Absolutely right. I've shown a few spectra over the last few posts - here's an example:AKSA said:
However, because I'm still not quite happy with my testing procedure (due to using wire-wound resistors in the dummy load / voltage divider, which are not low-inductance), I haven't yet shown more details than that. When I've got to a point where I feel some more comfort with the results, I'll post more details. (I'll also be able to take my amp back to the lounge room and actually listen to it again then! 😀 )
Hugh: other than showing the graph of the spectra, THD, THD+noise, and IMD (all across a variety of frequencies and power levels), are there other pictures or measurements which you feel may be useful?
JP, these are all very interesting. Could you please post a schematic for dummy how to setup the measurement? I want to measure my GC.
Is it using left-out and right-in channels so that the signal is looped back, and then feeding 1Hz tone?
My software is ARTA.
Is it using left-out and right-in channels so that the signal is looped back, and then feeding 1Hz tone?
My software is ARTA.
Will do Gainphile - gimme a day or two to confirm it all works OK first, then I'll post full detail on the setup and process. What sound card have you got?
BTW, take a look at Visual Analyzer as well - ARTA has lots of stuff for speakers too, but Visual Analyzer has many more features suitable for amp measurement and analysis.
Of course, none of this will match what an Audio Precision machine can do, but for 100x less money, it's a reasonable compromise I think! 😉
(PS: I've received all the necessary bits for my Plutos now, except for the plumbing parts. I'll be sure to let you know when it's done so you can hear how it sounds.)
BTW, take a look at Visual Analyzer as well - ARTA has lots of stuff for speakers too, but Visual Analyzer has many more features suitable for amp measurement and analysis.
Of course, none of this will match what an Audio Precision machine can do, but for 100x less money, it's a reasonable compromise I think! 😉
(PS: I've received all the necessary bits for my Plutos now, except for the plumbing parts. I'll be sure to let you know when it's done so you can hear how it sounds.)
Hi Jeremy,
No, that would be enough for me.
Since most amps suffer more distortion at HF, I try to simulate distortion profiles at 10KHz and 20KHz. You may not be able to properly do this with a digital sound card which I'd assume has a sampling rate of 44.1KHz, but you certainly can with a competent simulator. The AP should be able to do it, of course, but it's exxy.
The only other thing of real importance is a Bode plot to assess stability. This again can be done in a simulator, though it is reliant on good component models and often they are flawed. Of course, in real life you can drive the amp with a provocative square wave and check the leading edges for ringing; this is as good as anything and can be assessed with a CRO.
The final test has to match the consumer test - which is a listening session, and by its subjective nature this is very controversial, not least because people's audio tastes do vary a lot.
Great progress!
Hugh
No, that would be enough for me.
Since most amps suffer more distortion at HF, I try to simulate distortion profiles at 10KHz and 20KHz. You may not be able to properly do this with a digital sound card which I'd assume has a sampling rate of 44.1KHz, but you certainly can with a competent simulator. The AP should be able to do it, of course, but it's exxy.
The only other thing of real importance is a Bode plot to assess stability. This again can be done in a simulator, though it is reliant on good component models and often they are flawed. Of course, in real life you can drive the amp with a provocative square wave and check the leading edges for ringing; this is as good as anything and can be assessed with a CRO.
The final test has to match the consumer test - which is a listening session, and by its subjective nature this is very controversial, not least because people's audio tastes do vary a lot.
Great progress!
Hugh
AKSA said:
Then it's a pointless test unless it is done blind otherwise who's ears do you use as the reference point ??
Let's not get into that discussion - everyone has their own opinion about that and I doubt we're going to change anyone's minds here! 🙂Ray Chuck said:
jp_howard said:
I'm not interested in changing peoples minds but can we stick to the objective measurements ?? I am interested to see the measurements for high power 😉
I'm interested in that too, which is why I'm adding a higher level of attenuation in my new voltage divider. Until that's working I can't do measurements at high power.Ray Chuck said:
Jeremy,
Are you using 24 bit 192KHz wav files for your testing?
If you are not, my argument about sampling rates is likely valid.
