Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 16, 2009 9:52:41 GMT
Class A Preamp-DC coupled , with JLH. N.B. The JLH uses Current Limiting , set at approx. 600mA. 2 Offset Correctors are required.(1 for each channel.) A Jaycar KC5418 Headphone Power Supply kit may be used to supply power , preferably with resistor values tweaked to give +-15.5 to +-16V. (e.g. a 2K2 resistor soldered across each 100 ohms MF resistor) Similar modifications should be made as in the Jaycar HA thread. A 15-0-15VAC 30VA Toroidal transformer may be used to supply power, but should be either in a separate case or well shielded from the rest of the preamp. The function of the series 1N5819 (now SF12, see later schematic) and adjusted 500 ohm trimpot, is to drop the same voltage as the differences in voltages shown in Blue (from my own Class A preamp)The vertically mounted 1N5819 provides thermal tracking, so as to keep the collector voltages of the differential pair transistors within close limits. URL=http://imageshack.us/photo/my-images/40/akoffsetcorrectorrevise.jpg/]  [/URL] |
|
robertkd
Been here a while! 
Electronics Engineer from sunny Queensland
Posts: 111
|
Post by robertkd on May 16, 2009 11:55:18 GMT
mmm interesting what sort of preamp is this?
and if your wanting to go the extra distance and get 16V @ 600mA from the PSU I would suggest to got to the dual 18V transformer and some serious heat sinking.
Robert
(edited by mistake)
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 16, 2009 12:03:42 GMT
Robert
Not necessary to go any higher. This is essentially my Class A preamp, without input switching and delay switch on , as it is in this guise only driving headphones. It is virtually identical to what Leo posted in his photo.
Bias is set at approx 100mA-110 per channel. The 600mA current limit of the JLH ensures a slightly quicker rise to full voltage without normally having switch on thumps. It also gives a little lower impedance from the JLH than with a lower current limiting setting.
Alex
|
|
robertkd
Been here a while!
Electronics Engineer from sunny Queensland
Posts: 111
|
Post by robertkd on May 16, 2009 12:21:17 GMT
Alex
I was wondering about the 600mA reference as that wouldn't work so well you should probably get away with dual 15V secondaries at only 100mA.
but to remove any chance of regulator drop out would recommend 18V secondaries for headroom.
Robert
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 16, 2009 12:33:24 GMT
Alex I was wondering about the 600mA reference as that wouldn't work so well you should probably get away with dual 15V secondaries at only 100mA. but to remove any vhance of regulator drop out would recommend 18V secondaries for headroom. Robert Robert The 30VA will do 1A each winding. Output voltage before regulation is a little over 21V. The transformer is during use quite lightly loaded at <250mA per supply rail.The suggested parallel 2K2 across the 100R in the dual reg. supply gives around +-15.6V. This basic preamp has been in use for many years, but with occasional updating, and never any problems. Alex |
|
leo
Been here a while!
Team wtf is it?
Posts: 3,638
|
Post by leo on May 16, 2009 12:36:58 GMT
This thing still outclasses the SCHA by quite a lot imo when used as a HA The input pair mods are really needed tbh to get the most out of it Have you built yours yet Fritz?? come on man get your finger out  |
|
Will
Been here a while!
Ribena abuser!
Member since 2008
Posts: 2,164
|
Post by Will on May 16, 2009 13:15:49 GMT
I've got about a third of the components in place, but need to etch a PSU board, JLH and a DC corrector as well.
It'll get there soon, I hope.
Would a 15-0-15 15VA traffo per board be OK?
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 16, 2009 13:25:10 GMT
I've got about a third of the components in place, but need to etch a PSU board, JLH and a DC corrector as well. It'll get there soon, I hope. Would a 15-0-15 15VA traffo per board be OK? Will How are you going to work them in with a single JLH ? You could of course use twin PSUs and twin JLHs, but I doubt there would be any really noticeable SQ improvement. Alex P.S. Initially you could get it working without offset correctors by using 4.7uF input caps and 100-220uF feedback electrolytic capacitors where there are links fitted in this version . |
|
Will
Been here a while!
