|
|
Post by brmslash on Aug 20, 2015 5:09:25 GMT
So I've finally done all the rev 1.2 updates to my rev 1.0 board. Although I'm still tweaking the 100pF added to the negative leg of OPA551. That 100pF cap definitely gets rid of that odd treble glare I had been experiencing and tried alleviating with the attenuation module. However I think it maybe takes away too much treble. I'm currently using a 22pF which takes the edge off, but still needs to be more. I've got a 47pF on order so hopefully that will be the sweet spot for me  Also in regards to the latest photos above, I'm curious as to what the grey coloured caps are that have replaced the previously red WIMA MKS4 (i.e. the C4's)? |
|
solderdude
Administrator
measureutternutter
Posts: 4,886
|
Post by solderdude on Aug 20, 2015 5:40:23 GMT
Those are the stock 1uF caps.
Should ask Jeremy why he changed them.
Thanks for the feedack.
The 100pF was more of a calculated thing and determined based on squarewave response under capacitive loads.
Possibly the 47pF will do fine.
Used 10pF in my experiments but this wasn't nearly enough and 100pF looked optimal in the plots.
|
|
|
|
Post by brmslash on Aug 20, 2015 10:26:59 GMT
Thanks for the response. It's nice to have an insight into the design and is much appreciated.
|
|
Crispy
very active
Madrigal music is playing - Voices can faintly be heard, "Please leave this patient undisturbed."
Posts: 779 
|
Post by Crispy on Aug 22, 2015 10:07:39 GMT
The shipment currently is in transit, it should arrive by the end of the nextcoming week. Can't wait to listen at it!  Hi Nicola, I have just checked the transit status of the Polaris and it is still stuck in the delivery depot at your end, it was posted on the 10 Aug and it arrived at your end on the 13th. This means it has been stuck at your end for 9 days, and I thought royal mail was bad  Lets hope it arrives soon, as I am begining to wonder if it has got lost? lets hope not eh  |
|
z3d
quite active
Posts: 170
|
Post by z3d on Aug 24, 2015 18:52:41 GMT
Hi Chris All is fine, Polaris was delivered and the package was in good conditions. I will compare it with my Ember soon, just waiting to have some spare time to play with |
|
Crispy
very active
Madrigal music is playing - Voices can faintly be heard, "Please leave this patient undisturbed."
Posts: 779
|
Post by Crispy on Aug 24, 2015 20:11:19 GMT
Hi Chris All is fine, Polaris was delivered and the package was in good conditions. I will compare it with my Ember soon, just waiting to have some spare time to play with Thank **** for that Nicola, I was getting worried that it had got lost - I checked again today and it still said it was at the depot  Your a lucky lad to be able to compare the Ember & Polaris together, I had to go by memory but from what I can remember they sound very similar. Look forward to your comparisons. |
|
|
|
Post by brmslash on Sept 6, 2015 4:32:25 GMT
So I tried the 47p, but still needed to be more. I tried a low voltage (50v) 100p and to my surprise it actually sounded different to the first 100p I tried (1kV). Both were c0g, but I guess the physically larger size of the 1kv cap must have some different electrical characteristics that causes it to roll off more gradually and thus cut lower. Now with the lower voltage part, I think 100p is pretty good.
I've also been trying different electrolytics in C5 and C15. I tried Elna Cerafines and Silmic II's. Both are a big step up in clarity and physical size. I'm currently preferring the Silmics, and they keep getting better as they burn in (mainly the mids). The only problem is there isn't quite enough space to lay down the C15 caps down flat to fit under the cover. The C5's just fit. I used 16v for C5 and 35v for C15. I may end up drilling some holes in the cover or just leave the cover off.
For reference, my playback chain is Windows 10 (24bit, 44100Hz) > Foobar2000 (DS:Primary Sound Driver) > Schiit Bifrost Uber USB2 (via usb cable) > Project Polaris (R1.0 modded to R1.2 mellow BW, w-A 1.21k, mid BW, mid G, low-R) > Sennheiser HD598.
|
|
solderdude
Administrator
measureutternutter
Posts: 4,886
|
Post by solderdude on Sept 6, 2015 6:23:38 GMT
From a technical POV the 100pF just 'starts' to do something at 100kHz (-3dB point = 500kHz)
It is there to keep the OPA551 (in the 2x gain mode) more stable under (unrealisticly) high capacitive loads combined with a high impedance headphone.
With headphones like the HD598 the 100pF can even be completely removed.
