Output Transformer Sizing Help

[bjosephs]

Hi All,


I have a 50W (6L6 push/pull) guitar amp with some undersized iron on the output. I've built a few amps, have several transformers in inventory for/from various projects, and from past experience I'm fairly confident that upgrading it won't be waste of money if I pick right. Nothing I have on hand is a fit electrically and very few of the transformers I can find in the wild will fit mechanically given the way they laid out the chassis. So I'm talking with a small shop about a custom that would be a variant of a HiFi transformer they make.

The original is a 4k4:8/4 probably only a couple lbs at most with rough dimensions in the attachment. This is a Mesa Single Rectifier Series 2 for the curious.

The goal is to get closer to the rated power of the amp at 60Hz instead of all the bass flattening out with the volume up. I have a 2xEL34 amp on the bench using a 15W Hammond 1650E that thumps more at half the nominal power. I attribute this to a core that can handle the energy.

The custom is going to be 4k2:8/4 rated nominally at 60W on a 3x3.5" EI core (that's outside dims, described in the quote as 1.125x1.5" which I believe is the size of the middle arm of the "E"), 2.5" over the windings, 25H inductance, between 4 and 4.5lbs. I'm hoping someone with better instincts or experience could tell me if the quoted specs sound like they might get me in the ballpark of my goal. I have testimony from the designer but I'd like some outside advice.

The alternative is using a smaller wattage hifi Hammond unit that will fit in this spot, perhaps after rotating 90 degrees. A couple in the 15W range would fit and I *think* that 15W at 30Hz means around 30W capability at 60Hz give or take.


Thanks,
Brian

 

[My3gger]

I have a 2xEL34 amp on the bench using a 15W Hammond 1650E that thumps more at half the nominal power. I attribute this to a core that can handle the energy.

I believe guitar OT for PP EL34 needs more than 15W at 60Hz. I'm just curious, let's see what experienced git amp builders say.

 

[ruffrecords]

25H at 60 Hz is a reactive impedance of just over 9K ohms which should not affect the 60Hz response much in a 4K4 reflected plate load so the inductance is fine. Next thing is are there enough primary turns so the core does not saturate at 60Hz. For that you need to know the number of turns, the core cross section area and the core material (presumably M6)

Cheers

Ian

 

[bjosephs]

I believe guitar OT for PP EL34 needs more than 15W at 60Hz. I'm just curious, let's see what experienced git amp builders say.

The 1650E is rated at a lot of bandwidth, full power at 30Hz, at least an octave below my lowest frequency of interest. Based on some rules of thumb I’ve seen regarding core size and power handling I think that makes it good for 30W at 60Hz.

 

[bjosephs]

25H at 60 Hz is a reactive impedance of just over 9K ohms which should not affect the 60Hz response much in a 4K4 reflected plate load so the inductance is fine. Next thing is are there enough primary turns so the core does not saturate at 60Hz. For that you need to know the number of turns, the core cross section area and the core material (presumably M6)

Cheers

Ian

It is M6 and the lamination stack will be about 1.5” thick. I presume that makes the area ~2.25in2. It’s the weekend now so I can’t ask about the primary turns but it doesn’t sound like any red flags are up just yet.

 

[guitarmike2107]

The 1650E is designed for 6AQ5, 6V6, 6BQ5, EL84, SV83 tube.
The datasheet does not give any max winding currents, but from the tube suggestion above I would assume the primary and secondary wires are selected based on those tubes currents and the 15w power output. So even though the bandwidth is high, you may burn out the winding wire. Although it may work I don't think it is sound engineering practice to to use that OT in an amp running 6L6's or EL34's without checking with Hammond about the currents.

It doesn't look like you have the space for a bigger OT in that space, so the only way to up its bass performance on the same size core would be to increase the primary windings. which might mean reducing the primary wire size to get more turns. or you might be able to get a bigger lamination in there with the same stack height.

What I would do before swapping anything out is get the amp on a scope and a signal generator and see what the frequency response of the power section actually has, you might find the OT is fine and to get the "thump" you actually need to change something else in the amp circuit.

 

[bjosephs]

The 1650E is designed for 6AQ5, 6V6, 6BQ5, EL84, SV83 tube.
The datasheet does not give any max winding currents, but from the tube suggestion above I would assume the primary and secondary wires are selected based on those tubes currents and the 15w power output. So even though the bandwidth is high, you may burn out the winding wire. Although it may work I don't think it is sound engineering practice to to use that OT in an amp running 6L6's or EL34's without checking with Hammond about the currents.

Why would the primary currents be higher with EL34 or 6L6 if there are no changes to reflected impedance or B+ potential?

