New Build

a 3W single ended EL84 amp for home, work & school

New Build

Postby AmadeusMozart » Sun Dec 22, 2013 6:35 pm

I've just ordered the budgie board so now the collection of miscellaneous parts starts. I'm not completely clear on the power transformer, did some modelling with Duncan PSU OII Designer and it looks like the 120uF is creating some large peak currents. I am wondering if the XPWR013 is a better option and will run cooler. Another option I've been thinking about is using one of the higher voltages transformers (e.g. the XPWER106 275-0-275 or the XPWR014 with 300-0-300) and use solid state series voltage regulation instead of using a choke. Main reason for the latter is to keep shipping costs from the US down - these exceed the cost of the actual item. Has anyone an opinion on replacing a choke with a "gyrator" (electronic choke) as advertised on eBay Germany?
http://www.ebay.de/itm/Elektronik-Dross ... 1c330356fd

http://www.ebay.de/itm/Elektronische-Dr ... 19b61ed293 ?
AmadeusMozart
 
Posts: 46
Joined: Thu Dec 10, 2009 9:06 pm

Re: New Build

Postby Shannon Parks » Mon Dec 23, 2013 10:02 am

Not to discourage you as I think this is a fun DIY project to try and suitable since you can tweak it to your specific rectified voltages, but...

I spent a few weeks researching something like this (a cap multiplier circuit specifically) during the Budgie SE initial design. Bottom line was that is that the circuit starts getting ever complex as you try to improve it (SS electronics in a nutshell) and became much more expensive than an $8 choke and not more effective. But the deal breaker was the burning up lots of power and possibly going *poof* - a definite possibility depending on the transformer sourced and local mains AC levels. And then there's the start up voltages and the possibility of running without tubes. The KISS method just works well for tube amps, with the unregulated but heavily filtered power supplies.

Shannon
User avatar
Shannon Parks
Site Admin
 
Posts: 3764
Joined: Tue Mar 18, 2003 5:40 pm
Location: Poulsbo, Washington

Re: New Build

Postby AmadeusMozart » Mon Dec 23, 2013 3:47 pm

I was researching the same when I build the SE 6L6-GC amplifier and in the end decided to stay as far as possible away from solid state regulation especially because of the "poof" factor. Unfortunately shipping that $8.00 choke will cost me $ 44.00...... Other reason is that solid state components may be available today but not long term. I know from my HAM days that there was an expensive transceiver and three (or was it two years) later the transmitter power output transistors were unobtainable making it in a 10K US doorstop.

I'll be looking at the XPWR013 (255-0-255V @ 125mA, 6.3V @ 4A) transformer and a XC75 (1.5H @ 250mA) choke - don't like transformers running too hot (had this with the James stuff, the choke especially got rather got) which will give me some flexibility to configuration change in case things do turn out to be less than optimum.

It will take several months before I make much progress but eventually I hope to post some pictures. I want to get these metronomes singing again..

edit: As mentioned before I've been rather fond of plate-to-plate feedback but have not been able to model the Budgie in that configuration (too much medication making me dopey - everything takes me now five if not ten times as long as only a few years ago.) I had saved some schematics from the Motorola SH12E / HS-704 record player' amplifier which does not use feedback and has cathode resistors that are not bypassed with a capacitor. I've always considered it a bit of a "sleeper" amplifier of approx. 4 watts. If I look at it then I can see that the Decware Zen is something of a descendant from that design. Did a bit of modelling in the past and it looks pretty good but at the time decided to go with the 6L6-GC. It would be interesting to figure out how it would stack up against the Budgie. One thing I have always found that if the OPT is not included in the feedback loop that then one can get away with a far lesser quality (= cheaper) for the OPT. Perhaps this is why so many like triode mode without feedback.

At this stage the plan is to put it in triode mode with feedback but I'll have to measure the output transformers first to work out the impedance ratio and see if they can be used in that manner. But as plans go - they are apt to change at the last moment so nothing is set in concrete.

I've decided to bite the bullet and order the XPWR013 transformer and XC75 1.5H / 250mA choke - unfortunately freight will be a couple of dollars more than the actual goods if I let Edcor ship them directly. Even asking them nicely if they would ship it through a cheaper way did not make a dent in their attitude. I'll see if I can ship them through a freight forwarder or other US reseller.

