UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Other Discussions > Homebrew Equipment

Notices

Homebrew Equipment A place to show, design and discuss the weird and wonderful electronic creations from the hands of individual members.

Closed Thread
 
Thread Tools
Old 6th Jul 2011, 1:24 pm   #61
G8HQP Dave
Rest in Peace
 
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
Default Re: Mullard 5-10 homebuild project

Quote:
Originally Posted by Herald1360
Wrong anode voltage on EF86 shouldn't unbalance the dc conditions on the phase splitter,
Yes it will. This is because the LTP is, ideally, finely balanced between grid cutoff and grid current. Too high EF86 anode means ECC83 takes too much current at too low anode voltage, so grid-cathode voltage is too low so grid current flows and unbalances the LTP due to the high resistance feeding the left hand triode.
G8HQP Dave is offline  
Old 6th Jul 2011, 1:51 pm   #62
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,638
Default Re: Mullard 5-10 homebuild project

Hi Dave,whats LTP? Ive still got a problem.I thought I was on the way to sorting it out,but whereas the EF is about right anode 53v cathode 1.5v and grid 69v the ECC anodes are way out ie the first at 175v the second at 253v.
I thought I,d sussed it yesterday,as I had them about matched,now I,m back to square 1.What I cant twig is that with a supply voltage @ 320v and everything checked as per all your suggestions its all so far off the values it should be.

Could it be down to wrong cathode resistors on the EL84,s or wrong OPT a-a load?Herald suggested changing the EF cathode R,Andy.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline  
Old 6th Jul 2011, 2:34 pm   #63
Herald1360
Dekatron
 
Herald1360's Avatar
 
Join Date: Feb 2007
Location: Leominster, Herefordshire, UK.
Posts: 16,528
Default Re: Mullard 5-10 homebuild project

LTP is Long Tailed Pair. A pair of valves with commoned cathodes fed via a high resistance (or constant current source) which is the "long tail".

They act as a cathode follower driving a common grid stage. In operation an increase in the current in one half is balanced by a corresponding decrease in the other which ideally results in equal and opposite (ie antiphase) signals at each anode. In practice things are never ideal...... but it's near enough.

I reckon the EF86 anode volts would have to be seriously high to unbalance the DC conditions, though I can see that adding signal as well could result in dynamic unbalance caused by the scenario described. Surely it's the right hand triode that suffers from the high impedance feed, though?
__________________
....__________
....|____||__|__\_____
.=.| _---\__|__|_---_|.
.........O..Chris....O
Herald1360 is offline  
Old 6th Jul 2011, 2:44 pm   #64
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,638
Default Re: Mullard 5-10 homebuild project

Thanks,Ive tried a 1.5k in place of the 1.8k k R it increases EF anode voltage by 2v.Ive also tried to up the supply voltage and used a different valve.The anode voltage in all cases sits at 53-55v.One thing changing the k R did change tho,was to reduce the hum.

Both EL anode,s etc are pretty ok despite the inbalance in the above.But the signal still distorts easily,which makes me think its maybe the OPT.Maybe,Andy.

Just measured the ECC anodes again they are 217v and 219v.I,m going to leave it on for a bit.Its very frustating.Bl**dy electronics!
__________________
Curiosity hasn't killed this cat...so far.

Last edited by Diabolical Artificer; 6th Jul 2011 at 2:53 pm. Reason: New info.
Diabolical Artificer is offline  
Old 6th Jul 2011, 3:12 pm   #65
Herald1360
Dekatron
 
Herald1360's Avatar
 
Join Date: Feb 2007
Location: Leominster, Herefordshire, UK.
Posts: 16,528
Default Re: Mullard 5-10 homebuild project

Bit confused now. Dropping the EF86 cathode R from 1K8 to 1K5 should decrease the EF86 anode volts not increase it, since the valve should pass more current. Are you measuring anode volts wrt 0V or measuring the voltage across the anode resistor?

