LME & Maxim = hum in Quad 34?

  • No one logged in.

07-Jun-2018 06:06 PM

The Quadfather

The Quadfather

Posts: 110

Hi there.

I’ve just gone and replaced all the op-amps in my grey Quad 34 with sockets, and have installed a full set of LME op-amps which I purchased from DaDa some time ago.

The RIAA/Phono section op-amps were replaced with the ‘tin hat’ versions of the LME op-amps a few years back, which have functioned perfectly. Also back at that time (2015) I replaced the 4066 switches with the MAXIM types.

I have added the recommended 100nF decoupling caps between pins 3 and 4 and 3 and 7 on IC9 and IC10.

It works and sounds great on CD, but when I switch to phono I now get some hum on both channels. Curiously, if I alter the position of the balance control away from centre the hum vanishes in both channels.

If I switch to the Radio input (there’s nothing actually connected to the 34 here) I get the low level hum there too.

I’m at a loss. Is the combination of all LME op-amps AND the MAXIM switches too much for the 34s power supply, or does this hum mean more decoupling is needed? Strange how the CD input is hum free though.

Any suggestions?

Thank you.

John.


  Reply

08-Jun-2018 06:09 AM

EJP

EJP

Posts: 1505

John

The bypass capacitors across IC9/10 are at the volume control, and affect all channels, curing a hum which is at its worst at zero volume.

If you only have hum on a couple of channels, the problem is elsewhere, and earlier. It may be mechanical. Check the flat ribbon connector from the phono PCB, both ends; check that the screws securing the phono insert card are tight; and check the soldering on the Phono and Radio input socket(s). These are known to crack over long periods of time, and develop a ring of poor or non-conduction between the pins and the PCB.

The 4066s don't have anything to do with it, and the power supply on my 34 hasn't cracked up under the strain either.

EJP

  Reply

12-Jun-2018 05:50 PM

Joost Plugge

Joost Plugge

Posts: 962

In some cases it is also necessary to decouple the op-amps in the phono section. It is a inverting stage so connect the 100nF between pin 3 and 4 on Ic7 and between pin 3 and 7 on Ic8. Two caps in total will do the trick. Quad used long and complicated routed power lines for the Ic's and ignored the advice of every op-amp manufacturer, even in those days!

Joost

  Reply

13-Jun-2018 03:15 PM

The Quadfather

The Quadfather

Posts: 110

Thanks for the replies EJP & Joost - I appreciate your help :)

Checked the phono connectors and all fine there. Then I noticed that I'd put 100nF caps ACROSS pins 4 and 7 on each phono op-amp (following the advice on another audio forum 3 years back - I'd forgotten about this). In light of Joost's comments about the inverting phono stage, I simply removed those two 100nF caps, and the hum has gone. It wasn't an issue before I added the other LME op-amps, but obviously was once they were in place. So it was probably never working 'by the book' before. So thanks for the advices. I'll likely follow Joost's advice and add the single decoupling cap on each phono op-amp just as an extra precaution.

By the way, in addition to the upgrades in your Quad 34 manual, and capacitor bypasses and op-amp changes, I can recommend another (cheap!) tweak. Replace C77 and C78 electrolytics with Panasonic 6.8uF 250V ECWF polypropylene caps. A worthwhile and very cost effective modification, so those caps would never have to be touched again so good for longevity as well as sound. Somebody did the maths for me and the bass response is not degraded by the far smaller capacitance of the film caps. I can confirm this by listening too :)

Best wishes,

John

  Reply

13-Jun-2018 03:17 PM

The Quadfather

The Quadfather

Posts: 110

https://uk.rs-online.com/web/p/polypropylene-film-capacitors/7270054/
  Reply

13-Jun-2018 04:48 PM

The Quadfather

The Quadfather

Posts: 110

Joost I had a look at the RIAA stage on the Quad 34 and cannot see the inverting part of the phono stage. If it’s not too much trouble would you be able to indicate this area on the schematic for me? Thank you.

  Reply

13-Jun-2018 07:00 PM

The Quadfather

The Quadfather

Posts: 110

Curiouser and curiouser!

I thought the hum was cured, but that was listening with the cover off. Wih the chassis cover on, I get hum on phono and radio (not on CD). If I start to slide the cover off, the hum disappears when the cover is about 1/3 off. Or completely removed.

Any ideas on this weird issue? Usually it’s testing with the cover off that can cause a little hum. This is the opposite.

