The Hammond catalogue here http://www.hammondmfg.com/pdf/5c0050.pdf gives a figure for "Maximum DC Bias". Does anyone know what this means?
Looking at the suggested tube applications, it would make some sense if this was a near-to-saturation peak current and the transformer was gapped for an optimum idle current of around half this value. When it comes to the 1638SEA, though, 45mA would not be optimum for an 845 or a 211 with a 10k load.
(I have tried asking Hammond about this, but technical questions are diverted to local sales staff who have no technical knowledge about the products they sell, and who are unable to find anyone in the Company who does.)
Looking at the suggested tube applications, it would make some sense if this was a near-to-saturation peak current and the transformer was gapped for an optimum idle current of around half this value. When it comes to the 1638SEA, though, 45mA would not be optimum for an 845 or a 211 with a 10k load.
(I have tried asking Hammond about this, but technical questions are diverted to local sales staff who have no technical knowledge about the products they sell, and who are unable to find anyone in the Company who does.)
The 90mA is the maximum -static- DC current (for example 211 SE static anode current).
Simplified, this, and maximum AC current generates excitation within the core, which can't exceed the desired -gapped- value.
Call their offices in Guelph or use their "contact us form" for "technical information"...
Phone: (519) 822-2960
EDIT: I see you are not in Canada, and long distance might be expensive.
The contact form is here: https://www.hammfg.com/contact
euro21 is correct though as far as I know. The design maximum is usually higher than design centre and in this case, the idle current shouldn't be higher than 90mA... Think a 211 amp running with lower plate voltage running in class A2 vs a standard class A1 circuit.
It's also the DCR against the idle current turns into heat so you can only put so much current through it without overheating the coil.
Phone: (519) 822-2960
EDIT: I see you are not in Canada, and long distance might be expensive.
The contact form is here: https://www.hammfg.com/contact
euro21 is correct though as far as I know. The design maximum is usually higher than design centre and in this case, the idle current shouldn't be higher than 90mA... Think a 211 amp running with lower plate voltage running in class A2 vs a standard class A1 circuit.
It's also the DCR against the idle current turns into heat so you can only put so much current through it without overheating the coil.
Last edited:
Thanks for the reply, euro21.
So, does that mean the 1638SEA could have an idle current of 90mA and still swing ±50mA without saturating?
Thanks, but the problem is that without using a VPN, the form is diverted to local sales, whatever heading you use.
If you use SE with 90mA static anode current, it's capable swinging ±50mA, but the estimated power will be less than maximum.
I have no experience with the 1638SEA, but I must assume that the optimum idle current should be less than 90 mA. This would support a swing from 0 mA to 180 mA in an "ideal" situation. Of course ideal tubes and circuits don't exist, and perfect symmetry in a real SE tube amp is also a myth.
I have a pair of 1628SEA's which are speced at 120 mA. That does seem to be the best choice for extended bass response without saturation. They will do pretty well up to almost 200 mA before saturation gets real ugly. Saturation in an OPT is also dependent on the driving impedance. A pentode with minimal feedback will show saturation earlier than a triode, or pentode with heavy local feedback on the same OPT. My experiments with the 1628SEA have been done with a BIG TV sweep tube operating with lots of local feedback which brings out the best in most OPT's.
I have a pair of 1628SEA's which are speced at 120 mA. That does seem to be the best choice for extended bass response without saturation. They will do pretty well up to almost 200 mA before saturation gets real ugly. Saturation in an OPT is also dependent on the driving impedance. A pentode with minimal feedback will show saturation earlier than a triode, or pentode with heavy local feedback on the same OPT. My experiments with the 1628SEA have been done with a BIG TV sweep tube operating with lots of local feedback which brings out the best in most OPT's.
It was because the specs looked so different from other manufacturers' products, and so far from the normal current range for the suggested tube applications, that I asked the question. Using a 160mA transformer for a single 2A3 didn't make sense to me, but it does seem that this range of transformers are just very over-specified, and rather a bargain.
