advertisement


Tannoy HPD 315 Questions

ToToman.. interesting to note that the HF units are often mismatched.. I understand the pepperpot machining also plays into this. Will also do some distortion measurements as you suggest.
The distortion measurements should already be there. In the main window of REW you should see the measurements you took as a list of tabs on the left. Then to the right there should be a row of buttons, SPL & Phase, All SPL, Distortion, etc. If you click Distortion you should see a graph with the Frequency Response on top and the Distortion measurement below. Make sure all of the harmonic checkboxes are ticked apart from Noise Floor. Also you may need to change the Y-axis scale, I recommend 90dB as the upper limit and 0dB as the lower limit.

PS - To give you an idea of typical frequency response variations between Tannoy drivers, here are some I've measured over the years:

40254939593_3b3cfb80f3_b.jpg


40254939763_d63673908c_b.jpg


47219697991_79cdf52fe4_b.jpg


40254685543_cde59026e0_b.jpg


This one would have been a good candidate for HF re-alignment:

36757649485_a15f8d2bcd_b.jpg
 
ToToMan, thanks these examples are great as a reference!
Today I decided to take the DIY crossovers out of the equation and did new measurements using one of my recapped factory crossovers hooked up. The matching between drivers now looks possibly worse than before.. :(

HPD315 Factory Xovers by Jonathan Chiles, on Flickr

Must look into how to interpret, but below are the distortion plots:

CAB 2 DISTORTION by Jonathan Chiles, on Flickr

CAB1 DISTORTION by Jonathan Chiles, on Flickr
 
ToToMan, thanks these examples are great as a reference!
Today I decided to take the DIY crossovers out of the equation and did new measurements using one of my recapped factory crossovers hooked up. The matching between drivers now looks possibly worse than before.. :(
This is odd to me. If you measured both drivers with the same DIY crossover, and then measured both drivers with the same factory crossover, I would expect the difference in frequency response between the two drivers to remain constant, even if crossover changes the absolute frequency response of the drivers.

BTW - The Distortion plots look fairly innocuous to me. No major issues of note, though ideally you want both drivers to have similar distortion profiles. I'd say the first graph is more 'normal' looking than the second. The hump at crossover frequency (1kHz) is pretty typical IME (at least it is for MGs and HPDs), - Tannoys aren't as distortionless as some might have you believe....
 
I was trying a few things, one of which was swapping HF diaphragms between drivers, I think I had them flipped around when I did that last measurement. I also later tried flipping the phase of the HF units to make sure I had them the right way, I did fortunately.
I put everything back as original and carefully recentered again this morning and seem to have achieved a better match now using a single factory crossover for the measurements. Distortion plots now show a THD peak around 8k though..

HPD315 FCTRY XVR 1304 by Jonathan Chiles, on Flickr

CAB1 DISTORTION PLOT 1304 by Jonathan Chiles, on Flickr

CAB2 DISTORTION PLOT 1304 by Jonathan Chiles, on Flickr
 
You have much better pair matching now, I am in fact very surprised by how much you've managed to improve it! The distortion peak at 8kHz isn't ideal but is inevitable to some degree due to the higher SPL at this frequency. You also still have a 5dB difference in SPL at 12kHz. Did this move with the HF swap? (I'm trying to work out if the difference at 12kHz is due to the condition of the pepperpot or due the condition of the HF unit).
 
Last edited:
Good question, would be interesting to know if this is pepperpot machining or HF variations that cause that divergence at 12k. Let me look at my mess of plots here (should have kept better notes of what I was doing:()

At least I am hopefully in the ballpark of a reasonably matched pair of HPDs now!? I am almost scared to mess with them further...

I also need to revisit my cabinet tuning which is sitting a little high I think, at around 40Hz for 100 liter net volume. Also have been chatting to Paul Coupe (of RFC) re my DIY crossovers and looks like I'll need to make some changes there as well if I want to successfully replace the factory units. I would truly be lost without the guidance here!
 
What's up with the pepper pots?
If the significant divergence between 12kHz-16kHz doesn't move to the other driver when he swaps HF units then isn't it plausible that the pepperpot and/or horn throat is responsible for the difference?
 
If the significant divergence between 12kHz-16kHz doesn't move to the other driver when he swaps HF units then isn't it plausible that the pepperpot and/or horn throat is responsible for the difference?
I can see there is a 5 dB difference, I'm asking what is physically / visually wrong with the pepperpot(s)? Do both cones sit at the same position relative to the throat mouth, are the dust caps clean, is the phase plug clean, is the hf gap absolutely clean, have you tried this hf top plate with +/- washer, have you reflowed the solder at the frame socket solder point etc etc.? Welcome to the rabbit hole :)
 
As I have never had the cones out I havnt ben able to properly check the pepperpots visually other than with a torch through the dustcaps. I did do a thorough cleaning of the HF magnet gap, and have also carefully checked solder joints and actually ditched the Tannoy 4 pin plug and soldered carefully straight onto the tags. Also noticed one of the drivers has 3 paper washers and the other 2. My understanding is the washers "belong" to the driver not the HF unit, to account for small differences in machining there.

Anyways this rabbit hole is deep! I documented my build of the cabinets here all the way back in 2016...

https://pinkfishmedia.net/forum/threads/tannoy-hpd315-build.188346/

Back to measurements.. Please bear with me on the learning curve! It looks like I had the microphone too nearfield at 12 inches from the dustcap (was hoping to eliminate the room) Below is a plot with the mic at 1 meter. The differences in the high end remain, but the deep notch around 2k is gone and there is now quite a deep notch at about 120Hz.

