Menu

Show posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Show posts Menu

Messages - jfetter

#16
Excellent work, proud of you my man.
#17
Preamps and Effects / Re: dspic-codec
December 31, 2011, 10:58:40 AM
Here is an example of a one knob super tone control. aka STONE control...
The control voltage is reduced to 64 discrete points and the tone MACs use the multipliers as "gain" for lo mid and hi. The center position gives an enormous scoop. It sounds very musical.
Adjustments can be made for LOW and HI too. What I like is the soldering iron is never plugged in.

volatile fractional STONE[64][3]={
    {Q15(0.570),  Q15(0.1),   Q15(0.235)},   //00
    {Q15(0.560),  Q15(0.1),   Q15(0.205)},   //01
    {Q15(0.545),  Q15(0.1),   Q15(0.195)},   //02
    {Q15(0.540),  Q15(0.1),   Q15(0.175)},   //03
    {Q15(0.535),  Q15(0.1),   Q15(0.165)},   //04
    {Q15(0.530),  Q15(0.1),   Q15(0.170)},   //05
    {Q15(0.525),  Q15(0.1),   Q15(0.175)},   //06
    {Q15(0.520),  Q15(0.1),   Q15(0.180)},   //07
    {Q15(0.515),  Q15(0.1),   Q15(0.185)},   //08
    {Q15(0.510),  Q15(0.1),   Q15(0.190)},   //09
    {Q15(0.505),  Q15(0.1),   Q15(0.195)},   //10
    {Q15(0.500),  Q15(0.1),   Q15(0.200)},   //11
    {Q15(0.495),  Q15(0.1),   Q15(0.205)},   //12
    {Q15(0.490),  Q15(0.1),   Q15(0.210)},   //13
    {Q15(0.485),  Q15(0.1),   Q15(0.215)},   //14
    {Q15(0.480),  Q15(0.1),   Q15(0.220)},   //15

    {Q15(0.475),  Q15(0.1),   Q15(0.225)},   //00
    {Q15(0.470),  Q15(0.1),   Q15(0.230)},   //01
    {Q15(0.465),  Q15(0.1),   Q15(0.235)},   //02
    {Q15(0.460),  Q15(0.1),   Q15(0.240)},   //03
    {Q15(0.455),  Q15(0.1),   Q15(0.245)},   //04
    {Q15(0.450),  Q15(0.1),   Q15(0.250)},   //05
    {Q15(0.445),  Q15(0.1),   Q15(0.255)},   //06
    {Q15(0.440),  Q15(0.1),   Q15(0.260)},   //07
    {Q15(0.435),  Q15(0.1),   Q15(0.265)},   //08
    {Q15(0.430),  Q15(0.1),   Q15(0.270)},   //09
    {Q15(0.425),  Q15(0.05),   Q15(0.275)},   //10
    {Q15(0.420),  Q15(0.05),   Q15(0.280)},   //11
    {Q15(0.415),  Q15(0.025),   Q15(0.285)},   //12
    {Q15(0.410),  Q15(0.025),   Q15(0.290)},   //13
    {Q15(0.405),  Q15(0.001),   Q15(0.295)},   //14
    {Q15(0.400),  Q15(0.001),   Q15(0.300)},   //15

    {Q15(0.395),  Q15(0.001),   Q15(0.305)},   //00 mid
    {Q15(0.390),  Q15(0.001),   Q15(0.310)},   //01
    {Q15(0.385),  Q15(0.025),   Q15(0.315)},   //02
    {Q15(0.380),  Q15(0.025),   Q15(0.320)},   //03
    {Q15(0.375),  Q15(0.05),   Q15(0.325)},   //04
    {Q15(0.370),  Q15(0.05),   Q15(0.330)},   //05
    {Q15(0.365),  Q15(0.1),   Q15(0.335)},   //06
    {Q15(0.360),  Q15(0.1),   Q15(0.340)},   //07
    {Q15(0.355),  Q15(0.1),   Q15(0.345)},   //08
    {Q15(0.350),  Q15(0.1),   Q15(0.350)},   //09
    {Q15(0.345),  Q15(0.1),   Q15(0.355)},   //10
    {Q15(0.340),  Q15(0.1),   Q15(0.360)},   //11
    {Q15(0.335),  Q15(0.1),   Q15(0.365)},   //12
    {Q15(0.330),  Q15(0.1),   Q15(0.370)},   //13
    {Q15(0.325),  Q15(0.1),   Q15(0.375)},   //14
    {Q15(0.320),  Q15(0.1),   Q15(0.380)},   //15

