what's causing this non-flat frequency response curve for my turntable setup?

0 Members and 1 Guest are viewing this topic. Read 6126 times.

claytontstanley

Hey everyone,

Here is a plot of the frequency response curve of my turntable using Dr. Feickert's Adjust+ software to measure. I'm wondering why this curve is non-flat (peaking more at the high end, with the downward slope from 20-10k). I would figure that a properly aligned turntable with a decent cartridge & preamp could produce a relatively flat curve. Any ideas what device / configuration might be causing this?

To measure the curve, you plug the RCAs from the preamp's output into the 1/8 line input on your computer. This means that the cause has to be somewhere between the turntable source and the preamp's output (can't be main amp, speaker wires, speakers, etc.).

Gear:
-Grado Gold cartridge/stylus
-Project Xpression III turntable
-DH Labs BL-1 interconnects
-Graham Slee Reflex preamp
-Standard power supply for preamp
-Pro-ject speed box
-Gear for calibrating turntable
  -Feickert protractor
  -Feickert adjust+ software (to align azimuth)
  -Pro-ject digital force tracking gauge
  -A bubble level for VTA
  -A bubble level for the turntable

As far as I can tell, the turntable is calibrated correctly, but maybe I'm missing something in the calibration that's causing the non-flat frequency response curve, or maybe I need to upgrade a component.

Any thoughts / help?
Thanks,
-Clayton






Wayner

The graph was hard to read, but I think the bottom section states Max. Deviation, which I assume is from 0 db reference. So your 5db up from flat on the bottom, slowly coming down somewhere around 7k and then have a nice rounded spike at ? 12-15K. It's not terrible. It may not be any one particular components fault either. Some add, some subtract.

Since you are running a Grado Gold (not the Gold1), just curious what VTF you are running it at, and what do you think the capacitance of your interconnects are?

Wayner

claytontstanley

Wayner,

My mistake. I do own a Grado Gold1. VTF is 1.5g

Also, I tried to find the capacitance info for the DH Labs BL-1 interconnects, and this is the closest piece of info I could find:

"if the BL-1 is used in an unbalanced application (e.g., most home installations), the shield should float at one end. On the source end, connect both the black center conductor and the drain wire to the shield of the RCA connector. On the load end, leave the drain wire unconnected. In this configuration, the BL-1 has a capacitance of about 44pF/ft."

The length of the interconnects from the turntable to the preamp is 1M.

I guess that I'm using the interconnects in an unbalanced application, but I am not doing anything fancy with the drain wire (mentioned above; whatever that is). I'm just plugging the RCAs into the equipment.

-Clayton

claytontstanley

A quick update after messing with a few of the calibration parameters and then re-running the frequency sweep.

-Adding .5g of VTF doesn't drastically change the shape of the curve (just makes it a bit more jagged/noisy)
-Moving the azimuth +- .5 deg from center doesn't change the overall shape of the curve (does change which channel is above the other, but both channels still produce the same overall shape)

So, still looking...

martin_eg

Hello Clayton, I met a Dr.Feickert guy at the Stockholm HighEnd last weekend, and they seem to be very helpful and speek english much better than I do :)

Why not send them a mail, they are really experts in this area. IMO something is wrong, you do not want to start your setup with +- 4 dB already from the source.
Good luck/Martin

felixscerri

G'day mate, there is a lot of anecdoctal evidence that increasing the resistive loading to 10 to 20 k from 47 k produces a substantially flatter overall frequency response from the Grado's.  Regards, Felix.

Wayner

I have a custom built load box (I made myself) and the only real thing that other loads did was cut off the high end, not smooth it out.

 

However, some suggest 10K is the right loading, I might go farther out, like 20K. Loading doesn't seem to change the lower end at all. Maybe if I had a scope.......And the slope could be caused by your phono preamp's RIAA curve. It is linear, just downhill. It would be interesting if you had another phono pre to try.

Wayner

dlaloum

  • Full Member
  • Posts: 710
Grab Luckydog's experimental spreadsheet (VinylEngine near zero capacitance thread) - plug in the cartridge spec inductance, and your load parameters (C, R, I) - remember to set it to 35% reduction factor (the model is still being refined and that seems to be about the right factor for those cartridges measured to date .... which has yet to include a Grado)

The result will be a theoretical plot of electrical response showing the electrical resonant peak.

You can then experiment interactively with the model, varying C and R - when the "Normal" line is as close to flat as possible - try out those values on your setup.

The model appears to work at this stage for a 2M ortofon an AT15SS and a Shure 1000e....

The Model does NOT take into account mechanical cantilever resonances - but the Grado's with the transmission line cantilever should have that resonance well outside the audio band, so it should not be an issue

Having +/- 5 db is not at all unusual for MM/MI cartridges - the manufacturers often do not tell you what loading/setup was used to generate the beautifully perfect F/R charts.

A lot of the classic MM/MI carts show perfect flat response which is unachievable unless you get a total Capacitance of 100pf (and with most MM stages having 220pf loading built in - you would need a special phono stage to achieve that!).

