UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio 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 22nd Dec 2015, 11:26 pm   #121
bikerhifinut
Octode
 
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,591
Default Re: EL34 power amp project

Well,
Got the transformers back yesterday.
They have had an additional vacuum impreg of varnish and a bake and I could see the evidence.
I was also provided with a "Simon" Frame with them and a thick paper gasket. It looks like a steel tray that fits under the base. Not sure what it's supposed to do though.
The feedback I got from Majestic also said in an email that the transformers had "an acceptable noise level". Whatever that means. I have to say that for me "Acceptable" means Silent or nearly silent. i.e. inaudible at distances over a couple of feet away.
Anyway in they went.
Oh Dear, just as bad. I tried a tweak of using some double sided foam tape to give a bit of insulation round the base and wellied it down. Not a lot better, audibly buzzing from the other end of the utility room around 10 feet.
Now here's the thing:
I accidentally touched one of the mounting bolt heads............ Ouch! hot, very hot, to the touch. Also found another the same diagonally opposite. transformer off load at this point and i was detecting the faint smell of something getting Hot, possibly the windings.
Now I have used the fibre washers supplied by the makers, and the frame is isolated by the gasket from the laminations, and the gasket is isolated from the top of the painted chassis by the foam tape, so as far as I can see, I have done what I can to avoid a shorted turn effect.
But obviously I haven't.
And I wonder if this could be the underlying cause of the vibes?
I got curious at this point and wired an IEC socket to the original Danbury transformer. I just dropped it onto the foam tape round the drop through opening, admittedly upside down so the tags werent in danger of shorting on stuff underneath. It was a lot quieter than I remember it and just about ok resting unfixed, upside down. AHA I thought, it just needs a bit of the squashy foam tape round the base. So i puts a bit round and bolted it in. Switched on and BUZZZZZZZZZZZ.
So it has to be a mounting issue, but its doing my head in. The transformer manufacturers tell me it should work silently as a drop through in a steel chassis. But it doesn't.
I have never used a steel chassis before, I have mounted large drop throughs in aluminium and other non magnetic alloys and not had a vibration issue or hot bolts. Logic tells me the steel chassis is having an effect but if that's the case, why do people still use them?
A long post, I apologise. Somewhere the solution lies with mountings. I cant see how I can realistically use "top hats to space the mounting bolts as the 5mm holes would suggest 4mm or 3mm bolts which don't seem to me to be man enough to hold down a 4.5+ kg transformer. the bolt heads would probably drop through the top holes apart from anything else.

A.

Last edited by bikerhifinut; 22nd Dec 2015 at 11:36 pm. Reason: clarification
bikerhifinut is offline  
Old 22nd Dec 2015, 11:56 pm   #122
joebog1
Octode
 
joebog1's Avatar
 
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 1,096
Default Re: EL34 power amp project

If there is sufficient clearance between the mounting bolts and the holes in the laminations, I sometimes use thin heatshrink along the length of the bolts, and use steel washers at each end ( under the head of the bolt, and directly against the nut) BUT with a fibre washer directly against the laminations/cover. I leave the heatshrink slightly overlength, so that I actually crush it a small amount when the nuts are done up tight.

I have now run out of ideas.

Joe
joebog1 is online now  
Old 23rd Dec 2015, 12:01 am   #123
joebog1
Octode
 
joebog1's Avatar
 
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 1,096
Default Re: EL34 power amp project

hhmm rereading the last line of your post, 3 or 4mm bolts will be big enough!! After all your not going to be kicking the amp around. Large washers can be obtained from car shops, called mudguard washers, and can be had with 15mm outside dia with 3 or 4 mm holes a pack of ten costs about $2 here. You can also use "Unbrako" Allen head bolts which are grade 12 or 13 high tensile steel and you WONT break them !!!! even if you do kick the amp around.

j.b.
joebog1 is online now  
Old 23rd Dec 2015, 12:44 am   #124
bikerhifinut
Octode
 
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,591
Default Re: EL34 power amp project

Cheers Joe.
Got to be worth a try.
If I get time tomorrow in between the festive chores I will nip down to the fastener store locally and see about some high tensile allen bolts in 4mm, with heatshrink that will get me enough clearance.
I guess i shrink it on with the hairdryer errr "heat gun" first. I was thinking of using stud bar and dome heads or nylocs (they can look quite effective if done up just flush to the top of the nut thread.) and nylocs on the bottom as I have a phobia about mounts unscrewing with heat and vibes.
Do you think the vibration is connected with this?
There's definitely some sort of interaction with the chassis either as a resonator or transducer effect or perhaps magnetic coupling causing the steel plate to vibrate with the mains frequency?
if it means a new pair of chassis then so be it. I can get a nice hardwood surround with anodised ally top plate made by Hammond for not too much and they do look nice. Also the guy that sells them will cut the mounting holes and valve base holes etc for a nominal sum if given a scale drawing of what's needed.
I'd rather not go that route but at least with my "sub assembly" tagboard technique, I would save an amount of soldering as I could simply remove the boards and reinstall them.