    {Q15(0.315),  Q15(0.1),   Q15(0.385)},   //00
    {Q15(0.310),  Q15(0.1),   Q15(0.390)},   //01
    {Q15(0.305),  Q15(0.1),   Q15(0.395)},   //02
    {Q15(0.300),  Q15(0.1),   Q15(0.400)},   //03
    {Q15(0.295),  Q15(0.1),   Q15(0.405)},   //04
    {Q15(0.290),  Q15(0.1),   Q15(0.410)},   //05
    {Q15(0.285),  Q15(0.1),   Q15(0.415)},   //06
    {Q15(0.280),  Q15(0.1),   Q15(0.420)},   //07
    {Q15(0.275),  Q15(0.1),   Q15(0.425)},   //08
    {Q15(0.270),  Q15(0.1),   Q15(0.430)},   //09
    {Q15(0.265),  Q15(0.1),   Q15(0.435)},   //10
    {Q15(0.260),  Q15(0.1),   Q15(0.440)},   //11
    {Q15(0.275),  Q15(0.1),   Q15(0.445)},   //12
    {Q15(0.285),  Q15(0.1),   Q15(0.460)},   //13
    {Q15(0.295),  Q15(0.1),   Q15(0.470)},   //14
    {Q15(0.325),  Q15(0.1),   Q15(0.480)}   //15    16 X 4=64
};

then in the DMA handler
// do the macs
        for(i=0; i<16;i++){
        ACMA = __builtin_clr();
        ACMA = __builtin_mac(ACMA,LOW, lowpassout, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL);
        ACMA = __builtin_mac(ACMA,MID, midpassout, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL);
        ACMA = __builtin_mac(ACMA,HI, hipassout, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL);
        sumadd = __builtin_sacr(ACMA,-3);
        }
#18
Preamps and Effects / Re: dspic-codec
December 30, 2011, 03:36:41 PM
I!m bettin devaluation of usd very soon. Getting speakers will be a problem so a head is probably the way to go.
#19
Preamps and Effects / Re: dspic-codec
December 29, 2011, 09:24:39 PM
Only plans are a pedal or head, maybe a combo. (I want to have something ready for production when the cheap imports go away). I will try to get a clip posted this weekend. Lost three days troubleshooting a nasty sound on one string. Turned out to be my headphone amp. I could not see it on scope so should have known. I'm now working on the WET-DRY mix having a lot of fun with it.
Its using the dspic33fj256gp710, wolfson WM8731 and/or ti TLV320AIC codec. I have 4gb sd flash card for reverb tank. The flash should last several years with leveling(see microsoft "readyboost" http://blogs.msdn.com/b/tomarcher/archive/2006/04/14/576548.aspx ). DMA is utilized with the codec so processor has about 20% loading. The circuit itself is pretty simple. The codec takes care of anti-alias.
Schematic attached..
As mentioned I now do all work on OSX
#20
Preamps and Effects / dspic-codec
December 23, 2011, 12:50:48 PM
I finally got the dspic preamp board done. Is any one else doing 16bit (_Q15) dsp?
It sounds great with no noise whatsoever. Running three filters now low,mid,hi.
Not much on the internet or from Microchip on pro-audio apps. Some of the old Analog Devices code works. The 16 bit works great for this app.

I need a good integer q15 reverb and a soft limiter.

This board is not a finished gadget. It is only for developing dsp code. Has six control voltages. Audio path is 100% digital.
#21
Amplifier Discussion / Re: Class D Amp Recommendations?
September 08, 2009, 10:15:20 AM
I'm giving it a go with the TPA3106D1 and dsPIC33 for a combo unit. The dsp will let me do in software what I've been trying to accomplish in analog. The target is a smooth overdive with mild reverb. The dspic33fj128mc804 has a 16bit dac and 12 bit A2D. I'm planning on trying oversampling to get 14bit. We'll see.

Running on 24vdc I guess about 35w.

Strange no one is using this part.

#22
Amplifier Discussion / Re: Class A solid state
September 04, 2009, 11:02:51 AM
Quote from: phatt on January 15, 2009, 09:31:52 AM
Quote from: Boprikov on January 15, 2009, 05:32:52 AM
http://www.vintagehofner.co.uk/selmer/schematic/ssmerc.html

This is an interesting vintage design. If i understand correctly, the primary of the output transformer must stand 400 mA. I have no idea what the impedance ratio should be. Is it possible to somehow estimate it? And what should the power supply voltage be? I think 10-20 Volts is not enough as efficiency is poor in class A.

Sorry my english... but i´m quite interested about what type of output transformer was used at those SS Selmers.

