26th Oct 2015, 1:16 am | #21 |
Octode
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,993
|
Re: EL34 power amp project
Just to clarify,
the transformer has a 5V heater winding nominally rated at 2A. To reiterate, I had a phone conversation with Danbury transformers who assured me i was fine running a 3Amp heater off the winding as long as I didnt exceed the total VA rating of the transformer because the wire gauge used for the 5V winding was up to the current. And by the way Joe, if I use a GZ34 the cathode is internally connected to one side of the heater so that also is at HT potential. I do have a modicum of knowledge, just to reassure people. A. |
26th Oct 2015, 6:11 am | #22 |
Banned
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
|
Re: EL34 power amp project
My apologies
I wasnt challenging your knowledge. I didn't catch the details on the 5V winding. Joe |
26th Oct 2015, 8:36 pm | #23 |
Octode
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,993
|
Re: EL34 power amp project
No worries Joe.
Anyway plenty of time for me to do something stupid. Here's how far I have got. I am waiting for the chokes to arrive and I will also need to get a few resistors and odd small value capacitors. I had to use a couple of resistors in series on the HT+ rail as I couldnt easily find a 27k x 2W. I think I will redo it before firing the amp up, I should be able to blag a couple of 27k 2W. Heaters wired, then mains in. Voltages tested on heater circuits. Output valve anodes and screen grid wiring to transformer. I used shrink sleeving to insulate any exposed mains connections on the back panel. Nothing else now until I get a couple of bits and also have a good old think about my plans for the wiring. The plan is to put a couple of extra tag boards in for the input valve and phase splitter components. Run a return wire from each stage to the "star earth" |
27th Oct 2015, 6:17 am | #24 |
Dekatron
Join Date: Jul 2009
Location: Sleaford, Lincs. UK.
Posts: 7,662
|
Re: EL34 power amp project
Good to see your making progress Andy. drop me a PM with your address and I'll pop some resistors/caps in the post if I have them. I have 27k 2W and should have any caps you need too.
Andy.
__________________
Curiosity hasn't killed this cat...so far. |
28th Oct 2015, 12:50 pm | #25 |
Octode
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,993
|
Re: EL34 power amp project
Pictures will follow as soon as i get my phone sorted out............ On the final stages. I hope my Earthing strategy has been the sensible way with "Star" returns from the various stages.
A. |
28th Oct 2015, 12:54 pm | #26 |
Octode
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,993
|
Re: EL34 power amp project
Anyone know off the top of their heads of an Indirectly heated Rectifier that drops more volts than a GZ34? I know I can use a 5U4G but its directly heated although they are cheap as chips if you shop around, but I also have a nice virtually unused GZ32 which I think loses more volts than the later 34. It was a "tweak" I believe on Leak EL84 amps that hadn't been modded for lower HT prior to the change of output transformer.
Last edited by bikerhifinut; 28th Oct 2015 at 12:55 pm. Reason: clarification. |
28th Oct 2015, 2:55 pm | #27 |
Dekatron
Join Date: Nov 2011
Location: Surrey, UK.
Posts: 4,395
|
Re: EL34 power amp project
5V4 is an indirectly-heated rectifier with 5V 2A heater and quite good cathode current rating, ISTR often quoted at 175mA or somewhat more than the typical 125mA associated with the 5Z4 family. Forward drop is likely to be more than the GZ34- but it wouldn't surprise me if it was subject to development over the years and different suffixes, so might be difficult to pin down precisely for a particular circumstance.
|
28th Oct 2015, 4:39 pm | #28 |
Pentode
Join Date: Feb 2015
Location: Conwy, Clwyd, UK.
Posts: 246
|
Re: EL34 power amp project
5Y3GT drops at least 50v, but is another directly heated one. GZ34 seems to drop voltage very little, pretty much anything else will give a lower HT I think.
Lovely neat job so far! Glyn |
28th Oct 2015, 11:47 pm | #29 |
Octode
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,993
|
Re: EL34 power amp project
Thanks guys,
Ok, this is where i get clever.'; 5V4 is a GZ34 by any other name. 5Y3 is similar to 5U4 as far as I can ascertain. sorry guys.............. I did my homework on those types. Thanks for the input, and its showing what I always suspected, there arent a right lot of valve rectifiers out there. hardly surprising if you think about it. all they have to do is rectify............. So what about GZ32? as fitted to mr walkers quads? Andy. |
28th Oct 2015, 11:52 pm | #30 |
Octode
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,993
|
Re: EL34 power amp project
Pics to follow.
The amp is finished. I had a couple of "oh sh**" moments but so far the filaments all light up. I will wait until tomorrow before applying the HT. Heres where I wish i had a variac. andy. |
29th Oct 2015, 12:06 am | #31 |
Dekatron
Join Date: Sep 2007
Location: Oxfordshire, UK.
