all e-caps in the signal path with wire links as indicated by the DADA revision manual (C30, C31, C48, C54, C89, C90, C91, C91). Some have experienced clicking noise when turning the volume pot due to a possible DC component across the pot, I won't waste further thoughts on these issues unless I should observe excessive noise.

 

All remaining e-caps in the signal path have been replaced with non-polar caps. These did not come with the 2015 DADA standard kit, but Esmond did provide them in his "deluxe" kit. Non-polar caps have been discussed controversially. Effectively, a non-polar e-cap is just two polar caps, each twice the desired capacity, put in series. Consequentially, the resulting equivalent series resistance (ESR) is that of two bigger caps summed up. However, with modern low ESR caps there is one big advantage: Polarization of non-polar caps does not matter at all. I ordered a couple more non-polar e-caps for renovating my second Quad 34.

 

I did not make any changes to the power supply, yet. Also see my other thread "Quad 34 Mk II PS mods when using LME opamps and Maxim switches?". I just replaced the e-caps with the new parts from the DADA kit.

 

I used high quality sockets for all opamps.

 

I changed the output level from 500mV to 775mV by replacing R118 and R121 with 1k5 resistors and R119 and R122 with 1k5 as well. Now I'm keeping my fingers crossed for the first test run.



  Reply

03-Mar-2019 05:02 AM

CloseApproach

CloseApproach

Posts: 15

Thanks for the progress so far. We'll be interested in how it turned out!

  Reply

03-Mar-2019 01:57 PM

Fabian

Fabian

Posts: 72

So, here I am with more news on my first Quad 34 revision. Fist things first: The amp is up and running, no smoke, no funny smell, all buttons working as they should, crisp and clean sound from both channels (only PHONO and CD input tested so far).

 

I forgot to mention yesterday that I also opted to change C18 and C22 in the phono section from 2.2uF to 3.3uF for a less aggressive (or less effective, however you want to put it) subsonic filtering. I did not attempt any comparison with the unmodified unit yet (time is short, it's Sunday and I rather listen to the fully revised rack right now), so I can't say if there is a noticeable difference. However, contrary to common believe rega turntables arenot very prone to subsonic noise anyway, so this was a no-brainer for me.

 

More details on the tiny decoupling caps around IC9 and IC10: Soldering them into place is not a pleasant task for a nearsighted person over 50. Glasses on and off and on and off and on again. In the end I decided to keep all legs as short as possible. The cap between pins 3 and 7 resides directly between pins 1 and 8, it touches pin 3 from the left side. The other cap connected between pin 3 and pin 4 sits right in the center between pins 3, 4, 5 and 6. It touches pin 3 from the right side. I cut and bend the legs into place and than fixed the caps to the pcb using a drop of solvent-free glue. Solvent-free is synonymous with non-adhesive, by the way. :-P I think, this was the most time-consuming job in the whole rebuild for me.

 

My 34 does not suffer from clicking noise at all when moving any of the three pots. This should mean that there is no noteworthy DC present, as one should expect from modern day opamps. Consequentially all wire links will be kept in place as recommended by the revision manual. I will approach my second 34 in the same way then.

 

I'm still waiting for the non-polar caps to arrive that I will be using for C10, C11, C16, C17, C77 and C78. I had used those provided by EJP for the "standard " kit, but I won't put the polar ones that are left now in the other pre-amp. I went with Nichicon VP 16V types for no better reason than convenient size, availability and brand name. Let's wait and see.

 

Lessons learned during desoldering the old components: Too strong a desoldering pump is at least as unhelpful as a too weak one. I reverted to my older one (looks like I'm buying everything in pairs, but that's not exactly true) during the rebuild process. The rebound from the new one was so high that effectively more solder spread over the PCB. Also the solder pad on pin 8 of IC20 got sucked in completely. Luckily this pin is not connected to anything in the 34, so there was nothing to repair. Cutting IC off from the component side pior to desoldering is impossible to do on the 34 due to space restrictions. I did not find this to be a problem at all. In fact, desoldering the ICs was almost the easiest part of the whole rebuild, when I had expected it to be the hardest part. A good old Weller soldering station and a (not too strong) solder sucker were more than appropriate for the job. I did cut some of the e-caps prior to desoldering them. This was a mistake, I will not do that again. A permanent hunt for tiny leftovers of the legs was the only result. Not helpful.