Is that a "70V" line transformer? You think about 400ma bias idle current is through primary? That is about 30w to dissipate at idle with 35v supply.
Hi Boprikov,
I searched years back trying to find such a circuit and you found one,, Good find.
Here is one I built some years back similar in concept to Selmer Mercury.
I never went on to complete it because of the bass issue but insanely loud even though it was driving a small 100mm speaker.
I have no idea where the OT came from, just found it in some trash.
The output Tr has to be a fairly high voltage as the collector voltage can get quite high when using transformer coupling.
Q3 only needs a clip-on heat sink as it only runs lukewarm.

Jfetter, I do wonder about Exotic transistors and all that heat ,, my thoughts are go with transfomer coupling as it puts a lot less heat into the Amp,,heck line transformers are easy found.  Teemu is on the mark there,,
just my take on it. :)
Phil,  Nambour Australia
#23
Amplifier Discussion / Re: Pics of my last build
February 19, 2009, 04:46:00 PM
Thats a nice build!
#24
Do you think they may be a NOS buyout (new old stock)?


Quote from: dsmnoisemaker on January 22, 2009, 01:48:55 PM
Quote from: jfetter on January 22, 2009, 11:42:59 AM
I attempted to use them but found they were difficult to obtain. They also did not seem stable as far as future availability.

Well here in chile seems to be plenty of them and they are pretty cheap..in fact LM3886/4780 are very hard to get, and they cost the same or more than a STK chip of the same power..

It seems that the main reason for not using STK is availability..and not quality..but i´m not planning to make them massively..just 1 or 2 of them..
#25

What distributor or seller do you use?

Quote from: dsmnoisemaker on January 22, 2009, 01:48:55 PM
Quote from: jfetter on January 22, 2009, 11:42:59 AM
I attempted to use them but found they were difficult to obtain. They also did not seem stable as far as future availability.

Well here in chile seems to be plenty of them and they are pretty cheap..in fact LM3886/4780 are very hard to get, and they cost the same or more than a STK chip of the same power..

It seems that the main reason for not using STK is availability..and not quality..but i´m not planning to make them massively..just 1 or 2 of them..
#26
I attempted to use them but found they were difficult to obtain. They also did not seem stable as far as future availability.
#27
Amplifier Discussion / Re: Class A solid state
January 16, 2009, 08:10:24 AM
that is an interesting design,
i was thing about zero watt idle by capacitive coupling the drain (or collector npn / emitter pnp) to PS and biasing using a "calibration resistor" which is then removed and hence a virtual 1/2 PS. within one half audio period the circuit is running true to the load. no pop or thump but no wast heat at all. A "cold" class a amp. I'm looking at the 2n297A and the ruskis either single or parallel. PS would be 30-40v. no feedback. not much dampening either.




Quote from: Boprikov on January 15, 2009, 05:32:52 AM
http://www.vintagehofner.co.uk/selmer/schematic/ssmerc.html

This is an interesting vintage design. If i understand correctly, the primary of the output transformer must stand 400 mA. I have no idea what the impedance ratio should be. Is it possible to somehow estimate it? And what should the power supply voltage be? I think 10-20 Volts is not enough as efficiency is poor in class A.

Sorry my english... but i´m quite interested about what type of output transformer was used at those SS Selmers.
#28
Amplifier Discussion / Re: Class A solid state
January 04, 2009, 09:38:43 PM
I'm thinking about using a RTD sensor on the load resistor for fast RMS detection.
Google and found this hifi example using the Russ T813 Ge.

http://www.hifi-forumi.com/forum/viewtopic.php?f=18&t=10229&st=0&sk=t&sd=a

I think I'll be able to do a dsPIC bias driver that will keep the waste heat down.




#29
Amplifier Discussion / Re: Class A solid state
December 17, 2008, 09:35:12 AM
yes the 70v output transformer is an option. I have a CBS champ and have looked at possibly using a IGBT or HV mosfet just for testing the concept. It has the se output transformer. The (genuine)class a sound is stunning.

I was thinking about using a automatic 'idle' bias ckt to reduce pwr dissipation when not running. Idle i think is about 30-40% of time on stage. 

Also thinking the input to final should be xfmr coupled.

By the way, on my old TDA2050 design, the air core first order output filter IS the way to go. Clipping sounds sweet. I used i think 330uH into 8 ohm. second order may be better don't know.
#30
Amplifier Discussion / Class A solid state
December 15, 2008, 05:56:23 PM
hi im back....

doing a new design, i once biased a germanium doorknob power resistor 1/2ps and ran simple class a. it has a special sound for sure. ebay has lots of russian germanium transistors.
the physics may be more complex then simple spice circuit analysis. not sure but it sounded really good. problem is dumping all that heat. my test was a 10 watt unit, gets too hot.

has anyone tried ss class a guitar? ive studies the jlh and variants , think simple class a is best. mosfets better?

ps ive switched to eagle layout.