Hello everyone,
The attached schematic shows the SE amplifier I have built. Sorry if I drew it in a hurry.
Since the circuit is very simple, I used good quality parts (Cerafine and Audyn Cap Plus caps, Kiwame Resistors, a GE JAN black plates 6SN7, Golden Dragon 2A3 tubes, etc), but I bought very cheap OPTs for a first run.
I did some measurements and while the results in the bass region are very satisfactory on paper and in listening tests, the treble are quite heavily rolled-off. My measurements (I did them with the speaker connected to the amp, not a simple resistive load) indicate a -3dB point in the 8-9KHz region, that is very low. It is not a dip, since the response doesn't rise again, going higher in frequency. I think this depends on the output trannies.
Since I am on a budget, I was thinking about upgrading them with James 6112HS transformers. Will these trannies help me getting a more linear response or is it not worth it? I have read praising reviews about them.
Feel free to suggest any possible advice about my schematic.
Thank you!
The attached schematic shows the SE amplifier I have built. Sorry if I drew it in a hurry.
Since the circuit is very simple, I used good quality parts (Cerafine and Audyn Cap Plus caps, Kiwame Resistors, a GE JAN black plates 6SN7, Golden Dragon 2A3 tubes, etc), but I bought very cheap OPTs for a first run.
I did some measurements and while the results in the bass region are very satisfactory on paper and in listening tests, the treble are quite heavily rolled-off. My measurements (I did them with the speaker connected to the amp, not a simple resistive load) indicate a -3dB point in the 8-9KHz region, that is very low. It is not a dip, since the response doesn't rise again, going higher in frequency. I think this depends on the output trannies.
Since I am on a budget, I was thinking about upgrading them with James 6112HS transformers. Will these trannies help me getting a more linear response or is it not worth it? I have read praising reviews about them.
Feel free to suggest any possible advice about my schematic.
Thank you!
Attachments
more bass
LED matrix in 2A3 cathodes
It may be of your interest - I once built SE amp with Russian 6C4C (almost equiv. to 6B4G). It was used for a couple of years. After I had built SY's Red Light District amp, I decided to do the same trick with the former amp - I replaced RC network in output triode with red LED matrices. The improvement in bass was astonishing - it became a new amp. If you are interested - let me know.
Total workload, calculation + soldering, was 3 hrs.
It may be of your interest - I once built SE amp with Russian 6C4C (almost equiv. to 6B4G). It was used for a couple of years. After I had built SY's Red Light District amp, I decided to do the same trick with the former amp - I replaced RC network in output triode with red LED matrices. The improvement in bass was astonishing - it became a new amp. If you are interested - let me know.
Total workload, calculation + soldering, was 3 hrs.
I'm no expert, but I don't see anything obviously wrong with your circuit. I'm a little surprised the bandwidth of the amplifier is poor. I would hope that even the most inexperienced transformer winder could have supplied something that would perform a little better than what you are seeing.
Has your winder made output transformers before? Did he make any attempt to interleave the primary and secondary windings? It's not a "split bobbin" type, is it? (primary wound on the lower half of the bobbin, completely separate from the secondary which is wound on the upper half).
Has your winder made output transformers before? Did he make any attempt to interleave the primary and secondary windings? It's not a "split bobbin" type, is it? (primary wound on the lower half of the bobbin, completely separate from the secondary which is wound on the upper half).
The circuit itself is performing very well indeed. I have the 6SN7 halves correctly biased at 6mA, the 2A3s at 60mA, 50V cathode bias (250V A-K),
~2mV 50Hz ripple at the speaker.
Waveforms are correct, the output power is 3W before clipping in the center of the band, transformer saturation effects take place only under 70 Hz, from 1.5/2W upwards.
Excluding this roll-off, it's very lush sounding, vocals and guitars are beautiful.
So I would think I have to blame the OPTs for that.
