OK, folks. We're in the home stretch here, but still some problems to solve.
First, I added a 580pF cap across the input transformer secondary to kill some ringing from the input transformer. I found that value by using a capacitor substitution box and a 100K pot and twisting / selecting until the front-edge of the square wave flattened out. It was flattest when I dropped the resistance to zero, so I left the resistor out entirely.
Maybe it's a mistake to exclude the resistor?
Whatever happened, it resolved a lot of the high-frequency hash I was experiencing, but I'm still hearing some sibilance and harshness in the higher frequencies. For example, I'm listening to 'Birdland' (Patti Smith - Horses) as a test (not only is it one of my favorites by her, but she sprays 'S's around all over the place so it's a great sibilance test) and sometimes her S's tweet and leap out of the speaker.
Once you start hearing sibilance you can't stop, so it's possible that I'm mentally over-emphasizing their presence but I feel like the high-end is harsh. That could also be the result of single-ended ears tuning in to push-pull highs but I don't think so. Maybe there's still some oscillation to attend to?
I'd really appreciate advice on this one. It's driving me out of my mind.
Kofi
First, I added a 580pF cap across the input transformer secondary to kill some ringing from the input transformer. I found that value by using a capacitor substitution box and a 100K pot and twisting / selecting until the front-edge of the square wave flattened out. It was flattest when I dropped the resistance to zero, so I left the resistor out entirely.
Maybe it's a mistake to exclude the resistor?
Whatever happened, it resolved a lot of the high-frequency hash I was experiencing, but I'm still hearing some sibilance and harshness in the higher frequencies. For example, I'm listening to 'Birdland' (Patti Smith - Horses) as a test (not only is it one of my favorites by her, but she sprays 'S's around all over the place so it's a great sibilance test) and sometimes her S's tweet and leap out of the speaker.
Once you start hearing sibilance you can't stop, so it's possible that I'm mentally over-emphasizing their presence but I feel like the high-end is harsh. That could also be the result of single-ended ears tuning in to push-pull highs but I don't think so. Maybe there's still some oscillation to attend to?
I'd really appreciate advice on this one. It's driving me out of my mind.
Kofi
Thanks for responding. I get it. So the secondary of the transformer may be providing enough resistance.
Thanks!
Anyone else have input on the harshness / sibilance in the upper frequencies?
Kofi
OK-- so I changed out every variable today. I first listened to records that I found to have excessive sibilance on my single-ended 300B amp and noted that the sibilance was almost gone. I changed speakers, turntables, room positions and everything else I can think of and I'm still experiencing the effect with the Uniamps.
It could still be the speakers, of course, but I'm starting to think a filter would help. It could also be oscillation, but I need to figure out where it's coming from.
I've read through a number of posts and it seems like I could apply some resistance across the speaker terminal or perhaps come up with a zobel to attenuate the offending frequencies.
For the record, the highs sometime spit and other time whistle at me, so there's a lot of energy at what I'm assuming is the 5K to 8K frequency band.
Any DIYers have experience with a filter or another solution that might help?
Kofi
It could still be the speakers, of course, but I'm starting to think a filter would help. It could also be oscillation, but I need to figure out where it's coming from.
I've read through a number of posts and it seems like I could apply some resistance across the speaker terminal or perhaps come up with a zobel to attenuate the offending frequencies.
For the record, the highs sometime spit and other time whistle at me, so there's a lot of energy at what I'm assuming is the 5K to 8K frequency band.
Any DIYers have experience with a filter or another solution that might help?
Kofi
I've noticed sometimes when speakers show up sibilances the effect is reduced with certain programme material but there is always a problem that can be hard to identify. The issue could be brought on by a dip in the response below the peak, which could be the result of a tweeter crossover.
This in turn could be because of cancellations or it could be due to a mis-match in directivity of the drivers in the box. If you reduce the level of the tweeter it may show this problem, it may also make the speaker sound lacking in the treble when the correction needs to be more specific.