Also many classic cartridges of the 70's and early 80's were designed with the CD4/Quad R loading in mind = 100k, as opposed to the now "standard" 47k. And there seems concensus that Grado's prefer somewhere between 10k and 30k..... so much for "standard" loading.

Good luck with it

David

BaMorin

  • Full Member
  • Posts: 407
  • AR turntable rebuilder/modifyer
Grab Luckydog's experimental spreadsheet (VinylEngine near zero capacitance thread) - plug in the cartridge spec inductance, and your load parameters (C, R, I) - remember to set it to 35% reduction factor (the model is still being refined and that seems to be about the right factor for those cartridges measured to date .... which has yet to include a Grado)

Good luck with it

David

OH MY...........good luck indeed.

BaMorin

  • Full Member
  • Posts: 407
  • AR turntable rebuilder/modifyer
Hey everyone,

Here is a plot of the frequency response curve of my turntable using Dr. Feickert's Adjust+ software to measure. I'm wondering why this curve is non-flat (peaking more at the high end, with the downward slope from 20-10k). I would figure that a properly aligned turntable with a decent cartridge & preamp could produce a relatively flat curve. Any ideas what device / configuration might be causing this?

To measure the curve, you plug the RCAs from the preamp's output into the 1/8 line input on your computer. This means that the cause has to be somewhere between the turntable source and the preamp's output (can't be main amp, speaker wires, speakers, etc.).

Gear:
-Grado Gold cartridge/stylus
-Project Xpression III turntable
-DH Labs BL-1 interconnects
-Graham Slee Reflex preamp
-Standard power supply for preamp
-Pro-ject speed box
-Gear for calibrating turntable
  -Feickert protractor
  -Feickert adjust+ software (to align azimuth)
  -Pro-ject digital force tracking gauge
  -A bubble level for VTA
  -A bubble level for the turntable

As far as I can tell, the turntable is calibrated correctly, but maybe I'm missing something in the calibration that's causing the non-flat frequency response curve, or maybe I need to upgrade a component.

Any thoughts / help?
Thanks,
-Clayton






Just for clarification, what exactly is the cart tracking? 

neobop

  • Full Member
  • Posts: 3448
  • BIRD LIVES
Hi Clayton,
I really think the scale of the chart makes it look worse than it really is. Bass not withstanding, response goes down to about -2dB. That treble peak looks to be less than +3dB. Bass response would prob sound great on some mini monitors or speakers lean in the bass dept.

I notice you said that you played with azimuth and VTF. Is VTA adjustable? That would be my first thought of something to mess with. With the very low inductance of your Grado, I doubt if capacitance has anything to do with it. I suggest you first try VTA adj, if possible. Is the bass subjectively too warm? If so, perhaps internal damping might help. John or Wayner could help with that if you want to try. VTA might fix most of that as well.
neo

claytontstanley

Neo,

Yes VTA is adjustable. And... I was going back through some plots produced before setting the VTA using a bubble level, and the plot was much worse before (the treble peak was much more pronounced). So maybe I improved the VTA by calibrating it using a level, but didn't get it quite dead on.

Now that I know that azimuth and VTF aren't drastically affecting that treble peak, and remembering that VTA did....

Tomorrow, I'll try calibrating the VTA by aiming for a flat frequency response curve (instead of using a bubble level) and get back to you.

-Clayton

neobop

  • Full Member
  • Posts: 3448
  • BIRD LIVES
Clayton,
One thing I forgot to mention. After you get the VTA nailed down, try 1.6g VTF. Sometimes you have to go back and forth with these things. They're all interrelated and changing one might effect the other.
neo

Wayner

I think that a check on SRA (stylus rake angle) may be in order. Grado states that the front face of the cartridge be at 90° to the record surface, but after Mikey Fremer's article in Stereophile regarding SRA, I talked to the folks at Grado (we are a dealer) and they agreed that the 92° SRA would be beneficial. That means, the tail end of the tone arm is up in the air. The inclusive angle from the front of the cartridge to the record surface now should be at 88° (92° the exclusive angle). On a normal 9" tonearm, 1 mm of VTA adjustment (raising the tonearm base) equates to .25°. So once you have put the arm in a parallel position, and using that position as a point of reference (it becomes zero), you would have to raise it an additional 8 mm. Looks kind of funny, but it does work. It may flatten your response, but at the least will give your stylus a better contact area with the record grooves (less wear too).

Wayner  :D

Wayner

Neo brought up another thing that reminded me of some other offerings. As Neo stated, the rise in low frequencies is down there in the mechanical noise area, as well as the evil motor hum (60 cycle) and looking at the curve, those elements could easily be the culprit, as well as hum induced noise into the Grado (tho in this case, my be minimal). Using plasticlay will lower the fundamental resonance frequency of the table plinth, of course, other "loose" elements may be also vibrating either in harmony with the motor noise or bearing noise or worse, what ever is being played on the TT. I have found that placing a 1/8" diameter, then flattened out to about 1/16" thick piece of plasticlay about 2" down from the tonearm pivot, on the tonearm tube itself, really reduces inducted tonearm noise, from either end it seems. It's a very cheap tonearm damper, but will not rob the tonearm of needed response or energy intended for the cartridge. Just to make it even more confusing......

