[PinkFloyd]

Quote:

Originally Posted by l_simon_l Wow mike, this sounds exciting. What mods are done to your WNA in addition to changing the opamp to AD8065 and getting rid of output and input caps? Zobel Network? Removed lamps? My WNA is currently all stock. Is the difference significant? If so, I just might have to make use of my soldering iron again

Hi Simon,

Zobel fitted, removed lamps (replaced with ferrite inductors or zero ohm links) removed both input and output caps and yes IMO (to my ears) there is quite a vast improvement. The zobel is probably unnecessary with FET opamps so if you just whip out all three LM6171's and remove the output caps you're good to go. I find something like an opa 627 is good in the rail splitter.

Mike.

 

[l_simon_l]

Quote:

Originally Posted by PinkFloyd Hi Simon, Zobel fitted, removed lamps (replaced with ferrite inductors or zero ohm links) removed both input and output caps and yes IMO (to my ears) there is quite a vast improvement. The zobel is probably unnecessary with FET opamps so if you just whip out all three LM6171's and remove the output caps you're good to go. I find something like an opa 627 is good in the rail splitter. Mike.

Thanks! No I just have to find a source for the components... I think I live in the wrong country for DIY

 

[Paco]

Hello all,
The AD8065 also was the best one in the MKI version to me (I was using it without input and output caps). I've finished with the “capless” MKII to use it as a preamp a couple of days ago and I have a question. I've measured TR2 and TR4 and they are at about 100ºC with case open and with an external temperature of about 25ºC. TR1 and TR3 are at about 90ºC. My question is... should I use heatsinks? I've seen that those transistors are rated at 150ºC but I think that once the case is closed and external temperature raise to 30ºC, they will be close to the maximum temperature... what do you think?

Best regards,
Paco

 

[Henry Morgan]

I recently tried the SOIC mounted AD8610 without caps in my WNA Mk2 & felt it was not quite as good as the 6171 with output caps. Tangent on Headwize seems to be able to supply the adapters for SOIC chips either mounted or unmounted. Shame I don't think he uses the AD8065 as I would like to give it a try!

 

[Alick]

Quote:

Zobel fitted, removed lamps (replaced with ferrite inductors or zero ohm links) removed both input and output caps and yes IMO (to my ears) there is quite a vast improvement. The zobel is probably unnecessary with FET opamps so if you just whip out all three LM6171's and remove the output caps you're good to go. I find something like an opa 627 is good in the rail splitter.

Bias current back to original level Mike? I've got the AD sample kit lying in a drawer at work and I'm hoping there's a pair of 8065's in there. I knew if I waited long enough I could ditch the output cap's ...

 

[PinkFloyd]

Quote:

Originally Posted by Alick Bias current back to original level Mike? I've got the AD sample kit lying in a drawer at work and I'm hoping there's a pair of 8065's in there. I knew if I waited long enough I could ditch the output cap's ...

Yes Alick, bias back to normal. Things were getting pretty erm "warm" with the upped QC and AD843..... you're perfectly ok with the QC increase as long as the trannies are sinked and with winter coming up you may be glad of the extra heat

Mike.

 

[Alick]

I seem to recall it was a hell of a struggle to get the heatsinks in, there being not much room for them. Hopefully, they'll come out easier, but I'll leave them in place until I know I'm happy with the sound (as I'm confident I will be) ... just in case. No output cap's - I can't wait.

EDIT: I just reread your post (properly this time). I'll probably try it with the increased QC to start with as it's already set up that way.

 

[PinkFloyd]

Quote:

Originally Posted by Paco Hello all, The AD8065 also was the best one in the MKI version to me (I was using it without input and output caps). I've finished with the “capless” MKII to use it as a preamp a couple of days ago and I have a question. I've measured TR2 and TR4 and they are at about 100ºC with case open and with an external temperature of about 25ºC. TR1 and TR3 are at about 90ºC. My question is... should I use heatsinks? I've seen that those transistors are rated at 150ºC but I think that once the case is closed and external temperature raise to 30ºC, they will be close to the maximum temperature... what do you think? Best regards, Paco

Hi Paco,

I've measured mine in the past with ambient temperatures ranging from 20C to 30C and the trannies remain around the 80C - 85C mark irrespective and that's without heatsinks in place.

I measured earlier with 2 meters and the results were pretty much the same:

Left Channel with ambient temp. of 22C results:

TR1 - 84C
TR3 - 82C
TR2 - 82C
TR4 - 74C

Right Channel with ambient temp. of 22C results:

TR1 - 80C
TR3 - 82C
TR2 - 79C
TR4 - 78C

Your 90C doesn't seem "excessive" but 100C is certainly pushing it a bit Paco. What values of resistor do you have in R5 & R6? They should be R5=120R R6=75R

If you've upped the quiescent current by replacing R5 & R6 with 160R and 100R then you'll have to heatsink the transistors as the increased QC will cause TR1, TR2, TR3 and TR4 to run a lot hotter. If this is the case then heatsinks are essential. If you are running the amp at stock QC (R5=120R & R6=75R) and the temperature is in the high nineties then It's essential you heatsink them to preserve the working life of them.