I'll have to make my own chassis and have been leaning towards just making an aluminium top plate with a wooden base instead of the approach I took with the 6L6-GC. Problem with the latter was that the paintwork could get easily damaged and then is hard to repair whereas with wood such is not the case especially if the top plate is not painted.
AmadeusMozart
 
Posts: 46
Joined: Thu Dec 10, 2009 9:06 pm

Re: New Build

Postby AmadeusMozart » Fri Dec 27, 2013 2:30 am

Change of plans....

Have been using Duncan Amp PSU Designer II to model the power supply. At this moment I've come up with a R-C-L-C configuration that looks flexible in voltage adjustments. Due to the high freight I am leaning now towards the Edcor XPWR106 275-0-275V @ 125mA with two 6.3V @ 2.5 A and a CXC100 5H 200mA choke. First resistor following the rectifiers between 330 and 400 (depending on current draw), C at that location 20uF (not critical) then the choke and then a capacitor around 100u - 220uF (once again not critical). The resistor will absorb around 3 ~ 4 watt (estimate) so a 10W wirewound will do. This will have much lower peak currents through the transformer.

I am wondering if the first capacitor should be a motor start capacitor in order to cope with the relative high currents in there - does anyone know if they'll last longer?

edit: There are several different types of motor capacitors - one wants one of the later film capacitors designed for "run" application, the "start" have a limited life, the "run" are often designed for 60 000 hrs.

Hopefully the choke is able to cope with a "swing" configuration (L-C-R-C) - this configuration drops the final voltage considerable.
AmadeusMozart
 
Posts: 46
Joined: Thu Dec 10, 2009 9:06 pm

Triode or.....

Postby AmadeusMozart » Sat Dec 28, 2013 9:41 pm

Have just been looking at all the different graphs and came across the specifications for the 6BQ5 from STC (Standard Telephone Company, Sydney). In there was some detailed information when the tube is run as a triode and as a pentode.

In pentode mode at 200V on plate and g2 the output is 3.2 Watt with less distortion than in triode mode. Normally a pentode has more D3 distortion than D2 however at 200V the D2 is higher than D3 up to approx. 2.75 Watt. Why is it that many are running these tubes at higher values than is suggested in the tube manuals - cannot find any distortion figures for those higher B+ numbers.

In a triode the D2 is higher than D3 which is a natural behaviour for instruments and which is for me a major reason to go to a triode mode. However the pentode mode at 200V is interesting - more so since my OPT looks to be designed for approx. 3 watt. Better get a choke that is capable of running in "swing"mode (e.g. the 5H or 7H, there is a minimum load they'll need) so I can change the smoothing configuration and have different B+ values available for the different configurations. (L-C-R-C versus C-L-C-R-C)

Plenty of time to ponder over this - I have first to finish another project: making a small tube tester for selecting pre-amplifier dual triode (12A?7 etc) - while working on that I have started to collect parts for the amp.
AmadeusMozart
 
Posts: 46
Joined: Thu Dec 10, 2009 9:06 pm

Re: New Build

Postby EWBrown » Sat Dec 28, 2013 10:37 pm

In the 1950,s it was the common practice to run the 6BQ5 / EL84 / 6P14P series, at lower B+ voltages, 200 to 250VDC was the standard operating practice.

in the 1960s and 1970s, it was "discovered" ;) (lol) that they run very well on 325 to 380V B+, (just keep the plate dissipation below 13 watts).

Lower B+ and higher plate current (250VDC, 45 mA) makes for a lower plate resistance, higher values (350VDC, 35 mA), the rP is somewhat higher,
both work well with a 5K OPT in SE mode, and 6.6 to 8K A-A for push pull.

In the Budgie SE, the 6BQ5s are run fairly "gently", and should have a good long operating lifetime. In my SE designs, I tend to push them to the PD limit, and sometimes slightly above. [:) p[

/ed B in NC
Real Radios Glow in the Dark
User avatar
EWBrown
Insulator & Iron Magnate
 
Posts: 6389
Joined: Wed Mar 19, 2003 6:03 am
Location: Now located in Clay County, NC !

Re: New Build

Postby AmadeusMozart » Sun Dec 29, 2013 3:39 am

When I modelled the 6L6-GC amplifier I found that at lower voltages D2 is larger than D3. When an instrument is played then the harmonics have an decay as D2 > D3 > D4 etc. Mullard mentions 1.5W output in triode mode, all the others mention 1.9W ~ 1.95W. Pentode mode at 200V gives 3.2 watt which is a nice gain.

I have got an APPJ miniwatt amplifier where one channel goes into oscillation after the CCS was removed. It was marginally stable before but that unstable side always measured higher current draw / higher cathode voltage (approx. -9.4 V at 260V B+ ). The output transformers are Japanese and (supposedly) rated for 3 Watt.