Making transformers is their main activity. I don't know if it's a split bobbin type. I accepted the fact they were cheap OPT, and I was prepared for somewhat weak bass or not very sparkling highs. In the end, bass is quite satisfactory, but the high-end roll-off is definitely unacceptable.
Last edited:
Do you have scope and some sort of signal generator?
I would measure the output voltage from 20Hz - 20kHz at the plate of both the 6SN7 and the 2A3 before buying new OPTs. Check that there is not appreciable audio across each bypass cap at higher frequencies.
What sort of cathode bypass cap are you using on the 2A3?
Not likely to account for all of your FR problems, but I have run into problems with some motor run types that were somewhat inductive, paralleling them with a smaller film cap helped a lot.
For some hum cancelling if needed you could take a 22uF cap directly from the cathode of the 2A3 to the B+ side of the OPT. This will also help the HF response provided that the cap used is a reasonably good one. Leave the 100uF cap in place. (Western Electric type connection)
It probably is the output transformers, but it would be best to make absolutely sure that something else isn't amiss before spending your limited funds.
I would measure the output voltage from 20Hz - 20kHz at the plate of both the 6SN7 and the 2A3 before buying new OPTs. Check that there is not appreciable audio across each bypass cap at higher frequencies.
What sort of cathode bypass cap are you using on the 2A3?
Not likely to account for all of your FR problems, but I have run into problems with some motor run types that were somewhat inductive, paralleling them with a smaller film cap helped a lot.
For some hum cancelling if needed you could take a 22uF cap directly from the cathode of the 2A3 to the B+ side of the OPT. This will also help the HF response provided that the cap used is a reasonably good one. Leave the 100uF cap in place. (Western Electric type connection)
It probably is the output transformers, but it would be best to make absolutely sure that something else isn't amiss before spending your limited funds.
I had the same problem with a 300B amplifier. If the signal at the grids of the 2A3 is strong at 20kHz, the load impedance of the OPT is correct for the tube, and the tube is biased strong enough, then it is almost certainly the OPT.
At the end of the day, it was the OPT. Swapping out with a Lundahl solved it.
At the end of the day, it was the OPT. Swapping out with a Lundahl solved it.
I'd have to agree. Checking that there is no audio across the cathode bias network is the only other thing you should do before ordering better transformers. I am a fan of the cerafine and would expect its performance to be quite good even at 20kHz - so cap performance is probably ruled out.
I would save your money for a bit and buy an even better transformer. I would also talk to Bartolucci and see what they have available at a reasonable price - imho something comparably priced from them almost has to be better than the comparably priced James with the shipping costs factored in. And they make very good transformers.
I can confirm there is no audio across cathode bypass caps, just some millivolts at very low (100 Hz or lower) frequency, as the reactance of the cap increases.
I also measured the response directly at the grids of the 2A3s. Absolutely flat up to 20 KHz.
I understand James transformers are nothing fancy. But I'm fine if they are free of such major frequency limitations, until I can afford very good OPTs. I would do a switch now, because as it is now, the amplifier is too much rolled-off and I really can't enjoy it at all.
I also measured the response directly at the grids of the 2A3s. Absolutely flat up to 20 KHz.
I understand James transformers are nothing fancy. But I'm fine if they are free of such major frequency limitations, until I can afford very good OPTs. I would do a switch now, because as it is now, the amplifier is too much rolled-off and I really can't enjoy it at all.
I did some more measurements.
I used a 8 ohm resistive load at the output instead of my loudspeakers (B&W DM602). The roll-off at 20KHz is only 0.7dB under the response in the pass-band.
How is this possible?
Is it normal with DHT SE amps to expect such a heavy HF roll-off with the speakers connected (-3dB @ 8-9 KHz)?
Or are my OPTs not able to drive my speakers correctly?
I used a 8 ohm resistive load at the output instead of my loudspeakers (B&W DM602). The roll-off at 20KHz is only 0.7dB under the response in the pass-band.
How is this possible?