Ribena abuser!
Member since 2008
Posts: 2,164
|
Post by Will on May 16, 2009 13:38:57 GMT
True, Alex. My idea is to build as a dual mono, each board with it's own PSU and JLH PSRR, partly because it appeals to my OCD. Thats the end result, but I'll get it working on a single JLH PSRR/PSU first. Mentioning JLH PSRR, Jon Clancy and I have re-done version coming soon, which includes the additional resistor that was originally missed out. Full detail in the JLH PSRR for sale thread!!  ;D |
|
leo
Been here a while!
Team wtf is it?
Posts: 3,638
|
Post by leo on May 16, 2009 13:53:59 GMT
BTW, tried HD650's with this amp  don't match at all in my opinion, its like having cotton wool stuffed in your ears |
|
robertkd
Been here a while!
Electronics Engineer from sunny Queensland
Posts: 111
|
Post by robertkd on May 16, 2009 14:02:05 GMT
Alex I was wondering about the 600mA reference as that wouldn't work so well you should probably get away with dual 15V secondaries at only 100mA. but to remove any vhance of regulator drop out would recommend 18V secondaries for headroom. Robert Robert The 30VA will do 1A each winding. Output voltage before regulation is a little over 21V. The transformer is during use quite lightly loaded at <250mA per supply rail.The suggested parallel 2K2 across the 100R in the dual reg. supply gives around +-15.6V. This basic preamp has been in use for many years, but with occasional updating, and never any problems. Alex well that would be a design choice if you do the maths and feel the input ripple is ok than fine, however I don't see the point of falling back to 15.6V over 15V?? The 1A secondary (let face it it's in series) makes little difference given the loss in the diodes means your scraping the bucket at some 19.5V into the regulators, you may measure 21V average but have a look with a cro, all I am saying is it's border line at best and for 16V you have little to no head room. Robert |
|
robertkd
Been here a while!
Electronics Engineer from sunny Queensland
Posts: 111
|
Post by robertkd on May 16, 2009 14:04:03 GMT
This thing still outclasses the SCHA by quite a lot imo when used as a HA The input pair mods are really needed tbh to get the most out of it Have you built yours yet Fritz?? come on man get your finger out Leo, perhaps you haven't tried the right modifications yet Robert |
|
Will
Been here a while!
Ribena abuser!
Member since 2008
Posts: 2,164
|
Post by Will on May 16, 2009 14:07:13 GMT
If anyone is interested, I have an as-yet un-tested PSU layout to etch, for a dual rail PSU.
Basically the jaycar PSU board, but with variable output and a position for the traffo tuning cap on the input of the board.
|
|
leo
Been here a while!
Team wtf is it?
Posts: 3,638
|
Post by leo on May 16, 2009 14:20:33 GMT
This thing still outclasses the SCHA by quite a lot imo when used as a HA The input pair mods are really needed tbh to get the most out of it Have you built yours yet Fritz?? come on man get your finger out Leo, perhaps you haven't tried the right modifications yet Robert Perhaps what you got in mind? BTW if you read all my posts on the SCHA thread you'll see I have not been lazy with trying various things out already  |
|
leo
Been here a while!
Team wtf is it?
Posts: 3,638
|
Post by leo on May 16, 2009 14:23:43 GMT
If anyone is interested, I have an as-yet un-tested PSU layout to etch, for a dual rail PSU. Basically the jaycar PSU board, but with variable output and a position for the traffo tuning cap on the input of the board. That would be nice Will, I'm arsing about with soo much at the moment its hard trying squeeze everything in the small free time  Anyway, theres a lot I want to try out with this thing yet, a suitable regulated supply is certainly one of the things |
|
robertkd
Been here a while!