  Reply

13-Jun-2018 08:31 PM

The Quadfather

The Quadfather

Posts: 110

Joost, I just tried your suggestions of the 100nF between pin 3 & 4 of IC7 and pin 3 and 7 of IC8. It sadly hasn't made any difference the the differing levels of hum on the inputs, IF the cover is in place.

I think some Quad 34s just aren't suited to a combination of the extremely fast LME op-amps with the MAX4066 switches, even with the 100nF caps. My unit is a 1986 grey edition, PCB version is M12730-666.

As I do not have any test gear I guess all I can do it try installing my old set of OPA134 and 2134 opamps and see if the problem vanishes.

  Reply

13-Jun-2018 11:23 PM

The Quadfather

The Quadfather

Posts: 110

My apologies for another update, but I’ve just noticed around IC12 and IC13 there’s two bypass capacitors (yellow, looks like a resistor) soldered across R66 and R70 that looks like a late addition by Quad, soldered to the underside of the PCB. Could this capacitance perhaps be causing some problem when used with an LME instead of the original TL071? Those caps are the only extra components I can see on my PCB compared to earlier versions.

  Reply

14-Jun-2018 03:29 AM

EJP

EJP

Posts: 1505

6.8uF looking into a power amplifier load of at least 10k is OK, but I wonder if you realize how big polyproplyene capacitors get at these values? The ones you linked to are 23x23x17mm, which is a good chunk of a cubic inch. You'll never fit that into an early 44, and you'll have trouble on a late one.

EJP

  Reply

14-Jun-2018 08:11 AM

Joost Plugge

Joost Plugge

Posts: 962

Pin3, the positive op-amp input is connected to ground, so Ic7 and 8, part of the phono stage are inverting, like op-amp Ic9 and 10, the volume Ic's.  The hum issue with cover on or off is strange, first solve that. Are all the screws on the chassis and Pcb mounts tightented? The extra caps accros R66 and 70 limit the upper bandwith, should give no problems.The LME's draw substantial more current than TLO's, you could try increase C58 and 84 too 220u. To limit the current you could remove Ic25 as a test, it is only there to drive the tape input. Also check the pcb for cold joints esspecially the connecting pins for the upper and lower Pcb tracks.

I use 100uf bi-polars these days for interconnecting the op-amps. The distortion created by the elektrolytes is proportional to the signal voltage across them. By increasing the value you drop the signal voltage, a trick Quad allready mastered. So the 6.8u foil is working ok, but is a bulky and expensive component, as EJP mentioned. Beware of hum pickup by such large caps.

But first, fix the cover hum issue, it is not ok!

  Reply

14-Jun-2018 01:58 PM

The Quadfather

The Quadfather

Posts: 110

EJP - Yes it sounds big from the dimensions, but they both fit into the C77/78 positions in a Quad 34 nicely and are not massive like many ‘audiophool’ brands. I do not own a Quad 44.


Joost - Yes the lid issue is very perplexing, and I only discovered it by chance. My soldering work is factory neat, I can see no errors anywhere, the pcb is in fine condition. I will check all the through-hold pins as you suggest. I’m wondering if I may have a defective op-amp somewhere. I can detect no op-amps getting very hot though, just slightly warm. i’ll try your other suggestions. To simplify things first I may just pull out all the LMEs and try my OPAs.

  Reply

14-Jun-2018 07:45 PM

The Quadfather

The Quadfather

Posts: 110

Removed all the socketed LME opamps (with the exception of the directly soldered tin hat LMEs in the phono positions) with OPA134s and 2134s. No humming. No weirdness at all with the cover on or off.

I can only assume therefore that one of the LME opamps was faulty and turning the 34 into an oscillator that was sensitive enough to the positioning of the chassis cover.

OR that the issue 6 PCB plus complete set of LME plus Maxims is too much. Looking through the archive here I reminded myself I had issues here 3 years back too. It would be interesting to know if there’s a correlation between pcb revisions (and changesof components Quad used) and the stability of using LMEs and Maxims together. 

  Reply

15-Jun-2018 09:14 AM

clearsound

clearsound

Posts: 27

Hi John,

I had a similar problem with the dip LME's back in March 2015. This was just hum on the channels and I don't know if it was anything to do with the cover. If I remember correctly was  partially solved / reduced by fitting the metal can versions to all of the placements. These 34's also had the Maxim fitted. Since then both 34's have been changed substantially but if you have any of the T0-99 ( ? ) version, you might want to give them a try before going back to the OPA? I think RS do them at a reasonable cost.

