BMF
contributing
Posts: 99
|
Post by BMF on Dec 11, 2013 18:47:08 GMT
I am on the same page with this as keith is. When there is a lot of damping materials (wool, rockwool e.t.c.) in the cups a closed or opened port didn't seem to affect the lows anymore. I also found that closing the port (no wool) has a distinctly different effect than stuffing it with wool. In Jukes case (to get L and R to measure similar) I had to resort to closing the port on the 'bass heavy' side with a piece of felt (with the glue still on it) and using a certain amount of wool to lower the bass levels to where I want them. On the other (bass light) driver side I had to do the same as described by Keith by taping felt over the bass port with a small 'gap' of a few mm cut out. Had to use less wool to make it the same as the other channel. Till now (the few I have modded) mostly L and R cups were treated similarly and the amount of wool needed to get R and L differed only slightly. The end results was always close. It is 'concerning' that replicating 'the trick' doesn't necessarily translate to even bass response even between L and R. It is true that it will become flatter and better controlled. To make one into reference material I suppose you really need to measure it. Perhaps we could join forces in making (and supplying) cheap kits for measurements. Thinking of a small 'plank' with a mic in it. On the backside it is connected to a simple fantom supply with a 9V battery clip and a wire with RCA plug. Strap it on the earpiece with rubber bands. Essentially what Keith makes but on a plank instead of an etymotic flange. Also we can supply REW (or you can download that yourself) and we can provide a correction file for the 15kHz peak. a full measurement kit for the price of a cheap headphone. Takes the guessing out of it. Perhaps make a thread for it or a tutorial on the the website. Hermetically sealing it will be hard to do and is not recommended. I know Luis (Paradox) claims his versions are hermetically sealed but acc to owners they are not. You would also have to close the hole where the headbandwire enters and completely seal of the cable socket with a blob or something. Also the baffle port would need to be fully closed. and the edges of the baffle would need a sealant as well. The membrane (if truly hermetically sealed) would be pushed inwards and outwards (like a barometer works) depending on the air pressure that day. Of course you can completely close the port on the back completely and this will lower the bass (as in standard T40 opposite standard T20/T50) but the wool and anti reflection materials are better suited to 'control' the lows. Frans, I have been busy and did not realize how much time has passed since I've visited, here. I hope all is well with your and yours. I built over 10 measurement kits and sent them to guys for actual costs (and in some cases for free). At least one has posted, somewhere, within this forum. Your phantom power supply is superior to what I had been using. Thank you, again, for sharing the design! I credited you for its design within Fostex Incremental Mods and Measurements (FIMM) and linked back to your post, here. I think having a measurement kit is essential for determining what works and insuring channel balance. I use REW and my measurement kit for FR graphing and channel balancing. I have seen many TOTL current production headphones differ with regard to Channel Balance, more than 1 or 2 dB. When making modifications with damping materials, the L-R difference can be even greater despite making what may appear to be such insignificant modifications. We must remember we are dealing with tiny transducers. A single mm difference in the size of modified bass ports and a few 10ths of a gram difference in the weight (and thickness) of cotton or rock wool can make a vast difference in SQ and/or SPL. This is why I favor a FR measurement kit, digital calipers, and digital scales that can be calibrated and measures to 0.01 gram. Having these tools will help anyone to 'approach duplication' of someone else's modification configuration. Without these tools, there will be too much variation in each component to truly recreate anyone's modification configuration. Multiply the potential variance for each component in a config by the number of components and it's easy to understand why following a step-by-step guide is not enough to approach duplication of the configuration. I still favor the mic in my ear because I have the tactile cues I need for mic and headphone placement. I simply remember to make the offsets between 2 kHz and 5 kHz to compensate for error variance when comparing graphs generated from my (real dummy) head vs my wooden dummy head. Keith
|
|
|
Post by imstimpy on Mar 11, 2014 3:29:06 GMT
Well, I think I'm sold on the T50RP. Is it really a $100USD starting price!?
