• 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.

Experience with AnTek OPTs?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Let's see, 3dB down would be .707 of +/-3 div or +/-2.12 div. Looks like the AT-0358 is hitting 2.5 div at 10 KHz and 2.0 div at 15 KHz. Interpolating, that would put the 2.12 div at 14 KHz. So 3 dB down at 14 KHz likely.

The AT-1008 looks like its hitting the 3 dB down at 20 KHz.

Did you notice any peak (increase of V above the low freq. case) at any freq. on these?

I usually think of the primary resistance from P to P, but not sure what (if) any standard is.
Obviously, the center to plate R is the figure of interest as far as comparing to the tube plate Z or reflected load Z.
 
Last edited:
It depends. Pentode tubes operated as such are best "adjusted" by way of changing the screen grid voltage, since it primarily determines how much current a pentode can draw in a given circuit. Changing the plate voltage can adjust the tube's required loading as well, but not nearly as much as an equivalent change in screen voltage will.

For triodes, there is little you can do, since the tube's parameters dictate the external load almost absolutely, with exception to the variables listed below.

With UL operation, the screen tap allows for some variation, as in general, a lower tap % requires a lower plate to plate load, while a higher % generally raises the load requirement. But since the tap is fixed within the transformer itself, which is what you are trying to optimize the circuit for, it really offers little help except to help dictate which tubes might work best based on the screen % a given transformer represents.

However, for all output configurations, the type of bias scheme will affect loading requirements as well. Fixed bias operation allows more current to flow, lowering the required load impedance, while using cathode bias requires a higher load since it places more limits on maximum possible current flow. Likewise for class of operation. Class AB1 operation does not allow for any flow of grid No.1 current, where as Class AB2 does. As a result, more plate current flows in Class AB2 designs requiring a lower load impedance, relative to Class AB1 designs where higher loads would be used.

This is obviously a very general description relating primarily to push-pull circuits, but is none the less accurate as to how different issues effect loading. In all examples, it is offered based on the idea of "all else being equal".

Dave
 
Sure.
The Rp mainly depends on the tube current, but since the max plate dissipation controls the product of the two, you can alter the OT primary Z within bounds (max cathode current, max plate voltage, max plate and screen dissipations) For pentodes, you will need to adjust the screen voltage to get the load line matched up with the "knee" of the tube curves for the chosen load Z. (for example, going to a higher primary Z and higher B+ will cause the load line to intercept the plate curve knee at a lower current, so it is advantageous to drop the screen V until the near zero Vg bias curve's knee is on the load line) If you follow Tubelab's experimenting online, you will notice that he often changes OTs along with plate and screen voltages and grid biasing to optimise power out or distortion. Usually higher voltage/higher Z (until max B+ rating anyway) will give you a little more efficiency and less distortion. But electrolytic cap voltages, and power transformer price, and OT price and performance (and safety considerations too) will put the lid on the B+ around 500 V. Some tubes, especially "sweep" tubes are more optimised for low Z outputs since they were designed to operate off the line voltage in TV sets. Some tube data sheets will also give curves for distortion (usually just THD or 2nd and 3rd H. only though) versus primary Z and power out, usually some optimum there. The best sounding Zpri and biasing, by consensus, often gets called the "sweet spot" for that tube.

edit: just noticed dcgillespie's response. UL operation does greatly complicate matters since you are constrained by screen voltage set equal to B+ unless it is some rare OT with a separate UL winding.
 
Last edited:
Oh.., don't forget damping factor too. A higher Zpri will usually give you a better damping factor for a given plate Rp. But you can change the plate Rp via plate operating current to compensate. A higher Zpri might also limit the amount of global NFDBK that can be applied to the amp. due to phase stability (OT phase shift, leakage L, shunt C). But local FDBK, like pentode Schading, should not be bothered by OT issues. (exception: cross coupled feedbacks to driver grids may be, since feedbacks then depend on signals that traverse from P to P thru the OT in class B)
 
Well, I finally took my AT-040L into work to bench test it. The baseline of my measurment equipment is flat to 120kHz.
With the 8 ohm secondary tap terminated into 8ohms and driven with a balanced source with an impedance = the primary Z (4k8) it's -3dB @ 11kHz :(

Something I find interesting is that the primary inductance measures almost 400H! It could posibly be an error with the measurement equipment, but the measurements of some other manufactures OPT's I own are close to manufacture specs.

They could probably get acceptable high frequency performance and keep a good bottom end by pulling some windings off of that doughnut. As it is now, they would make a killer bass amp OPT, but they aren't suitable for full range unless driven by a very low source impedance.
 
Administrator
Joined 2004
Paid Member
Well, I finally took my AT-040L into work to bench test it. The baseline of my measurment equipment is flat to 120kHz.
With the 8 ohm secondary tap terminated into 8ohms and driven with a balanced source with an impedance = the primary Z (4k8) it's -3dB @ 11kHz :(

Something I find interesting is that the primary inductance measures almost 400H! It could posibly be an error with the measurement equipment, but the measurements of some other manufactures OPT's I own are close to manufacture specs.

They could probably get acceptable high frequency performance and keep a good bottom end by pulling some windings off of that doughnut. As it is now, they would make a killer bass amp OPT, but they aren't suitable for full range unless driven by a very low source impedance.

Try driving the transformer with a source impedance representative of the rp (in triode - see below) of the tubes you would use to drive this transformer and you should see significantly better results. (Still will not be great, but might make >20kHz at least.) Does seem to indicate that this transformer has massive amounts of interwinding capacitance as opposed to leakage inductance so it looks like pentode mode operation would not be a good idea. Triode mode might work though.
 
Try driving the transformer with a source impedance representative of the rp (in triode - see below) of the tubes you would use to drive this transformer and you should see significantly better results.

I was initially planning on a UL 6550 amp, so Zsource = Zprimary should be in the ball park. Any requests for a desired source impedance measurement?
 
Out of curiousity I ran some more tests:
Simulated 300B drive (700R + 700R) = -3dB @ 28kHz
Simulated 6c33c drive (80R + 80R) = -3dB @ ~ 45kHz

The transformer has a resonance at 50kHz. The frequency response drops until 50kHz and then goes back up again. Changing the source impedance seems to be changing the Q. Lowering the source impedance raises the cutoff frequency, but it rolls off really steep and comes back up steep. Higher source impedance rolls off sooner, but it's more gradual and it comes back up gradual.

Looks like bass amp duty for this guy. It's a shame. The build quality is very good.
 
I have found the same resonance behaviour when I have used mains toroidals.
Typically it is much reduced by reversing the polarity of the primary side windings.

Shoog

Unfortunatly, this transformer has a center tapped primary rather than dual primaries. And it is potted, so it is not possible to unwrap and split the CT.

The primary inductance is outrageous. I have a few other 5k transformers from reputable winders and the primary inductance's measure about 1/10th that of this Antek unit. The low end response is probably outstanding, but they could likely gain some top end performance by sacraficing some of the low end (Using far fewer windings, but maintinging the same ratio).
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.