HPD315 AT 1 METER by Jonathan Chiles, on Flickr
 
As I have never had the cones out I havnt ben able to properly check the pepperpots visually other than with a torch through the dustcaps. I did do a thorough cleaning of the HF magnet gap, and have also carefully checked solder joints and actually ditched the Tannoy 4 pin plug and soldered carefully straight onto the tags. Also noticed one of the drivers has 3 paper washers and the other 2. My understanding is the washers "belong" to the driver not the HF unit, to account for small differences in machining there.

Anyways this rabbit hole is deep! I documented my build of the cabinets here all the way back in 2016...

https://pinkfishmedia.net/forum/threads/tannoy-hpd315-build.188346/

Back to measurements.. Please bear with me on the learning curve! It looks like I had the microphone too nearfield at 12 inches from the dustcap (was hoping to eliminate the room) Below is a plot with the mic at 1 meter. The differences in the high end remain, but the deep notch around 2k is gone and there is now quite a deep notch at about 120Hz.

HPD315 AT 1 METER by Jonathan Chiles, on Flickr

That looks like a much more normal response from a Tannoy driver now!

Ignore the notch at 120Hz, that is caused by the room (floor-bounce cancellation). The lower in frequency you go, the more the room comes into play, so as long as both drivers are well matched in the low frequencies I wouldn't worry about the peaks and dips you see below 400Hz as these will vary with mic distance.

Looking at the above graph I'd be tempted to try turning the Green driver's Energy control down one notch and post another graph to show how the matching changes. Even if it doesn't improve the matching it will illustrate the effect of the Energy control.

Also try putting the Rolloff control up a notch on the Green driver (but only after returning the Energy control to Flat position), I'd be interested to see the steepness of the slope.
 
That looks like a much more normal response from a Tannoy driver now!

Ignore the notch at 120Hz, that is caused by the room (floor-bounce cancellation). The lower in frequency you go, the more the room comes into play, so as long as both drivers are well matched in the low frequencies I wouldn't worry about the peaks and dips you see below 400Hz as these will vary with mic distance.

Looking at the above graph I'd be tempted to try turning the Green driver's Energy control down one notch and post another graph to show how the matching changes. Even if it doesn't improve the matching it will illustrate the effect of the Energy control.
I was going to suggest the roll off -1....
 
I was going to suggest the roll off -1....
Yes that might sort it too. My gut feeling is that the Rolloff slope may be a little too steep and cause a mismatched response in the other direction. It'll be interesting to see the effect of both.
 
Interesting exercise visualizing what the switches do. For anyone that doesnt know the "Energy" control boosts or cuts the entire range of the HF unit (1kHz - 20kHz) while the Rolloff attenuates 5kHz and above in 3 steps.
Below:
Cabinet 1 measured flat / energy -1 / rolloff 1
versus
Cabinet 2 flat

ENERGY AND ROLLOFF COMPARISON by Jonathan Chiles, on Flickr
 
Interesting exercise visualizing what the switches do. For anyone that doesnt know the "Energy" control boosts or cuts the entire range of the HF unit (1kHz - 20kHz) while the Rolloff attenuates 5kHz and above in 3 steps.
Below:
Cabinet 1 measured flat / energy -1 / rolloff 1
versus
Cabinet 2 flat

ENERGY AND ROLLOFF COMPARISON by Jonathan Chiles, on Flickr
I'm struggling to see which trace is closest to CAB2 FLAT, but it looks like ENERGY-1 is too heavy-handed below 2.5kHz while ROLLOFF1 is too heavy-handed above 5kHz. Can you upload two separate graphs please, one showing CAB2 FLAT with CAB1 ENERGY-1, and the other showing CAB2 FLAT with CAB1 ROLLOFF1 ?
 
Sorry things were getting a bit crowded there with 4 measurements on one chart... here are 2 separate comparisons. Both with Cabinet No2 all controls FLAT as reference. Plotted against Cabinet 1 at Energy setting -1 and then plotted against Cabinet 1 at Roll-off setting 1. Much of a muchness between them really, but I'll give them both a proper listen when time permits..

HPD MATCHING 2 by Jonathan Chiles, on Flickr

HPD MATCHNG 1 by Jonathan Chiles, on Flickr
 
Sorry things were getting a bit crowded there with 4 measurements on one chart... here are 2 separate comparisons. Both with Cabinet No2 all controls FLAT as reference. Plotted against Cabinet 1 at Energy setting -1 and then plotted against Cabinet 1 at Roll-off setting 1. Much of a muchness between them really, but I'll give them both a proper listen when time permits..

HPD MATCHING 2 by Jonathan Chiles, on Flickr

HPD MATCHNG 1 by Jonathan Chiles, on Flickr
Depending on the kind of music you listen to and the acuity of your hearing you may find that either the first or second graph sounds more closely matched, so have a listen and let your ears be the judge. Either way it's pretty exceptional pair matching as far as Tannoys go so you should consider yourself fortunate! :)
 
Yep, I think I can happily live with these as is.. not a bad pair in the end and a reminder that a bit of persistence with the HF alignments will pay off! Thanks for bearing with me through my REW learnings.. and hope this has been of interest/benefit to other HPD315 owners. :) Will take some time experimenting with positioning in my room as well as the roll-off and energy settings of course.
 


advertisement


Back
Top