A.
bikerhifinut is offline  
Old 23rd Dec 2015, 3:14 am   #125
joebog1
Octode
 
joebog1's Avatar
 
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 1,096
Default Re: EL34 power amp project

What prompted my previous post was your comment re: screws getting hot!! That tells me its a shorted turn! NO WONDER the iron is growling. I did do a rough reverse engineering math on the link you sent me, and that amount of iron, coupled with your present current usage, does NOT need a shorted turn.
The shorted turn is only about a 1/5th shorted turn as the screws lie "mostly in the corner" of the lamination stack. Its similar to a shaded pole ( motor) in fact.
I have had the problem myself.

Joe
joebog1 is online now  
Old 23rd Dec 2015, 7:57 am   #126
Ed_Dinning
Dekatron
 
Join Date: Jan 2004
Location: Newcastle upon Tyne, Tyne & Wear, UK.
Posts: 5,816
Default Re: EL34 power amp project

Hi Andy, try using a sleeve around each bolt. Thin sellotape should do as well as your insulated washers. If the bolt is getting hot it looks like eddy currents etc.
Brass bolts might also assist.
If it is not magnetostriction you will not need to clamp the transformer tightly.

Ed
Ed_Dinning is offline  
Old 23rd Dec 2015, 1:39 pm   #127
bikerhifinut
Octode
 
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,591
Default Re: EL34 power amp project

Thanks guys.
Ed I hope you are enjoying your birthday break in the sunshine.
Ok first of all the hot screws were my doing, its so cramped in there trying to work around wiring that I fitted post original fit of transformers that I inadvertently "nipped" one of the heater wires off the transformer under the mounting nut and washer. Bizarrely i also managed to do the same on the opposite side where the transformer chassis/mains earth is bolted down. I couldnt see any obvious insulation damage but it was obviously enough to get a leakage to earth via the 2.2 0hm earth lift resistor. I think that saved the day as the 0.1 uF cap in parallel with it had disnitegrated with the usual smells, it was that which alerted me. Just goes to show that Joes advice about walking away from a fault and looking at it afresh the next day is good.
Anyway that issue is resolved and I have done as much as I can for now to ensure the wiring runs are safe and away from any problem areas.
So I am left with the noisy transformer issue still.
Steel chassis and i am left with little else to think other than it must me a magnetic coupling effect.
I shall try the insulated bolt theory. I am minded to see if Sharps fasteners sell stainless stud bar in 4mm, otherwise its get a set of 65-70 mm long stainless allen bolts, which will look good anyway and tighten down nicely. I think I am right in that stainless steel is non magnetic.
It's an odd one for sure and I am hoping I don't have to go down the route of another chassis. I think its a double whammy, interaction with magnetic materials possibly, and the mild steel chassis ringing like a bell. My experience with alloy has been that it seems to be better "damped" accoustically but I have no solid science based evidence for that.
bikerhifinut is offline  
Old 23rd Dec 2015, 2:44 pm   #128
Ambientnoise
Heptode
 
Join Date: Feb 2006
Location: Marlborough, Wiltshire, UK.
Posts: 534
Default Re: EL34 power amp project

A small point but not all stainless steel is non-magnetic. You need to do a bit of research on the grades and their properties.

Ken
Ambientnoise is offline  
Old 23rd Dec 2015, 9:13 pm   #129
bikerhifinut
Octode
 
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,591
Default Re: EL34 power amp project

Cheers Ken,
I had an idea that may be the case.
Anyway the problem is mostly solved
I used 4 of "Andy's patent isolation washers" under the "simon frame" between it and the steel chassis. They are actually a hard rubber tap Washer from the plumbing spares shelf at the DIY shop................
Result! There's a residual "tizz" amplified a bit by the top cap and It sounds to me like the windings rattling. The transformers had an extra vacuum impregnation and bake so I reckon there's no improvement to be had now.
I will continue to fine tune the remaining buzz, I think I could damp the top cap a bit and wonder if a pad of some sort under the top cap would be possible.