Posts: 4,311
|
Re: EL34 power amp project
Don't forget to short the amp's input and to put a load on the output before powering up. In the absence of speakers or an equivalent high-power resistor it's possible to do quite a lot of expensive damage if the amp should do something unexpected. Don't use your best speakers though. They might not appreciate a very loud signal applied to them. If by any chance you've got the feedback polarity reversed (you wouldn't be the first to do this) then the amp will quickly become a very high power oscillator. So keep your finger on the mains power switch for the first minute or so and turn it off as fast as you can if there's any sign of howling or components overheating or worse. Bear in mind too that the oscillation might be ultrasonic, in which case you wouldn't be able to hear it. When I first switch an amp on I hook a voltmeter across the output valve(s) cathode resistor(s) as this can give a very early indication of excess current flowing.
Cheers, GJ
__________________
http://www.ampregen.com |
29th Oct 2015, 12:26 am | #32 |
Octode
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,993
|
Re: EL34 power amp project
All great advice GJ.
I was aware of all of it but it is good to be reminded especially as there are folk reading this who may be inexperienced. It's always a heart in mouth moment when applying HT for the first time. Sorry I didn't remind myself to you at Tonbridge, You were having a natter with Terry at the time and i didn't want to interrupt. Regards. A. |
30th Oct 2015, 12:06 am | #33 |
Octode
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,993
|
Re: EL34 power amp project
Ok heres a pic of the hammond choke, and the innards of the (almost) completed unit. I added a "Bleed" resistor of 470K across the HT supply so that the capacitors didnt hold a nasty surprise after powering down.
A. |
30th Oct 2015, 12:20 am | #34 |
Octode
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,993
|
Re: EL34 power amp project
Now for the scary stuff.
I tested that the valves lit up ok minus the rectifier, and they did. I then put a GZ34 in minus the rest of the valves and measured the voltage rise. As it climbed above 500Volts i quickly switched off as my reservoir caps are only rated at 500V. I then put a pair of EL34 in, Having made sure the input was shorted out and the dummy load was wired up along with a "Disposable" small speaker. No bad smells, no magic smoke, and no scary noises. But HT up at 475V! which is well above the desired 450v on the output transformer CT. I tried a 5U4G and that brought the supply more or less on the button at 445V. Bit scary watching the HT rise until the EL34's warmed up as the 5U4G is a directly heated job. So I think the solution lies with the 5U4 although I would rather stick with an indirectly heated rectifier if I can find one with a 40V drop across it. Or is there a way with resistors on the anodes? Anyway, in for a penny etc, and i put the ECC81 and EF40 in and switched it back on. As the volts rose i heard it come to life, with a bit of audible hum and at about 250V all hell broke loose and I realised i have built a 25watt oscillator! So as its late I decided to call it a day and i will swap the feedback and speaker connections round to reverse the phasing tomorrow. But at least it didnt go "pop" on me! A. |
30th Oct 2015, 5:01 am | #35 |
Banned
Join Date: Jun 2015
Location: Mareeba, North Queensland, Australia
Posts: 2,704
|
Re: EL34 power amp project
tut tut lad !!!
I have learned finally that late night amp "finishes" are NOT for switching on yet. Breakfast, a walk with the dog, THEN just sit and look at what you have created. The phase reversal I will forgive, I dont think ANYBODY here has got it right every time!! read the circuit! is that REALLY 270K or like me only a week ago a 474 OHM resistor 2%, BUT reading it at night I didnt see the black in the late evening!! CHECK is it REALLY red, purple, yellow, red? or red, purple yellow, BLACK, red!! My little 6BW6 amp sounded delightful, EXCEPT for the hum that I couldnt trace!! Turned out that feeding the RIAA filter with 274 ohms IS NOT THE SAME as feeding it with 270K ohms NICE WORK!! and the person that has never made a mistake, has never made anything Joe Edit: The resistors I use are Rodestein carbon film .6 watt, and are painted dark brown. The black is nigh impossible to see. Last edited by joebog1; 30th Oct 2015 at 5:03 am. Reason: error |
30th Oct 2015, 6:07 am | #36 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,876
|
Re: EL34 power amp project
Andy,
You've just discovered one of the issues with choke input filters. Without load current, the HT voltage climbs to the rectified peak of the transformer voltage... 580v less rectifier drop, and the rectifier drop will be low at low currents. As a result, with this form of filter it is usual to use rather high voltage capacitors everywhere so that removal/failure of the output valves or their cathode resistors doesn't start fireworks. At least you don't need as many microfarads for a given ripple voltage. The choke is an energy store, contributing to smoothing. It is also usual to look at the power supply design and calculate the minimum choke inductance needed to keep the rectifier in continuous conduction for a chosen minimum load current (meaning also allow for valve ageing). If the load current drops below this, the HT voltage starts to rise, heading for well over 500v in this case. Counter intuitively, beefier amps can get away with fewer Henries, low power amps need more. The second issue with choke filters is to determine what the peak current is that the load can draw, and then to get a choke of the necessary inductance which is physically big enough to not saturate (well, not lose too much inductance) at this current. What this means is that there tends to be a ratio between the volume of the mains transformer core being used in a design and the volume of the core of the associated choke. Only tends to be, proper calculations are the real decider. However, that Hammond choke looks significantly smaller than is usual given the size of the mains transformer. You've quoted a current rating for it, but not mentioned whether this is its thermal rating or its saturation rating (and different people consider saturation ratings at different percentages of loss of inductance) There have been significant improvements in electrolytes and the etching of foil which have dramatically shrunk modern capacitors, but nothing significant has happened to chokes. The iron isn't much changed, they still need the same gap sizes and they still need to be the same weight. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
30th Oct 2015, 1:02 pm | #37 |
Dekatron
Join Date: Nov 2011
Location: Surrey, UK.