I can reuse those 3 Watt OPT when running at 200V in pentode mode if it can cope with 50mA or going into triode mode at 250 - 270V. Drive voltage is a lot larger for the triode mode. Both voltages (for triode @ 250V and pentode @ 200 V ) can be achieved with the same transformer, just needs slightly different configuration with the resistors in the (R-)C-L-C-R-C.

Speaking about the transformer - I am confused about this. When modelling with PSU Designer I get a current value for the transformer which is higher than the Edcor transformer specified for the Budgie (60mA). The PSU Designer comes up with a value of 90 - 110 mA rms depending on first capacitor and internal resistance in the transformer, choke and load.

Going to the Radio Designer Handbook version 4 it mentions on page 236 that with a condenser input: (quote) "The r.m.s. current in each half of the transformer secondary may be taken approximately as 1.1 times the direct current to the load."

I have had a transformer burn out once before that -should- have been properly sized hence my apprehension.

Any clarification on this rms subject will be greatly appreciated. I am planning to play it safe and am looking at a 250-0-250 @ 125mA transformer for triode operation. (XPWR068) I might get a choke with higher inductance to keep the ripple down.
AmadeusMozart
 
Posts: 46
Joined: Thu Dec 10, 2009 9:06 pm

Re: New Build

Postby EWBrown » Sun Dec 29, 2013 3:04 pm

There isn't a problem with having "too much" power supply, as long as the B+ voltages are correct, more current available can be a good thing, to assure that the amplifier will be less likely to go into clipping / distortion due to insufficient current. Keep the EL84s at a max of 325 VDC B+ and 45 mA, each, and everything will be good to go.

There is the Chubby Clementine, now I'm developing a Pudgy Budgie, let's just say that it will use two of the red circuit boards....
And a heftier power transformer, each channel's board will have its own dedicated power supply. [:) (y) O:)

/ed B
Real Radios Glow in the Dark
User avatar
EWBrown
Insulator & Iron Magnate
 
Posts: 6389
Joined: Wed Mar 19, 2003 6:03 am
Location: Now located in Clay County, NC !

Re: New Build

Postby AmadeusMozart » Sun Dec 29, 2013 9:27 pm

Situation just improved a lot. Decided to take another look at the Miniwatt APPJ amplifier and found a SMD resistor far out of spec and a capacitor out of spec. Replaced both, added a couple stopper resistors and got the amplifier stable (without parasitic oscillation). It is running in triode mode and has a pretty fine performance except it runs at a higher temperature than I like. I have added ventilation holes and increased the size of existing ones - dropped by about 15 C but the enclosure is still running at approximately 50C on the outside and the inside will be even hotter. So it will go on the local internet auction website and help to fund some decent OPT. Thinking about transcendar or tomiko here and then build the Budgie as per design but with a slightly bigger mains transformer. Liking this triode mode a lot - more bass presence and I am perceiving more detail. Sound staging with my FF105WK metronomes is about the best I've heard. This is the way to go - I feel this little amp will be giving the Decware Zen (ES84ZS) a run for its money.

I have little doubt that the little Budgie with good iron will be just as good (as the Decware SE84ZS), all the ingredients are present.
AmadeusMozart
 
Posts: 46
Joined: Thu Dec 10, 2009 9:06 pm

Output transformer

Postby AmadeusMozart » Tue Dec 31, 2013 5:24 pm

I have started to do some calculations on what the specifications have to be for the output transformer. According to the voltages given in the Budgie manual and the values of the components the amplifier will be better off with an output transformer that has a 7000 Ohm primary. I believe the Edcor GXSE 10-8-8K (7000 => 8 ohm) OPT is a better choice than the GXSE 10-8-5K (5K => 8 Ohm) if one is looking for a transformer on a budget.