Electronics Engineer from sunny Queensland
Posts: 111
|
Post by robertkd on May 16, 2009 14:34:49 GMT
Leo, perhaps you haven't tried the right modifications yet Robert Perhaps what you got in mind? BTW if you read all my posts on the SCHA thread you'll see I have not been lazy with trying various things out already indeed and all my cards have been laid out here (mostley) rockgrotto.proboards.com/index.cgi?board=chitchat&action=display&thread=3622&page=7still need to get +/-20 to 22V on the collectors of the output devices mind you this FET module is sounding pretty damn good as well, will have to do some Noise and Distortion tests on them when I can get hold of the LinDos again  Robert |
|
Will
Been here a while!
Ribena abuser!
Member since 2008
Posts: 2,164
|
Post by Will on May 16, 2009 16:47:44 GMT
If anyone is interested, I have an as-yet un-tested PSU layout to etch, for a dual rail PSU. Basically the jaycar PSU board, but with variable output and a position for the traffo tuning cap on the input of the board. That would be nice Will, I'm arsing about with soo much at the moment its hard trying squeeze everything in the small free time Anyway, theres a lot I want to try out with this thing yet, a suitable regulated supply is certainly one of the things For general use by RG members: rockgrotto.proboards.com/index.cgi?action=display&board=m&thread=4201&page=1Feel free to critique, in the PSU thread. Hopefully, this'll help those building this amp as well. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 17, 2009 3:01:29 GMT
Robert The 30VA will do 1A each winding. Output voltage before regulation is a little over 21V. The transformer is during use quite lightly loaded at <250mA per supply rail.The suggested parallel 2K2 across the 100R in the dual reg. supply gives around +-15.6V. This basic preamp has been in use for many years, but with occasional updating, and never any problems. Alex well that would be a design choice if you do the maths and feel the input ripple is ok than fine, however I don't see the point of falling back to 15.6V over 15V?? The 1A secondary (let face it it's in series) makes little difference given the loss in the diodes means your scraping the bucket at some 19.5V into the regulators, you may measure 21V average but have a look with a cro, all I am saying is it's border line at best and for 16V you have little to no head room. Robert Robert I suggested 15.6V only to allow for a couple of hundred mV drop in the Current Limiter section of the JLH. Mine is actually adjusted for a little less, to give +-15V after the drop through the CL section.I have had a good look with a CRO at the DC rails. The SF12 diodes that I use also have a little lower forward voltage drop than the 1N4004 that I have tried. BTW, the dropout voltage of the LM317/LM337 is quoted as less than 2V at 500mA, with an ambient temperature of around 13C, and a little lower again at 25C. At around 200mA , the drop out voltage is only approx. 1.75V. I will however recheck the IN/OUT Voltage differential when the opportunity arises. If I ever go higher, it will only be to give added headroom, and it would most likely be at +-20V , as the 15W Class A is running at +-21V from twin dual regulated PSU boards. Alex |
|
robertkd
Been here a while!
Electronics Engineer from sunny Queensland
Posts: 111
|
Post by robertkd on May 17, 2009 8:07:25 GMT
Alex,
what is the purpose of the offset corrector? is it to compensate for thermal drift?
Robert
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 17, 2009 8:34:52 GMT
Alex, what is the purpose of the offset corrector? is it to compensate for thermal drift? Robert Robert No, it is purely to permit DC coupling via a potentiometer/attenuator. Without the FB capacitor ,(and it sounds a tiny bit better without it) the DC offset will vary with the setting of the volume control.The smaller initial time constant helps to smooth out switching steps when using an attenuator, but not really needed for a dual gang potentiometer.The overall frequency response when using the offset corrector was measured by a friend at about +1dB at a couple of HZ. He didn't notice any loss of transparency with it, in comparison to his own Class A preamp.I think that Leo currently has the FB electro in situ without the offset corrector, but no input capacitor, as he hated the SQ deterioration with even a Wima 4.7uF MKS2. The main PCBs are modified PCBs from the SC 15W Class A design.(RCS Radio) Alex |
|
Will
Been here a while!
Ribena abuser!
Member since 2008
Posts: 2,164
|
Post by Will on May 19, 2009 14:06:20 GMT
Just in case anybody needs it, this is a list of components for the SC Class A board. I've left off resistors, trimpots (in-line type) and heatsinks.