Posts: 4,395
|
Re: EL34 power amp project
Is this a choke input arrangement, or CLC pi-type filter? It's difficult to be certain, but it looks as though the black lead-outs from the Hammond choke are connected between the +ve terminals of a dual-section capacitor. I wouldn't be too concerned about slight exceeding of nominal voltage until the valves all warm up and start drawing current- on the basis that voltage rating is a statistical thing, rather than an absolute wall- there have been quite a few commercial designs that pushed things a little here and got away with it for long enough. It's especially difficult with reasonably grunty amps where the design considerations say "get as close to 450V running as you can" but electrolytic capacitors, presumably for some fundamental reason that combines chemistry and physics, rarely appear with much choice above 500V. Even 600V ones are disproportionately bigger, more expensive and generally poorer-specced than 500V types.
520V initial I wouldn't worry overly about, 550V I'd want to do something about it.... Oh, and I'll go along whole-heartedly with what joebog1 says about late-night syndrome- sometimes, one can get perserverance (good!) crossing over into perserveration (bad!) and things can slide backwards somewhat. Very possible when burning midnight oil. Tempting as it might be to do otherwise, draw a line under things and re-visit the next day, when that erroneous connection or mis-read component is only too obvious with a fresh mind. A lesson I've had cause to remember |
30th Oct 2015, 2:17 pm | #38 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,876
|
Re: EL34 power amp project
CLC after the rectifier would tend towards peak rectification and the choke would have a lot easier time. LC after the rectifier is true choke-input filtering and needs big chokes but gives a very smooth HT at mean voltage (with the nasty rise at low current)
David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
30th Oct 2015, 4:30 pm | #39 |
Octode
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,993
|
Re: EL34 power amp project
Ok gents.
It's a CLC filter 50uF-5H-50uF. If I use a GZ34 the Ht doesnt rise above 475V, the warm up times of the valves seems to be nicely in sync. The actual maximum using a 5U4G just about keeps within the 10% momentary overvoltage of the caps. I would have Used 630V jobs but couldnt find any can types, plenty wire ended though. Joe, I had spent a whole day checking and checking again, tracing all the wiring before attempting a full switch on. (I tested the heater wiring before building the rest of the amp). So I was happy to give it a go, fully expecting the oscillation as I had a 50-50 chance of getting it wrong. Anyway so far so good. I have a lot of hum which was easily diagnosed as bad input wiring from the phono socket. My idea of the earthing run was not a good one. Nuff said. I'll get a bit of twin core microphone cable out of the toybox and do a "semi-balanced" connection, i.e. only the 2 internal conductors wired at the input of the EF40 and the screen connected at the socket. That should fix it. i also have the option of opening out a spare hole on the front of the chassis directly adjacent to the EF40 and put the RCA socket right next to the input valve. This approach appeals to me on a number of scores. Mainly because its a very short bit of wire and its as far from the mains transformer and PSU as possible. indications so far are that its a nice quiet amp but when i plug the hi fi speakers in I will know for sure. A. |
30th Oct 2015, 4:52 pm | #40 |
Octode
Join Date: Nov 2013
Location: Penrith, Cumbria, UK
Posts: 1,993
|
Re: EL34 power amp project
So, heres the question. Do I put a bit of resistance between the anodes of the rectifier and the transformer in order to drop a few volts? assuming a 145mA current draw, the nearest preferred value would be 180 ohms to lose 25 Volts. And I think I really ought to be using a bit of resistance on the anodes anyway. Leak seemed not to bother but i figure it was a bit of cost engineering, 2 fewer power resistors.
The other baffling thing is the voltages across the cathode resistor are a bit high, at 35V which means the EL34's are pulling 74mA rather than 68 as specified. I dont know if it matters that much. the cathode resistors are 470 ohm rather than the 440 specified in the circuit but I didn't think it mattered that much as that's well within a 10% tolerance and not far from 5%. I could easily get it down to 440 ohms with a 6.8k in parallel. this all just tweaking though. I'm rather chuffed so far, I was expecting a lot of debugging especially as if I had been able to specify the chassis holes I would have done them differently. Thanks for the hints and tips gents. I hope anybody else thinking of having a go is getting the benefit of my experiences. A. |