• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

New or used output xformers to play around with?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I'm looking for suggestions for PP output xformers for, mainly EL84s, but also EL80s since I have a few of those hanging around.

I've scanned eBay for many months looking for either fairly good or good xformers but think the prices, when compared to buying new ones from Edcore, are too high?

The amp I plan to build and play around with will not be for use as a amplifier in my main system, but rather for use in either my kitchen or my play room where I play around building things.

Currently, I plan on starting out with Eli's El Cheapo design, but using a power supply using xformers I have on hand.

My inclination is to keep saving up and buying full range Edcore xformers: 21W CXPP21-MS-8.6K for $57, 25W CXPP25-8-7.6K for $48, 25W XPP25-8CT-7.6K for $31.

But, I lack experience to judge if used xformers, like from Hammond organs, would be just as good, at a slightly lower cost? Or, are older xformers, like those in Heathkit EA3s, better than current production Edcore's

I'd appreciate any comments.

Best from Tucson
Bob
 
Hey Bob,

Edcore makes fine transformers, and you will also fine good ones in vintage hi-fi gear. One of the problems using organ parts is that you need identical pairs and that can be problematic when salvaging from mono gear. Junk shops often have old stereo consoles that can have pretty good iron. I salvaged a nice PP EL-84 Stromburg Carlson from a console that had very nice OTs.

So, if you want to re-use OTs, there are a few things to consider other than quality. They must fit your application too. Some OTs will just have three primary wires and two secondaries, meaning they only have one secondary winding. Others will have four, meaning they have taps for perhaps 4, 8 and 16 ohms. Obviously, the latter is usually better.

Also, the primary impedance is important as it must match your power tubes impedance. You can measure it by applying a low voltage AC signal to the primary and measuring the output voltage. A variac or an AC wall wart are both suitable sources. The ratio of voltages is equal to the winding ratio. The square of the winding ratio times the secondary impedance is the primary impedance. Let's say you have a winding ratio of 27 when measured against the middle tap (8R). The primary impedance calculation is: 27*27*8R=5832R.

I hope that helps. Happy hunting.

Dave
Phoenix
 
Last edited:
Cheap Output Trannies (and lots of other stuff) from here.
Reliable Supplier.
How good is the EL84 PP Output Tranny - good but not great -when I pulled it out of an amp and replaced it with a Hammond 1608 I noted an immediate improvement BUT price is 1/4 of the Hammond.
mableaudio.com

The Mable trannies pulled form this HiFi Amp are now doing a fantastic job as output trannies in a pair of 6V6 push pull Guitar Amps.

Since you are in the US you may be better going with the Edcore, good trannies and once you factor in delivery costs (although Mable are pretty good in this regard) then prices will not be all that much different.

I've just bought a set of SE Output trannies and the power tranny for a stereo 300B SE from Mable. I use their EL34 amp power tranny in my latest Baby Huey build.

Cheers,
Ian
 
Last edited:
Bob,

If you restrict yourself to triode wired "finals", Edcor's GXPP15-6-8K will allow you to try both 8 Ω and 4 Ω speakers.

To try UL mode "finals", buy Edcor's CXPP25-6-7.6K and you get the same flexibility with speaker impedance.

Do what you like in the B+ PSU dept., as long as at least 200 mA. are available and the rail comes in at about 355 V.
 
I appreciate the time everyone took to give information to me.

Finding used tubed gear here at Goodwill places has been a fruitless endeavor; been looking during the past 5-7 years.

Using my Variac to test output transformers is a great tip.

Eli's suggestion seems to resolve my dilemma.

Best from Tucson
Bob
 
I've bought five or six consoles for the amps in them.

I've also bought several transformers and sets of transformers off Ebay in the past.

I no longer buy old transformers for use in amps as I consider them (1) a risk due to age, and (2) overly expensive compared to new production transformers (3) inferior to new production transformers.

In an apples to apples comparison, the Edcor XSE series 10W transformers are as good as or better than the original Magnavox (and other mfg) transformers I've seen in 6BQ5 pp systems. At $18us plus shipping, I don't see how buying used transformers is worth it.

The original Magnavox (and other mfg) transformers I've seen are no larger than the Edcor 10W transformers.

The GXPP for $26 are certainly a step up, and one could go to the 15W version and improve bass for $37 each.

Stepping up to the CXPP you are in a totally different class and on par with (or better than) old component system amp transformers.

My latest amp is a single end stereo amp with 6P41S output tubes. They are selectable for Triode strapped or UL.

I used the Edcor GXSE15-8-3.5K transformers, and driving a set of Klipsch Heresy speakers they sound fantastic.

