1.5W top boost-style amp

[oxbow_lake]

Hello! First time posting here, so forgive me if I missed any forum etiquette. I could really use some help troubleshooting an amp I recently designed and built based on the AC30 top boost preamp with a push-pull output section using a 6BN11 compactron dual pentode. Each pentode is rated for 3.1 watts maximum output, and I was aiming for about 1.5 W average output power. I also threw a generic tremolo circuit in which varies the grid bias of the output tubes. The amp appears to work, and sounds pretty good, but the output is very quiet. I have a couple of (homebrew) ~5 watt single-ended amps, and at full volume I would say the vox-style amp is not quite half as loud as the 5-watt amps turned a quarter up.

Scoping a test signal at each stage is unremarkable - the gain and waveforms are what I expect going all the way to the pentode control grids. The DC bias points seem to be on target. The volume is the same regardless of what speaker I plug into - either a Warehouse ET65 or a Celestion greenback.

Using the 1 ohm cathode resistors as a surrogate for measuring cathode current, I get 14 mA in each cathode. By my reckoning, should should give me a quiescent plate dissipation of ~2.5 watts for each pentode, and is bang-on for my load line diagram. Measuring the AC voltage across those resistors I get 6mV (corresponding to 6mA RMS), which based on my load line should translate to a plate voltage swing of at least 100V? I don't have a great way of measuring the actual AC current/voltage going to the OT (without blowing up my scope), but when I stuck a big capacitor between one primary lead and my scope probe, I got a peak-to-peak signal of about 4V. Measuring the AC voltage on the secondaries I end up with a signal on the order of millivolts. These seem far too small.

Something is not adding up here. I have the output transformer wired to an 8R speaker, which should give a primary impedance of 22.5k. So my questions are:
1) Is 1.5 watts just way quieter than I expected?
2) Why am I generating such low voltages on the OT primaries? Is my output transformer shot? How do I test that other than buying a new one and seeing if it works? I really don't want to spend another $50 if I don't have to.
3) Is there some other mistake in my design or implementation that is crippling the power output?

I really appreciate any help or insights anyone can offer. Attached are the schematic (with measured DC voltages in blue) and my load line diagram (with colored pencil!). I changed the output cathode resistor to 36R instead of 56, but otherwise it's more-or-less accurate.

Thanks again,
~o

[bepone]

bypass R15 34R with 47uF 25V and check again,
btw what is the coltage on R15?

[oxbow_lake]

Thanks for the suggestion! I only had a 100uF cap available, but bypassing R15 with it made no difference in volume. The voltage over R15 is 0.98V.

[oxbow_lake]

To clarify, that 0.98V on R15 is what I measured just now, and I get the same voltage (plus ~14 mV) at the each cathode as well. When I labeled the schematic voltages a few days ago I was reading just over 0.8V. Not sure if that drift is significant.

[bepone]

max AC voltage (peak to peak) to output tube g1?
do you have a scope?
also double check output transformer connections just in case

[oxbow_lake]

Maximum AC signal to G1 is 5.1V. The input signal starts to clip at around 0.9V p-p.

I was able to use the scope to measure the voltage at the OT primary. With the 0.9V p-p input signal going to G1, the OT signal is 6V p-p.

Transformer connections look right. I also tried connecting a different pair of secondaries to try a lower impedance - this did yield slightly higher volume but not much.

[martin manning]

Your op point and load line calc's look reasonable.
I would try disconnecting the OT and verifying its impedance ratio with some low voltage AC.

[Stevem]

If you scope the output of the amp with it just idling do you see anything other then 60 and 120 hz noise?

[bepone]

6Vpp between anodes?
from explanation, i think secondary and primary on the OT are swapped, ie tubes are working in almost short circuit between anodes

[oxbow_lake]

Thank you again for all of your suggestions. Here are some answers to your questions and some data:

For the Hammond 125B, the primary is colored blue and brown on either end and the CT is red. Per the manual, to get a primary impedance of 22.5k using an 8 ohm speaker, you use the yellow and orange secondary connections.

So, I measured the OT, with the power tube removed, the primary CT unconnected, and with no speaker attached to the secondaries. One side of the secondary is grounded (yellow) and one side of the primary grounded (brown). I fed a 1V sine wave into the un-grounded lead of the primary (blue). The primary voltage peak-to-peak (blue lead) is 990mV, and the secondary voltage (orange) is 25mV. So I get an impedance ratio of (990/25)^2=1500. So for an 8 Ohm speaker this yields a primary impedance of 12.5k, which is about half of what the datasheet for the 125B says. Does my procedure/math seem right?

The 6Vpp OT voltage is what I see with the amp turned on, a ~1V AC signal at the pentode grids, when probing the blue primary lead (which is attached to the upper pentode's plate) with the scope probe ground clipped attached to the chassis. It's that 6V AC signal superimposed on ~280V DC voltage.

I see your point about the transformer being backwards/shorted, but the nature of the Hammond 125B wiring is that it's almost impossible to wire it backwards since the number of leads on each side is totally different and the color coding is very clear! I did double check, even so.

When I scope the output of the amp just idling I just see a ~3Vpp 120Hz noise. Which is maybe a little weird because I don't hear a lot of hum through the speaker, but i do hear the test tone whose amplitude is only 6V p-p? I've attached images of the scope traces idling and with the 1 kHz test signal (going into the input jack). Sorry my finger apparently got in the way of the second photo a little.

[martin manning]

So, I measured the OT, with the power tube removed, the primary CT unconnected, and with no speaker attached to the secondaries. One side of the secondary is grounded (yellow) and one side of the primary grounded (brown). I fed a 1V sine wave into the un-grounded lead of the primary (blue). The primary voltage peak-to-peak (blue lead) is 990mV, and the secondary voltage (orange) is 25mV. So I get an impedance ratio of (990/25)^2=1500. So for an 8 Ohm speaker this yields a primary impedance of 12.5k, which is about half of what the datasheet for the 125B says. Does my procedure/math seem right?

The data sheet shows 16.7k using Black and Green and a 7.58 load. You should be using the Black and Green leads to get 17k6 primary for an 8 ohm load.
From the impedance table I can get 11.5k for an 8 ohm load Yellow to Orange, which I would say confirms your measurement.
https://www.hammfg.com/files/parts/pdf/125B.pdf

[oxbow_lake]

Ok, I might not be understanding the datasheet here. Using the "Hookup Data" table at the bottom, since I'm using an 8 ohm speaker, won't I get 22500 ohms on the primary (from blue to brown) using secondary leads 2 and 4, (orange and yellow)? And then the impedance that one tube would see in a push-pull configuration would be half of that running in class AB...? That's what I drew the load lines for, anyway.

[martin manning]

Using the "Hookup Data" table at the bottom, since I'm using an 8 ohm speaker, won't I get 22500 ohms on the primary (from blue to brown) using secondary leads 2 and 4, (orange and yellow)?

Yes. I used the inductance table, which failed for some reason. Your measurement should have been much closer to 22k5, using 8 x (V(Blue-Brown) / V(Orange-Yellow))^2

And then the impedance that one tube would see in a push-pull configuration would be half of that running in class AB...?

No, each tube sees 1/4 of that in Class B, when the opposite side is in cut-off.

[oxbow_lake]

Right, sorry, I meant class A. Thanks!

[martin manning]

BTW, what did you do with the suppressor grids? It doesn't appear that there is any internal connection to the cathodes.