I believe this will lead me down the measurement kit road. It will be wonderful to take the guess work out of what I think I'm hearing. I'm trying, unsuccessfully, to get frequency plots for my modified D2000s. As best I can tell, the modifications were generated entirely by sound o.0
|
|
solderdude
Administrator
measureutternutter
Posts: 4,881
|
Post by solderdude on Mar 11, 2014 6:06:40 GMT
T50RP can be found second hand for much cheaper sometimes, new as well in the US but more difficult to find in Europe. The T40RPmkII is the same IF you plan to modify it as it only differs in the colour of the metal braces (silver vs bronze) and the fact that the 'vent' on the back is closed with a piece of plastic instead of a paper filter. www.thomann.de/nl/fostex_t40rpmkii_kopfhoerer.htmIt is cheaper though. Making a test rig isn't that hard but you need an original WM61(a) mic as the 'replacements' aren't really 1 on 1 replacements. Also there is the question of making the needed corrections to get fairly 'absolute correct' measurements. It does take the guesswork out though. I tried to EQ by ear correctly and found it difficult to get it as good as by measurements. Some headphones, even though made flat FR wise, still have problems. Mostly these are resonances or distortion (cone break-up) which cannot be removed by any EQ. Some headphones also cannot be corrected with a 'simple' EQ. The more elaborate the compensation must be the less 'improvements' can be had. Of all the headphones I had and still have I like the modified and filtered Fostexes the best. I prefer it over my DT990/600, HD650 (even a filter corrected one), modified K551, filtered DT1350. It simply is the most realistic one but needs to be fed high quality recordings. It is ruthless on lesser recordings. DT990 and HD650 are more 'forgiving'.
|
|
|
Post by imstimpy on Mar 11, 2014 13:13:44 GMT
Looks like a WM61(a) can be had on eBay for $10 from Taiwan or Hong Kong, or a 20 pack for $50.
The price between the T40RP and T50RP is negligible so I'll read into it a bit more on which set to buy.
|
|
solderdude
Administrator
measureutternutter
Posts: 4,881
|
Post by solderdude on Mar 11, 2014 13:47:06 GMT
I have been told fake WM61's are being sold as they are no longer in production. You need only one of them.
The difference between T40 and T50 is sometimes very small. If you are going to mod them it doesn't matter that much which one you pick (except for look perhaps) They have the same drivers, perhaps the T50's have their drivers more closely matched but haven't seen real evidence of it. There does seem to be quite some variance in drivers but always have been able to make them closer by adjusting the mods or amount of wool. This, of course, is only possible with a measurement rig.
|
|
|
Post by imstimpy on Mar 12, 2014 14:11:42 GMT
I just purchased 6 pieces from a fellow in Japan for $20USD. The story is they were found in a shop in Akihabara. Having been there myself, this is no surprise... It has a guarantee through eBay so if I can somehow verify them I should be in the clear. Unfortunately, they will be coming on a slow boat.
|
|
solderdude
Administrator
measureutternutter
Posts: 4,881
|
Post by solderdude on Mar 12, 2014 14:29:12 GMT
That'll give you time to make the coupler and front-end for the mic. Mine is very simple and consists of 2 planks at roughly the same distance as my ears and are covered with self-adhesive rubber foam to mimic skin. I made a corrected amplifier for it which works well for FR but is a bit noisy for THD measurements. The advantage is I don't have to correct the input so have no need for correction files. There is a thread that discusses DIY rigs. diyah.boards.net/thread/361/
|
|
|
Post by imstimpy on Mar 17, 2014 14:58:37 GMT
I have two big gripes with the T50RP so far: 1) The stock pads are hot, uncomfortable, and do nothing for isolation or sound. Different pads are a must here. I ordered Shure 840 pads with the headphones and installed them almost immediately. 2) The locking mechanism on the stock cable locks the cable to the headphone, but doesn't provide a good audio connection. My right channel is cutting in and out. I'm ordering a V-Moda right now...
With Shure pads, the sound isolation far exceeds my Denon D2000s with the Lawton pads; the Shure pads are hotter and, therefore, not as comfortable. Base appears to be bigger and lighter in the D2000, but the T50RP with Shure pads and a base port offer better bass than the AKG K701. In the current, mostly stock, configuration, the rolled off highs need some love. The character of the highs is similar to the D2000, but the airiness is missing.
I'm currently amassing supplies to further mod the headphones as well as measure those mods.
I've looked around a touch, but I haven't seen anybody mention amplifier output impedance for the Fostex. People seem to indicate more power is better, but that is about it. I have read Benchmark's 0-Ohm amplifier paper and see merits to going nearly 0 ohm, but switching between 120 Ohm and 0 Ohm on my Panda didn't produce much audible change.
|
|
solderdude
Administrator
measureutternutter
Posts: 4,881
|
Post by solderdude on Mar 17, 2014 16:05:30 GMT
The stock pads are indeed uncomfortable and sweaty. I use SRH-940 pads instead (velour versions of SRH-840 pads)
Cleaning the connectors seems to help. I didn't like the connection and found the cable to be too stif and microphonic and replaced it with a more supple cable.
Didn't like the stock sound one bit. It needs EQ (or correction filter) and then it sounds great and that may be an understatement.