Several of my granddaughters friends who have listened to the system are extremely impressed with the bass, which far exceeded my expectations.

This is at moderate listening levels.

My wife and granddaughter both have SE 6P1P stereo systems using Edcor XSE10-8-5K transformers in the bedrooms. These drive speakers built with FE103EN. They both can play them loud enough to drive me to the den. The poor little FE103s are going to be trashed and I'll have to build new speakers with bigger drivers.
 
Last edited:
If you do plan on buying from Edcor, they have a long wait period. I ordered 3 transformers for my project on Feb. 23 and the order is still in the "processing" stage. Of course, the money came out of my bank account 3 days after the order, but since they make each order individually, there is a long lead time. Just keep that in mind as you plan your project(s).
 
Thanks to both of you for your time and the additional information.

Your prose flows, is well organized, concise, and to the point Mr. Gimp.

Knowing that the Edcore xformers are at least equal to or better than the output xformers I have been chasing relievs me of the urge to keep checking eBay for my planned project.

Best from Tucson
Bob
 
The original El Cheapo used Triode Electronics TF110-48 UL output transformers. For playing around, they are fine at $29.95 apiece, but the amp itself deserves better if you can spring for better OPTs. The Z-565 is a good option, as are the Edcors.

I wanted to interject a note regarding B+ on the El Cheapo since I'm currently in the process of breadboarding for a final chassis build and doing a LOT of experimenting: I have listened to this amp for three or four years now, and the best sweet spot I have found for B+ is actually about 305-312 VDC. This is easily attainable as originally laid out using the N-77U isolation transformer specified, or alternatively, going to a PA-774 transformer and 5R4GYA with a cLCLC supply using 4-5 muF input cap followed by 10H/150 muF/3H/220 muF (this was worked out in partnership with Eli, but was not in the original "frugal" plan. Detail and impact with the tube supply are superb, and the original UF5408 suppy is great, too.

The CCS is set at 3.5 mA per 12AT7 section or a total of 7 mA per channel. This is per Jim McShane's suggestions for the 12AT7. The 12AT7 anodes under these conditions are running at about 132-137 VDC each after the 50K anode resistors.

In my opinion and after many sessions, the higher voltages attained using a booster transformer with the N-77U or a tube rectifier with less forward drop all sounded either less focused/detailed or a little less dynamic on kick drums and bass. I have yet to try going in the other direction and running the 12AT7s hotter, say at 4.5 or 5 mA with MUCH higher B+ needed. In that case, the voltage on the anodes and screens of the 6/12AQ5s could be a little to high, and I would likely add some to the screen resistance. The 'AQ5s start to glow with that wonderful blue color once you get about 345 on their plates.

Eli knows I'm overly cautious, but for my ears, the 3.5 mA at 305-312 B+ sounds fantastic.
 
The Z565, either original or "cloned", is (IMO) the "Gold Standard" PP O/P trafo for use with "12" W., multi-grid, power O/P tubes. You will spend a good deal more to get anything resembling a genuine improvement. The Z565 is a price/performance champion. :D The guitar amp trafo shown on the schematic is no longer the bargain it was a few years back. Go with Edcor, if the budget can't tolerate buying Z565s.

Jim McShane clued me into the 'T7 triode sounding good at 3 mA. IB, with 200 to 220 V. on the plate. You can see where I got the 355 V. rail voltage number from. As the rail voltage rises, increase the value of the O/P tube cathode bias resistor to 350 or even 390 Ω.

Jeff, if you want to go with a lowish B+ rail voltage, drop the 'T7 load resistance to 33 KOhms and buffer the plates with DC coupled ZVN0545A source followers. Buffering is necessary to keep both gain and linearity up, when the load resistance is reduced.
 
Thanks for the mention of the Z565-48 xformers Randy. I had read about them but had forgotten about them.

I appreciate the time your post took Steve. Your experience provides a treasure trove of jewels for people.

I have a few copies of Eli's schematic. Unfortunately none of them are clear enough to make out the xformer numbers.

Once again Eli, many thanks for your time, the update information, and of course for your creation.

Having both the PA xformer and the circuit for the power supply enables me to play around hearing what lesser configurations sound like.

Best from Tucson
Bob
 
Bob,

I've uploaded the biggest "El Cheapo" graphic in my possession. I think you can make the various part #s out.

Touch base with Jim McShane and get the EL84 cathode bias resistor value. The 'AQ5 shown in the schematic is a member of the 6V6 "clan".