However a tube amplifier is all about what iron is being used and if I look at the sporadic information that I can find about the Edcor it appears that the inductance is just not sufficient to give a very low bass performance. THE RDH 4 manual specifies that at 7000 Ohm 48 Henry is required to reproduce 50 Hz properly and 80 Henry to reproduce 30 Hz properly. The 1 or 2 dB drop in output is not the major consideration, it is the phase shift that happens with the resultant distortion that plays a big role. Now the hunt is on to find some good iron. Time to send some emails off.

edit: Have not yet received the PCB but I like what I can deduct from the pictures. Initially I was a "purist" and wanted only point to point wiring (certainly better for long term maintenance) however I do like the PCB for ease of assembly. And even some very expensive amplifiers are using PCB these days, it is not like the days of the old phenolic boards that would warp and tracks would come off (especially in the televisions with vertical boards and many tubes on them).

edit2: Bought some 10 watt Trancendars OPT. Will do the trick.
AmadeusMozart
 
Posts: 46
Joined: Thu Dec 10, 2009 9:06 pm

Re: New Build

Postby AmadeusMozart » Fri Jan 03, 2014 6:19 pm

EWBrown wrote:In the Budgie SE, the 6BQ5s are run fairly "gently", and should have a good long operating lifetime. In my SE designs, I tend to push them to the PD limit, and sometimes slightly above. [:) p[

/ed B in NC


FYI

Just came across a Dutch design that was well regarded. Measurements at the 6BQ5: pin 7 (plate) 280V, pin 3(cathode) 8.5V, plate current 58mA (not a typo here, perhaps in the book?) Cathode resistor: 2 x 270 Ohm, 1 watt parallel (135 Ohm).

By the way if you are into lots of feedback (40dB) then I have a well regarded Dutch design (stereo, complete with point to point layout so simple reproduction) with an ECL84 driving an EL84. I do not like too much feedback so not much use to me.
AmadeusMozart
 
Posts: 46
Joined: Thu Dec 10, 2009 9:06 pm

Re: New Build

Postby AmadeusMozart » Sat Jan 04, 2014 4:55 am

Ordered beefy power transformer and a choke: Hammond 370FX and Hammond 156R. (Freight on Edcor was just too much and over a certain amount another well known firm will sent me without charging freight so the Hammonds work out just as expensive as the Edcors). Will need some resistors to drop the B+, was not much difference in price between the 370FX and the 370EX. Now if I want I have the opportunity to change the circuit. OPT are 10 Watt rated so I could conceivably go with the 7027A (touted as the "HiFi" version of the 6L6-GC).

On the subject of the 7027A: if you are prepared to use a top cap then use the 6BG6, needs some rewiring but the NOS are still very cheap and they are a quality tube. http://www.vacuumtubes.com/6BG6.html

5 January - ordered the bulk of the parts to populate the PCB with Mouser. Am shipping through a freight forwarder, direct shipping from Mouser is ridiculously priced. A couple of capacitors were not available but I'm sure that I have those values somewhere in the junk box.

6 January - mailman delivered the PCB, yeah. Decent quality PCB, nice silk screen. Waiting to get confirmation on choke and power transformer shipping. Ordered a 300 Ohm 30 W adjustable resistor from ePray to adjust the B+ (not many decent ones around any longer) some years ago I got some nice lockable Ohmite for the 6L6-GC amp as cathode resistor - wanted those so I could do tube rolling). Mouser no longer has the Alps pots - obsolete. Managed to find one on ePray. Transcendars have been shipped. Time to start giving some thought towards layout, likely to use some 1/8" aluminium.

7 January - Transcendars are in the country, expecting delivery tomorrow. That's fast from the US, suspect it has to do with the high shipping costs being some urgent delivery. Have heard that if it is valuable stuff it is best to ship that way otherwise you run a chance that it falls off the back of a truck along the way.

7 January - did not have to wait - the transformers arrived, yippee! Weighing at 1870 grams each, still on the small side but OK for me, don't know how these are rated at 10 watts, they are a different shape but about the same size as the ones I bought as youngster for 4 ~ 5 Watt (specified for EL84). My rule of thumb calculations give -2dB at 25 Hz for these which is sort of OK for not too much phase shift at 40 Hz. Anyway good value for the money, I am going by the notion that it is always the cheapest buying the best straight away. (otherwise you'll buy twice - first the junk and then the good stuff)

8 January - have last few days been searching the web for the ultimate answer on to use screen resistors or not and if so what values. Finally found the definitive answer in a publication from Langford -Smith (1955 Radiotron) that recommends 47-220 Ohm and a cap of 1 or 2 N between screen and plate while g1 should have 10K and 470K to earth. Even the plate can benefit from a 100 Ohm resistor (often seen in transmitter outputs) but I am skipping that one. Saw a Tango U-789 advertised on oBoy but closer examination showed it was 3.5K and had only 15H inductance. Thanks but no thanks, someone else may have those. Originally I had selected a more expensive Hammond that would have given me 250V-0-250V instead (same current, apparently a more popular model) than the one that I have on order. Would have been slightly easier to implement but less flexibility down the line. Removed the cloth tape around the Transcendars, the adhesive was already sticky and it will collect dirt in a hurry. Have written to Gery asking what varnish he used so I can give the core an extra coat. If oil based I may even put on a semi-gloss black.
AmadeusMozart
 