Another thing worth trying is measuring the voltage across the 10R resistors... according to WNA it should be around 500mV (measures 394mV on mine) also make sure the 10R resistors "are" 10R and not around 8R / 9R. If all the 10R resistors are lowered to say 8R this will also increase the QC.

If you have upped the QC then you either have to heatsink the transistors or revert back to the original stock QC (R5=120R & R6=75R)

Hope this is helpful.

Mike.

 

[PinkFloyd]

Quote:

Originally Posted by Alick I seem to recall it was a hell of a struggle to get the heatsinks in

Not much room to manoeuvre in the MKll with components being pretty close together... heatsinking is "doable" but not with off the shelf sinks (Maplin. Raid etc.) hopefully if there is ever a MKlll David will allow a bit more breathing space and allowance for larger caps such as Cerafine or Black Gates etc.

 

[PinkFloyd]

Measuring DC offset:

1: Multimeter set to 200mV on the DC range

2: Turn amp on... nothing connected (no 'phones or interconnects etc.)

3: probe between Sleeve and tip (ground & Right) on the headphone socket as shown:

The reading should ideally be 0.00mV but seldom is.... anything below 20mV is perfectly acceptable (closer to 0.00mV ideally). Once you are happy that the offset is within limits then repeat the test but this time with the source connected....... play a CD and then hit the pause button... now measure the offset with the CD player connected and disc on pause. If both measurements are pretty close to 0.00mV (say something like 3.74mV) then you're good to go! Another thing..... turn the volume control up and make sure the offset remains the same at all levels (it will if it's a FET opamp

Repeat the same for the left channel by probing between Sleeve and ring (ground & left) as shown:

The reading between channels will invariably be different ie: L = 3.74mV and R= -0.05mV nothing to worry about here, perfectly acceptable.

When probing connect the black lead to COM on your multimeter and the red lead to V. Always probe the sleeve (ground) with the black lead.

Something along those lines and most opamps like OPA627BP generally measure around 0.07mV (or thereabouts) and some measure a perfect 0.00mV. LM6171 will measure something crazy like 800mV hence the need for DC blocking output caps or trimpots. The AD-8065 I have just inserted into the WNA measure 0.01mV (left channel) and 2.06mV (right channel) so the offset is VERY low and well within limits.... that's not to say that all 8065's will measure the same and you'd have pick through a few of them to select two that measured 0.00mV.

That's the general idea and I'm sure there are more elaborate ways of measuring DC offset and better ways of explaining the procedure but this is all I've got in me at the moment (knackered) hope it helps.

Mike.

 

[Paco]

Hi Mike,
I've build the kit as it comes from WNA so I suppose that it has the standard quiescent current. I’m going to check this afternoon. In fact I've checked again and all the transistors are close to 100ºC (between 93 and 98 ºC). The only special think in my preamp is that the power supply has been fully developed to double monophonic (2 toroidals, two WNA Cascode power supplies and 2 rail splitters, one for each channel). I don’t know if this extreme power supply would cause this overheating but it shouldn’t as the Cascode power supplies are delivering +-24V. I’m going to put the heatsinks ASAP.

Best regards,
Paco

 

[Alick]

Paco, in the assembly instructions, David mentions that if you use a PSU providing greater than +/-15V, you will need heatsinks on Tr1 - Tr6. Sounds like you definitely need to fit them; check your instructions. I presume you're not using the rail splitter?

 

[Paco]

Hi Alick,
Yes, it looks like it's the way to go. I've seen that some of you use clip tipe heatsinks (I think Mike do it)... do you have farnell or RS references for them? I'm using two rail splitters, one for each channel. By the way, I like the LM6171 in the rail splitter.

Regards,
Paco

 

[12796]

I have tried opa627, ad825 and lm6171 in the rail splitter and preferred the latter. Has anyone tried ad843, ad847 or ad8065?

 

[PinkFloyd]

Quote:

Originally Posted by biovizier I have tried opa627, ad825 and lm6171 in the rail splitter and preferred the latter. Has anyone tried ad843, ad847 or ad8065?

Hi Nick,

I've tried 627, 6171, 825, 8610, 8605, 843, 134, 811 etc. in the rail splitter. One of them smoked like a chimney I think it may have been the 811 (or was it the 843) I can't remember.

ATM I've got 627 in the splitter along with your 8067 (gain of 10) and am just going to try the LM6171 with the 8067.

8067 sounds very promising I'll leave it in the circuit for a couple of days then have a darned good listen to it once the dust has settled....... the house is on the market and I'm busy beavering away regrouting, painting, jet washing, filling black bags, hoovering, nailing, screwing, fitting lights, wallpapering and a host of other equally mind numbing tasks.... this is a pretty hectic period of activity (for me) and I can't wait till it's all finished so I can have a good session or two with the 8067.

It "appears" stable with 10K resistors in R3 and sounded ok (listened to a couple of tracks) off out to grab an LM6171 and a bite to eat now and will report back probably early next week.

All the best.

Mike.

PS: the offset of your two chips is -0.00mV and 0.07mV