For reference, what are the voltages at C15+, C14+, C7+ and C4+? (All the various HT supply voltages for the different stages).
__________________
....__________
....|____||__|__\_____
.=.| _---\__|__|_---_|.
.........O..Chris....O
Herald1360 is offline  
Old 6th Jul 2011, 3:15 pm   #66
G8HQP Dave
Rest in Peace
 
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
Default Re: Mullard 5-10 homebuild project

Sorry, yes it is the right hand triode which gets its grid supply from the big resistor. The EF86 anode voltage does not have to go very high before this problem occurs, especially if the HT supply is a little low as well. I suspect that Mullard was pushing the valves to their limits in this design, so a slightly gassy ECC83 would suffer more. I hit this problem with a modified 5-20 with the first stage anode only about 20V higher than expected - which would normally not be worth worrying about in a valve design.

This is a DC issue. Signal voltages are OK because the anode impedance of the EF86, although high, is still low enough to provide a little grid current to the left hand triode. The big cap to ground can supply plenty of AC grid current to the right hand triode.

As an experiment try feeding the LTP from a potential divider instead of the EF86. Measure (and plot?) the two anode voltages and the input voltage. With a well-balanced ECC83 the anodes will stay together at first, but as the input rises they will diverge. This is actually quite a good test for residual gas in the valve!

Last edited by G8HQP Dave; 6th Jul 2011 at 3:18 pm. Reason: add experiment
G8HQP Dave is offline  
Old 6th Jul 2011, 3:22 pm   #67
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,638
Default Re: Mullard 5-10 homebuild project

Ahh,silly me,its been a long day.Voltages at @ C15 etc are 351,341,283,144.C15 was @ 320V but Ive left it on for the last hour.So if I increase cathode R to say 2.2K which is the next biggest value I have?

Sorry,I,m measureing wrt 0v IE off the bus bar.As of typeing ECC anode voltages are 240v and 245v.Thats left to right on a schematic.
__________________
Curiosity hasn't killed this cat...so far.

Last edited by Diabolical Artificer; 6th Jul 2011 at 3:27 pm. Reason: New info.
Diabolical Artificer is offline  
Old 6th Jul 2011, 4:28 pm   #68
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,638
Default Re: Mullard 5-10 homebuild project

Last post today,its turning into a bit of an epic.Ive replaced the EF cathode resistor with a 2.2k.All measurements are pretty damn close,but I now have hum and the signal still distorts at very low volume and there isnt much volume either.Which is where I started.So I,m going to tidy the wiring up and then I,m not sure.

The OPT,s I have are 6k ish anode to anode.The Sterns plans call for 8k,so I think the cathode resistors for the EL84,s are the wrong value.They seem to be running hot.Andy.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline  
Old 6th Jul 2011, 5:50 pm   #69
Top Cap
Octode
 
Top Cap's Avatar
 
Join Date: Mar 2006
Location: Watford, Hertfordshire, UK.
Posts: 1,270
Default Re: Mullard 5-10 homebuild project

What concerned me was that the voltages seemed correct until C9 was connected. Could there be HF oscillation or something similar giving the run around here?
In situations like you have, I have always tried to isolate the stages, getting each one operating correctly then interfacing the lot together. The bias for the phase splitter could perhaps be provided temporarily with a resistor network to enable you to confirm that the phase splitter is operating correctly. Then do some tests on the front end and confirm all is well there. Sometimes just having the wrong phase of feed back from the output transformer can cause hours of grief trying to find the 'Fault'. Keep perservering, I am sure you will get there and the whole project will give you more pride when it's fully up and running.
__________________
Whether the Top Cap is Grid or Anode - touching it will give you a buzz either way!
Top Cap is offline  
Old 7th Jul 2011, 12:00 pm   #70
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,638
Default Re: Mullard 5-10 homebuild project

Today Ive tried changing the cathode resistors on the EL84,s.In the Mullard 5-10 plans it says to use 437 ohms for a OPT with a primary of 6.6k.I had 270 ohms.Ive managed to makkle 425 ohms.The only difference it seems to have made is a slight increase in cathode voltage.As things stand I have all DC conditions pretty smack on.