An equaliser is a handy tool for finding problems like this.
An amp can sound thin when the supply doesn't sink lower frequency currents well (small capacitance).
You could experiment with the listening angle of the speakers, or turning them away from the walls.
Thanks for responding. Yeah. I just added some resistance across the speaker terminals (like, 3 minutes ago) and it sounded pretty dull and STILL had the sibilance. In contrast, I changed from a vinyl version of a record to the digital version and found that the sibilance disappeared.
Sounds like it's somewhere in the analog chain. I just put an Aikido Tetra phono pre together, so it looks like I'll be testing that to see if I've run afoul of the RIAA curve.
Only have an oscilloscope and an inverse RIAA filter, but I should be able to use them together to do some basic voltage tests, right?
I guess it could be the table setup, but bouncing between two tables and the sibilance is there on both.
Thanks again for your input. Feels good to have some guidance.
Kofi
Last edited:
Here's Broskie's text on the matter:
The Tetra phono stage’s passive equalization network holds an optional high frequency correction. Correction? The passive network assumes that the LP’s high frequencies continue to climb at 6dB per octave from 2,122Hz to infinity. But do they?
Not likely, as the cutting heads and cutting power amplifiers used to make record master have high frequency roll offs that fall closer to 30kHz to 60kHz than to infinity. In other words, the passive equalization will over attenuate the high frequencies beyond the record’s intrinsic high frequency corner frequency.
Adding the optional resistor, R9, will place a countervailing stop to this excessive attenuation. Is it really needed? It depends who you talk to; I have tried it both ways, with the extra resistor and without, and I preferred leaving the resistor out.
But, by all means, do feel free to experiment; start with 100 ohms, which will impose a frequency transition at 46.4kHz. Frequency = 159155/R/C, where C is in μF.
The Tetra phono stage’s passive equalization network holds an optional high frequency correction. Correction? The passive network assumes that the LP’s high frequencies continue to climb at 6dB per octave from 2,122Hz to infinity. But do they?
Not likely, as the cutting heads and cutting power amplifiers used to make record master have high frequency roll offs that fall closer to 30kHz to 60kHz than to infinity. In other words, the passive equalization will over attenuate the high frequencies beyond the record’s intrinsic high frequency corner frequency.
Adding the optional resistor, R9, will place a countervailing stop to this excessive attenuation. Is it really needed? It depends who you talk to; I have tried it both ways, with the extra resistor and without, and I preferred leaving the resistor out.
But, by all means, do feel free to experiment; start with 100 ohms, which will impose a frequency transition at 46.4kHz. Frequency = 159155/R/C, where C is in μF.
OK-- I think I get what you mean. So, I've to a Denon DL-103R (moving coil) and I have a 100R resistor in parallel with the output. I've changed the parallel load quite a bit to see if that makes a difference with the sibilance and no dice.
Is this what you're referring to?
It's been a while since I've done it so I'm not sure how much I remember. I also used to use MM cartidges so maybe this doesn't apply in the same way.
What I found is that loading with the standard 47k can leave a peak. When a capacitor was also used (adjusting for cable capacitance) the response I ended up with that worked was a couple of dB down at the point where the peak used to be, and the rolloff was then coming from a slightly lower frequency than the cartidge would naturally roll off.
If you use a MC step up transformer, is it possible this will also apply to that transformer?
What I found is that loading with the standard 47k can leave a peak. When a capacitor was also used (adjusting for cable capacitance) the response I ended up with that worked was a couple of dB down at the point where the peak used to be, and the rolloff was then coming from a slightly lower frequency than the cartidge would naturally roll off.
If you use a MC step up transformer, is it possible this will also apply to that transformer?
Yes. I have both a transformer and JFET step up and I have applied a resistive load to both. I wonder if messing with parallel capacitance could change the peak frequency. Looks like additional capacitance would move the peak up in the frequency range?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.