Quad 2805 power supply malfunction?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
The right channel of my beloved Quad 2805 speakers is distorting heavily. I invested a lot of time to try out different amps and cables to be sure that the quad speaker is the perpetrator.

In my search I discovered that when I pull the AC plug on that speaker the distortion almost immediately disappears. So to me it looks that something is wrong with its power supply. Can someone shine some light on this matter please ??
 
I had a very similar problem (in one of my 2805s), and i suspected arcing or severe leakage in one of the panels. However the neon lamp indicator was blinking only about once/second, which seemed to contradict that (no difference in volume compared to the OK speaker channel, no rustling or hissing noises, no blue arcs in darkness, nothing)

In the end I found the protection circuit being the root cause -- replacing the 1000 uF electrolytic capacitor after the rectifier bridge solved everything.
(This capacitor going bad is also mentioned by others in ESL forums.)


The fact that your problem disappears when you turn power off (i.e. disable the protection circuit) also indicates a problem somewhere in this area .... my 5 cents.

/M
 
I appreciate your 5 cents very much.
I couldn't find the service manual for the Quad 2805. Instead I found one for the 2905. Checking this schematic i found the capacitor you mentioned. I searched the web for instructions how to dismantle the esl and specifically how to get access to the power supply. Again no success. At the back of the base I see 4 phillips screws. Is that the way to get access.
Perhaps a weblink or some hints for me ??
 
That 1000uF capacitor is a 16V unit in the ESL63's, 988's and even some of the 2805's. The power supply runs at about 16.5volts so those caps are under some pretty heavy stress and can go bad.

IF you have a failing panel, the clamp circuit can sense the arcing from the bad panel even if it's not very audible and clamp frequently. One way to tell is if the clamp circuit only goes off if you turn the volume up but is much more well behaved if the volume is low.

Sheldon
quadesl.com
 
When I turn up the volume only the distortion increases cq becomes more audible. As I said when I switch off the AC of the faulty ESL, distortion disappears and the sound volume seems to increase a little bit. Both esl’s are equally loud. The panels themselves sound okay. No pops, cracks or rustling noises. I am wondering if this sounds like clamping.

So perhaps a check of the 1000uF capacitor is a good way to go forward.
I like to attempt the repair myself but before breaking things I would appreciate to know what’s the best way to access the power supply.
 
Sheldon Stokes is the pro, I'm just an amateur...

I used a small sharp screwdriver to carefully lift either end of the wooden lacquered piece at the top. Lift it carefully about 1 cm (without scratching anything) , then you can grip it with your fingers and remove it. In the middle you find a big bolt. Unscrew it to carefully remove the tensioning rod, without damaging the threads at the bottom of the rod (no bending). After completely removing it, there are only four screws to remove to take off the lid covering the electronics. Pull the top carefully just slightly up and then towards the back.

Unfortunately the later QUADs (IMHO) are not built for easy servicing -- to save production cost there is only one big circuit board that contains both HV stuff and protection part and you need access to the solder side to remove components. With some wiggling you would be able to replace the capacitor, without unsoldering any connecting cables, taking care to not stretch a number of HV cables too much.

Agree with stokessd that it may still be a faulty panel, but once you got this far, you will have a better idea of what's going on, e.g. the frequency of the neon lamp is invaluable in diagnosing the root cause.

Hope for a happy ending - a capacitor/protection circuit fault is MUCH nicer than a panel fault :)

/M
 
Super, this is exactly what I was hoping for. Now I've got the confidence to proceed with my attempt to repair. I have build some amplifiers, so I have a basic knowledge of electronics. And soldering is no problem. But if one of the panels is bad I'm in trouble.

Elsewhere I've read that it is not advised yo let the newer Quad ESL's under power permanently. My former Quad ESL 63 where always powered on for almost 25 years. Never a problem !

Christmas and the last days of 2018 are for the family. The first week of January is for the 2805. I will come back here with pictures of the disassembly and hopefully the fix. Perhaps useful for other inexperienced DIYers like me.
 
Sounds good, I'll be glad to help if you send pictures or questions.

You'll want to play music and watch the neon bulb. When nothing is playing it should blink quite slowly, potentially many seconds between blinks depending on cleanness of the panels and humidity. When you play music you will get an increase in flash rate, but it should still be slow (<1Hz). What you will likely see is that you get a flash and the clamping action at the same time.

Sheldon
quadesl.com
 
As mentioned in my thread starter one of my Quad 2805's is distorting heavily. As to the repair I'll give it a shot. As a newbie however I need some guidance. Nowhere on the net I could find pictures of the disassembly process. Diyaudio member mats31 was so kind to give me some hints. For other newbs like me I took some pictures.
First thing is removing the top bar of the ESL. Following mats31's instructions I used a flat piece of metal (picture) and lifted the left hand side of the top bar a centimeter or so. The same for the right hand side. Then you are able to remove the bar completely (picture).
In the resulting cavity (picture) you can see the nut that holds the tension brace at the backside of the speaker. Loosen the nut but hold the brace firmly to prevent damaging the threaded downside of the brace. Loosen the tensioner at the bottom and remove the brace (picture). Now you can remove the four screws at the backside of the base. Lift the cover of the base a few centimeters, slide it backwards and remove it (picture).
The electronics is located at the base of the speaker (picture).