Hugh
Are you using 24 bit 192KHz wav files for your testing?
If you are not, my argument about sampling rates is likely valid.
Hugh
AKSA said:
jp_howard said:
Hugh, Jeremy,
It's worth doing a FR test on the sound card to see if/where the
analog low pass filters are, mainly on the ADC side of things.
These modern sound cards are quite amazing spec wise and even
my old EMU1212 gets close to -120dB DR (A wtd / 20kHz BW).
Provided the soundcard gets somewhere near 80kHz I think
probably 10kHz is a good compromise between stressing the amp
and seeing enough high order spectral components. 10kHz into
4ohms at high power should be enough to get the amp well above
the SC's residual.
Does the soundcard have balanced I/O's?
Jeremey, sorry if this has been asked, what software are you using?
cheers
Terry
It goes up over 80kHz with absolutely flat response - both shown in the Audio Precision report it comes with, and in my testing.Terry Demol said:
No.
ARTA, Visual Analyzer, and RMAA. See earlier in the thread for spectra charts output from each, FYI.
Many distractions over the past couple of days, but some may find the distractions of interest so I figured I'd summarise what I've been up to...
Evelyn and I attempted to do the front panel art for the amp this morning, using a stencil and silver paint. The result was not that great - with much effort we could probably eventually get to a point where we had a decent result, but it would be fiddly:
So I've decided to trial other approaches, firstly creating a simple "front panel" for the new dummy load / voltage divider. My goal was to find an approach that allows me to do all the art on the PC, with minimal fuss in transferring that onto the actual front panel. First issue then was finding suitable software. Thus far I've only completed a simple panel in one piece of software, which is Visio. Here's what the result looks like, which shows drilling marks, plus front panel art, for the dummy load:
Frankly, this simple result took a lot longer than it should have - Visio looks like a powerful tool, but it's not very intuitive, and behaves in very odd ways when it comes to scaling groups, snapping shapes to each other, etc. It also doesn't have (AFAICT) some fairly basic symbols, such as drilling marks - the ones on the panel above I had to draw manually.
I've also looked at the software at frontpanelexpress.com, but it's very limited in terms of creating printed artwork - it's focussed on creating files for feeding into their engraving/drilling/etc tools.
Most hopeful is FrontDesigner, at http://www.abacom-online.de/uk/html/frontdesigner.html . I have thus far held back on purchasing it, since I wanted to see if I could get by with software I already have. However I have used the demo (save/print disabled), and it's dead easy to use. I think I'll splash out the small amount of money they're asking - unless someone here can suggest something as good or better...
OK, so that's the artwork generation piece of the puzzle. Next is transferring it onto the panel. My first (and so far only) attempt is to simply ink-jet print onto white glossy adhesive "sticker paper". The printout looks great, although I haven't yet tried to actually stick it onto the front panel. I'll wait until Evelyn is up to do that - she's much better than me at that kind of thing! Once it's stuck on, I'll look at sealants/covers. I bought some clear resin today, which apparently can create a great finish (described on the pack as "equivalent to 50 coats of urethane"). I suspect something like that is a good idea, since otherwise the panel will get damaged over time.
I've also ordered some clear adhesive printable film, which will result in similar results to the "sticker paper", but will allow the metal underneath to show through (and will be more robust, so may not need a sealant).
The problem with both of the above methods is that the finish ends up being rather plasticy. However, I'm very hopeful about another technique, which is to use Lasertran. I read about this at this great page, which shows three methods for front panel artwork - check out the results at the bottom to see how nicely the Lasertran approach works. Here's an example from there:
So that's where I'm up to... If anyone has any other ideas for easy and inexpensive ways to do front panels, please let me know! I hope some folks find the above information of some use. 🙂
Evelyn and I attempted to do the front panel art for the amp this morning, using a stencil and silver paint. The result was not that great - with much effort we could probably eventually get to a point where we had a decent result, but it would be fiddly:
- Sealant required to ensure the paint sticks
- Stencil must be stuck down very carefully to ensure no paint leakage
- Stencil must be cut carefully to match artwork
- Application of paint requires careful masking, and multiple coats
So I've decided to trial other approaches, firstly creating a simple "front panel" for the new dummy load / voltage divider. My goal was to find an approach that allows me to do all the art on the PC, with minimal fuss in transferring that onto the actual front panel. First issue then was finding suitable software. Thus far I've only completed a simple panel in one piece of software, which is Visio. Here's what the result looks like, which shows drilling marks, plus front panel art, for the dummy load:
Frankly, this simple result took a lot longer than it should have - Visio looks like a powerful tool, but it's not very intuitive, and behaves in very odd ways when it comes to scaling groups, snapping shapes to each other, etc. It also doesn't have (AFAICT) some fairly basic symbols, such as drilling marks - the ones on the panel above I had to draw manually.