Joe

  Reply

30-Jun-2018 10:27 AM

Joost Plugge

Joost Plugge

Posts: 962

Hello Joe,

Could you check your mailbox, no messages are coming trough, Demon replies.

Joost

  Reply

30-Jun-2018 11:41 AM

clearsound

clearsound

Posts: 27

Hi Joost,

Just looked at them - nothing from you so far and nothing in main BTMail box - which is odd as I got the message from the forum to check the post?

I'll send you another email from this account in case there's a problem with my address.

Cheers

Joe

  Reply

30-Jun-2018 12:36 PM

Joost Plugge

Joost Plugge

Posts: 962

Hello Joe,

Still thesame error, sorry Joost

  Reply

30-Jun-2018 02:53 PM

clearsound

clearsound

Posts: 27

Hi Joost,

 just to clarify - you can't send to me without it being bounced back ? If you did get the emails, did you try contacting me via the wifes email - the one with the 8's in it?

I can't think whats happened, as I'm still getting notifications from the forum and all other emails.

I guess if all else fails  - its back to the post office :-)

Joe

  Reply

30-Jun-2018 04:34 PM

Joost Plugge

Joost Plugge

Posts: 962

hello Joe,

Send me your 06 via the mail, so I can text you.

Joost

  Reply

10-Jul-2018 02:48 PM

The Quadfather

The Quadfather

Posts: 110

Hi Joe. Thanks for sharing your experiences. Buying a new set of tin-hats for every op-amp is too expensive for me really, plus there’s no guarantee it would help. I need to narrow this problem down to the cause.

I actually tried another set of standard (not tin hat) LME opamps but the Q34 still buzzed. It’s perfectly OK when using the set of OPA2/134s, or original TLs, so this suggests to me all soldering is fine.

There is one thing I am going to try next - the original 7918 regulator is a brand I do not recognise - looks like SSS? Perhaps it’s not actually good for 1A current. I have bought a new On Semi 7918CT and will see if that improves matters. Fingers crossed!



  Reply

11-Jul-2018 08:30 PM

The Quadfather

The Quadfather

Posts: 110

Installed a new -18V 1A regulator, and this Q34 still hums like before. Louder when the lid is on! Switch back to the OPA op-amps, and all is 100% fine (lid on or off!). It truly is a bizarre problem.

I have a friends Q34 here which has an issue 5 pcb that I’ve just installed the DaDa mods on, and his runs perfectly well. There are two differences between his Issue 5 and my Issue 6 pcb 1) His has the DIN sockets. In all other respects the pcb layout is identicle (it’s modern looking DIN version pcb, with the modern solder mask instead of that awful light green coat). 2) The other difference is his is using the original CMOS switching chips.


i am absolutely stumped. The only other thing I can try is to remove the MAXIMs on my unit, add sockets and try the original switches. This is driving me crazy!

  Reply

11-Jul-2018 08:35 PM

The Quadfather

The Quadfather

Posts: 110

P.S. C77 & C78 are 100uF caps again by the way (took the films out incase they were picking up any hum too). No improvement alas.

  Reply

12-Jul-2018 02:27 AM

EJP

EJP

Posts: 1505

The buzz in my experience comes from IC9/10 and is solved by appropriate decoupling of those opamps. If you really can't make that work, just replace those opamps with OPA134 or whatever you prefer.

It certainly has nothing to do with the MAX4066s. Don't even consider wasting your time replacing those.

  Reply

12-Jul-2018 07:53 AM

The Quadfather

The Quadfather

Posts: 110

I have done the decoupling on those two chips, neatly and as per DaDa’s instructions.

I want to find out what the issue is, not revert back to the OPAs. As I said, the revision I did for an issue 5 Q34 works perfectly well, and the pcb is identical (I have both in front of me) with the exception of the din sockets.

The only other difference between the two IS the Maxims...

  Reply

12-Jul-2018 07:57 AM

clearsound

clearsound

Posts: 27

Hi John,

I'm still pretty new to this so if you've already tried the following , apologies:

Just thinking about the 'tin hats' why not swap them around the board at the various IC Pair locations and see if that doesn't show anything up?

I had a problem with the turntable earth lead (connecting TT earth to pcb) once - so you might want to check that, and I don't know if you've re soldered your din/ RCA connections to the PCB. I seem to remember having a loss of continuity in the ground on the main board around this area and also at the two caps by the metal Transformer box. It was a while ago though.

