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 9th Jun 2011, 12:01 am   #41
glowinganode
Octode
 
glowinganode's Avatar
 
Join Date: Apr 2009
Location: Aylesbury, Buckinghamshire, UK.
Posts: 1,522
Default Re: Mullard 5-10 homebuild project

Maybe severe overloading, at 1V (rms?) input.
You are running the amplifier into a dummy load aren't you?
For us to make sense of the scope traces, we really need to know (at least) the V/cm settings in the photo's.
What's the power output?
Rob.
__________________
We have done so much, for so long, with so little, we are now qualified to do anything with nothing.
glowinganode is offline  
Old 9th Jun 2011, 11:07 am   #42
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,637
Default Re: Mullard 5-10 homebuild project

Scope is set to .5v/div-time/div = 5ms.The input is set to 1v.I have it connected to an old 8 ohm speaker.With a music input,there is distortion present when I turn the volume control past a quarter turn.Otherwise its fine.At first I thought I,d connected the anode and taps to OPT wrong,but Ive checked my wiring.The output voltage seems low,I measured .2v.Andy.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline  
Old 9th Jun 2011, 1:00 pm   #43
glowinganode
Octode
 
glowinganode's Avatar
 
Join Date: Apr 2009
Location: Aylesbury, Buckinghamshire, UK.
Posts: 1,522
Default Re: Mullard 5-10 homebuild project

I think by the looks of things, the clipping is occurring in the first (EF86) stage, the phase splitter is just reproducing it.
Can you scope the input to the phase splitter?
Have you tried injecting a signal directly into the phase splitter, though remember to feed it through a coupling capacitor (0.1uF) as it is around +80Vdc from memory.
Rob.
__________________
We have done so much, for so long, with so little, we are now qualified to do anything with nothing.
glowinganode is offline  
Old 11th Jun 2011, 1:06 pm   #44
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,637
Default Re: Mullard 5-10 homebuild project

Since my last post Ive changed the rectifier resistors to 100 ohm.I did this as the voltage readings were down by about 20v according to Mullards specs.Also my own and mostly Robs(Glowing etc) calculations called for 100 ohm R,s.Originally they were 300 ohm,as in the Sterns layout.Therefor here are the new voltage readings-C1 353v, C2 332v,C3 274v, C4 143v.
EF86 a-56v, grid 72v,k-1.664v, ECC a1-208v,g1-55.5v,k57.9,a2-231v,g2-50.9v, EL84 V3-a-326v,g-339,k-11.74v, EL84 V4,-a-328v,g-339v,k-10.46v.
I have some more osc pics,sig gen set to 0.1mV/1khz;I measured it at 0.109v on my DMM;osc set to 2ms/div-.5v div.Pic #1 is signal before amp (0.150v),pic #2 is just starting to distort (0.476V),pic #3 is distorting badly (0.689v),pic #4 is (at Robs suggestion) input via 0.1uf cap into ECC phase splitter at pin 2 (grid).
Whilst trying to get to the bottom of this problem Ive been looking at my OPT specs;in the Sterns plan it calls for a 8k anode to anode resistance/impedance,the OPT,s I,m using are 6.6k a-a etc.However looking at the Mullard 5-10 schematic it uses the same 47 ohm R,s in the UL configuration.Ive tried to measure the DC resistance of the OPT primarys,but the DMM display flickers,therefor I cant get a stable reading.
One more question I have is,as I,m housing two 5-10 monoblocks in the same housing with two mains transformers;will there be a problem in the two sharing the metal chassis as part of its earth,s/-HT? Thanks for all your help,Andy.
Attached Thumbnails
Click image for larger version

Name:	scope 5-10 #3 012.jpg
Views:	133
Size:	66.0 KB
ID:	52265   Click image for larger version

Name:	scope 5-10 #3 009.jpg
Views:	129
Size:	56.7 KB
ID:	52266   Click image for larger version

Name:	scope 5-10 #3 004.jpg
Views:	131
Size:	49.4 KB
ID:	52267   Click image for larger version

Name:	scope 5-10 #3 014.jpg
Views:	124
Size:	48.4 KB
ID:	52268   Click image for larger version

Name:	scope 5-10 #3 015.jpg
Views:	136
Size:	97.1 KB
ID:	52269  

__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline  
Old 12th Jun 2011, 10:50 am   #45
glowinganode
Octode
 
glowinganode's Avatar
 
Join Date: Apr 2009
Location: Aylesbury, Buckinghamshire, UK.
Posts: 1,522
Default Re: Mullard 5-10 homebuild project