But,I still have this distortion problem.One thing I,ve noticed is that when I touch the metal part of the volume control is a change in sound,the humming gets a little louder.Another thing Ive noticed is a glow at the top of the EL84,s? Does the heater go to top?Lastly could the problem be parasitic oscillation?And could that be in the ECC stage.As I understand it pentodes were developed (suppressor grid) to stop this.

Ive included a few pics.I know it looks a bit all over the shop at present but I,m very carefull.I,ll get the scope hooked up this PM and have a poke around.
Thanks everyone for sticking with me.TC had another look at FB,dont think thats the prob,Andy
Attached Thumbnails
Click image for larger version

Name:	Mull 5 etc 006.jpg
Views:	206
Size:	114.1 KB
ID:	53555   Click image for larger version

Name:	Mull 5 etc 009.jpg
Views:	186
Size:	86.3 KB
ID:	53556   Click image for larger version

Name:	Mull 5 etc 007.jpg
Views:	184
Size:	104.0 KB
ID:	53557  
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline  
Old 7th Jul 2011, 1:01 pm   #71
Herald1360
Dekatron
 
Herald1360's Avatar
 
Join Date: Feb 2007
Location: Leominster, Herefordshire, UK.
Posts: 16,528
Default Re: Mullard 5-10 homebuild project

With the long lengths of straggly wiring between valves and components, I might be inclined to agree with the parasitic oscillation idea!

For a quick bodge, try adding some 4.7K or thereabouts resistors at the valveholder in series with the wire feeding each valve's grid pin. Also, disconnect the negative feedback until everything's looking good dc wise.

It may be a PITA but perhaps rebuild it following the Mullard original's layout?
__________________
....__________
....|____||__|__\_____
.=.| _---\__|__|_---_|.
.........O..Chris....O
Herald1360 is offline  
Old 7th Jul 2011, 1:07 pm   #72
G8HQP Dave
Rest in Peace
 
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
Default Re: Mullard 5-10 homebuild project

You could be suffering from instability. RF oscillation will cause bias shifts and distortion.

There are two ways to build the 5-10. Follow the Mullard design exactly, or use conventional point to point wiring with short leads and your own grounding scheme. You seem to have rather long leads, which is asking for trouble.

BTW your posts would be easier to read if you used ' for apostrophe instead of , and added a space after each full stop.
G8HQP Dave is offline  
Old 7th Jul 2011, 1:35 pm   #73
ms660
Dekatron
 
ms660's Avatar
 
Join Date: Apr 2011
Location: Cornwall, UK.
Posts: 13,454
Default Re: Mullard 5-10 homebuild project

Looks a bit "loose".
As suggested, try and tidy up anway, it's always good practice.
Keep input/output leads to the various stages well seperated but as short as is reasonably possible if not screened, and close to the chassis, avoid to many earthing points.
Parasitics can run around HT leads as well.
I have not really followed this thread from the begining but one thing eliminated is usually something less to think about.
Trying to trouble shoot homebrew with to many self inficted variables can sometimes be a bit of a nightmare, I have done it myself many a time in a rush to get things up and runing!

Good luck with the project.
ms660 is offline  
Old 7th Jul 2011, 3:49 pm   #74
Sideband
Dekatron
 
Sideband's Avatar
 
Join Date: Oct 2003
Location: Croydon, Surrey, UK.
Posts: 7,550
Default Re: Mullard 5-10 homebuild project

When I built a couple of 5-10 back in the 70's I followed the Mullard layout as accurately as possible. I built the versions without controls so I could use them with the stereo pre-amp, also built exactly to the Mullard layout. I never experienced any problems, it all worked first time (barring some tweaks to the attenuation in the pre-amp) and it all continued to work when I (stupidly) gave everything away to another enthusiast about 5 years later.

You need to tidy up that layout. Believe me, Mullard knew what they were doing when the design was published. The earthing is particularly important with a single busbar earthed at one point only and everything returned to that.

With the EF86 running at a very high gain, you don't want long wires running about.

This is where a 'scope would come in handy to see if it is parasitic oscillation causing all your problems.