The maximum total current: of the DC quiescent current, plus the Peak Swing current; which can cause saturation is completely different:
at 1 kHz than it is at 30Hz
Right?
Global negative feedback tends to increase saturation when the transformer is already saturated.
Right?
at 1 kHz than it is at 30Hz
Right?
Global negative feedback tends to increase saturation when the transformer is already saturated.
Right?
I am sure someone will correct me if I am wrong, but I think the current required to saturate the core will be a constant. The impedance will vary according to frequency, so as the frequency gets lower more current will be required, and distortion will increase depending on how much current the driver can supply. Feedback will lower the effective driver impedance.
Your intuitions and assumptions sound very correct to me. To re-frame your question, a single-ended OPT operates on its B/H curve, where signal transfer is more-or-less linear from slightly up-and-right of 0,0 to a region where it starts to flatten out ("saturate"), more righter and upper. Where on this curve does the manufacturer spec their device? Output valve idling current will be half way up the curve, cannot be spec'd too low (lost sales and wasted core iron, including larger than necessary parasitic capacitances and leakage inductance - smaller core is better), cannot be spec'd too large (distortion from core saturation - larger core is better).
These actual curves could be provided for single-ended OPTs, at least for a single or a couple of frequencies (they're unfortunately also dynamic with measurement frequency, but that's a measurement issue), but are they? Then you could decide for yourself where on the curve your idling current would sit, with your own choice of how much saturation bend is acceptable. And I am Marie of Romania.
All good fortune,
Chris
These actual curves could be provided for single-ended OPTs, at least for a single or a couple of frequencies (they're unfortunately also dynamic with measurement frequency, but that's a measurement issue), but are they? Then you could decide for yourself where on the curve your idling current would sit, with your own choice of how much saturation bend is acceptable. And I am Marie of Romania.
All good fortune,
Chris
Chris,
I am working on the assumption that the figure given is the maximum idle current that will still allow enough headroom to deliver the specified power at 30Hz, so the centre point on the B-H curve will be lower than this. As you say, it would be nice to know what it is, which I was hoping Hammond could tell me.
The only manufacturer I know of which supplies comprehensive specs is Lundahl. I strongly suspect that the SEA transformers were not designed in-house, as they are very different from the original and poorly-regarded SE range.
Note that the B/H curve itself is independent of frequency, in the sense that it's generated by joining the pointy ends of all possible hysteresis curves, which are each dynamic. Iron core dynamics are so non-linear that they're almost impossible to discuss in text, even if anybody understood them, and nobody even claims to. Measurement issues are another, separate issue.
If, and that's a very big If, we had a believable B/H curve for a single-ended OPT, then the remaining idling point decision would be: how much bend, on the righter and upper ("saturation") end is too much? That's the right/upper-most end, and the usually much steeper (and smaller) lefter/lower-most 0,0 flattening burble sets the other end (if perfect, the limit would still be 0,0 for ideal valve model Class A). Idling point is half way, and Bob's yer MAGA uncle at Thanksgiving.
Everything about audio transformers is shrouded in mystery. Partly from prospective purchasers' inability to specify their requirements, and partly from the very, very high actual costs to fulfill those idealized requirements. We all expect microprocessor-type modern miracles, but ain't happenin'.
All good fortunes,
Chris
If, and that's a very big If, we had a believable B/H curve for a single-ended OPT, then the remaining idling point decision would be: how much bend, on the righter and upper ("saturation") end is too much? That's the right/upper-most end, and the usually much steeper (and smaller) lefter/lower-most 0,0 flattening burble sets the other end (if perfect, the limit would still be 0,0 for ideal valve model Class A). Idling point is half way, and Bob's yer MAGA uncle at Thanksgiving.
Everything about audio transformers is shrouded in mystery. Partly from prospective purchasers' inability to specify their requirements, and partly from the very, very high actual costs to fulfill those idealized requirements. We all expect microprocessor-type modern miracles, but ain't happenin'.
All good fortunes,
Chris