Now to the other bit.
I am a bit miffed that the transformer was wound to a 230V primary. Our mains was 246V last night and it's caused issues on the secondaries. I had a word with majestics and they say they wind to the standard for the mains which is 230V, apparently in Poole thats as high as it gets. No good to me though. In future if I get a transformer wound to my spec i shall ask for a tap on the primary so it can be used 230 or 240 V.
The issue is that the EL34 heaters are running at 6.8V measured on the new DMM and they are visibly brighter at a yellow shade of orange which looks a bit hot to me. The other valves are no problem as I have them on a different secondary and regulated at 6.3V DC.
The Ht needs to be managed down, with a 220R in each anode of the GZ34 I am getting 483V off the cathode and after the choke its still up at 463V. I dont want more than 450V on the CT of the output transformer.
I gave them a quick run up after checking the rest ov the voltages and i thought the amp wasnt working, total silence at the speaker with ear pressed against it. But a quick unplug and re insertion of the input RCA plug proved otherwise. So the REGulated DC on the EF40 and ECC81 has been a definite advantage, although the amp was already pretty darn quiet.
BUT............. I put a signal through it and as the music started i saw the dreaded "Red Plate" effect begin.
The only parameter I have changed is the mains transformer so I am assuming that this particular circuit configuration of EL34 with cathode bias of 470 ohms in UL with 43% taps was already "On the limit" at the design voltage of 450V at the CT. I reckon the valves were passing between 76 and 78mA which is about 10mA more than specification.
So its bung another 100 ohms on each anode, i rearranged the tagboards so that i could use different resistors either singly or series and parallel to get the correct voltage on test. That will make the tweaking easier.

Anyway fingers crossed that i have got over the hump, and thanks to all for their very useful input as many of the hints and tips have been put into use.

Happy Xmas everyone!

Andy.
bikerhifinut is offline  
Old 24th Dec 2015, 2:20 am   #130
bikerhifinut
Octode
 
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,591
Default Re: EL34 power amp project

well it urns out that the EL34 itself was a duffer. annoying as its only had a hundred hours or so use. Sovtek EL34G one of a matched pair.
So its a new pair of valves for that amp and probably a quad to do both.
Those old Sovteks usually are ok so maybe i was just unlucky.
I'll probably give the Brimar lads a call after the hols.
And here's another thing, i measured my mains again tonight as the voltages on the heaters and off the transformer were crazy, 7+ V on the heaters and 463V for the 410V secondary. Turns out our mains rockets to 253V after about 11pm. No wonder we keep blowing light bulbs regularly. I think the grid wallahs want a rocket up em.
in view of that I shall revert to the 470 ohm anode resistors rather than the 330 Ohm and live with a reduced HT during the day. I was putting 460V on the anodes.
I think this goes a long way to explain the noisy transformers and why i have had trouble in the past with overvoltage on projects.
the transformers are still a bit noisy but I will have to learn to live with it.

A.
bikerhifinut is offline  
Old 24th Dec 2015, 6:13 am   #131
joebog1
Octode
 
joebog1's Avatar
 
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 1,096
Default Re: EL34 power amp project

The voltage probably is a tad high, but remember what ALL the valve data manuals say!!
Voltages +/- 20% should not affect performance or life of the valves.
The brand you are using is about par for the course!!
I absolutely HATE em !!
But thats personal opinion. I bet if you ask around someone here will drag out some good used Mullards at the right price. I have em but Australia Post buys the 747 jet at your expense then adds the weight of the parcel!!

Anyway its xmas eve here so everyone have a wonderful peaceful holiday

Best wishes for the season
Joe
joebog1 is online now  
Old 24th Dec 2015, 7:51 am   #132
Ed_Dinning
Dekatron
 
Join Date: Jan 2004
Location: Newcastle upon Tyne, Tyne & Wear, UK.
Posts: 5,816
Default Re: EL34 power amp project

Hi Andy, no problem them winding the transformer for ratio based on 230v, but they should wind it to operate at the desired flux density based on at least 240v. If not you will be running up towards saturation if it is a Stalloy material, not quite so bad if it is Unisil.

Brass studding should be perfectly adequate stress wise and no danger of getting a magnetic grade of alloy.