Andy, can you be a bit clearer with your voltage measurements and let us know which grid you're measuring and leave out the dashes which confuse matters. Also referring to a scope as an osc doesn't help. I assume all voltage measurements are wrt chassis.
The EF86 anode voltage seems too low, and shouldn't be lower than the screen (g2) voltage, check the value of the anode load resistor.
Pic#4 shows the sig-gen coupled directly to the phase splitter (0.1V, not 0.1mV?), what happens to the trace if you increase the signal to a few volts, does it still distort?
Having two seperate amps on the same chassis is fine, just use a common earth busbar earthed at the input socket (or volume control if easier).
Rob.
__________________
We have done so much, for so long, with so little, we are now qualified to do anything with nothing.
glowinganode is offline  
Old 13th Jun 2011, 5:01 pm   #46
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,637
Default Re: Mullard 5-10 homebuild project

Thanks Rob for sticking with me.I,ll try to answer your questions.First I apologise for the my use of (-,s) they dont mean negative voltage.The voltage measurements shown are taken with the common(negative) lead of my DMM on the negative terminal of C1 and 2,(is that what you meant by wrt chassis?) and the positive lead on the various anodes,grids etc I,m trying to measure.However,having read up last night on how a pentode operates,would I be right in thinking that taking readings this way,wont show a proper reading at the grids(g1,input) of the EF86 and EL84,s,as they are negative in relation to the cathodes.In that case,would I put my positive lead on cathode and negative lead on g1 to get a proper reading/

Regarding the input voltage I meant 100mv.I lowered it as you said in a previous post that 1v input voltage may have been too much,and could have been contributing to the distortion problem.

Sorry for the delay between posts,I,m single parent and find it hard to get the time to get in the shed,but with your help we seem to have narrowed things down to the EF86 input stage.i will increase the voltage direct into the ECC phase splitter (pin 2),check the value of the EF86 load resistor and try to get you some better readings.

Lastly regarding using two mains transformers on one chassis/one earth,did I understand you correctly to use one bus bar for both mono amplifiers?Andy.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline  
Old 13th Jun 2011, 7:26 pm   #47
G8HQP Dave
Rest in Peace
 
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
Default Re: Mullard 5-10 homebuild project

The two ECC83 grids should be within about 0.1V of each other - you have 4.6V difference. This could mean that the LTP is not working. Check C7 and R5. I suppose it could be your DMM loading the grid, but the anode voltages are quite a long way apart too.
G8HQP Dave is offline  
Old 15th Jun 2011, 3:18 pm   #48
glowinganode
Octode
 
glowinganode's Avatar
 
Join Date: Apr 2009
Location: Aylesbury, Buckinghamshire, UK.
Posts: 1,522
Default Re: Mullard 5-10 homebuild project

That's what I thought Dave, with a 10Mohm dmm and 1Mohm between the grids, you'd expect the earthy one to read about 10% low, although there is quite a difference in the anode voltages suggesting inbalance.
Andy, are you using known good valves or pulls?
With a stereo source, you're bound to have a common earth, so best to make this at the input socket. One common bus-bar is fine, earthed in the middle. Don't mix up the two channels, and try to connected the input stage nearest to the earth with the following stages progressively towards the ends.
Take your time Andy, we're not going anywhere and quite understand.
Rob.
__________________
We have done so much, for so long, with so little, we are now qualified to do anything with nothing.
glowinganode is offline  
Old 15th Jun 2011, 5:00 pm   #49
G8HQP Dave
Rest in Peace
 
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
Default Re: Mullard 5-10 homebuild project

Put the DMM between the two ECC83 grids, in parallel with R5. What voltage difference do you see?
G8HQP Dave is offline  
Old 16th Jun 2011, 11:44 am   #50
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,637
Default Re: Mullard 5-10 homebuild project

I took some more readings Weds. I checked the EF a load R,its on spec 121k.I increased the signal at ECC grids-still distorting.As per your last posts voltage between ECC grids = 0.170v.I have some more scope pics,but they show the same conditions as previous ones.However I did take voltage readings of EF anode when distorting = 84.4v.

I,m cross referenceing to the Mullard table for DC readings in the 5-10.In them for instance it gives EF scr grid should be 72v and anode at 61v.Surely the scr grid voltage should be slightly less than anode voltage?