I've also looked at the software at frontpanelexpress.com, but it's very limited in terms of creating printed artwork - it's focussed on creating files for feeding into their engraving/drilling/etc tools.
Most hopeful is FrontDesigner, at http://www.abacom-online.de/uk/html/frontdesigner.html . I have thus far held back on purchasing it, since I wanted to see if I could get by with software I already have. However I have used the demo (save/print disabled), and it's dead easy to use. I think I'll splash out the small amount of money they're asking - unless someone here can suggest something as good or better...
OK, so that's the artwork generation piece of the puzzle. Next is transferring it onto the panel. My first (and so far only) attempt is to simply ink-jet print onto white glossy adhesive "sticker paper". The printout looks great, although I haven't yet tried to actually stick it onto the front panel. I'll wait until Evelyn is up to do that - she's much better than me at that kind of thing! Once it's stuck on, I'll look at sealants/covers. I bought some clear resin today, which apparently can create a great finish (described on the pack as "equivalent to 50 coats of urethane"). I suspect something like that is a good idea, since otherwise the panel will get damaged over time.
I've also ordered some clear adhesive printable film, which will result in similar results to the "sticker paper", but will allow the metal underneath to show through (and will be more robust, so may not need a sealant).
The problem with both of the above methods is that the finish ends up being rather plasticy. However, I'm very hopeful about another technique, which is to use Lasertran. I read about this at this great page, which shows three methods for front panel artwork - check out the results at the bottom to see how nicely the Lasertran approach works. Here's an example from there:
An externally hosted image should be here but it was not working when we last tested it.
So that's where I'm up to... If anyone has any other ideas for easy and inexpensive ways to do front panels, please let me know! I hope some folks find the above information of some use. 🙂
Jeremy,
I had some luck using materials from this supplier:
http://www.pulsarprofx.com/DecalPRO/index.html
Restoration example:
http://www.pulsarprofx.com/decalpro/Vertical/2_EXAMPLES/2c_Customer_Photos/Ham_Radio/Ham_Radio.html
I had some luck using materials from this supplier:
http://www.pulsarprofx.com/DecalPRO/index.html
Restoration example:
http://www.pulsarprofx.com/decalpro/Vertical/2_EXAMPLES/2c_Customer_Photos/Ham_Radio/Ham_Radio.html
jp_howard said:
Jeremy
I have always found it a lot easier to do the front panel separate from the case/box etc
This way its a lot easier to work on & you can choose a different material than the actual enclosure. I have used both plastic & aluminum extrusions & pieces of sheet to good effect both painted, stenciled & engraved.
I have also used front panel-express although it can get a bit expensive
but the results are very good.
Below is a Panel i did with Front Panel Express a few years ago.
An externally hosted image should be here but it was not working when we last tested it.
Mick
Very interesting - thanks for the link. After doing a bit of reading about this, I see that Lasertran is a "water slide transfer" whereas Decal Pro is a "dry transfer". Waterslide transfers can use all the colours of your laser printer, so they're good for more complex coloured graphics. However dry transfers can be done in metal foils, so are good for single-coloured panels, particularly where a metal 'colour' is desired.dudaindc said:
With Lasertran, you definitely have to do the front panel separately, since the panel has to go into the oven to get bonded - you wouldn't want to put your whole amp in there! 😱mickyh said:
- Status
- Not open for further replies.