SB
__________________
There are lots of brilliant keyboard players and then there is Rick Wakeman.....
Sideband is online now  
Old 8th Jul 2011, 10:05 am   #75
yestertech
Nonode
 
yestertech's Avatar
 
Join Date: Dec 2002
Location: Coulsdon, Surrey, UK.
Posts: 2,109
Default Re: Mullard 5-10 homebuild project

Don't mean to be unkind, but with a wiring layout like that, the amp is never going to work properly.
I'm afraid I learned this the hard way a good few years ago. Even a few inches of unnecessary or incorrectly routed wiring can cause havoc.
I didn't believe it ( but subsequently proved it ) that even moving earthing points a few inches up and down a bus-bar earth can make big differences to the hum levels !
At minimum, ensure there is one earth for the bus-bar to chassis, near to the input socket,
Shorten anode and g2 leads as much as possible to the EL84s and route all input wiring well away from output stage. stopper resistors in the EL84 anodes and g1s may improve and stabilise things.


Andy
yestertech is offline  
Old 10th Jul 2011, 12:34 pm   #76
bonesetter
Triode
 
Join Date: Jun 2011
Location: Potters Bar, South Herts
Posts: 23
Default Re: Mullard 5-10 homebuild project

Magnetic, I'll upload some pics of my 5 10 which may help with wiring layout. The 0v busbar is earthed at small lugs at the fixing screws for the input socket and this is a common star earth point. wiring generally as short as practical with the tag bds mounted v close to valve bases. Checkout the Mullard data/ diagrams.

I found with my amp best to disconnect the coupling caps between ECC83 and the EL84's which separates the o/p stage and this makes it easier when fault finding with the dc setup. I don't want to hijack your thread but in my case the old caps where leaking (volts) and caused the o/p stage to be be biased on with the cathode resistors volts at over 30v, at one point the solder on the base tags of EZ81 melted! Now replaced all caps apart from main 50/50 and it works fine although HT volts still too high - need to juggle with R values to correct.

Mike
bonesetter is offline  
Old 10th Jul 2011, 12:38 pm   #77
bonesetter
Triode
 
Join Date: Jun 2011
Location: Potters Bar, South Herts
Posts: 23
Default Re: Mullard 5-10 homebuild project

One more thing I was getting hum with the input open cct, no hum when input earthed to chassis, and radio breakthrough caused by long speaker leads and input lead from guitar, this because I live near Brookmans Park radio stn.

M
bonesetter is offline  
Old 10th Jul 2011, 1:14 pm   #78
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,638
Default Re: Mullard 5-10 homebuild project

Thankyou all for your replys. I never considered the wiring might be a problem. But having looked at how other valve equipment is wired, I can see Ive made a big mistake. I think your all right in regards to the earthing. In the pics you can see I have one earth point, but I dont think I have it right.

I think my biggest mistake was to try and combine two 5-10,s and a preamp in one box,and combine the Sterns and Mullard plans. I can now see theres a lot more to building an amplifier than plonking a lot of components together. A valuble lesson. Mike,can I take you up on your offer of some pics please? So back to the drawing board,Andy
__________________
Curiosity hasn't killed this cat...so far.

Last edited by Diabolical Artificer; 10th Jul 2011 at 1:16 pm. Reason: Legability
Diabolical Artificer is offline  
Old 10th Jul 2011, 2:43 pm   #79
bonesetter
Triode
 
Join Date: Jun 2011
Location: Potters Bar, South Herts
Posts: 23
Default Re: Mullard 5-10 homebuild project

Magnetic, will do later got a bit of F1 to watch yet.

M
bonesetter is offline  
Old 10th Jul 2011, 3:19 pm   #80
G8HQP Dave
Rest in Peace
 
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
Default Re: Mullard 5-10 homebuild project

If you want to learn about audio grounding then there is lots on the web. The basic rule is to make sure currents go where you send them, by ensuring that there is no alternative! Try a search on DIY Audio site, as well as google.

The suggestion to use stopper resistors on grids is a good one, and can sometimes rescue a poor layout. This trick is often used (to excess!) in guitar amps.
G8HQP Dave is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 10:17 am.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.