Posts: 46
Joined: Thu Dec 10, 2009 9:06 pm

Modifications

Postby AmadeusMozart » Fri Jan 10, 2014 9:02 pm

Some modifications have been decided upon. First smoothing capacitor is a polypropylene in oil motor run capacitor of 30 uF / 370 VAC (not start, they have less life expectancy, sometimes as little as 1000 hrs, the run are often 60 000 hrs) and it will be preceded by an 500 Ohm 50 W adjustable resistor after the rectifier. This because that capacitor is normally the first to fail due to the high ripple and the resistor is to adjust the final DC voltage at the OPT to 257 V (250 + cathode voltage making net 250V between plate and cathode).

Cathode resistors of the EL84 will be 140 Ohm - this cannot be done with the recommended mains transformer because the tubes will draw a lot more current. Plus the output transformers need to be beefier than the small Edcors.

I spend several days to have a definitive answer if UL needs a screen resistor or not. According to Langford-Smith of the Radio Designer Handbook fame it needs a screen resistor even more so then when in pure pentode mode and the value should be between 47 and 220 Ohm. At the same time there should be a capacitor between plate and screen between 1 nF and 2 nF. There is also a recommendation to have a g1 series resistor of 10k.

Since I will be using this amplifier with either my FF105WK metronomes or the FF125WK metronomes I will use the headphone resistor location to put a zobel in place. The value for this is a 9 Ohm resistor with a 0.4uF capacitor in series.

The 12AX7 will be replaced by an ECC803S from Tung Sol.

Have not yet made a decision if I will replace the cathode resistors of the driver with 1.5K and at the same time replace the 220K grid resistor with 330K. If I do change to 1.5K then I might have to fiddle around with the feedback loop and frankly I prefer to stay away from that since it can easily lead to instability - I would need an oscilloscope that I do not have.

14 January

Running into a snafu -my usual aluminium supplier has imposed minimum lengths of aluminium profiles which makes making my own chassis very expensive. Only way around this is just to get a top and bottom plate and then get a wooden surround but I am not enthralled by that Decware Zen look. Have written to Watts Tube Audio to find out their shipping costs - I want an aluminium chassis and theirs is 0.09" thick while Hammond is only 0.04" thick. The Transcendar OPT need a larger chassis than the standard Budgie - this was expected also with having a larger than normal mains transformer.
AmadeusMozart
 
Posts: 46
Joined: Thu Dec 10, 2009 9:06 pm

Re: New Build

Postby MacL » Tue Jan 14, 2014 8:24 am

FYI - I used a Edcor XPWR013 and two GXSE15-8-5k on a 12x7 chassis for my budgie build-you can a photo on the "it's alive" post.

I see you are not in North America so it hard to recommend chassis parts- but I started building guitar amps before I began HI-FI stuff and you may want look around/goggle "guitar amp chassis" - I used a blank guitar aluminum chassis. This guy used to make cheap custom chassis but he is out of business due to a fire - there must a be local sheet metal guy (HVAC) that can bend you a box. This chassis cost me $20- and our $ doesn't buy anything anymore. Of course I like a "industrial" look anyway...
MacL
 
Posts: 52
Joined: Wed Mar 06, 2013 11:01 am
Location: new to Austin, TX

Re: New Build

Postby AmadeusMozart » Tue Jan 14, 2014 11:33 am

I had already been exploring ideas "outside the box". Unfortunately I am living in the countryside and when I include the cost of transport then I might as well ship from the US to New Zealand. Additionally workmanship normally leaves much to be desired and I do not want to finish up with a chassis that still needs more work and does not sit not flat, it is better to bite the bullet and do it right. Hence the blank Watts guitar chassis which is 0.09" thick (http://www.turretboards.com/) rather than the Hammond which is 0.04" aluminium (Hammond also has steel but I do not want that). I have to get friends / family to help since I have limited movement / power in my right hand due to rapidly deteriorating arthritis. This also means that this is a last effort at getting something meeting my requirements rather than to have to spend a small fortune for something ready made which still has shortcomings. It will be the last time I'm making something so I want to be a showpiece. (c)
AmadeusMozart
 
Posts: 46
Joined: Thu Dec 10, 2009 9:06 pm

Next

Return to budgie se

Who is online

Users browsing this forum: No registered users and 11 guests