What are the downsides to a Split-Bobbin construction for use in tube amp plate power supply?

I've been trying to find a definitive answer to the question but I can't seem to find anything directly relating to my intended use. The intended us is... .

  • I have a stereo, dual-mono constructed, 2-stage push-pull 2A3 power amp that requires a separate power transformer for its driver stage.
  • Each channel of push-pull 2A3s uses its own 500VCT 150mA power transformer. Those are on the chassis, ready for action.
  • Each channel will use a 6N6P LTP with its plates DC-coupled to the 2A3 grids. Each 6N6P triode will be drawing about 15mA, so I'll need 60mA for four triode.
  • I need a +150VDC plate supply for the 6N6P tubes which will be 'stacked' on top of the +325VDC plate supply for the 2A3s.
  • I already have the 6.3VCT 1.2A transformers necessary for the 6N6P heaters.

I don't have a lot of room on the chassis. It's an old project that I didn't lay out very well. Oh well, it is what it is.

I was thinking of using a pair of Triad Magnetics isolation transformers (117VAC:117VAC) and rectifying the 117V secondary with CRC filtering to bring the rectified voltage to +150VDC.

My choices are:

The VPS230-110 is a bit smaller, so I'd like to use that if possible. But it uses split bobbin construction. I understand that has the lowest capacitive coupling of the various transformer configurations, but does that come at the cost of a lot of uncontrolled electromagnetic radiation? I don't want to inject hummmmm into everything.

The N-48X is an unknown quantity, but I know Eli Duttman has used that series of transformers in some of his amp designs. Do single-bobbin EI core transformers control their electromagnetic radiation better than split bobbin transformers?

I understand toroid transformers have the tightest control over their radiation, but they also couple line noise to the secondary more readily, and they also have a bigger current surge at power on. I could go with a toroid if it's absolutely necessary, but they're all pretty large. The minimum diameter in a transformer with 120V secondary from Antek is 3,6" for a donut rated at 50VA (120V at 400mA). That's way more than I need.

What's better in this particular usage situation? Split bobbin? Or EI core? Or does it not matter?
Last edited:
I always put a simple DC blocking circuit on the incoming AC of the toroid. It’s just a bridge rectifier and two caps. There’s lots of info on line. It cures some of the problems related to wall AC having a DC component that toroids don’t like. That plus a RF emi filter and you’re set.
I'm not sure if I was clear in the OP. This is for the B+ supply to the 6N6P LTP driver stage. It's not for an output transformer.

The question is, are there any great disadvantages to using a split-bobbin construction power transformer for the B+ to a line level audio stage?

You see lots of tube amplifiers using EI core (single bobbin) power transformers, and lots using toroids. The high end ones even use C-cores and R-cores. But I never see split-bobbin power transformers being used, and I'm wondering why that is.
The main disadvantage of split-bobbin is the higher leakage inductance. Can't see how that's a problem here except with greater leakage inductance there would normally be more 'leaked out' flux. The lower pri-sec capacitance though would be an advantage in reducing mains-borne noise.
Thanks for the responses.

Nice to know there's no huge downside to a split-bobbin transformer in this use.

I went 'round and 'round with this question and couldn't find anything conclusive. I even checked in RDH4. So I went ahead and made a decision based on physical practicalities. The N-48X has wire leads while the VPS230-110 has 'quick-connects'. The wire leads won me over, since that makes it possible to mount them on top of the chassis, even though they're ugly. I ordered four of the N-48X (they're cheap) and a few other odds and ends.
The spilt-bobbin issues can be fixed if required: with a single, main secondary, the leakage inductance can be addressed by inserting a compensating capacitor in series with the rectifier bridge (normally a 1 to 10µ cap is suitable, to be determined by a measurement + calculation), and the stray field caused by heavy secondary loading can be minimized by a Gauss band.
However, if capacitance and size are not a problem, simply opt for single bobbin
I'm not sure if I was clear in the OP. This is for the B+ supply to the 6N6P LTP driver stage. It's not for an output transformer.

The question is, are there any great disadvantages to using a split-bobbin construction power transformer for the B+ to a line level audio stage?

For audio applications , IMO the key is that split bobbin transformers have low capacitive coupling between the primaries and secondaries .

The Hammond 229 series may be of interest - which is effectively an r core transformer with dual secondaries.
But ....... these require mounting on a circuit board , and the 229B88 wired in series would only give a plate voltage of 124 Vdc .

But for 6n6p a better option looks like a Hammond 185D230 43 VA . Using one secondary would give a plate voltage of 162 Vdc .
I've bought a number of these , and not one of them has hum issues .

Then in the power supply , the voltage could be dropped a bit with an CRCRCRC filter.
Have a look at the schematic of the famous Marantz model 7 - you'll see the power supply has quite the RC ladder .

Have you seen this load line calculator ?

For 6n6p with a plate voltage of 155 Vdc a good bias point looks to be 8.0 mA

Also , if you have the real estate , consider connecting a DIY EMI filter called a fo felix to the plate voltage power supply only .
I can't measure any difference , but I can sure hear it .

Last edited:
The only problem with split bobbin transformers is the poor regulation. If a regular concentric wound trafo is 10%, the same size split bobbin will be 20-25%. If you can live with that, for whatever reason, it’s actually better. Power factor is better so its easier on the rectifiers. If youre using a choke input supply, a little extra inductance in the trafo is welcome. CRC filtering can always be adjusted to what’s needed to get the desired volts under load. Just not a good choice for a 1000W sand amp where good regulation effectively buys output stage SOA.

Gotta be careful using one for a filament transformer. I’ve got a couple 72 VA 6.3 volt Triads that put out just shy of 9V without a load.
  • Like
Reactions: 1 user