Lastly I,m using EF,ECC that were pulled from a working amp,and have tried several different valves.The EL,s are NOS russian "6n14n-EB".In a new built amplifier,is it normal to have to alter components to fine tune the circuit? Should I be trying to get the EF working properly,by changing cathode resitor etc to set the bias,and get this working.Or should I concentrate on the PSU side of things?Andy.Thanks again Rob and Dave.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline  
Old 16th Jun 2011, 12:41 pm   #51
neon indicator
Retired Dormant Member
 
Join Date: Jun 2010
Location: Co. Limerick, Ireland.
Posts: 1,183
Default Re: Mullard 5-10 homebuild project

No, something is either not wired where you think it is or is faulty.

Many valve circuits worked fine with 20% resistors and 50% capacitors.

Unless a schematic has "AOT" (Adjust on test), Select or 1% etc it's not critical.

The ECC83 grids with no signal should be same voltage. The Anode voltage will differ slightly.

Unplug ECC83 and remove R8 the 1M Ohm resistor that's between its grids.
Disconnect feed back R11 C6 from T1 to EF86. They are the only variable parts according to Loudspeaker type.

Get the EF86 stage working
http://www.r-type.org/static/5-10-b1.jpg

The voltage on G2 may be higher than on Anode of EF86.

Check from EF86 socket Cathode to Ground (with power off and no valve) that you have about 1.9k (1.7K to 2.2K is likely OK).
C2 from G2 to Cathode should have no leakage and be 47nF approx (0.05uF on schematic) with at least 250VDC working.

If EF86 working after fixing whatever is wrong, leave out the 2 x EL84 and reconnect R11, C6, R8.
Check AC decoupling capacitor on 2nd ECC83 grid is OK (no leakage). It can even be 220nF (0.22uF). It is open circuit at DC, so both ECC83 halves are DC biases by EF86, but at AC ground.

ECC83 cathode resistor MUST NOT be bypassed by a capacitor and minimum of 6k8 Ohms or you don't get "phase splitter".

On scope the two ECC83 anodes should be same AC level but 180 degree out of phase on dual beam scope. Once that's correct you can put back the EL84s and all should be fine.

Note that input drive is is lower for output on EF86 and ECC83 with either R11 & C6 missing or the EL84s missing as you have no negative feedback.

If you have winding polarity wrong or ECC83 to EL84s crossed over the gain and distortion INCREASES when you connect R11 & C6. They go to which ever side of output gives lowest gain. The other end of output of T1 must be earthed (maybe near V1 EF86 cathode resistor). The R25 1K Ohm on o/p T1 is for safety if speaker is unplugged. It limits the peak voltage on primary avoiding insulation breakdown. Common on transmitter outputs too for limit SWR on open circuit.

Last edited by neon indicator; 16th Jun 2011 at 1:05 pm.
neon indicator is offline  
Old 16th Jun 2011, 1:32 pm   #52
G8HQP Dave
Rest in Peace
 
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
Default Re: Mullard 5-10 homebuild project

0.17V difference between ECC83 grids is OK.

If the EF86 anode voltage rises significantly with signal then this means that the bias is shifting, probably due to second-order distortion. Something not right here, as you seem to have the anode a bit low without signal. How, exactly, have you wired the EF86 control grid?
G8HQP Dave is offline  
Old 16th Jun 2011, 2:08 pm   #53
neon indicator
Retired Dormant Member
 
Join Date: Jun 2010
Location: Co. Limerick, Ireland.
Posts: 1,183
Default Re: Mullard 5-10 homebuild project

Yes, < 200mV for ECC83 grids is roughly "the same". If it was much more I'd suspect a leaky capacitor. Depends on impedance of meter as it's a 1 M Ohm resistor, unless you measure between the grids rather than to ground. My "good" meter is 10M Ohm, my cheaper one is 1M Ohm and the AVO is for measuring when schematic says voltages measured assume 20k ohm per volt.
neon indicator is offline  
Old 5th Jul 2011, 12:18 pm   #54
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,637
Default Re: Mullard 5-10 homebuild project

Ive rewired everything and have also made sure I have 320v @ C1 by putting two 300r resistors between the cathode of the rectifier and C1.I now have voltages nearer what they should be,but the EF anode voltage is 6v too low.

In the Mullard 5-10 bumf it says-"the anode voltage of the EF86 determines the operating conditions of the phase splitter" (as Mr Neon etc have said),also it says "if EF anode voltage too low,the phase splitter bias will be too high"

I still have something odd happening at the ECC anodes,IE 140v on the first,250v on the second when they should be 200v or so.