Well now I am again in newbie land. At the left and right hand side I see the audio transformers. In the middle the mains transformer and just behind it the main board. On the main board (picture) pointing to the power supply the blue caps and the diodes of the high voltage supply (??). Just behind it you can see the cooling fins with the triac and the timer ic of the clamp section (??). Right at the backside of the board two electrolytic capacitors. One of them is the 1000u 25V member mats31 mentioned as a possible suspect. To the left on the main board the diodes which are feeding the audio to the different sections of the stators (??). Please correct me if I'm wrong, I am here to learn, to pass on my findings and in the same time repair my esl with hopefully the help of more experienced members.

Okay now back to the diagnose:
-- the esl is distorting;
-- no strange noises in the panels;
-- the defective esl is as loud as my non defective esl;
-- when I pull the power plug the distortion is gone;
-- when powering up the neon bulb flashes 3 or 4 times;
-- when powered the bulb flashes just once every 35 seconds.

Replacing the 1000u capacitor mentioned by member mats31 seems a realistic move to me in the view of the foregoing or are there other opinions.
 

Attachments

  • IMG_1806nw.JPG
    IMG_1806nw.JPG
    40.4 KB · Views: 483
  • IMG_1807nw.JPG
    IMG_1807nw.JPG
    36.6 KB · Views: 436
  • IMG_1809nw.JPG
    IMG_1809nw.JPG
    51 KB · Views: 420
  • IMG_1812nw.JPG
    IMG_1812nw.JPG
    61.9 KB · Views: 425
  • IMG_1815nw.JPG
    IMG_1815nw.JPG
    55.4 KB · Views: 431
  • IMG_1831nw.JPG
    IMG_1831nw.JPG
    68.5 KB · Views: 238
  • IMG_1817nw.JPG
    IMG_1817nw.JPG
    74 KB · Views: 289
...Replacing the 1000u capacitor mentioned by member mats31 seems a realistic move to me in the view of the foregoing or are there other opinions.

Before removing the board and replacing the 1000uF (C28) LV power supply capacitor, you might consider measuring the supply voltage with a digital volt meter. It is most easily accessed from pins 1 and 4 of the 555 timer IC. Measure both DC and AC voltage. You should get about 16VDC, and very low < 0.1VAC. If you get lower DC voltage and greater AC voltage, this would confirm that the 1000uF capacitor needs to be replaced.
 

Attachments

  • Quad_ESL_LVsupply.png
    Quad_ESL_LVsupply.png
    576.6 KB · Views: 247
The voltage I measured on pin 1 and 4 of the ne555 is 15,5 VDC. I can’t confirm the <0,1 VAC. The VAC on my dvm is constantly changing in the rythm 38,15,6,2,0 V. Isn’t that the ripple of the capacitor charging and discharging process ? Or am I totally wrong here?
 
Yes, the VAC reading is measuring the ripple of the capacitor charging from the bridge rectifier output pulses. It should be fairly steady and certainly less than 1VAC. If you are seeing large AC voltages like 38 or 15 then replacement of the 1000uF capacitor is in order.
 
Thank you bolserst that was really helpful.
Now I'm wondering how to proceed. The most obvious way is to desolder all the wiring, unscrew the main board, take it to the workbench and exchange the capacitor.
But perhaps it's possible to desolder just a part of the wiring, unscrew the board and flip it to get access to the underside and exchange the cap.
Maybe it is even possible to do the job from the upside of the board by wiggling the cap until it comes loose and solder in the new one leaving the leads of the new cap somewhat longer. Does anyone has ideas here?
When working on the board and the wiring are there pitfalls to avoid ?
When I'm at it is it advisable to renew other components on the board? For instance the big 220u non polar cap (c26).
Your help is much appreciated.
 
I have never worked on the later iterations of the Quad ESL with the large consolidated board. But, it does look like you could flip the board enough to replace the capacitor without needing to unsolder most of the wires. stokessd or mats31 may be able to provide further guidance if you haven’t already figured out a path forward.
 
Since it's a radial cap you need to access the solder side.
Yes, you can flip the board around to access the solder side (at least to a 45 degrees angle) without removing any cables. Some cables from the left transformer may have to be loosened from a plastic cable wrap if I remember right.

Note that the if you remove any HV cables you need to use corona lacquer or bees wax to insulate the solder points once done. All HV components on my 2805s were nicely covered with beeswax. (I used beeswax on other ESLs (ESL 57) -- messy but kinda fun to see how well it works.)

I actually have replaced the bipolar 220 u cap. Brings "audiophile peace of mind" :) and I think the QUADs are worth it. Can't say the difference was immediately audible for me (never did any A/B or blind comparison), but many report real improvements in the midrange/treble.
Used a short wire soldered in parallel with the existing axial bipolar (thus easy to remove from the top if I ever want to do that) and then fitted higher quality and bulky components in a plastic box outside the 2805s (along the speaker cables). Used 100+100+22 uF Jantzen caps in parallel with 10 x 15 ohm 1/2 W metal oxide resistors.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.