"El Cheapo" was started with a very low budget in mind. We hoped for a decent, if unspectacular, end result. Our ambitions were greatly exceeded. :spin: The design seems limited by O/P trafo quality. The best reason I can come up for that is our refusal, from day 1, to compromise PSU quality. While highly cost effective, the Greinacher doubler PSU is also very competent. The Greinacher doubler was good enough for McIntosh, Hegeman, Fisher, and Marantz. :D So, it's definitely good enough for all of us.
 

Attachments

  • EC big.gif
    EC big.gif
    38.8 KB · Views: 270
Yes. That schematic came out nice and clear Eli.

My HeathKit SB221, dual triode 3-500Z amplifier, also uses a voltage doubler.

If I remember correctly I bought it used towards the end on the 1970s.

I became interested in rebuilding tubed audio gear after I built a CW transmitter and power supply, 6AG7-6L6, I found in a late 40s ARRL HandBook.

For some reason i wanted to see what state of the art was when I was born, 1948.

Best fromTucson
Bob
 
Eli and Bob,

Bob, it's fun to hear your history. My journey is similar, but more through following my brother's interest at the start. We used to drag all kinds of junk home. I picked up electronics as a hobby later, following an interest in loudspeaker building. My early searches on the Internet led me to the El Cheap thread.

Okay, I decided to give Eli's and Jim's parameters another try. I now have 350 VDC on B+ and reduced the CCS current to 3.0 mA. There are 200 VDC on the 12AT7 plates. You need a good booster transformer to help the N-77U. A LV control transformer on the order of 14 to 24 Volt with at least a 3A current capacity would be my starting point. My Radio Shack 12 V won't quite get there but stood up well. A Signal 14V was making a little noise, so I'm still digging around.

I also achieved this with the PA774 by using two 6CA4/EZ81 rectifiers (each with plates paralleled, one per secondary lead) and a 24 muF initial capacitor in cLCLC. I could get a few more volts out by taking out or reducing the second choke, but then I would need to do as Eli suggested and put in higher value cathode resistors. I may try that. I'm using 350 Ohms on the shared PP cathodes now, and am getting a 23.5 V drop on average (~33mA per 'AQ5). NOTE: the amp finals are hooked up in ultralinear mode.

This is a new "sweet spot" for El Cheapo and my listening. Once the B+ got to this level, the sound improved remarkably. What was pleasantly euphonic at the lower rail voltages is now accurately detailed, well-staged, crystal clear without any listening fatique. Bass is clean and deep. The Z-565s are amazing at this operating point.

Thanks to Eli for the gentle push to keep at this.

--Jeff
 
Last edited:
FWIW, the power "iron" set I currently recommend for "El Cheapo" is a Triad N-77U for B+, a Triad VPS24-1800 for "12" V. heater power plus B+ boost (1 secondary each), a Triad C-24X as the B+ filter choke, and an Allied 6K27VF for B- plus 'T7 heater power. Order 'em all at once from Allied Electronics and hold shipping costs down.

Put ferrite beads on the leads providing the "12" V. heater power in an attempt to keep "hash" associated with the big doubler stack caps. from sneaking into the signal path via the "back door".
 
Bench test results on El Cheapo power supply per Eli's "iron" recommendations:

I easily achieved 350-353 VDC on the B+ rail using the VPS24 as a booster with the N-77U. I got 348-350 VDC using one secondary and 351-353 VDC with the secondaries paralleled. So you can get a few volts higher on the rail if you dedicate the dual 12 volt secondaries in parallel and choose not to use one for heater duty. If your mains voltage is healthier than mine, one set of secondary output should be fine. My mains voltage ranges some during the day in my neighborhood.

The 12AT7s now have 200 - 207 on their plates, which is more in the ballpark for what Eli says that Jim McShane recommends (see earlier in this thread).

This sound at this rail voltage is phenomenal, both for the original solid state and the tubed versions of the power supply.

Question to Eli or anyone else: With this full voltage rail, I measure ~350-352 Volts on the screens when using the UL taps on the Z-565 and 1KOhm resistors on the screen pins. This is higher than the plate voltage (plates are at 323V plate-to-cathode or about 347 plate-to-ground). Is this a problem, or is this how UL works? I realize that UL taps won't drop as much as the full primary path, since they are set at a percents of the full primary. Eli has mentioned the fragility of the 'AQ5 screens, so I wonder if I need to up the screen resistors.

Thanks,
Jeff
 
Last edited:
Jeff,

353 V. of B+ should be just enough. :) Less than 2 V. get dropped in the "check valve" decoupling setup.

I think the 1 KOhm screen grid resistors will be OK. What damages screen grids is excessive power dissipation. If screen current goes up, the voltage drop across the resistor increases and that causes the current to drop back. I would not be at all comfortable with 100 Ω parts in that role.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.