To get the EF anode voltage up to 61v ish,should I change the anode load resistor or raise the supply voltage? Its all driving me bit nuts,hey ho,Andy.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline  
Old 5th Jul 2011, 1:14 pm   #55
Herald1360
Dekatron
 
Herald1360's Avatar
 
Join Date: Feb 2007
Location: Leominster, Herefordshire, UK.
Posts: 16,527
Default Re: Mullard 5-10 homebuild project

Wrong anode voltage on EF86 shouldn't unbalance the dc conditions on the phase splitter, it only acts as a longtailed pair for ac signals, though it will change the overall bias level. Maybe the ECC itself is somewhat unbalanced.

Try increasing the 1.8K cathode R on the EF86 to reduce its anode current a bit.

Beware that the grid of the EF86 is dc coupled to the input- any stray dc at the input will also upset bias conditions.
__________________
....__________
....|____||__|__\_____
.=.| _---\__|__|_---_|.
.........O..Chris....O
Herald1360 is offline  
Old 5th Jul 2011, 2:54 pm   #56
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,637
Default Re: Mullard 5-10 homebuild project

Ive got the EF working ok,and after disconecting C9 (coupling cap) I have both anodes on the phase splitter at the same voltage now .So I guess C9 is a dud,which is wierd as they were new caps.I,ll change the cap and see how I go.Unfortunatly I cant check the phase-shift as I only have one good scope test lead.Anyway,at least I seem to be getting somewhere,Andy.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline  
Old 6th Jul 2011, 9:16 am   #57
Diabolical Artificer
Dekatron
 
Diabolical Artificer's Avatar
 
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,637
Default Re: Mullard 5-10 homebuild project

Caps ok,but when C9 reconnected anode voltage on the earthy side of ECC shoots up.I checked the new caps by connecting to a 9v battery.I thought caps were supposed to block DC.But these dont.I checked them against some mustard caps of the same value,these also allow DC thro.I,ll have to look into this more.

Anyway,I,m progressing and getting better at this fault finding,am starting to read the circuit better.Still have hugh gaps in my knowledge tho,Andy.
__________________
Curiosity hasn't killed this cat...so far.
Diabolical Artificer is offline  
Old 6th Jul 2011, 9:52 am   #58
neon indicator
Retired Dormant Member
 
Join Date: Jun 2010
Location: Co. Limerick, Ireland.
Posts: 1,183
Default Re: Mullard 5-10 homebuild project

What voltage rating are the caps?

You need a nano-ammeter for 9v! Even with 250V DC, you could have 0.1uA leakage which would change bias. Most meters can't measure that.

My 1950s leakage tester + bridge uses between 30V and 120V across the capacitor. 100M Ohms leakage flashes the "Magic Eye" once per second. That's a fail. Good caps it doesn't flash at all. 30V and 100M ohms is 0.3uA! At 200V Across a cap from anode to grid, 100M Ohm would be 2uA, with 1 M Ohm grid "leak" resistor that would raise the grid by 2V, a huge change.

The capacitor leakage at 200V needs to be less than 0.1uA. If it was linear with voltage at 9V you would see 0.0045uA That's 4.5nA. Likely it's not linear so 9V current is even less.

All of my perfectly good Electrolytic capacitors have too high a leakage for valve grids. On tests I'm doing on sub-miniature Russian pentodes I'm using 2 x 220nF ceramic in parallel, as I don't have any 0.47uF ceramic, the Tantalums are too high leakage and the foil capacitor's legs are too fat for breadboard.

Last edited by neon indicator; 6th Jul 2011 at 9:58 am.
neon indicator is offline  
Old 6th Jul 2011, 10:27 am   #59
Sideband
Dekatron
 
Sideband's Avatar
 
Join Date: Oct 2003
Location: Croydon, Surrey, UK.
Posts: 7,549
Default Re: Mullard 5-10 homebuild project

Quote:
Originally Posted by magnetic View Post
I thought caps were supposed to block DC.But these dont.I checked them against some mustard caps of the same value,these also allow DC thro.

Maybe they are not capacitors.... I was caught out by some inductors once that I thought were capacitors. They were cylindrical, colour coded like resistors (in this case red,red,red) which I assumed was 2200pF and they came out of a bag marked 2n2.... However when I fitted one as a decoupler (screen grid), it burnt up. Checking further they all measured s/c. It was then I realised they were 2.2mH chokes.....!
__________________
There are lots of brilliant keyboard players and then there is Rick Wakeman.....
Sideband is offline  
Old 6th Jul 2011, 11:13 am   #60
neon indicator
Retired Dormant Member
 
Join Date: Jun 2010
Location: Co. Limerick, Ireland.
Posts: 1,183
Default Re: Mullard 5-10 homebuild project

That would show up on 9V battery and a meter though!

I do have some inductors and caps that don't want to be mixed.
neon indicator is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 2:34 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.