advertisement


AD815 preamp build thread

Very smart stuff Avi. Using the "left" side of the ad815 as a buffer, and the "right" as the preamp with a shunt-to-ground attenuator. Very cool :)

Where did you get your 1nF polystyrene caps?

Regards,
Carl
 
Very smart stuff Avi. Using the "left" side of the ad815 as a buffer, and the "right" as the preamp with a shunt-to-ground attenuator. Very cool :)

Where did you get your 1nF polystyrene caps?

Regards,
Carl

The Caddock resistor between the two halfed AD815 is threefold purpose ;)
a- output resistor for the first buffer
b- upper resistor for the stepped attenuator divider
and c- An input RC filter for the second half gain stage (with the help of the small input capacitance of the AD815 this avoid RF getting into the gain stage (of course one can add another a small (few pF) polystyrene cap at the gain input).

Polystyrene is from RS of Farnell:
http://il.farnell.com/9520074/passives/product.us0?sku=lcr-components-fsc-160v-1000pf-2-5
 
Is the value of the 15K resistor determined by the impedance of the pot?

Yep, but you can set this resistor according the volume you need (depends on your loudspeaker sensitivity) (I use 15K with stepped attenuator of 100R-4k). any value between some 5K to 30K is reasonable.
 
Does using the buffer before the pot affect / improve the DC offset seen at the final output?

No, DC offset depends in the input impedance difference between the positive and negative inputs of the bipolar op-amp. one could lower DC offset by reducing the 15K shunt to say 1K, using lower value resistors in the stepped attenuator. (without the buffer you will not able to use such low impedance volume potentiometer/attenuator).

Regards,

Avi
 
No, DC offset depends in the input impedance difference between the positive and negative inputs of the bipolar op-amp.

Right, so is that the difference in impedance between each of the inputs and ground?

one could lower DC offset by reducing the 15K shunt to say 1K, using lower value resistors in the stepped attenuator.

I realise that the offset is varies according to the position of the volume control, as this varies the impedance. But wouldn't it be possible to lower the typical offset by using higher resistor values in the feedback loop of the right-hand buffer to bring the impedance up to the same order as that of the attenuator?

Dr. Matt...
 
wouldn't it be possible to lower the typical offset by using higher resistor values in the feedback loop of the right-hand buffer to bring the impedance up to the same order as that of the attenuator?

Dr. Matt...

It's not recommended to change Rf in the buffer (or Gain) in the AD815 (Specs recommends about 500R).
I don't consider a small DC offset a problem since power-amp usually have a capacitor at the input.

Avi
 
Hi Avi

I like your idea of using two boards that way. I should have taken one more.

What's surprinsing me, is that your input buffer is stable with an 1nf cap connected the way it is. I can make my setup oscillate spontaneously with just holding a 220pF mica cap across the pulldown resistor, R2 or R7. At +-12V psu, I see a nice 210Mhz oscillation of 70mV peak to peak amplitude. It's measured with a 50R output resistor, a x1 probe and the 50R inbuilt feature of my scope enabled, wich is a 400Mhz BW one. Does the AD815 perform differently at unity gain and with your decoupling scheme ? I don't know. I thought base stoppers were mandatory.

Happy listening,
Francis
 
Hi Francis,

I'm using only +- 10.5v PSU voltage (lower PSU means less prone to oscillations usually).
Sorry but I don't understand why an extra capacitance at the non-inverting input (not the output) had to influence the OP-amp regarding oscillations. can you explain more about this point?

Regards,
Avi
 
Hi Avi,

Thanks for answering, but I can't explain. It was a surprise for me too. I've reached my limits with such high speed opamps.
Btw, using lower psu voltage is safer and I couldn't hear a difference with my headphones.

Best regards,
Francis
 
It seems I get far too little time to actually work on things these days, but I've got a small update - a working PSU ;)

It's the same physical footprint as the preamp case and is properly fused and switched. Safety first! The idea with this setup is to minimise power consumption, so to that end I've rigged up the main PSU transformer (an Avondale EI core) via a relay. The relay is triggered by the Jos Eindhoven attenuator board which is powered by a very tiny 6V supply that's been recycled from an old Nokia phone charger. When the main switch is on at the back and the system has been powered down by the attenuator (via remote control or by holding the knob in for 5 seconds), the relay is unpowered which means the 115V supply for the Avondale trafo is disconnected and the PSU consumes only a few mA.