FWIW one of my 34's had  some extra decoupling on IC20 and 24, but that was Joost's work so I can't remember exactly why it was done. IC24 is decoupled differently to the others.

Ref your other post about volume and switch clicking, I wonder if it might just be more noticeable in some systems? In my 303 system its more pronounced than in the 909 system.

I hope you solve it.

All the best

Joe

  Reply

12-Jul-2018 11:00 AM

Joost Plugge

Joost Plugge

Posts: 962

Hello All,

I did a lot of work on  the two 34's of Joe. Too much to explain here, and in most cases not relevant. The 34 has a power supply system, and also important, earth rail designed for the TL071 low power op-amps. All the current trough the earth rail has to be delivered or sinked by Ic23. The psu lines are long and not well designed due to the Pcb layout. Also Quad did not decouple the Op-Amps as advised in every data sheet, even in those days. Normally as EJP stated, decoupling Ic9 and 10, as explained in the upgrade manual does do the trick. In some 34's, I also had to decouple all of the inverting op-amps, Ic7, 8, 12 and 13. Also place a 100nf across pin  4 and 8 of Ic24. I also increased C58 and 84 to 220u, to lower the dynamic impedance of the earth rail. Also checking the earth solder pins on the Din connectors, as Joe advised could be solving some issues.

Joost

  Reply

12-Jul-2018 01:24 PM

The Quadfather

The Quadfather

Posts: 110

Thanks for your suggestions Joe & Joost. I think it’s probably helpful to mention pcb revision numbers when discussing these mods in the future, it might create a ‘pattern’ when encountering problems, and why some Quad 34s need additional decoupling, and others don’t.

Obviously there’s no ‘black magic’ involved, so it must be due to either different conponents used on the pcb revisions, or slightly different pcb trace routes, which seems the most likely difference, logically. Otherwise we’re in the dark as to why one upgraded unit is fine, and another hums/oscillates.

I’ll try your suggestions Joost. Today I’m removing all wire  links, checking nine are too close to pcb traces, and the through-hole pins near IC9 and 10. I’m absolutely determined to solve this ;)


  Reply

12-Jul-2018 01:27 PM

The Quadfather

The Quadfather

Posts: 110

Joe - this hum is with turntable leads & earth wire connected or disconnected. Makes no difference. Less hum on CD input but still very much there. More on the unused inputs. It’s a phono socket version, and all soldered sockets are fine. Have checked for any dry or cracked joints. As I said it’s fine with OPAs so it must be an LME oscillation issue really. Times like this I wish I had an oscilloscope!

  Reply

12-Jul-2018 06:04 PM

clearsound

clearsound

Posts: 27

Hi John,

Ref the turntable connection - I didn't have a Player connected so it might still be worth looking at it. I think the wire joins the chassis to the pcb? I'd check the other earth connection too the one from the mains IEC. It was a long time ago but if I remember correctly these were 'dirty / corroded' and needed cleaning. What difference it would make when the cover was on or off is beyond me. maybe there's a chance of a loop? I'm not sure but I know I had Hum. FWIW, mine are 12000  - 15000 range in serial numbers. I see what you mean about the cost of the HA version - they've gone up a bit.

Joe

  Reply

12-Jul-2018 06:07 PM

clearsound

clearsound

Posts: 27

Sorry John - one last thing,

Have you re-soldered or checked the though pcb links? Probably not necessary - but worth a check before you remove any components.

Joe

  Reply

12-Jul-2018 09:34 PM

The Quadfather

The Quadfather

Posts: 110

Hi Joe.

Thanks for your helpful suggestions. Yes have checked all earth connections. All good. Have checked the through holes around the opamps for continuity using my multimeter and all seems fine there too.

I’m confident it’s nothing to do with the turntable either connected, or not connected, as I’m using my friend’s DaDa’d Q34 as a control. It’s fine with zero hum and has a complete set of non tin-hat LMEs, and the same sockets as my own. It only has bypasses on IC 9 and 10.

It truly is a mystery. As I said earlier other than the din socket area of the pcb vs the phono sockets, both his and my own have absolutely identical pcbs (his issue 5, mine issue 6). I am going over them both again section by section to see if anything jumps out, but the only obvious difference between the two is mine also contains Maxim switches, whereas his are the standard Motorolas. I know EJP says they’re not a factor, but they are the only obvious difference...