 

Sound wise it might be too early for a final verdict, but the 34/306 combo now sounds great to my ears, cd and phono. The amp is really extremely quiet now. Absence of hum, noise and unpleasant distortion might not sound like a high-end goal to some. For my taste these are just the attributes that will let music come alive in your living room without distraction. These features will provide the necessary acoustical "darkness" for placing musicians in the right place, for providing deep but well-controlled bass response and for clean and hiss-free sibilants. So far, all I can say is: Why didn't I do this 3 and a half years ago already? ;-)

 



  Reply

07-Mar-2019 08:27 PM

Fabian

Fabian

Posts: 72

Next steps completed.

The otherwise unmodified Quad 34 has been set to the same output level as the modified one for easier comparison (775mV, R118 = R121 = R119= R122 = 1k5Ohm).

In the modified one I uprated the main PSU smoothing cap C74 from 1000uF to 2200uF and the rail decoupling caps C58, C84 from 22uF to 220uF, even if there is nothing to be gained from this mod, At least the transformer and regulator did not complain.

I also replaced the output caps C77, C78 with Nichicon MUSE ES (100uF, 6.3V, BP). In the other 34 (the one that will be equipped with EJPs deluxe kit), all 6 remaining coupling caps will be of this type.

Finally I decided to also fit additional 100nF ceramic decoupling caps to IC7 and IC8 from +ve to ground and -ve to ground, just because it can't hurt.Or can it?  Well, literally in the very last moment I recognized that I had soldered one of them between pins 3 and 6 instead of pins 3 and 7. :-P Error corrected, no harm done.

Next up:

  • Thorough comparison between the revised and the unrevised unit.
  • Sourcing components for reduced input sensitivity of radio input
  • Replacing the 3.3uF electrolytic capacitors (after revision) C18, C22 with Wima MKS2 film caps, they should fit just perfectly.
  Reply

19-Mar-2019 10:54 PM

Fabian

Fabian

Posts: 72

Today I changed C18 and C22 again, from 3,3µF electrolytics to 3.3µF WIMA MKS2. I did not even bother to try a before/after comparison. Even a cheap polyester film cap is technically way better than the best electrolytic, plus it will last forever. That alone was enough of a reason for me to make the change.

The 3.3µF MKS2 is available in a rather tiny case with 5mm lead spacing. It's still just a tad too thick, so C18 sits partly on top of R33. C22 also won't make it all the way down to the PCB, at least when using sockets for the ICs (as I did). Still very easy to put them in place. The legs are exactly the right length to solder them. No cutting required.

As I said, I did not test the effect of this modification alone. In fact, I did not yet take the time to compare the revised preamp to the still unmodified (except for output level) unit. But I can say for sure that my current setup sounds amazing to me: Rega Planar 8 with AT VM540ML, Quad 34 (revised as described above, phono MM input capacity reduced to 47pF), 2 x Quad 306 (revised with Dada kits) bi-amping a pair of KEF R7. I just listened to the famous Opus 3 recording "Tiden bara går", performed by Thérèse Juel (Yes, I really DO love this weired music!). I don't think I have ever heard this masterpiece sound any better. Period.

  Reply

22-Mar-2019 08:09 PM

Ton

Ton

Posts: 52

Hi Fabian,

your posts about refurbishing your quad 34’s are great fun to read.

I have a question about one of your replacements.

You  replaced the output caps C77, C78 with Nichicon MUSE ES (100uF, 6.3V, BP). 

In the dada upgrade are 10V electrolytics. Why didn’t you buy ES type 16V Muse’s to be on the safe side ?

Is it because of the lead spacing of the 6.3V (I think 3.5 mm) against the 5 mm of the 16V ?

Regards, Ton

  Reply