It is just there to ensure proper operation under unrealsitic extreme circumstances with signals that can not occur in any audiosignals.
The cap can only reduce the gain (above 1MHz) by 6dB and cannot cause any roll-off in nor well outside of the audible range.
From a technical POV a higher voltage rated ceramic (that is not C0G) will be less voltage dependent and produce less distortion when the same voltage is applied.
It is true that very small (physical) sized ceramics below X7R rating and low voltage ratings can produce distortion based on a voltage across it.
However, in this case it is in parallel to a low resistance.
Should you want to 'know for sure' than you can always replace ceramic caps with film type caps who do not show this behaviour (changing capacitance under different voltages).
The bandwitdh limiting parts are:
C9 + input cap of the OPA, C10, C11 with R17+R18 and the slewrate of the OPA which all already start to 'do' something before the 100pF does.
The fun partof amps like this is that it is easy to play with components (as the schematics are published).
For more dramatic changes in sound playing with the amplification FET and its operating point will bring substantially different sound.
It should be noted though that the minimum Vds must be taken under consideration and each FET will require other resistors and operating points.
They 'act' somewhat like tubes but have a HUGE amplification compared to tubes.
For this to succeed (we played with many FETs) you really need an oscilloscope, analysers and some knowledge about fow to get the FETs in the proper operational point.
Even the same FET but at a different operational point will give wildly varying results that can range from really nice to awfull.
|
|
|
|
Post by brmslash on Sept 6, 2015 11:05:03 GMT
Thanks for the detailed response. I like to tinker by ear using values or components that are close to original, so hadn't performed the calcs.
It's only been fairly recent that I've been exploring ceramics in my Marshall guitar amp, so I'm appreciative of your additional experience and knowledge. In my tinkering with class 2 ceramics, I found that the tighter the temperature range and tolerance, the better the sound. So I ended up with a mix of X5F's and Y5P's. X7R's sounded terrible, or at least the ones I tried.
Yeah it's refreshing and enjoyable when it's easy to play with components and refer to official schematics. That was part of the draw for me with the Garage1217 amps.
I was contemplating adjusting the jfet biasing to see what the change would yield. Thanks for the info. I've been reading up as I wanted to make sure by doing the calcs first to see which way I wanted to go on the characteristic curve. But like you say, to understand the full picture, an oscilloscope etc would be essential. Having said all that, my Polaris seems to be settling in quite nicely at the moment, so I probably won't end up changing anything.
|
|
solderdude
Administrator
measureutternutter
Posts: 4,886
|
Post by solderdude on Sept 6, 2015 16:39:24 GMT
In guitar amps the capacitors are usually intended to filter sound well within the audible range.
As most of these amps are tube amps that work on high voltages AND have a high voltage swing across them ceramics should not be used.
For these applications film-type caps are essential.
IF one has to use ceramic multilayers they must be at least NP0 / G0G
|
|
|
|
Post by luckbad on Sept 6, 2015 20:34:32 GMT
I have a Project Polaris incoming this week. Pretty excited to check it out with all of its features. I'll also do a head-to-head comparison of it against the Cavalli Liquid Carbon after that arrives.
Mine is prebuilt by Jeremy, but I might decide to tinker with it. I've been searching around for good guides/parts lists about how to get into electronics DIY (specifically for headphone/audio stuff).
|
|
solderdude
Administrator
measureutternutter
Posts: 4,886
|
Post by solderdude on Sept 6, 2015 21:51:47 GMT
When you are a novice at soldering I suggest to check every component twice or even more for polarity/orientation or value.
The reason for this is the unusually thick PCB's are a real nightmare to 'desolder' or to change parts on.
Once they are soldered in position they should stay there.
|
|
|
|
Post by luckbad on Sept 7, 2015 3:45:32 GMT
Thanks. I don't plan to tinker with amps until I'm confident in soldering.
I have an Audio-GD NOS-1704 on the way. Would you recommend medium bandwidth?
Thanks!
|
|
solderdude
Administrator
measureutternutter
Posts: 4,886
|
Post by solderdude on Sept 7, 2015 4:57:15 GMT
As the AGD is already rolled off I would recommend the wide BW setting.
You could also try the medium one though.
|
|
|
|
Post by crea78 on Sept 9, 2015 0:25:42 GMT
Frans, I will be getting the HD600 this week and looking at possibly getting the Polaris very soon. What are the optimal settings to start with. Thanks
|
|
Also in regards to the latest photos above, I'm curious as to what the grey coloured caps are that have replaced the previously red WIMA MKS4 (i.e. the C4's)?