I've included the farnell part numbers, so that you can use their descriptions for part sizes, to help source your own.
Order Code Description Mftr. & Part No.
1219466 CAPACITOR, RADIAL, 100UF, 25V PANASONIC - EEUFM1E101
1413944 CAPACITOR, KP, 100PF, 630V, 1% VISHAY - KP1830110631
1519291 CAPACITOR, 68PF, 1000V, 5% WIMA - FKP2 68PF/5/1000/5
1215508 CAPACITOR, 0.1UF, 63V VISHAY BC COMPONENTS - 2222 370 11104
1336533 DIODE, SCHOTTKY, 1A, 40V VISHAY - 1N5819
9844562 DIODE, ZENER, 3.3V, 500MW FAIRCHILD SEMICONDUCTOR - BZX55-C3V3
1467886 TRANSISTOR, PNP, TO-92 FAIRCHILD SEMICONDUCTOR - BC560C
1467885 TRANSISTOR, PNP, TO-92 FAIRCHILD SEMICONDUCTOR - BC559
1467883 TRANSISTOR, PNP, TO-92 FAIRCHILD SEMICONDUCTOR - BC558B
1467869 TRANSISTOR, NPN, TO-92 FAIRCHILD SEMICONDUCTOR - BC547
1017674 TRANSISTOR, NPN, TO-92 FAIRCHILD SEMICONDUCTOR - BC549C
1017668 TRANSISTOR, NPN, TO-92 FAIRCHILD SEMICONDUCTOR - BC337
9846875 TRANSISTOR, PNP, TO-92 FAIRCHILD SEMICONDUCTOR - BC327
1017678 TRANSISTOR, PNP, TO-92 FAIRCHILD SEMICONDUCTOR - BC557B
MPSA18, 2SC5171 and 2SA1930 can be bought from cricklewood electronics.
Remember to order more transistors than you need, as you will need to match them between boards. The BC560C/MPSA18 both need very close matching (hFe and VBE)
Feel free to offer better types for the function they provide.
Hope this helps some.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 19, 2009 22:46:03 GMT
Will
Thanks for the list, really helps ordering. Do you mean all transistors need matching or just BC560C/MPSA18?
Thanks,
Syd
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 19, 2009 23:05:25 GMT
Will Thanks for the list, really helps ordering. Do you mean all transistors need matching or just BC560C/MPSA18? Thanks, Syd Syd For best results, the other corresponding transistors in each channel, should have relatively similar characteristics. If possible, select the 2SA1930 and 2SC5171 o/p devices to have reasonably similar HFE on both output halves. Try not to go much higher in HFE than 500 for the lower transistor in the VAS stage.(BC549C) Also, try and select the BC327 and BC337 to have roughly similar HFE in each half of the channels. i.e. try to avoid using BC327/BC337 with widely different HFE. All these other devices are fairly cheap, so it would pay to get a few extra of each type for selection purposes. Alex |
|
Will
Been here a while!
Ribena abuser!
Member since 2008
Posts: 2,164
|
Post by Will on May 20, 2009 7:42:13 GMT
I must admit I went a step further, and matched resistors for each location between each board! I did this when building my SCHA (I only had the PCB, so had to buy all the components in). Should probably do the diodes as well (do we have a smiley for obsessives?)
What real difference it makes is debatable, but at least the box is ticked for doing the best I could.
I'm nearly complete on installing all the components, bar the transistors (still need to do the matching for them), but I'm getting a better idea of what is what on the pcb (I've no silkscreen to work with). I think that I should be able to update the original artwork to include the LTP balancing mod fairly soon, for those who are thinking on this.
I'll also do a layout drawing if people are interested.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on May 20, 2009 7:57:22 GMT
Will
If you go to a lot of trouble to match transistors, then it is very worthwhile closely matching the 4 front end emitter resistors in each channel. I also do that kind of thing routinely,plus the rest that you said as well, even matching the forward voltages of the LEDs in the SC HA etc.
Alex
|
|
;D
don't match at all in my opinion, its like having cotton wool stuffed in your ears