Edit: I mentioned a couple different scenarios and I now realize one was the notion of treating a 15W hifi as a larger power guitar OT. It’s been a while since I posted the OP and thought you were referring to the 15W EL34 amp I referenced. Apologies.

 

[guitarmike2107]

Yes, it wasn't clear to me that you understood that you can't just take a 15w OT and use it as a 50W OT because the core is big enough.

 

[bjosephs]

What I would do before swapping anything out is get the amp on a scope and a signal generator and see what the frequency response of the power section actually has, you might find the OT is fine and to get the "thump" you actually need to change something else in the amp circuit.

Attached are some images. I’d love some input on how to keep this scientific. This is a function generator straight into the Fx return so it’s the shortest path through the OT possible. The images should be 1khz cleanest level and onset of clipping, then followed by 64Hz at the same onset of clipping level. Last one is 64Hz cranked.

 

[bjosephs]

Measurement error above. Here is 1khz and 64Hz at the same output level with a 10X probe:

 

[cleohot]

If you can find place for a big transformer,
i can recomend Hammond 1650N.
Lots of bass in a amp i replaced it in.

 

[bjosephs]

Well I went ahead and did it. Let’s all learn from my experience - if there is anything to learn.

I tried to recreate my test scenario and got more scope shots. Like last time we have the largest clean 1k I can get and then the performance at 64Hz, then the same two frequencies with the signal cranked well into clipping. It looks like I was able to get a little more clean headroom on the 1K and the 64Hz is very different. Still plenty of what looks like core saturation but now a fully clipped signal is getting through. Maybe what we’re seeing is indicative of higher inductance, hence the larger low frequency signal, but still insufficient core/windings for the power?

 

[Caphead]

I have a 50W (6L6 push/pull) guitar amp with some undersized iron on the output.....

....The original is a 4k4:8/4 probably only a couple lbs at most with rough dimensions in the attachment. This is a Mesa Single Rectifier Series 2 for the curious.

Believe it or not the undersized output transformer is actually a deliberate design choice on the Mesa Recto. I would never have believed it until I heard it for myself. I did a mod once on a total POS Crate with the smallest OT I've ever seen in a 50 watt amp, the thing was about the size of a Deluxe Reverb OT! I was so concerned about it burning up I told him I wouldn't mod the amp unless he bought a new OT. In the meantime though I did the mod, a Recto/SLO type thing and turned it on to check if everything worked and holy crap, it fkn killed! It didn't sound like your run of the mill Recto either, it sounded like the best version of a pre-500 Recto I've ever heard. It had that massive 3D like low end early Rectifiers are known for as well as a nice high end quality that didn't get harsh or overly bright like the later revision Rectifiers. It really defied all belief at the time that such a small transformer could sound so huge. I realized later that without the corrective effects of a NFB loop around the PA, the transformer is able to saturate freely filling out the frequency spectrum with added harmonics and other artifacts that are normally canceled out when NFB is applied.

 

[bjosephs]

Believe it or not the undersized output transformer is actually a deliberate design choice on the Mesa Recto.

Seems a lot more likely that they are saving money. I agree that the lack of NFB in the lows adds something musical but the idea that saturating the core makes the sound better doesn’t match my experiences at all. And if the only example case is on an entirely different amp that also just happened to have a small OT? I’m skeptical if only because of how many other variables are at play.

 

[Wavelength]

Big help would be the schematic for the output section. First the specs on everything I have played with for years really doesn't mean much. You could have a big output like a AC30 and yet it doesn't go over 6Khz.
If you want the basic on primary the DCR*idc (so from CT to plate on each side) should never exceed 20V. Same thing for chokes and single ended stuff. Also the two sides of should have very close DCR.
Your outputs above look like you have some xover distortion. Are the tubes individually biased? People don't realize this but if one tube is drawing 5% more than the other the output is going to suck! Tried some regulated current biasing but the problem is then you have to go Class A and Class A and PP don't really make sense if you are looking for power vs build.
I primarily do single ended amps, but I do have access to all the Peerless/Altec and many other drawings over the years.
The Hammond stuff all came from RCA Canada and some of it is really good and others oh well.
When I do a PP amp first I use a tapped inductor for a phase splitter, 2 things low DCR on the grid and you never have an unbalance.
Another thing you have to realize is everything is 1/4 the Z so a primary of 4.4K looks like 1.1K on each plate. I typically use 8K primary on 6V6 and 7591A. Never been a fan of 6L6, just me never had good luck with them. Most of the Peerless/Altec stuff for 6L6 was 6K primary.

Gordon