When the power is switched on using the remote or knob, the relay kicks in and 115V is supplied to the Avondale trafo. This in turn powers the superregs which power the AD815. Powered up in this state it consumes 60mA at +/-15V on top of the attenuator's power requirements.

The really cool thing with this is I've setup a spare relay point with which I'm going to send a slave signal to my monoblocks. This will mean I can power everything on/off from the pre :)

Here's the PSU as it is just now. I've left room in there for more transformers to power the DAC circuitry that'll be getting added to the preamp later:
psu2.jpg


Here's the little 5V relay that switches the 115V supply:
psu3.jpg


Everything's designed with safety in mind. Once I'm finished with it there will be no exposed mains. In fact, the only exposed mains just now is on the side of the transformer, but I'm going to fashion a cover for that.
psu5.jpg


Power is supplied to the preamp by 2 7-pin locking connectors:
psu4.jpg


The Jos attenuator comes with its own rectifier/smoothing and expects a 6VAC supply, but I've bypassed that and soldered in the 6VDC supply from the phone charger. The 5V relay output uses the pink/white wires and goes out from the pre and back into the PSU:
psu6.jpg


Next up: build some power amps!

Carl
 
Hi All,

I got my new AD815 today and quickly got it soldered onto the board. You might recall I found what looked like a short between layers (or something) that was putting + supply voltage (or thereabouts) onto the cap 8 pin.

I don't think that was the (only) problem. I had 14VDC on the Output No 2 (Cap 7). I remeasured the 815 pins and got Pin 13 (15.6V) OK, and also pins 1-8 and 17-24, as the +Vs is connected to the thermal heat pads and they, in turn, are connected to the NC 1-4 and 21-24 pins on the PCB. What really confused me was the continuity test - OK between pins 13-16 (i.e. no connection). So where were these extra trons coming from? :confused:

I desoldered the 815 and checked the voltage on the pads. Pad 13 was OK and the rest seemed to measure fine, so I can only assume that there is another short either in mounting a component or in the pcb. I removed a few of the caps and smd resistors and all remained the same. Plopped the 815 on and tacked Pin 13 and the voltages reappeared on Pins 14-16 (Out 2, In -2, In +2). I've pored over the pcb drawings and just can't find a possible explanation.

So, the board is now desoldered (30 secs with the SMD tweezers) and I'm going to populate a new board. Before I do, I'd be grateful for any thoughts - If it was something I did, I don't want to make the same mistake again! ;)

Cheers

Jon
 
I was tidying up in the workroom today, and found this project boxed up. In it were the recycled parts and a new board. I think I'll have another stab at it.

Can someone just remind me if the BOM is the same as Carlos FM's original diagram?

Many thanks!!

Jon
 
H, Jon

Nice find :) I popped the BOM and schematic in a email just now. Let me know if you don't get it.

Cheers,
Carl
 
Hi Jon. I got one of these too (thanks Carl) so I'm very interested to hear how you get on. What amps will you be driving with this pre?

Keep us posted man :)
 
It's a pretty sweet pre. There are two mods I'd very much like to try, if I ever get the time.

1. Add a VBE to the regulator inputs so that the 317/337s don't have to contend with lots of HF noise.

2. Use large film caps for C2 and C4, instead of electrolytics. Wima MKS2 10uF fit... just. However, having a large film cap hanging off a 317 or 337 is a recipe for disaster, so to make this mod work you could use a 0R5 (two 1R in parallel) instead of wire links at LK2 and LK3. This would add just enough series R to keep the regs happy, while keeping lots of high quality C right at the opamp pins :)

I think the combination of those two mods should do wonders for the sound quality.
 
One thing I wish i'd tried with mine, was to replace the SMD feedback resistor with a Vishay dale. I know through experience that the feedback resistors in a Naim pre-amp can make or break it (to my ears anyway)
 
Hi Jon. I got one of these too (thanks Carl) so I'm very interested to hear how you get on. What amps will you be driving with this pre?

Keep us posted man :)

Hi Mike,

Currently ARCAM D290P (fully serviced, lightly modded and soon for sale), then MyRefCs when I have them cased up.

I have a variable TR power supply I can use for this, so was thinking about proving the op amp section then adding on-board regulation (or building another couple of TR) later on.

It would be nice to here this project in the system! :D

Cheers

Jon
 


advertisement


Back
Top