OT - too much top end?

[GainFreak]

Hi,

I have a power amp (2x6L6) producing too much top end. The OT is from unknown manufacturer. The FB is properly set (reversing the leads produces audible oscillation). I measured the leakage inductance of the secondaries and it looks like this OT can go above audio range (F=Z/2*Pi*LI). Below are the measurements of different taps and calculated HF response:

4 - 28mH (~23kHz)
8 - 23mH (~28kHz) - I'm using this tap
16 - 12mH (~53kHz)

OT primary impedance is 4k.

Could this extended frequency response be the reason for excessive (at least to my ears) top end or other factors may be involved?
I don't have a real scope (only PC via soundcard)) and can't look for any high frequency oscillations but measuring the output with no input signal gives me zero so I assume no oscillation at this time.
Any help would be appreciated.

[Jazz P Bass]

Personally I would look elsewhere.
Unless it is defective, the OT is only doing what it is told to do.
The circuit before the output stage is what I would question.

[GainFreak]

It's a simple power amp - one triode stage, PI, power amp.

[Jazz P Bass]

Is the first stage cathode resistor bypassed by a capacitor?
A 25uf/25v may help.
What coupling caps are being used?
.1 uf/450v may help.

[GainFreak]

Is the first stage cathode resistor bypassed by a capacitor?
A 25uf/25v may help.
What coupling caps are being used?
.1 uf/450v may help.

It's exactly like this. Maybe the problem is I'm running a high gain preamp into V30 loaded cab but my other similar amp isn't so bright.

[Chuck H]

Try swapping the OT from your "similar" amp. That'll tell you if it's the OT. Which, BTW, is not a problem. Good HF response is a ggod thing for an OT to have for some applications. Just not so much for guitar amps. For all I know most OT's can exhibit similar performance when not saturated so it may be that the OT has poor LF performance so the balance of frequencies sounds bad to you. Does the OT seem excessively large for the application? If yes then maybe your similar amp isn't so bright because there is some saturation of the OT rolling off HF where this problematic one doesn't. So changing the OT may be one solution. You could also try a shunt filter across the OT primary.

[greekie]

One thing I found super helpful is to place a high power 10K resistor in series with a 1nF or 2n2 cap able to handle 1000V, across the OT leads going to the plates of the output tubes. I use a 17W resistor because that's what I have most of - you can calculate the value you need by looking at the primary impedance of the transformer.

Jake

[Chuck H]

That's the shunt I was talking about. IIRC the resistor value should be something like 1.3X the primary Z. So if you have, say, a primary Z of 6600 a close resistor value would be 8.2k or 10k would probably be close enough. power should be at least 1.5 the amps rated power. Use this resistor in series with a capacitor shunted across the OT primary leads. I once burned up a 600V cap with this circuit with a Vp of 355. You need at least 2X the Vp and that doesn't account for spikes. Dr.Z amps are known to blow their "conjunctive filters" (same circuit) and I would think the part ratings are well considered in those amps. So I just use 2kV caps now and haven't had a problem since on about six amps built with this circuit. The cap value can be figured for it's knee but I would go by ear. The larger values use by most in this kind of circuit sound to me like a blanket in front of the cabinet. I used this circuit to flatten an overshoot spike visible on my scope under heavy power tube OD. In my case a 1500pf cap did the trick. That's comparably tiny for this circuit and even that small value was barely audible. Perhaps the difference I hear is just refinement of the square wave, which it did perfectly. So I would start with small values and work up until I found what sounded best.

BTW, all this with the consideration that there isn't some other problem with the amp causing the excessive top end. If the excessive HF seems kind of pitchy and spikey it could be a parasitic oscillation. If it just seems like the amp has a more generally high HF content then the shunt filter will probably help.

[GainFreak]

Try swapping the OT from your "similar" amp.

It's similar only up to the PI but I have another 50W OT, I can try that one.

...may be that the OT has poor LF performance so the balance of frequencies sounds bad to you.

It has plenty of LF. Size is EI96x36mm which is usual for a 50W OT. Primary inductance is 9H.

You could also try a shunt filter across the OT primary.