Is it normal with DHT SE amps to expect such a heavy HF roll-off with the speakers connected (-3dB @ 8-9 KHz)?
Or are my OPTs not able to drive my speakers correctly?
Neither actually, it is an indication that your speaker system has a strongly falling impedance at high frequencies and hence is not a great load for a small SE amplifier. Replacing the transformers actually won't fix this issue - you need to determine what the issue is with the speakers. (Probably impedance flatness with frequency) Do the transformers have a lower impedance tap you could try with these speakers as that may help at the expense of rather less output power.
Complex cross-overs are a problem with small SE amps as there is very limited current drive capability. Same can generally be said of low impedance speaker systems as well. Note that there are always exceptions and the above are generalizations.
I suspect the B&W speaker is not a good match to your current amplifier preference - finding an impedance plot over frequency should confirm this. Ideally you either want a large efficient 2 way speaker system with simple x-o, preferably first order or an efficient FR single driver system.
Complex cross-overs are a problem with small SE amps as there is very limited current drive capability. Same can generally be said of low impedance speaker systems as well. Note that there are always exceptions and the above are generalizations.
I suspect the B&W speaker is not a good match to your current amplifier preference - finding an impedance plot over frequency should confirm this. Ideally you either want a large efficient 2 way speaker system with simple x-o, preferably first order or an efficient FR single driver system.
Have you ever generated any impedance plots for your speakers? It's fairly common for them to have significant peaks, especially around port resonances and crossover frequencies. I do not recall seeing any feedback in your amplifier design, so high impedance peaks in the load would likely result in measurable changes in the power delivery.
http://i69.photobucket.com/albums/i43/Ty_Bower/Chorus II/SpeakerTest/ChorusIIleftPR.gif
http://i69.photobucket.com/albums/i43/Ty_Bower/Chorus II/SpeakerTest/ChorusIIleftPR.gif
The minimum impedance of the DM602 is listed as 4.3 ohms with a nominal impedance of 8 ohms, under FTC parlance here in the USA I believe this would/should probably be considered a 6 ohm speaker.
If you have a 4 ohm tap I would give this a try, otherwise replacing the OPTs with ones that do offer a 4 ohm tap would probably help somewhat. This reduces the effective source impedance of the amplifier to the speaker, and reflects a higher impedance back to the plate of the 2A3. Without knowing the impedance curve of the speaker system it is hard to know how distortion and power output will be effected - one or both may be improved or not.
The efficiency is on the low side for a small SE as well.
You might be as well off getting your hands on some inexpensive FR drivers and building something to go specifically with this amplifier.
Manufacturer Specs:
Frequency Response: 52Hz - 20kHz ± 3dB on reference axis
Frequency Range: -6dB at 43Hz and 30kHz
Sensitivity: 90dB spl (2.83V 1m)
Nominal Impedance: 8 ohms (minimum 4.3 ohms)
Power Handling: 25 - 120W into 8 ohms on unclipped program
From here: B&W DM602 Series 2 Loudspeakers
If you have a 4 ohm tap I would give this a try, otherwise replacing the OPTs with ones that do offer a 4 ohm tap would probably help somewhat. This reduces the effective source impedance of the amplifier to the speaker, and reflects a higher impedance back to the plate of the 2A3. Without knowing the impedance curve of the speaker system it is hard to know how distortion and power output will be effected - one or both may be improved or not.
The efficiency is on the low side for a small SE as well.
You might be as well off getting your hands on some inexpensive FR drivers and building something to go specifically with this amplifier.
Manufacturer Specs:
Frequency Response: 52Hz - 20kHz ± 3dB on reference axis
Frequency Range: -6dB at 43Hz and 30kHz
Sensitivity: 90dB spl (2.83V 1m)
Nominal Impedance: 8 ohms (minimum 4.3 ohms)
Power Handling: 25 - 120W into 8 ohms on unclipped program
From here: B&W DM602 Series 2 Loudspeakers
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.