  Reply

12-Jul-2018 09:34 PM

The Quadfather

The Quadfather

Posts: 110

The only other curiosity is it’s stable and hum free (on all inputs, cover on or off) with OPA opamps. So perhaps there is nothing wrong with it. That can then only indicate there must be manufacturing tolerance differences from batch to batch with the LMEs, so some Quad 34s are fine, on some they’re at their limit of stability and need extra decoupling.

So the hum plus variability in hum loudness depending how far on the cover is, can be narrowed to three possibilities:

  1. There’s still something not quite right with my Q34 when using LMEs, but not enough to cause issues (or the issues are masked) with OPAs.
  2. Differences in batches of LME opamps which mean some Q34s need extra decoupling, others just IC9 and 10. So not a Q34 problem as such.
  3. The Maxim switches ARE (for whatever reason) not a good team-player with the combination of a revision 6 pcb Quad 34 and LMEs.

As ESP said I could just use the OPAs, but I prefer to get to the bottom of any problem, plus I’m very intrigued about what this problem is, plus I’ve now paid for two sets of LMEs, so want to use them :)

The investigation continues. At least I can use my friend’s unit for the time-being.

  Reply

13-Jul-2018 09:50 AM

clearsound

clearsound

Posts: 27

Hi John,

That all seems logical. I've been tracking back in my brain and the Maxims had no negative  effect when I fitted them. I do remember originally purchasing a set of Dip LME's but they produced hum. I think the 34 went to Joost after that and he applied the extra decoupling. There's too much difference in my 34's from then to now as they run the  relays instead of the Maxims. So it would be hard to say when the hum finally disappeared, but I am sure it was also reduced by the Tin Hat version. Sorry if that doesn't help much.

Good luck with the solution

Joe

  Reply

13-Jul-2018 12:40 PM

The Quadfather

The Quadfather

Posts: 110

I think the extra decoupling is probsbly the answer.

My confusion arises over why my friends unit that I did the DaDa mods on doesn’t hum, and mine does. The pcbs are absolutely identical apart from the din vs phono areas. Components look identical too and his uses the same sockets & pdip opamps as my own.

There must be a logical reason. Other than Maxim switching, mine also uses the tin-hat LME in the phono position. I see no reason why that might make a difference, but I may change it to a pdip to rule that out too. Otherwise, it has to be production tolerances between LME. 

  Reply

13-Jul-2018 02:16 PM

The Quadfather

The Quadfather

Posts: 110

Just noticed the LME47920s look different.

On my friend’s upgraded unit there is just print on the chips, with a white dot inducating pin 1. All chips in his ate from DaDa.

On my unit, the 920s have a moulded in dimple, bottom left, which denotes pin 1. The style of the 920 is the same as all my 910s.

The style of the 910s in his unit also are with the dimples.

Admittedly I am clutching at straws now! I guess I could swap the chips over and see if mine stops humming.

  Reply

14-Jul-2018 02:40 PM

The Quadfather

The Quadfather

Posts: 110

Joe would it be possible to email you?

  Reply

14-Jul-2018 03:16 PM

clearsound

clearsound

Posts: 27

Hi John,

no problem, you sent me photos of your 34 a couple of years back so the email is the same. I think you can get my email from the forum?  As we submit email contact when we register?

Cheers

Joe

  Reply

14-Jul-2018 06:24 PM

The Quadfather

The Quadfather

Posts: 110

Ah you have a great memory Joe! :)

Actually Joost just emailed with an answer to a question I was going to ask you about which opamp was decoupled to stop the hum :)

  Reply

18-Jul-2018 03:40 PM

The Quadfather

The Quadfather

Posts: 110

Just wanted to say MANY thanks to Joost, and Joe's experiences with his Quad 34, for helping me to solve the hum issue with me Quad 34 (later revision 6 PCB version). Joost experienced the same hum issue when upgrading Joe's Quad 34. Turns out our units also benefit from having IC19 decoupled as well - I decoupled IC19 and 20, just to be thorough, though it's only really necessary to do IC19 or 20. As Joost said the psu is mono. If decoupling IC19 then use pin 10 of IC17 for the grounding point. If decoupling IC20 then use pin 1 of IC18 as the grounding point for the decoupling caps.

Following advice from Joost I also increased C58 and C84 to 220uF. I also changed R123 and R124 to 10K, and R84 and R85 to 220R to equalise psu voltages (do NOT do this if sticking with the original TL0 series op-amps).

This has solved the issues I was having with my Quad 34 - very grateful to all the helpful people here :)

John

  Reply