Ed
Ed_Dinning is offline  
Old 24th Dec 2015, 1:24 pm   #133
bikerhifinut
Octode
 
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,591
Default Re: EL34 power amp project

Ok chaps.
Thanks Joe and Ed.
I fettled the amps back to the 470 ohm anode resistors on the rectifier.
Mainly because Good ol' Maplins staff can't read and out of a set of 4 220R wirewounds they managed to give me a 2.2R one. It's not as if they have to read a colour code!
Anyway, everythings running more or less ok. I have managed to induce a bit of hum along the way on one of the monos. It's 50hz mains so i reckon its probably an earthing issue. I have moved the anode resistors to a spare tag on the driver boards, one end of the resistor direct off the valve base. I dont think thats the issue as both amps are identical and there would have been wiring there anyway. My gut feeling is the ground referencing for the regulated DC heaters, I did the Lazy thing and took it from the nearest earth point which was the earth return from the feedback screen at the amp input............. See where i am heading? I think I have done a bit of bad practice there and when next I get the urge (When I get the new PSU resistors sorted) I will run a separate earth return back to the Star earth point at the rear of the chassis. I should have thunk a bit as the heirarchy is to run your earths with the PSU and Output section first and run back up the line to the inputs. So referencing heaters at the input ground is probably a bit stoopid of me. Funny the other amp doesnt 'um. It's not loud so isn't audible 3 feet away.
One majestic transformer is reasonably quiet, the other is a bit too noisy still. I shall experiment with different tightness on the squashy rubber washers as its probably a resonance issue.

phew.........
bikerhifinut is offline  
Old 24th Dec 2015, 1:41 pm   #134
bikerhifinut
Octode
 
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,591
Default Re: EL34 power amp project

And finally.
Cos one EL34 is duff even though it seems to work and sound ok, measures proper and everything but has a "hot spot" on the anode, I had an idle notion and tried some GE USA made 6550A that are nearly new, maybe a couple of hundred hours use in my KT88 amp.
Measured the volts and they are pulling 76mA each (36v across 470R cathode resistors), anode Volts down to 385. Thats Low but at least the circuit DC conditions are safe for those valves. I know the grid leak should be reduced to about 330 or less, but no measured nasties as yet. and i'm running em well below their limits.
So into the system...........

OH YES! luverly sound. It's all there. As someone here says, "If your valve amp sounds warm, then its got a fault", well it's not warm and fluffy at all. So probably wouldnt please some bottle heads. But its not got any nasty HF surprises and i reckon its got good HF stability going by the sound. I havent scoped it yet so won't know for sure.

Anyway it's given me food for thought. I have a stash of KT88/6550 as ed is only too aware! (Thanks for running the stash through your valve tester ed). And these GE 6550 which are the real deal, got off Phil Taylor a couple of years ago, should meet Joe bogs approval!
So it's back to the books and figure out a suitable configuration for resistor and coupling cap values for 6550/Kt88 within the scope of the PSU (max 460V off the cathode of the GZ34). My own KT88 kit amp runs at a design HT of 408V for 35W so i should get into the ball park. That amp uses 5687 as the LTP Phase splitter/driver. Hungry little beggars for heater current. I'll stick with the ECC81, they were ok for Harold leak.

Ok I better put the toys back in the box and get the hoover out before the wife does something terminal to me.

Andy.
bikerhifinut is offline  
Old 5th Feb 2016, 6:12 pm   #135
bikerhifinut
Octode
 
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,591
Default Re: EL34 power amp project

I have a strange phenomenon on this amp.
A "hot spot" on one EL34 on each amp.
swapping valves round shows the hot spot stays with the socket.
its the same corresponding socket on each monobloc.
All voltages, anode, screen grid, cathode check out within spec. And no DC leakage from coupling capacitors.
The only thing I can think is that the grid leaks are quite high at 680k which is the value specified by HJ Leak for their TL25+ as these are Clones of that circuit as you all know.
I could go down to 470k or 330k i reckon without any noticeable loss of performance.
The other factor is the Heater voltages are high due to our high mains voltages at around 7volts. I wondered if excessive emission from the hot cathodes might be having an effect?
Bit of a mystery.
I have used KT88 and 6550 in the amp, the circuit has enough flexibility to handle these as the values I chose will keep a KT88 running well under its rated power and in permanent ClassA as far as I can ascertain. The mains transformers are well over specified so any extra current draw is catered for.
Using the higher rated valves I dont get this phenomenon.
Am I simply running the EL34's too hard? And would larger screen grid resistors make any difference? As i understand it they are really only in circuit to protect the output transformer if the valve goes S/C? Leak never even put them in their designs.
Ht to CT of output transformer 450V, anodes at 443V there or thereabout and cathode volts at 34V using 470R cathode resistors and 50uF bypass caps.
Puzzler this one as I can not see any reason why one valve would run hotter than the other as the circuit conditions are the same on each phase.
It's put paid to my conclusion that I had a bad quad of Sovtek EL34's, although I suspect they might all be a bit borderline and run, as these are, right on their limits, this could be why.
I have no qualms about reworking the circuit to use KT88's, the output transformers were designed to be suitable for either a pair of EL34 at 25-30W with 8 ohm output, or a pair of KT88 at up to 50W but 6 ohm output impedance, which is fine with my nominal 8 ohm speakers, or the spare pair of monitor audio 6 ohm speakers I have.
I wonder if the uprated mains transformers are to blame?