Wayner

BaMorin

  • Full Member
  • Posts: 407
  • AR turntable rebuilder/modifyer
Neo brought up another thing that reminded me of some other offerings. As Neo stated, the rise in low frequencies is down there in the mechanical noise area, as well as the evil motor hum (60 cycle) and looking at the curve, those elements could easily be the culprit, as well as hum induced noise into the Grado (tho in this case, my be minimal). Using plasticlay will lower the fundamental resonance frequency of the table plinth, of course, other "loose" elements may be also vibrating either in harmony with the motor noise or bearing noise or worse, what ever is being played on the TT. I have found that placing a 1/8" diameter, then flattened out to about 1/16" thick piece of plasticlay about 2" down from the tonearm pivot, on the tonearm tube itself, really reduces inducted tonearm noise, from either end it seems. It's a very cheap tonearm damper, but will not rob the tonearm of needed response or energy intended for the cartridge. Just to make it even more confusing......

Wayner

Ah, glad you brought that up.......as this can be done in a mathmatical formula (easy) for the odd order harmonics. Humans tend to find even order harmonics "sweeter"  AKA as "tube sound".  Still though, i have a question about this thread. Is this a test album the data is from?

claytontstanley

Yes, the data is from the test album that came with the Feickert Adjust+ software; it's a track that does a 20-20k sinusoidal frequency sweep.

I messed with the VTA over the past week, and I think I have the table dialed in about as good as I can get. I actually found that aligning the azimuth is pretty easy if you have a mono album and a symmetrical listening room. I was able to fine tune the results from the Adjust+ software using a Sinatra mono album and my ears. If the voice is in the middle, then the azimuth is aligned.

I could hear differences in the VTA as well, especially if you push the arm at the base down too low and have the arm raking upward; the bass response really starts to swamp the music.

In the end, I found that the best tools for alignment are your ears, a mono album, and a small bubble level to place on top of the tonearm to use as a reference for azimuth and VTA.

I'm including the plots after fine tuning the calibration this week. They aren't perfect, but they're pretty good, and the music sounds excellent, so I'm happy. If I ever upgrade a component, I'll definitely refer back to the plots, and see if the upgraded piece changes the response curve dramatically.

Thanks for your help everyone,
-Clayton

Log plot (like the original)


Non-log plot



neobop

  • Full Member
  • Posts: 3448
  • BIRD LIVES
In the end, I found that the best tools for alignment are your ears, a mono album, and a small bubble level to place on top of the tonearm to use as a reference for azimuth and VTA.

Now that's really interesting. Your ears, a mono record, and a reference point.

At first it looked to me that the new curve is much smoother. But I see the frequency scale is different. It does seem nicer though. Thanks for posting this.
neo

Tone Depth

  • Full Member
  • Posts: 635
  • Music Lover
    • SRLPE Wheel Works
It seems to me that the arm should be level, parallel to the surface of the record, and this is adjusted by raising or lowering the tone arm base to accommodate different cartridge heights.  Is there a problem with shimming a cartridge in the head to adjust the VTA/SRA?

I think that a check on SRA (stylus rake angle) may be in order. Grado states that the front face of the cartridge be at 90° to the record surface, but after Mikey Fremer's article in Stereophile regarding SRA, I talked to the folks at Grado (we are a dealer) and they agreed that the 92° SRA would be beneficial. That means, the tail end of the tone arm is up in the air. The inclusive angle from the front of the cartridge to the record surface now should be at 88° (92° the exclusive angle). On a normal 9" tonearm, 1 mm of VTA adjustment (raising the tonearm base) equates to .25°. So once you have put the arm in a parallel position, and using that position as a point of reference (it becomes zero), you would have to raise it an additional 8 mm. Looks kind of funny, but it does work. It may flatten your response, but at the least will give your stylus a better contact area with the record grooves (less wear too).

Wayner  :D

Wayner

The contact between cartridge and arm must be firm and a shim tends to lessen that. SRA is set at 92° at the record cutting lathe, in the record making process and to conform to the methodology of the vinyl groove, the playback stylus should "in theory" match the SRA of the cutting head. There is an article by Fremer on the topic, discussed a few issues back and the why's and how's to go about doing this. Of course, he wanted everyone to buy a $250 USB scope, but I proved (at least to myself) that the stylus could be viewed with a tripod mounted camera in zoom mode.

I'm not going to preach that all tables must be set to the 92° SRA, but there are some compelling info on the benefits of doing so. I have my Technics SL1200MKII set this way, as well as ARmod and ARCom. I think the other benefit may be better tracking (preventing stylus chatter because of the increased SRA), and that results in a cleaner presentation, while perhaps being kinder and gentler to the LP.

So, for those that want to try it, it may be worth it. I used to think that cartridge manufacturers built the 92° SRA into the stylus assembly so that when the arm was parallel, the SRA would be correct, but, as I have learned, and seen for myself, this is not the case, at least with the cartridges that I have.

Wayner