The Benchmark paper is a bit of nonsense really (marketting BS) which NwAvGuy used to make his 'point'. They aren't really measuring a higher distortion, they are measuring/showing the EMF generated by the driver. More specifically the part of the generated EMF voltage that falls over the (added) output resistor of the amp. They state this is distortion caused by the high output R and damping factor but... it isn't. If they measured the THD coming out of the amplifier section (thus BEFORE the output resistor) they would have found that the distortion that is added by the amplifier circuit itself would have been the same and there is no increase. What one is seeing in the plots is what the headphone driver generated acting as a microphone NOT distortion added by the active electronics or damping. Of course if one looks at the output of a black box (the amplifier with the added output resistor) than you measure across those output terminals. Would they have measured the actual (damping) CURRENT the differences would have been MUCH smaller.
It isn't damping, as much as people would like to see it this way cause it 'explains' things.
A driver is damped by CURRENT and this is determined by the drivers Ohmic resistance + amplifiers resistance, number of windings and magnetic field strength. Most damping for headphones is done by air surrounding the driver and membrane materials as well as membrane tension (in case of T50RP). Headphone membranes/diaphragms aren't comparable to heavy woofer cones.
For T50RP, which has a very linear resistance, a higher output resistance should not lead to a different sonic signature but does reduce the output level considerably. A lower output resistance is advised so you have more dynamics.
Simply place another socket on the Panda that has 10 Ohm (or even 0 Ohm) output R.
|
|
|
Post by imstimpy on Mar 21, 2014 19:32:45 GMT
I don't know where to post. There are at least two T50RP threads (one in the Headphones section).
* I cleaned the stock cable with some alcohol and that fixed the dodgy signal. * I added some cotton balls behind the driver. This, I think, improved vocal frequencies some. * I removed the output resistance from the Panda. I don't know if it has more "punch", but it definitely has more power.
A quick A/B with the Denon and the modified T50RP sounds surprisingly similar. The bass goes really low, but it isn't quite as big as the Denon. Against the K701, I'd say the vocals are little more subdued, but flatter and more realistic sounding. The frequency response does still sound a little downhill as it goes higher, but I don't intend to draw too many conclusions until I get my measurement rig going.
|
|
Rabbit
Administrator
Posts: 7,087
|
Post by Rabbit on Mar 21, 2014 20:16:36 GMT
If I compare mine to the D2000, the bass isn't as 'fruity' as the Denon and the treble is no way as extended until the filter is switched in. I do actually use it quite a lot without the filter as well. Noise floor is lower and I don't mind a smoothed off treble too much if I listen a bit louder. That way, the bass fills out with the added volume and it sounds really natural.
The connection problem is very common. I got rid of mine by just moving the plug back and forth until the crackling faded away. it never came back either.
I prefer them at low output impedance from the amp as well. They do seem more lively then.
One of my favourite headphones actually.
|
|
|
Post by videoguy on Aug 21, 2016 5:22:21 GMT
JLI Electronics in Harleysville, Penna. has a replacement for the WM-61A; it i9s the JLI-61A. Bought two of them for $2.16ea. plus cheap shipping to Fla. JLI website is <www.jlielectronics.com>. email address is: <info@jlielectronics.com>.
The frequency response plot is identical to the Pannys. They also have larger electret cartridges & condenser capsules.
|
|
|
Post by videoguy on Dec 3, 2018 21:25:32 GMT
I use my TRP 50's exclusively. either wit or without the original filter PCB. Built the AKHA from another forum & added an output impedance switch. The amplifier put out 3 watts the way I have it set up. Output impedance is somewhere in the area of 0.5 ohms to 1 ohm. Ortho-dynamic phones, Fostex TRP series, Audeze, Hi-Fiman, etc. perform best when fed from a very low impedance. I have added an impedance switch to the output of the amp.; the positions are 0, 65, 120 ohms. When feeding Ortho-dymamic phones the switch is set to o ohms, (there is a 2 ohm resistor on the amp board); the others are the listed values.
Whit the TRP 50's plugged in there is a very noticeable drop in level when switched to either 65 or 120 ohms. On the other hand, phones using standard drivers are not affected by a loss in volume as the switch is moved away from the zero ohms position.
|
|
solderdude
Administrator
measureutternutter
Posts: 4,881
|
Post by solderdude on Dec 27, 2018 16:43:00 GMT
The T50RP's are 50 Ohm so its easy to calculate how much softer they play on the 65 and 120 Ohm setting. 65 Ohm = -7.2dB 120 Ohm = -10.6dB Tonality won't change as the impedance is flat.
With a 32 Ohm headphone the numbers are: 65 Ohm = -9.6dB 120 Ohm = -13.5dB
A 300 Ohm headphone has less attenuation 65 Ohm = -1.7dB 120 Ohm = -2.9dB
|
|