A.
bikerhifinut is offline  
Old 5th Feb 2016, 6:58 pm   #136
G6Tanuki
Dekatron
 
G6Tanuki's Avatar
 
Join Date: Apr 2012
Location: Wiltshire, UK.
Posts: 8,261
Default Re: EL34 power amp project

It's not oscillating at above-audio frequency, is it? I've seen loads of 'strange' amplifier behaviour caused by this.

Try adding 100pF capacitors directly between the grids and cathodes, and maybe wiring 10K resistors in series with the grid-feeds.

See http://www.valvewizard.co.uk/gridstopper.html for more details.
G6Tanuki is offline  
Old 5th Feb 2016, 8:40 pm   #137
DangerMan
Heptode
 
Join Date: Jul 2009
Location: Stoke-on-Trent, Staffordshire, UK.
Posts: 700
Default Re: EL34 power amp project

It might be worth measuring the voltage on the grids of the EL34's as the amp heats up to try and see what's happening.
Probably best to measure at the other end of the 10k grid stopper, away from the grid pin so as not to cause parasitic oscillation.

The voltage should obviously be zero, but you might see it increase if there is any gas current within the valve as it gets hot.
Interestingly, Mullard quoted 700k as the maximum (limiting) value of grid leak with cathode bias but they used 470kohm in their own 5-20 design. Leak used 680k (+ 10k grid stopper) presumably to extract every last bit of open loop gain.

1 kohm screen grid resistors are a good idea as they reduce the risk of parasitic oscillation due to transformer leakage inductance, and in the Mullard book is a comment that they slightly reduce output stage distortion with a very small reduction in output and gain. They won't help if the fault is grid current from the valve though, except by slightly reducing the G2 voltage and running them a little bit cooler.
DangerMan is offline  
Old 5th Feb 2016, 10:33 pm   #138
joebog1
Octode
 
joebog1's Avatar
 
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 1,096
Default Re: EL34 power amp project

Methinks its due to the bottles!!! I did mention this before.
I have had the problem myself. I have had it much worse with the same brand KT88's!!!
If you look at the DC resistance of each half of your output transformers, I think you will find that the plate glow is caused by the lower reading, i.e. most of that section will be wound close the the centre leg.
See if you can borrow, beg, buy, steal an origional set of Mullards, and I will put money on it, that the red plating will go away.
An exact same discussion is being had now on an Australian guitar forum, but with 6550's The ancient yank made GE 6550 handle the 750 volts easily the exact same design ( sic) ruskie valves cant cut it and red plate.
The origional design was excellent as evidenced by the many Leak's and clones that were built, some still have 1950,s issue valves in. If they can be sourced it may be the answer to shell out once for REAL EL34/6CA7's and be done with it.

best of luck
Joe
joebog1 is online now  
Old 5th Feb 2016, 11:06 pm   #139
DangerMan
Heptode
 
Join Date: Jul 2009
Location: Stoke-on-Trent, Staffordshire, UK.
Posts: 700
Default Re: EL34 power amp project

I agree, Joe.

Mullard's reputation was (among others) well earned and modern "repro" valves don't seem to have the same materials quality and/or consistently high vacuum etc. as the originals, sometimes leading to runaway.
My Leak TL25plus shows all these traits: it's absolutely fine with old stock REAL valves yet is just a bit close to the edge of self destruction with some modern ones.
DangerMan is offline  
Old 6th Feb 2016, 12:55 am   #140
bikerhifinut
Octode
 
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,591
Default Re: EL34 power amp project

Thanks guys.
I'll scope the amps and see if they are oscillating at HF, seems odd though that both are doing it and exhibiting identical effects on the same phase. and not on the other.
However it cant do any harm to try 100pF caps from grid to cathode. Would ceramics do or do I need film caps?

I can also try 1k screen grid stoppers, nothing to lose by that. And maybe drop the grid leaks to 470k, I don't need so much gain.

I think i am running the EL34's way too hard though and in the absence of an alternative set of valves I can only try the above solutions. NOS or even used Mullards are so expensive that they never even figured in my plans. I dont believe all new manufacture valves are bad, just theres some out there not so good.

By the way I only spotted the hot spots because I had the amps on while the room was dark. In daylight or under normal artificial light I cant see any problem even close up to the valve bottles.

Onwards and upwards as they say.
bikerhifinut is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 3:03 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 - 2020, vBulletin Solutions, Inc.
Copyright ©2002 - 2020, Paul Stenning.