I was thinking about that but wanted to eliminate other possibilities.
Thanks for the detailed explanation how to calculate such a filter.

If it just seems like the amp has a more generally high HF content then the shunt filter will probably help.

It looks like this is the case and maybe a bit of "sand" on the top end otherwise no other abnormalities/artifacts. Unfortunately those V30s need at least 30-40W to produce "that" sound. At low volumes they just don't sound right.

[greekie]

Great, thanks Chuck! I've read alot about this and implemented it in a couple of amps. I actually used a 2kV cap, as that was what I had in stock! Phew, lucky me, huh?

Is it the Deluxe-like amps you do that you install this in? Do you have some tips? All advice is much appreciated!

Jake

[Chuck H]

I use the circuit in one of my designs. I've built two prototypes for Dean Markley for NAMM 2009 and three for personal customers and then I have one. The prototypes worked fine through the show with 1kV caps but I changed them out for 2kV when the 600V cap in my own amp shorted. So I've just stuck with the 2kV rating. In truth I think I'm using a 15W resistor and these amps are putting out more like 20+ watts overdriven and I've had no failures. I mean, how many watts could a 2Xel84 guitar amp possibly be producing above 7kHz. So the 1.5X amp power for the resistor was just me playing safe.

I used the shunt filter as a brute force method of taming a modest overshoot spike I was seeing on my scope. On this particular amp I also use the Paul Ruby mod and a zener across the cathode R for the power tubes. The PR mod reduces grid conduction and the zener across the cathode R fixes the bias right at the point of clipping. This amp was all about getting really refined power tube clipping from a pair of inexpensive tubes (el84's). It does this waaaay better than any "18 watt" out there. Some guys will argue that periphral circuits are aesthetically offensive and want only the most base schematic needed to make watts. But really base amps, IMHE, means that tubes, speakers, bias, OT's and even wall voltage needs to remain constant for the amp to sound "right". Take the Trainwreck Express for example. Incredible sounding, BUT, get even one thing wrong and that amp becomes unruly and bad sounding. These peripheral circuits I'm using allow for similar gain and more consistent performance with a wider range of tube, speaker, OT and voltage conditions. So... My experience with the circuit is limited to this amp design. But don't believe the hype that the shunt filter across the OT primary is a "bandaid" circuit. DR. Z seems to do just fine using them and aren't we building amps with the intention of manipulating the wave form in the first place?

IMHE the Deluxe amps shouldn't need a shunt filter. They aren't as extreme about power tube OD as these amps I build and the 6V6 is a much more graceful tube when clipping. Also, IMHO, a good Deluxe should have good fat tone but with a little chaos mixed in. Thats it's charm. Smooth as silk but brash as a hooker at the same time. But OT saturation does seem to be part of the Deluxe sound. So if you have an OT with higher specs than typical in a Deluxe type amp I think a shunt filter could be useful.

[Enzo]

How about what happens when you use a preamp in front of it for guitar play?

Powr amps tend to be full range, while guitar amps roll off in the 3k-5k region. Speaker too.

[greekie]

Exciting read, thanks Chuck! I too have no problems with doing these kinds of manipulations. Some of the tricks I somehow feel kills the "tone" or "life" of the amp in some way, like the snubber caps if the values are too extreme. But I really like this filter, as well as using the NFB (if the amp has this) to modify the amplifier's response. However, overdoing any of this is a bad idea anyways.

[GainFreak]

I tried a 8k2/2n2 shunt filter at low volumes and it took care of the problem but I don't like the idea of a 100W resistor heating up in my amp. I tried also a 470p cap across PI plates and it had almost the same effect so maybe I'll stick with it as a more simple solution. Have to try the other OT yet.

[Steve Conner]

Have you tried playing with the amount of NFB? Too much can cause a big peak in the HF response. The cheaper the OT the worse it will be, until the peak becomes infinite and the amp oscillates. From your measurements, yours is pretty bad.

Too little NFB also boosts the upper midrange and treble. The less NFB the higher the output impedance, and the less of an effect voice coil inductance has. This is quite noticeable if you're used to SS. But, it's usually considered to improve the tone.

So I suggest making the NFB adjustable and see if tweaking it improves things.